CHARACTERIZING FE(II)-POLYPYRIDINE EXCITED STATE EVOLUTION USING TRANSIENT ABSORPTION SPECTROSCOPY By Hayden Fendrick Beissel A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry – Doctor of Philosophy 2024 ABSTRACT Polypyridine based chromophores serve a key role in photo-induced electron transfer processes, with those based on iron(II) representing a promising earth-abundant alternative to the typical ruthenium-based complexes. However, the ultrafast deactivation of the metal-to-ligand charge transfer (MLCT) excited state in iron(II)-based complexes limits their ability to facilitate electron transfer. Studies have shown that the long-lived 5T2 excited state can induce electron transfer, yet there remain challenges in utilizing this excited state. To better understand how to manipulate the excited state dynamics of iron(II) polypyridyl chromophores, it is important to gain a fundamental understanding of excited state evolution in these complexes. A technique that is useful in this regard is variable temperature transient absorption (VT-TA) spectroscopy. By measuring the excited state lifetime as a function of temperature, both transition state and Marcus theory can be used to provide key information, such as the energy barrier or electronic coupling between states during a transition. Using VT-TA spectroscopy, the 5T2 excited state relaxation of two series of iron(II) polypyridines, one using substituted terpyridines and another using substituted bypiridines, was studied. VT-TA analysis on the terpyridine-based series provides experimental evidence that the relaxation from the 5T2 excited state occurs in the Marcus normal region. This provides important context on how changing the energy gap between the 5T2 excited state and 1A1 ground state will impact the lifetime of the 5T2 state. VT-TA analysis on the bipyridine-based complexes displays how substituent placement can be used to control electronic structure through leveraging resonance and induction effects. Furthermore, different substituent positions impact reorganization energy in different ways, which impacts the energy barrier for 5T2 relaxation. Lastly, ultrafast spectroscopy was used to study the vibrational coherence of several iron(II)-based polypyridine complexes during the early time post excitation. This reveals the vibrational motions that facilitate MLCT deactivation in these complexes. The information could be used to guide ligand design that will lengthen the MLCT lifetime. Copyright by HAYDEN FENDRICK BEISSEL 2024 ACKNOWLEDGEMENTS First, I want to thank my advisor, Jim McCusker, for his guidance and support. I would also like to thank former McCusker group members, especially Bryan, Sara, Jon, Matt, and Karl, for teaching me how to use all the techniques used throughout this work, from ultrafast spectroscopy to electrochemistry, and helping me begin my journey into graduate school. I would like to thank the current McCusker group members for all their help throughout the years. Atanu for helping to keep the laser systems running smoothly and being an excellent sounding board for ideas when I was stuck on a problem. Jyun and Bekah both for making the molecules for me to study and generally being helpful in lab. I want to thank my family for their unending support and always believing in me. Thank you to my parents for cultivating my love of science from a young age and always encouraging me to reach my goals. I thank my brothers, Noah and Ethan, for always pushing me to do my best and keeping my drive to succeed burning. Lastly, thank you to my amazing wife, Rebecca, for being with me every step of this journey. For keeping me sane, and my life full of joy. iv TABLE OF CONTENTS CHAPTER 1: EXCITED STATE DYNAMICS OF FE(II) POLYPYRIDYL COMPLEXES ….. 1 1.1 Introduction …………………………………………………………………………. 1 1.2 Photophysical Properties of Fe(II)-based Chromophores …………………………... 1 1.3 Electronic Structure/Symmetry Analysis …………………………………………… 7 1.4 Transient Absorption Spectroscopy ………………………………………………... 16 1.5 Contents of this Dissertation ………………………………………………………. 20 REFERENCES …………………………………………………………………………. 21 CHAPTER 2: COMPUTATIONAL AND SPECTROSCOPIC METHODS ………………….. 24 2.1 Introduction ………………………………………………………………………... 24 2.2 Computational Methods …………………………………………………………… 24 2.3 Electronic Absorption Spectroscopic Methods ……………………………………. 25 2.4 Nanosecond Transient Absorption Spectroscopic Methods ……………………….. 25 2.5 Ultrafast Transient Absorption Spectroscopic Methods …………………………... 26 REFERENCES …………………………………………………………………………. 31 CHAPTER 3: REVIEW AND ANALYSIS OF NON-RADIATIVE DECAY THEORIES …… 32 3.1 Introduction to Non-Radiative Decay Theories …………………………………… 32 3.2 Variable Temperature Analysis ……………………………………………………. 39 3.3 Determining the Energy Barrier …………………………………………………… 42 3.4 Resolving Discrepancies between Transition State and Marcus Theory ………….. 48 3.5 Concluding Thoughts ………………………………………………………………. 53 REFERENCES ………………………………………………………………………… 55 APPENDIX: SUPPLEMENTAL INFORMATION …………………………………….. 56 CHAPTER 4: VARIABLE-TEMPERATURE TIME-RESOLVED SPECTROSCOPY OF A HOMOLOGOUS SERIES OF FE(II)-TERPYRIDYL COMPLEXES ………………………… 61 4.1 Introduction ………………………………………………………………………... 61 4.2 Experimental Methods …………………………………………………………….... 64 4.3 Spectroscopic Results ……………………………………………………………… 68 4.4 Discussion …………………………………………………………………………... 91 4.5 Concluding Thoughts …………………………………………………………….... 98 4.6 Future Directions …………………………………………………………………... 98 REFERENCES ………………………………………………………………………... 101 APPENDIX: SUPPLEMENTAL INFORMATION …………………………………… 103 CHAPTER 5: LEVERAGING RESONANCE AND INDUCTION TO CONTROL THE ELECTRONIC STRUCTURE OF FE(II)-BIPYRIDYL COMPLEXES ……………………... 157 5.1 Introduction ………………………………………………………………………. 157 5.2 Experimental Methods ……………………………………………………………. 160 5.3 Experimental Results …………………………………………………………….... 164 5.4 Discussion ………………………………………………………………………… 181 5.5 Conclusions ………………………………………………………………………. 191 5.6 Future Directions …………………………………………………………………. 192 REFERENCES ………………………………………………………………………... 194 v APPENDIX: SUPPLEMENTAL INFORMATION …………………………………… 195 CHAPTER 6: USING VIBRATIONAL COHERENCE TO DEFINE THE REACTION COORDINATE OF MLCT DEACTIVATION WITHIN FE(II)-POLYPYRIDINE BASED CHROMOPHORES …………………………………………………………………………... 236 6.1 Introduction ………………………………………………………………………. 236 6.2 Experimental Methods ……………………………………………………………. 243 6.3 Results ……………………………………………………………………………. 251 6.4 Discussion ………………………………………………………………………… 272 6.5 Concluding Thoughts …………………………………………………………….... 279 6.6 Future Directions …………………………………………………………………. 280 REFERENCES ………………………………………………………………………... 281 APPENDIX: SUPPLEMENTAL INFORMATION …………………………………… 284 vi CHAPTER 1: EXCITED STATE DYNAMICS OF FE(II) POLYPYRIDYL COMPLEXES 1.1 Introduction This dissertation explores the photoinduced excited state dynamics of several six- coordinate transition metal-based chromophores, primarily focusing on those based on Fe(II). This is done with the goal of characterizing the excited state evolution within these complexes and understanding how to manipulate these processes with the goal of making these complexes more suitable for use in electron transfer processes, such as solar energy capture or photocatalysis. As the excited states in Fe(II)-based chromophores deactivate rapidly, the use of ultrafast transient absorption (TA) spectroscopy is necessary to examine these processes, and the technique is used throughout this dissertation. Therefore, in addition to explaining the electronic structure and photophysical properties of transition metal-based chromophores, this chapter will also discuss the basics of TA spectroscopy. 1.2 Photophysical Properties of Fe(II)-based Chromophores Chromophores are a class of molecules that can absorb a photon and store that energy in an electronic excited state where an electron is moved to an orbital of higher energy within the molecule. Chromophores based on transition metals are widely used for a variety of purposes, including use in solar energy conversion and photocatalysis.1-3 Transition metal-based chromophores can access several excited state transitions: such as d-d transitions where an electron is excited to a higher energy metal centered (MC) orbital, and metal-to-ligand charge transfer (MLCT) excited states, the excited state transition that most enables the above-mentioned processes. In this excited state, an electron is moved from a d-orbital on the metal to a π* orbital on a ligand. This leads to a charge separation that formally oxidizes the metal and formally reduces the ligand, which can then either act as an oxidant or reductant, respectively. This charge separated Figure 1.1: Schematic of a [Ru(bpy)3]2+ showing a metal-to-ligand charge transfer excited state transition. 1 state is illustrated for a typical transition metal-based chromophore, [Ru(bpy)3]2+ (where bpy is 2,2’-bipyridine), in Figure 1.1. The inverse electronic transition, a ligand-to-metal charge transfer (LMCT), is also possible. However, given that the molecules discussed in this dissertation engage in MLCT excited state transitions rather than LMCT transitions, this work will focus on the former excited state and not the latter. One of the benchmark transition metals used for transition metal chromophores is ruthenium, Ru(II)-based chromophores have been utilized in dye-sensitized solar cells (DSSCs) and as photocatalysts.1-3 Two things that contribute to the success of Ru(II)-based chromophores are that the MLCT absorption peaks rest within the visible region and that the MLCT excited state has a long lifetime, on the order of microseconds.4 However, there is a large concern with relying on ruthenium for solar energy capture: sunlight has a photon flux around 100 mW cm-2. This means that to capture large amounts of energy from sunlight, a large surface area is required which means that solar energy capture becomes a very material intensive process.5 Unfortunately, ruthenium is one of the rarest elements found in the Earth’s crust.6 This scarcity makes relying on ruthenium a non-viable option for globally scaled light harvesting. This creates a need for chromophores based on Earth abundant metals, like first-row transition metals. Iron immediately stands out as a candidate since it is Earth-abundant and its valence shell is isoelectronic to that of ruthenium. This means that a chromophore based on Fe(II) has access to the same electronic states as those based on Ru(II), including a MLCT excited state, which is primarily responsible for driving electron transfer interactions such as semiconductor injection. However, in 1998 Ferrere and Gregg found that a DSSC using an Fe(II) bipyridyl chromophore was notably less efficient than one using Ru(II).7 These results raise important questions: why is this happening and how can efficacy be improved? The culprit behind the reduced efficiency of Fe(II)-based chromophores is a drastically reduced MLCT excited state lifetime. The prototypical complex for Ru(II)-based chromophores is Ruthenium tris-bipyridine ([Ru(bpy)3]2+), so comparing that to [Fe(bpy)3]2+ makes a logical starting point for investigating the utility of first-row transition metal-based chromophores. After absorption, [Ru(bpy)3]2+ undergoes a spin allowed transition to the 1MLCT excited state which undergoes inter-system crossing (ISC) to the 3MLCT in 100 femtoseconds (fs),8,9 the 3MLCT state relaxes directly back to the 1A1 ground state in approximately 1 microsecond (μs) at room temperature.4,10,11 This means that [Ru(bpy)3]2+ remains in a charge separated state for about 1 μs 2 which lasts long enough for the complex to act as a photosensitizer.10 In [Fe(bpy)3]2+, after being excited to the 1MLCT excited state, the complex undergoes ISC to the 3MLCT excited state within 20 fs,12,13 from there the complex undergoes internal conversion (IC) through a 3T MC excited state followed by ISC to the 5T2 excited state within roughly 200 fs.14-16 The complex then relaxes from the 5T2 excited state in about 1 nanosecond (ns) at room temperature.4,17 This leaves [Fe(bpy)3]2+ with a charge separated state that only lasts about 200 fs, which is too short to be useful in electron transfer applications. Electron injection into a semiconductor from the 3MLCT state occurs on a picosecond (ps) timescale and molecular diffusion in solvent occurs on a ns timescale, meaning the ~200 fs lifetime of the Fe(II) chromophore MLCT excited state is too short to be useful for solar energy generation or photocatalysis.18,19 Regarding the identity of the MC 3T excited state the Fe(II) complex relaxes through, there are two possible states this could be: 3T1 or 3T2. A study by Vura-Weis and coworkers used ultrafast X-Ray spectroscopy supported with ligand-field multiplet (LFM) theory simulations to study the ultrafast excited state dynamics of [Fe(phen)]2+ (where phen is 1,10-phenanthroline).20 A feature with a ~40 fs time constant was identified to belong to the triplet intermediate. Vura-weis and coworkers suggested that the triplet intermediate is more likely to be 3T1 by comparing simulated Figure 1.2: Potential energy surface diagrams illustrating the excited state evolution of [Ru(bpy)3]2+ (left) and [Fe(bpy)3]2+ (right). Within the insets are a depiction of the valence electronic structure for each complex displaying the lowest energy excited state, 3MLCT for [Ru(bpy)3]2+ and 5T2 for [Fe(bpy)3]2+. 3 difference absorbance (ΔA) spectra with experimentally determined spectrum. The experimental spectrum displays two main peaks, a sharp feature around 57 eV and a broader feature near 64 eV. Both features are replicated in the simulated spectrum for 3T1. However, it appears that the experimental spectrum displays features present in both simulated spectra. A third, weaker a peak around 60 eV is present in the experimental spectrum, and this peak is only replicated in the 3T2 simulation (found in the supplemental information, SI, of the paper), only a weak shoulder is present in the 3T1 simulation. This is shown in Figure 1.3: Figure 1.3: Experimental (right) and simulated (left) ultrafast X-ray spectra published by Vura- Weis and coworkers.20 The experimental spectrum displays a low intensity peak near 60 eV, which is only simulated for 3T2, the 3T1 simulation exhibits a shoulder in that region. As such, it is reasonable to assert that the molecule undergoes a branching relaxation path where both triplet states serve as intermediates within the excited population. The relaxation pathways for both [Ru(bpy)3]2+ and [Fe(bpy)3]2+ are summarized below: [Ru(bpy)3]2+: 1MLCT 100 𝑓𝑠 𝐼𝑆𝐶 → 3MLCT 1 𝜇𝑠 𝑅𝑎𝑑𝑖𝑎𝑡𝑖𝑣𝑒 𝐷𝑒𝑐𝑎𝑦 → 1A1 [Fe(bpy)3]2+: 1MLCT 20 𝑓𝑠 𝐼𝑆𝐶 → 3MLCT ~150 𝑓𝑠 𝐼𝐶 → 3T1/3T2 40 𝑓𝑠 𝐼𝐶 → 5T2 1 𝑛𝑠 𝑁𝑜𝑛𝑟𝑎𝑑𝑖𝑎𝑡𝑖𝑣𝑒 𝐷𝑒𝑐𝑎𝑦 → 1A1 0.2 Å Fe-N bond contraction 0.2 Å Fe-N bond extension The MLCT excited state of [Fe(bpy)3]2+ deactivates so rapidly because Fe(II)-based complexes have a weaker ligand field (LF) strength compared to their Ru(II)-based counterparts. This stems from a phenomenon known as the primogenic effect. As a first-row transition metal, the valence shell of Fe(II) is the 3d orbital while the valence orbital of Ru(II) is the 4d orbital, which extends further from the nucleus than the 3d orbital thanks to the presence of a radial node.5 The valence electrons sitting closer to the nucleus in Fe(II) leads to weaker orbital overlap, and 4 therefore weaker covalent bonds, with the ligands in Fe(II)-based complexes. This results in a lower energy gap between the LF orbitals (t2g and eg*) than what is observed in Ru(II)-based complexes. The reduced energy gap between the t2g and eg* orbitals means that the metal centered LF excited states, which rely on the population of eg* orbitals, are reduced in energy; lower in energy than the MLCT excited states (which rely on the population of a π* orbital on the ligand). The LF state energy reduction, coupled with the Frank-Condon displacement caused by geometric changes accompanied with eg * orbital population, provides a cascade pathway for ultrafast relaxation to the 5T2 excited state. In Ru(II)-based chromophores, the LF excited states are higher in energy than the MLCT excited states, so no such deactivation pathway exists. The source of the primogenic effect and how it influences electronic structure is illustrated in Figure 1.4. Figure 1.4: Figures illustrating the source and impact of the primogenic effect published by McCusker.5 The left side shows the valence orbital radii for a first-row transition metal (A) and a second-row transition metal (B). The presence of a radial node in the 4d orbitals pushes the radius further from the nucleus. The right side shows a Tanabe-Sugano diagram for octahedral d6 complexes. Shown in green is the region Fe(II) polypyridines occupy, with LF states lower in energy than the MLCT state, while for Ru(II) polypyridines, shown in blue, the MLCT excited state is the lowest energy excited state. Having established why Fe(II)-based chromophores are less efficient photosensitizers than Ru(II)-based chromophores, now we can turn to the second question posed earlier: how can their efficacy be improved? There are two main approaches researchers have taken to do so: an electronic structure approach and a kinetic approach. The electronic structure approach attempts to increase the LF strength and in turn, raise the energy of LF excited states above that of the 5 MLCT excited states.5,21 There are several means in which researchers work towards this goal. One way is to design high symmetry ligands, such as 2,6-bis(2-carboxypyridyl)pyridine (dcpp).22 As will be explored in the next section, tris-bipyridine ligand structures reduce the symmetry of the complex from octahedral. This lower symmetry reduces the degeneracy of the t2g and eg* orbitals, splitting them further. This results in a lower energy gap between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), reducing the ligand field strength compared to a complex with octahedral geometry. By adding carbonyl groups between the pyridine rings on the ligand, dcpp grants complexes more octahedral geometry. This succeeded in increasing the LF strength of [Fe(dcpp)2]2+, but not enough to generate an MLCT excited state lifetime comparable to Ru(II)-based complexes. Another method of raising the energy of LF excited states is to use strongly σ-donating ligands, such as carbenes.23-25 These ligands bind to the metal with carbon, which increases the σ-donating nature of the ligand compared to something like bpy, which only binds through nitrogen. This is because carbon is less electronegative than nitrogen, causing the HOMO energy of the lone pair on carbon to be higher than that of nitrogen. Increasing the σ-donating ability of the ligand leads to stronger, more covalent bonds between the metal and ligand, which results in greater splitting of the eg orbitals, stabilizing the bonding eg orbitals and destabilizing the eg* anti-bonding orbitals. This results in a stronger LF, raising the energy of the LF excited states. This method has seen some success, greatly increasing the MLCT excited state lifetime, by hundreds of picoseconds in some cases.26 These results show that there is much to gain from characterizing the LF states and learning how to manipulate the electronic structure of Fe(II)-based chromophores. Furthermore, Fe(II)-based chromophores have seen some utility as photoredox catalysts.27,28 Quenching studies by McCusker and Woodhouse display that electron transfer is occurring from a LF excited state.29 This development could lead to photoredox reactions using Fe(II)-based chromophores that would not be possible using a Ru(II)-based chromophore. This provides additional incentive to characterize the electronic structure and study how to manipulate it. Lastly, the kinetic approach to increase the MLCT lifetime relies on identifying the reaction coordinate of MLCT deactivation and designing ligands to manipulate it with the goal of slowing the rate of relaxation from the MLCT excited state to the LF excited states. For example, as an Fe(II)-based chromophore enters the 5T2 excited state, two electrons enter the eg* anti-bonding 6 orbitals. This results in a symmetric 0.2 Å elongation of the Fe-N bonds.30,31 In addition, ultrafast X-ray studies by Chergui and coworkers on [Ru(bpy)3]2+ display a roughly 0.03 Å contraction of the Ru-N bonds.32 If this geometric change is comparable to what occurs in [Fe(bpy)3]2+, this means relaxation from the 3MLCT excited state to the 5T2 excited state is accompanied by a ~0.23 Å elongation of the Fe-N bonds. In 2020, McCusker and coworkers utilized a “cage” ligand designed to restrict the elongation of the Fe-N bonds, disrupting the population of the 5T2 excited state without interfering with the geometry of the MLCT excited state.33 This proved to be successful as the MLCT lifetime was increased to about 2.5 ps. In order to design ligands for this purpose, identifying the reaction coordinate is crucial. As such, a later chapter of this dissertation is dedicated to characterizing the reaction coordinate for several Fe(II)-based chromophores. 1.3 Electronic Structure/Symmetry Analysis The molecules discussed in this work broadly fall under one of two geometric classifications: bis-tridentate, such as [Fe(terpy)2]2+ (where terpy is 2,2’:6’,2”-terpyridine), or tris- bidentate, such as [Fe(bpy)3]2+. As mentioned earlier, neither of these molecular structures are truly octahedral, the symmetry of bis-tridentate complexes is D2d, and the symmetry of tris-bidentate complexes is D3. However, for the sake of simplicity, it is typical to treat these molecules as octahedral (Oh) when discussing the electronic structure for these molecules, and this dissertation will be no different. Treating all molecules as Oh will make drawing comparisons between different structures much clearer. That said, in this section the electronic structures of all three symmetries will be compared: Oh, D2d, and D3. This will serve as a reference to display more accurately what each excited state discussed in future chapters looks like. In addition, as studying the impact of varying σ and π effects from the ligand on excited state dynamics is a major focus on future chapters, this section will also serve as an introduction to how these ligands will affect the electronic structure. Figure 1.5: Illustration of the bis-tridentate complex [Fe(terpy)2]2+ (left) and the tris-bidentate complex [Fe(bpy)3]2+ (right). 7 Starting with the electronic structure for a d6 complex with Oh symmetry, using the ferrocyanide ion ([Fe(CN)6]4-) as an example. Analysis will initially focus on σ-interactions, with π-interactions addressed later. Character table analysis on the reducible representations of the lone pairs on the ligands reveals the symmetries for the ligands: a1g, eg, and t1u. The symmetries of the Fe valence orbitals are as follows: a1g for the 4s orbital, eg and t2g for the 3d orbitals, and t1u for the 4p orbitals. It is worth noting that all 3d orbitals have gerade symmetry, so LF states of these complexes will never have ungerade symmetry. Orbitals with matching symmetry mix, producing both a bonding and antibonding orbital. The Fe 3d orbitals with t2g symmetry have no match from the ligand, so they will not mix and become non-bonding orbitals. A molecular orbital (MO) diagram for the molecule is provided below: y g r e n E Figure 1.6: Schematic of a MO diagram for the σ-bonds of a low-spin d6 complex with Oh symmetry. This model is not to scale, as energy levels of orbitals are estimated. The frontier orbitals are highlighted by the green box. This model displays a low-spin electronic configuration; [Fe(CN)6]2+, as well as the Fe(II) polypyridyl complexes discussed in this dissertation, is a strong field complex. This means that the energy gap between the t2g and eg* orbitals, Δo or 10Dq, is greater than the energetic cost of pairing two electrons within the same orbital. As a result, the ground state in these complexes is the 1A1 state. 10Dq is greatly impacted by the strength of the σ-bond formed between the ligand lone pairs and the 3d σ-orbitals (in this geometry that is the 3𝑑𝑧2 and 3𝑑𝑥2−𝑦2 orbitals). The carbon atom in CN- is not very electronegative, making it a strong σ-donor. This leads to a strong bond formed between the CN- lone pairs and the Fe(II) orbitals, most importantly the eg orbitals. As 8 having a stronger bond causes greater splitting between the bonding and anti-bonding orbitals, 10Dq is increased as a result. Furthermore, the reverse is true for ligands that are weaker σ-donors. Take the hexaaqua complex ([Fe(H2O)6]2+) as an example, since oxygen is more electronegative than carbon, the lone pairs are held more tightly to the ligand than the carbon lone pairs on cyanide. This results in a weaker bond formed between the metal and the water molecules, and therefore a weaker 10Dq. Thanks to this, and weak-field π interactions which will be explored later, [Fe(H2O)6]2+ is known to be high-spin. Meaning that the ground state of the complex is 5T2, as illustrated in Figure 1.7. Figure 1.7: Schematic of MO diagrams displaying the effect of weaker σ-bonding between the MC and ligand. [Fe(CN)6]4- (left) has a strong σ-donating ligand and is a strong field complex with 1A1 as the ground state. [Fe(H2O)6]2+ (right) has a weaker σ-donating ligand. Until this point, only the effects that σ-bonds have on electronic structure have been discussed. However, π-interactions must also be considered. With an octahedral geometry, on-axis p-orbitals on the ligands can mix with the MC 3𝑑𝑧2 and 3𝑑𝑥2−𝑦2 orbitals, however this is not true for the remaining 3𝑑 orbitals. The ligand orbitals along the bonds rest within the nodes of the 3𝑑𝑥𝑦, 3𝑑𝑥𝑧, and 3𝑑𝑦𝑧 orbitals, as illustrated in Figure 1.8. Therefore, when only considering σ- interactions, those orbitals have t2g symmetry distinct from those of the ligand orbitals and do not mix. However, the off-axis ligand orbitals can mix with those 3𝑑 orbitals. Therefore, the t2g orbitals split upon interacting with those π-orbitals and depending on whether the ligand acts as a π-donor or π-acceptor, this splitting will either decrease or increase 10Dq. A π-donating ligand has filled, lower energy π-orbitals. After splitting, the anti-bonding t2g orbital is now higher in energy than the non-bonding orbital was, and since it is filled with electrons from the ligand π-orbitals, this 9 Figure 1.8: Schematic of Fe(II) 3𝑑 orbitals interacting will orbitals from a ligand. A: The 3𝑑𝑧2 orbital mixing with a ligand pz orbital. B: The 3𝑑𝑥2−𝑦2 orbital mixing with a ligand py orbital, forming a σ-bond. C: The same py orbital cannot mix with the 3𝑑𝑥𝑦 orbital, as it rests within an angular node of the orbital. The same is true for the 3𝑑𝑥𝑧 and 3𝑑𝑦𝑧 orbitals. D: The off-axis px orbital on the ligand mixing with the 3𝑑𝑥𝑦 orbital, forming a π-bond. becomes the new HOMO and reduces 10Dq. A π-accepting ligand has empty, higher energy π- orbitals. After splitting, the anti-bonding t2g orbitals are higher in energy than the eg* orbitals, while the bonding t2g orbitals fall lower in energy than the original non-bonding orbitals, increasing 10Dq. This is illustrated in the diagram below: y g r e n E Figure 1.9: Schematic showing the effects of π-donors and π-acceptors on ligand field strength. A computational study by Ashley and Jakubikova on [Fe(bpy)3]2+ revealed that the bpy ligand displays both π-donating and π-accepting character.34 This allows different substituents to 10 modulate the π-character of the ligand through resonance structures. Since the π-system of bpy is delocalized, an electron donating substituent, such as a methoxy group, leads to a resonance structure that pushes electrons into the π-orbitals on the nitrogen atom within the pyridine ring. In other words, the substituent increases the π-electron density at the bonding site of the ligand, Figure 1.10: Schematic displaying how substituents impact the π-donating/accepting nature of the ligand. An electron donating substituent gives the ligand more π-donating character (top), while an electron withdrawing substituent gives the ligand more π-accepting character (bottom). making it behave as a π-donor and reduces the 10Dq of the complex. Inversely, an electron accepting substituent, such as a cyano group, similarly pulls π-electron density from the binding site of the ligand, causing it to behave as a stronger π-acceptor. These resonance structures are shown in Figure 1.10. Given that analogous resonance structures are available to all polypyridyl ligands, such as terpyridine, it is reasonable to assume that similar behavior should be expected across various polypyridyl complexes, including those based on Fe(II). Similarly, substituents can alter the σ-donating strength of the ligand through inductive effects. This behavior is of key importance to the content of this dissertation. By altering the nature of the π-interactions, 10Dq can be fine-tuned while maintaining a low-spin (strong field) electronic structure. Since the 5T2 excited state involves two electrons populating the eg* orbitals, the energy gap between the eg* and t2g orbitals determines the free energy difference (ΔGo) between the 5T2 excited state and the 1A1 ground state. This means that ΔGo can be modulated through ligand- substitution, making it possible to examine various aspects of the 5T2 → 1A1 transitions as a function of ΔGo. Exploring such trends will be extensively discussed in later chapters of this dissertation. 11 The above discussion assumes that the complexes involved have Oh symmetry, however this is not truly the case for the molecules of interest in this work. As mentioned earlier, the symmetry point group for bis-tridentate complexes (such as [Fe(terpy)2]2+) is D2d and for tris- bidentate complexes (such as [Fe(bpy)3]2+) the point group is D3. As the symmetry of each structure model lowers from Oh, the manner each orbital interacts with the ligands alters and the electronic structure of the molecular orbitals changes. For example, with Oh symmetry all three MC 4𝑝 orbitals were degenerate (meaning they have identical symmetry, in this case t1u). However, with both D2d and D3 symmetry that degeneracy is reduced: the 4𝑝𝑥 and 4𝑝𝑦 orbitals are still degenerate, but the 4𝑝𝑧 orbital has different symmetry because of the shifted structure. This change in electronic structure also results in different electronic states being accessible. Both bis-tridentate and tris-bidentate polypyridyl complexes still have 1A1 as the ground state, and the 1MLCT and 3MLCT states are still accessible. The only states that differ are the LF excited states. In order to identify these states, the modified MO diagrams for each symmetry will be discussed, starting with bis-tridentate complexes. Figure 1.11 provides a model MO diagram for [Fe(terpy)2]2+. In the Oh diagram, it is simple to determine bonding and anti-bonding orbitals as pairs of degenerate orbitals mix into clear y g r e n E Figure 1.11: Schematic of a MO diagram for a low-spin bis-tridentate complex with D2d symmetry. As with Figure 1.4, this model is not to scale. The energy levels of orbitals are estimated based on DFT calculations performed by Gawelda and coworkers.35 The frontier orbitals are again highlighted by the green box. 12 bonding and anti-bonding pairs. However, in the case of D2d symmetry, many orbitals are mixing rather than just two. As a result, it becomes more difficult to predict whether some orbitals will be overall bonding or antibonding. The ordering of the MO’s in this diagram was influenced by a study on [Fe(terpy)2]2+ using density functional theory (DFT) calculations and X-Ray spectroscopy performed by Gawelda and coworkers.35 The reduction in orbital degeneracy also has a significant impact on the frontier orbitals, which is where the difference in LF excited states stems from. To start, in Oh symmetry, the LUMO is doubly degenerate which results in two electrons moving to the higher energy orbital in the weak field configuration leading the high-spin ground state being a quintet. However, in D2d symmetry, the LUMO is non-degenerate. So, in a weak field, only one electron moves to the higher energy orbital, so the high-spin ground state in a weak field bis- tridentate complex is a triplet. Regardless, the molecules discussed in this dissertation are low-spin (determined via NMR), so this distinction is not relevant to the work presented here. Regarding the LF excited states, the reduced degeneracy leads to structural differences between the states that form. With Oh symmetry, both “eg*” orbitals being degenerate means that the electronic population of those orbitals results in symmetric bond-lengthening. In bis-tridentate complexes the population of the higher energy orbitals now results in Jahn-Teller (JT) distortion. As can be seen in Figure 1.11, the 3𝑑𝑥𝑦 orbital has b2 symmetry. So, when the 3b2 orbital, which is anti-bonding with respect to the 3𝑑 orbitals, is populated as the molecule enters the 3A2 excited state, the bonds involving the 3𝑑𝑥𝑦 orbitals weaken. This results in those bonds lengthening, while the bonds involving the still unpopulated 3a1 orbital (which includes the 3𝑑𝑧2 orbital) do not lengthen to such a degree. The inverse is true for the 3B1 excited state. Therefore, if the complex is in the 3A2 state, equatorial JT distortion is expected and if it is in the 3B1 state, axial JT distortion is expected. It is also worth noting that with a bis-tridentate ligand structure, JT distortion also results in the Neq-Nax-Neq o angle changing; contracting with axial distortion and extended with equatorial distortion. Presently available data cannot distinguish between which of these two states are populated during excited state relaxation from the 3MLCT state, so similar to [Fe(bpy)3]2+, it will be assumed that both states serve as intermediates as the complex relaxes to the quintet excited state. DFT calculations performed by Gawelda and coworkers identified two viable quintet excited states: 5E and 5B1.35 The difference between these states is the energetic ordering of the nominally “t2g” orbitals, which leads to differences in how the bonds lengthen (all bonds in the 13 Figure 1.12: The electronic configurations of the ground state and the two lowest energy triplet and quintet excited states present in [Fe(terpy)2]2+.35 quintet states are longer than in the ground state, the difference is in how much they do so). In the 5E case, the 3𝑑𝑥𝑧 and 3𝑑𝑦𝑧 orbitals stabilize, resulting in equatorial JT distortion since bonds along the z-axis are weakened less that those along the xy-axes. In the 5B1 state, those same orbitals are now destabilized, which causes axial JT distortion as bonds along the z-axis are weakened more. Ultrafast X-ray data collected by Gawelda and coworkers revealed the quintet state exhibits equatorial JT distortion, suggesting that the long-lived excited state in [Fe(terpy)2]2+ complexes is the 5E excited state. The structures of the electronic states discussed are presented in Figure 1.12. With this information in mind, a bis-tridentate specific excitation relaxation pathway is constructed. [Fe(terpy)2]2+: 1MLCT 𝐼𝑆𝐶 → 3MLCT 𝐼𝐶 𝑤𝑖𝑡ℎ 𝐴𝑥𝑖𝑎𝑙 𝐽𝑇 𝑑𝑖𝑠𝑡𝑜𝑟𝑡𝑖𝑜𝑛 → 3A2 𝐼𝐶 𝑤𝑖𝑡ℎ 𝐸𝑞𝑢𝑎𝑡𝑜𝑟𝑖𝑎𝑙 𝐽𝑇 𝑑𝑖𝑠𝑡𝑜𝑟𝑡𝑖𝑜𝑛 → 3B1 0.2 Å Fe-N bond extension 𝐼𝐶 → 5E 𝑁𝑜𝑛𝑟𝑎𝑑𝑖𝑎𝑡𝑖𝑣𝑒 𝐷𝑒𝑐𝑎𝑦 → 1A1 0.2 Å Fe-N bond contraction The next ligand structure to discuss is the tris-bidentate model, which corresponds to [Fe(bpy)3]2+. As mentioned above, the symmetry of this structure is D3. A model MO diagram is provided in Figure 1.13. Similar to the D2d electronic structure seen in bis-tridentate systems, the D3 “t2g” orbitals have reduced degeneracy, however the “eg” orbitals retain their degeneracy. This means that as the tris-bidentate molecule populates LF excited states, the geometry will exhibit behavior like the Oh complex, rather than display JT distortion bonds will lengthen symmetrically. Therefore, the reduction in symmetry from Oh has much less of an impact on tris-bidentate complexes than it does for bis-tridentate complexes. The available LF excited states are shown in Figure 1.14. The lowest energy triplet LF excited state is 3A2 and the quintet state is 5A1. Higher energy states where an electron moves from the 3𝑑𝑧2 orbital, which has a1 symmetry, moves to the higher energy e orbitals, 3E or 5E, are possible and unique from states available to Oh complexes. However, they would be too high in energy to be considered as the identities of the triplet 14 y g r e n E Figure 1.13: Schematic of a MO diagram for a low-spin bis-tridentate complex with D2d symmetry. As with Figure 1.4, this model is not to scale. The frontier orbitals are again highlighted by the green box. Figure 1.14: The electronic configurations of the ground state and lowest energy LF excited states present in [Fe(bpy)3]2+. intermediate or the long-lived quintet excited state. A symmetry specific relaxation pathway is provided below. [Fe(bpy)3]2+: 1MLCT 20 𝑓𝑠 𝐼𝑆𝐶 → 3MLCT ~150 𝑓𝑠 𝐼𝐶 → 3A2 40 𝑓𝑠 𝐼𝐶 → 5A1 0.2 Å Fe-N bond extension 1 𝑛𝑠 𝑁𝑜𝑛𝑟𝑎𝑑𝑖𝑎𝑡𝑖𝑣𝑒 𝐷𝑒𝑐𝑎𝑦 → 1A1 0.2 Å Fe-N bond contraction Table 1.1 provides a reference tracking the lowest energy LF excited states alongside the nearest analog at varying symmetry models. As mentioned before, Fe(II)-polypyridines are largely treated as having octahedral symmetry. This assumption is largely reasonable, though it is important to note that population of the LF excited states in complexes with D2d symmetry leads 15 Table 1.1: A guide providing the lowest energy LF states present in Fe(II)-polypyridines ordered by symmetry. to JT distortions, which does not need to be considered for complexes with D3 symmetry. Regardless, throughout the remainder of this dissertation, each molecule discussed will be treated as octahedral; not to reduce the complexity of excited state dynamics, but so that functionally equivalent language may be used to describe these systems. The following terms will be used to describe the singlet, triplet, and quintet electronic states in these molecules: 1A1; 3T1,2; and 5T2. 1.4 Transient Absorption Spectroscopy Since the processes explored in this work deactivate rapidly, techniques capable of ultrafast time resolution (sub-picosecond) are needed to accurately measure the lifetimes and coherent features of the excited states studied. As a result, pump-probe transient absorption (TA) spectroscopy is used throughout this dissertation. This technique employs two laser pulses to study a sample: typically, the molecule of interest in an acetonitrile solution. The first, higher power pulse (called the pump) generates an excited state population within the solution. After a controllable time delay (Δt), a second, lower power pulse (called the probe) irradiates the already excited population, promoting a sub-set population to a higher energy excited state. By altering Δt, changes in the absorbance of the probe can be measured as a function of time.36 The experimental set-up is summarized in Figure 1.15. Since the differences in absorbance between the excited population and the ground population are often very small, TA data is presented as the difference in absorbance between the pumped sample and the unpumped sample (ΔA). 𝛥𝐴(𝑡) = 𝐴𝑃𝑢𝑚𝑝𝑒𝑑(𝑡) − 𝐴𝑈𝑛𝑝𝑢𝑚𝑝𝑒𝑑 = − log ( 𝐼𝑃𝑢𝑚𝑝𝑒𝑑(𝑡) 𝐼𝑈𝑛𝑝𝑢𝑚𝑝𝑒𝑑 ) 1.1 16 Figure 1.15: A: Schematic illustrating a TA spectroscopy experiment. Both pulses overlap within the sample holder to ensure the probe pulse irradiates an excited population. The pump is blocked after the sample to ensure only the probe reaches the detector, a photodiode in this case. To determine AUnpumped, the pump is blocked before the sample, and IUnpumped is then measured. B: Energy level diagram showing possible signal sources in a pump probe experiment. Excited state absorption is shown with the solid red line and stimulated emission in the dashed line. It is worth noting that the intensity of ΔA is dependent on the percentage of sample excited by the pump pulse, this is impacted by the power of the pump pulse. Additionally, as molar absorptivity is dependent on the nature of the electronic state, ΔA can also change in magnitude or sign as a complex transitions from one state to the next. There are three basic classifications for ΔA: excited state absorption (ESA), ground state bleach (GB), and stimulated emission (SE).36 ESA occurs when the pumped (excited) sample absorbs more light than the unpumped sample. For example, this occurs when probing [Fe(bpy)3]2+ at 600 nm. In the ground state, [Fe(bpy)3]2+ absorbs weakly at 600 nm, however the bipyridine radical absorbs strongly at 600 nm.37 Since the MLCT excited state involves transferring an electron from the MC to a ligand, creating a bpy radical, the resulting signal will be positive as the excited sample will absorb more light than the unpumped sample. However, as discussed previously, the MLCT excited state in [Fe(bpy)3]2+ is short lived (>200 fs), eventually relaxing to the 5T2 LF excited state. Absorbance in this state at 600 nm is weaker than the absorbance in the ground state. As a result, the signal in this state will be negative, emblematic of GB. This showcases how the signal of ΔA is not only time dependent, but also sensitive to the electronic state. These features of TA spectroscopy and ΔA are displayed in Figure 1.16. The spectrum included in Figure 1.16B additionally highlights why ultrafast TA 17 Figure 1.16: A: An illustration depicting ESA and GB. Above shows the absorbance for an excited state (red) and ground state (blue). At wavelengths where the excited state absorbs more than the ground state, ΔA will be positive. The inverse results in negative ΔA. B: TA Spectrum for [Fe(bpy)3](PF6)2 in acetonitrile solution, the pump was at 490 nm and the probe was at 600 nm. The data is in red with a fit shown in the blue dashed line, the features before the fit are caused by the solvent. Initially, ΔA is positive while the MLCT states are populated. As the excited population relaxes to LF excited states, ΔA becomes negative. spectroscopy is needed to explore the early excited state dynamics. The fit of the data gave a time constant for MLCT relaxation as 150 ± 10 fs. If the instrument being used had a time resolution slower than a picosecond, this relaxation would be utterly unobservable. Lastly, SE occurs when an excited complex interacts with a probe photon and relaxes to a lower energy state, releasing a photon in the process. Since the light is emitted in the same direction as the probe photon, this increases the amount of light reaching the detector, resulting in negative ΔA signal. As SE and GB both result in negative signals, distinguishing between the two is important. Regarding the molecules in this dissertation, SE is unlikely to be the source of negative ΔA in TA spectra. The long-lived excited states in Fe(II)-polypyridines are high spin LF states. The transition from the high spin states to the ground state is both Laporte and spin forbidden, therefore emission from the long-lived excited state is highly unlikely. The research described in this dissertation utilizes three systems: two ultrafast systems capable of producing sub-picosecond pulses and a nanosecond system with a roughly 5 ns instrument response function (IRF). The nanosecond system is used to study processes with longer lifetimes, but while the methodology in producing laser pulses is different, the principles behind TA spectroscopy remain the same. The two ultrafast laser systems (called WE and RR), differ in 18 the pulse durations generated and delay stages (the delay stage is a device that controls the distance the pump beam must travel, and therefore changes Δt between pump and probe). The system called WE generates laser pulses roughly 150 fs in duration and has a delay stage capable of creating Δt between pump and probe as long as 13 ns. RR has a shorter delay stage, and therefore cannot study longer processes, but can generate pulses as short as 40 fs. This is because RR utilizes a prism compression system to “anti-chirp” the pulse.38 Chirp refers to the phenomenon of laser pulses lengthening as they interact with optics (lenses, mirrors, etc.), because the index of refraction of any transparent medium is wavelength dependent. This results in some wavelengths lagging behind others in the pulse, making the pulse spatially and temporally longer. The prism compressors invert the chirping, placing the “lagging” wavelengths ahead far enough that by the time the pulse reaches the sample, the chirping compresses the pulse. A byproduct of generating temporally short pulses is that the pulses become spectrally broad. Sufficiently broad pulses can excite a population into multiple vibrational levels within an excited state.39 The superposition of these vibronic states results in what is called a “wave-packet,” which oscillates along the reaction coordinate. This vibrational coherence results in observable oscillations in the TA spectrum.33 The utility of these oscillations is important to the contents of Chapter 6 of this dissertation and will be explored further there. The last point worth exploring in this section is the combination of TA spectroscopy with variable temperature (VT) techniques. By using an optical dewar and a cryogen, such as liquid nitrogen, the temperature of the sample can be controlled while collecting TA spectroscopy data. By collecting ground state recovery (GSR) scans at multiple temperatures, it becomes possible to Figure 1.17: VT-TA ground state recovery for [Fe(4’-OH-terpy)2](PF6)2 in acetonitrile solution. 19 study excited state relaxation through the lens of temperature dependent non-radiative decay theories, such as Arrhenius theory. This analysis, and the relationship between non-radiative decay theories, will be explored in detail in Chapter 3. 1.5 Contents of this Dissertation The work in the remainder of this dissertation explores the excited state dynamics of several transition metal-based polypyridyl chromophores, primarily focusing on chromophores based on Fe(II). This is done with the goal of characterizing the transitions between electronic states and understanding how to manipulate those transitions. Chapter 2 will describe the experimental procedures that are utilized throughout this work. Chapter 3 will dive into the non-radiative decay theories that are used to describe the excited state dynamics of the molecules studied. It will begin with a review of Arrhenius, transition state, and Marcus theories before delving into how these techniques are used with VT-TA techniques and exploring which theories are best used to describe the molecules studied. The following two chapters utilize VT-TA spectroscopy to analyze excited state relaxation through the lenses of transition state theory and Marcus theory. Chapter 4 will study a series of substituted [Fe(terpy)2]2+ complexes through VT-TA spectroscopy. The VT-TA spectroscopic data will be analyzed to characterize ground state recovery. In Chapter 5, a similar series of substituted [Fe(bpy)3]2+ is studied with VT-TA spectroscopy. This analysis focuses on exploring how altering the position of the substituent impacts excited state transitions. The chapter will begin by exploring resonance and inductive effects of substituents and how changing the position of the substituent changes the resonance structures of the ligand π-system. Then the VT-TA spectroscopic data of the series will be analyzed similar to what was done in Chapter 4. Chapter 6 will use ultrafast vibrational coherence to characterize the reaction coordinate of excited state relaxation for several Fe(II) polypyridines, with a focus on exploring differences between bis-tridentate and tris-bidentate models. The chapter will begin with an explanation of how vibrational coherence manifests and how the resulting data is analyzed. With the aid of DFT calculations, a roadmap of the reaction coordinate during the early ultrafast processes will be constructed. 20 REFERENCES 1. Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L. and Pettersson, H. Chem. Rev. 2010, 110, 6595 – 6663. 2. S. Yun; Qin, Y.; Uhl, A. R.; Vlachopoulos, N.; Yin, M.; Li, D.; Han, X. and Hagfeldt, A. Energy Environ. Sci. 2018, 11, 476 – 526. 3. Arias-Rotondo, D. M. and McCusker, J. K. Chem. Soc. Rev. 2016, 45, 5803 – 5820. 4. Creutz, C.; Chou, M.; Netzel, T. L.; Okumura, M. and Sutin, N. J. Am. Chem. Soc. 1980, 102, 1309 – 1319. 5. McCusker, J. K. Science. 2019, 363, 484 – 488. 6. Haxel, G. B.; Hedrick, J. B. and Orris, G. J. Rare Earth Elements—Critical Resources for High Technology. https://pubs.usgs.gov/fs/2002/fs087-02/ (accessed Mar 11, 2024). 7. Ferrere, S. and Gregg, B. A. J. Am. Chem. Soc., 1998, 120, 843 – 844. 8. Damrauer, N. H.; Cerullo, G.; Yeh, A.; Boussie, T. R.; Shank, C. V. and McCusker, J. K. Science. 1997, 275, 54 – 57. 9. Yeh, A. T.; Shank, C. V. and McCusker, J. K. Science, 2000, 289, 935 – 938. 10. Adamson, A. W. and Demas, J. N. J. Am. Chem. Soc. 1971, 93, 1800 – 1801. 11. Van Houten, J. and Watts, R. J. J. Am. Chem. Soc. 1976, 98, 4853 – 4858. 12. Gawelda, W.; Cannizzo, A.; Pham, V.-T.; van Mourik, F.; Bressler, C. and Chergui, M. J. Am. Chem. Soc., 2007, 129, 8199 – 8206. 13. Lee, A.; Son, M.; Deegbey, M.; Woodhouse, M. D.; Hart, S. M.; Beissel, H. F.; Cesana, P. T.; Jakubikova, E.; James K. McCusker, J. K.; Gabriela S. Schlau-Cohen, G. S. Chem. Sci. 2023, 14, 13140 – 13150. 14. Zhang, W.; Alonso-Mori, R.; Bergmann, U.; Bressler, C.; Chollet, M.; Galler, A.; Gawelda, W.; Hadt, R. G.; Hartsock, R. W.; Kroll, T.; Kjær, K. S.; Kubiček, K.; Lemke, H. T.; Liang, H. W.; Meyer, D. A.; Nielsen, M. M.; Purser, C.; Robinson, J. S.; Solomon, E. I.; Sun, Z.; Sokaras, D.; van Driel, T. B.; Vankó, G.; Weng, T.; Zhu, D.; Gaffney, K. J. Nature. 2014, 509, 345 – 348. 15. Monat, J. E. and McCusker, J. K. J. Am. Chem. Soc. 2000, 122, 4092 – 4097. 16. Auböck, G. and Chergui, M. Nat. Chem. 2015, 7, 629 – 633. 17. Carey, M. C.; Adelman, S. L. and McCusker, J. K. Chem. Sci. 2019, 10, 134 – 144. 21 18. Kuciauskas, D.; Monat, J. E.; Villahermosa, R.; Gray, H. B.; Lewis, N. S. and McCusker, J. K. J. Phys. Chem. B, 2002, 106, 9347 – 9358. 19. Ardo, S. and Meyer, G. J. Chem. Soc. Rev. 2009, 38, 115 – 164. 20. Zhang, K.; Ash, R.; Girolami, G. S. and Vura-Weis, J. J. Am. Chem. Soc. 2019, 141, 17180 – 17188. 21. Wenger, O. S. Chem. - Eur. J. 2019, 25, 6043 − 6052. 22. Jamula, L. L.; Brown, A. M.; Guo, D. and McCusker, J. K. Inorg. Chem. 2014, 53, 15 – 17. 23. Liu, Y.; Harlang, T.; Canton, S. E.; Chábera, P.; Suárez-Alcántara, K.; Fleckhaus, A.; Vithanage, D. A.; Göransson, E.; Corani, A.; Lomoth, R.; Sundström, V. and Wärnmark, K. Chem. Commun. 2013, 49, 6412 – 6414. 24. Liu, Y.; Persson, P.; Sundström, V. and Wärnmark, K. Acc. Chem. Res. 2016, 49, 1477 − 1485. 25. Paulus, B. C.; Nielsen, K. C.; Tichnell, C. R.; Carey, M. C. and McCusker, J. K. J. Am. Chem. Soc. 2021, 143, 8086 – 8098. 26. Chábera, P.; Kjaer, K. S.; Prakash, O.; Honarfar, A.; Liu, Y.; Fredin, L. A.; Harlang, T. C. B.; Lidin, S.; Uhlig, J.; Sundström, V.; Lomoth, R.; Persson, P. and Warnmark, K. J. Phys. Chem. Lett. 2018, 9, 459 – 463. 27. Gualandi, A.; Marchini, M.; Mengozzi, L.; Natali, M.; Lucarini, M.; Ceroni, P. and Cozzi, P. G. ACS Catal. 2015, 5, 5927 − 5931. 28. Parisien-Collette, S.; Hernandez-Perez, A. C. and Collins, S. K. Org. Lett. 2016, 18, 4994 − 4997. 29. Woodhouse, M. D. and McCusker, J. K. J. Am. Chem. Soc. 2020, 142, 16229 – 16233. 30. Khalil, M.; Marcus, M. A.; Smeigh, A. L.; McCusker, J. K.; Chong, H. H. W. and Schoenlein, R. W. J. Phys. Chem. A. 2006, 110, 38 – 44. 31. Gawelda, W.; Pham, V.-T.; Benfatto, M.; Zaushitsyn, Y.; Kaiser, M.; Grolimund, D.; Johnson, S. L.; Abela, R.; Hauser, A.; Bressler, C. and Chergui, M. Phys. Rev. Lett. 2007, 98, 057401. 32. Gawelda, W.; Melanie, J.; de Groot, F. M. F.; Abela, R.; Bressler, C. and Chergui, M. J. Am. Chem. Soc. 2006, 128, 5001 – 5009. 22 33. Paulus, B. C.; Adelman, S.L.; Jamula, L. and McCusker J. K. Nature. 2020, 582, 214 – 218. 34. Ashley, D. C. and Jakubikova, E. Inorg. Chem. 2018, 57, 9907 – 9917. 35. Vanko, G.; Bordage, A.; Pápai, M.; Haldrup, K.; Glatzel, P.; March, A. M.; Doumy, G.; Britz; A.; Galler, A.; Assefa, T.; Cabaret, D.; Juhin, A.; van Driel, T. B.; Kjær, K. S.; Dohn, A.; Møller, K. B.; Lemke, H. T.; Gallo, E.; Rovezzi, M.; Németh, Z.; Rozsályi, E.; Rozgonyi, T.; Uhlig, J.; Sundström, V.; Nielsen, M. M.; Young, L.; Southworth, S. H.; Bressler, C. and Gawelda, W. J. Phys. Chem. C. 2015, 119, 5888 – 5902. 36. Berera, R.; van Grondelle, R. and Kennis, J. T. M. Photosynth Res. 2009, 101,105 – 118. 37. Brown, A. M.; McCusker, C. E. and McCusker, J. K. Dalton Trans. 2014, 43, 17635 – 17646. 38. Megerle, U.; Pugliesi, I.; Schriever, C.; Sailer, C. F. and Riedle, E. Appl Phys B. 2009, 96, 215 – 231. 39. Schrauben, J. N.; Dillman, K. L.; Beck, W. F. and McCusker, J. K. Chem. Sci. 2010, 1, 405 – 410. 23 CHAPTER 2: COMPUTATIONAL AND SPECTROSCOPIC METHODS 2.1 Introduction The following chapters utilize similar procedures regarding computational, electronic absorption spectroscopy, and transient absorption spectroscopy. Therefore, these procedures will be detailed in this chapter, to avoid excessive repetition. Information specific to each project, such as deviations from the computational methods described, will be detailed in the chapter dedicated to said project. 2.2 Computational Methods All Density Functional Theory (DFT) calculations were performed with the Gaussian 16 software1 on the Michigan State University High Performance Computing Center servers. Calculations began with ground state geometry optimization calculations on the unsubstituted complex, [Fe(terpy)2]2+ in Chapter 4 and [Fe(bpy)3]2+ in Chapter 5. Optimizations and subsequent frequency calculations were performed using modified B3LYP functional altered Hartree-Fock (HF) percentage. Additionally, a split basis set was employed, using SDD for the central Fe atom and 6-311G+(d,p) was used for the atoms comprising the ligand.2 Additionally, all calculations were done using the default solvent model for acetonitrile and with empirical dispersion = gd2 selected. Frequency calculations were checked for convergence and absence of imaginary frequencies. Once geometry optimization and frequency calculations were complete, Time Dependent DFT (TD-DFT) was performed using 6-311G+(d,p) as the sole basis set. This was done to simulate an electronic absorption spectrum to be compared with experimentally determined spectrum as a method of verifying the accuracy of the calculations. If the simulated spectrum was a poor match, the HF exchange percentage would be altered. It was found that 10% HF was the best fit for [Fe(terpy)2]2+ and [Fe(bpy)3]2+. After this, energy calculations were performed with singlet and quintet spin states using the aug-cc-pVTZ basis set. The different spin states were used to calculate the ground state energy and the energy of the lowest energy quintet state while in ground state geometry. Lastly, geometry optimization and frequency calculations were performed for the quintet state using the same modified B3LYP functional and split basis set as described above. When completed, energy calculations were done to determine the zero-point energy of the quintet state. This process was repeated for the remainder of the complexes in each series, using the optimized ground state and quintet state geometries as starting points. 24 2.3 Electronic Absorption Spectroscopic Methods Electronic absorption spectra were collected and analyzed as described previously.3 All electronic absorption spectra were collected on a double-beam Perkin Elmer Lambda 1050 spectrophotometer. All samples were prepared in acetonitrile. Sample concentrations were adjusted to ensure absorbance values near 0.5 AU for a given transition while using either 1 cm or 10 cm pathlength cells. Solvent blanks were collected for all samples and used to baseline correct each spectrum. Gaussian deconvolution of the d-d transitions was performed on each spectrum using IgorPro to determine the transition energies and molar absorptivities. 2.4 Nanosecond Transient Absorption Spectroscopic Methods For complexes with an excited state lifetime longer than ~4 ns, ground state recovery times needed to be collected on the nanosecond system, which has been described in previous reports.4 This system employs an Edinburgh Instruments LP980 spectrometer fitted with a Hamamatsu R928 photomultiplier tube and interfaced to a Tektronix TDS oscilloscope. Excitation pulses were generated by an Opotek Vibrant 355 LD Q-Switched Nd:YAG optical parametric oscillator (OPO). This OPO generated pump pulses with a near 5 ns duration, however this was wavelength dependent. Samples were prepared in a 1 cm pathlength optical quartz cuvette in spectrophotometric grade acetonitrile solution with concentrations yielding an absorbance ranging from 0.5 to 0.7 AU. The acetonitrile was kept in an argon atmosphere to minimize water contamination. The ground state absorption spectra of samples prepared for transient absorption (TA) were collected using a Varian Cary 50 UV-Vis spectrophotometer before and after TA data was collected. Pump and probe wavelengths were chosen depending on the sample being studied. The Instrument response function (IRF) was recorded by using a 1 cm quartz cuvette filled with an opaque solution (coffee cream), and recording the signal produced from the scatter from the pump pulse. The IRF changed with different pump wavelengths and was recorded for each wavelength used. Using the IRF spectrum, the TA data for each complex was fit using the DecayFit software developed by Søren Preus of FluorTools.5 This fits the data while taking the IRF into account, providing accurate time constants for excited state relaxation. Variable temperature TA (VT-TA) data was collected following previously described procedures.6 A Janis Research SVT-100 optical Dewar was placed at the sample region, and a continuous flow of liquid nitrogen was used to control the temperature of the sample. Temperature ranges from 235-295 K were employed, and the temperature was measured and recorded by a 25 temperature probe within the cryostat chamber connected to a Lake Shore 321 temperature controller. Part way through the study, our group had substituted use of the Janis optical Dewar with a Quantum Northwest TC-1 cooling system which used water as the coolant. The same temperature ranges were accessed using the Quantum Northwest system; however, an ice water bath was needed to reach temperatures below ~260 K. Information detailing which system was used for which data will be provided in future chapters. Regardless of the cooling system used, the sample was held at the desired temperature for 15 minutes before data collection began to ensure thermal equilibrium between the sample and its surroundings. 2.5 Ultrafast Transient Absorption Spectroscopic Methods Two ultrafast time-resolved absorption spectrometers were used to obtain the data presented in this dissertation: one which can generate ~150 fs pulses with a ~13 ns delay stage (called WE), and the other which can generate ~40 fs pulses with only a 1.5 ns delay stage (called RR). The experimental principles behind each instrument are the same, but they each use different machines to generate pulses. Each system has been previously described in detail.6-9 Regarding WE, a Coherent Mira 900 Ti:sapphire oscillator pumped by a Coherent Verdi V- 6 diode-pumped solid-state Nd:YVO4 laser operating at 5.0 W is used to generate a ~380 mW 800 nm modelocked beam. This beam is directed into a Positive Light Spitfire Ti:sapphire regenerative amplifier, which is itself pumped by a Coherent Evolution Nd:YLF laser. The Spitfire produces ~650 mW pulses of 800 nm light with a repetition rate of 1 kHz, which is split 70:30 into the pump and probe beams, respectively. The pump beam is directed into a Light Conversion TOPAS optical parametric amplifier (OPA), which can tune the wavelength in the visible region. The pump beam is then double passed through a retroreflected mirror mounted on a 1.2 m Aerotech delay stage, which is controlled by Soloist CP software. Then, the pump beam is passed through an optical chopper at a frequency of 453 Hz. The chopped beam is passed through a neutral density filter so that the beam has a power between 4-5 mW, and then is softly focused via lens through the sample before hitting a beam block. Meanwhile, the 800 nm probe beam is run through a waveplate, so its polarization is magic angle (54.7o) from the pump beam. The beam is then split 90:10, with the weaker beam directed to a Thor Labs Si photodiode to serve as a reference beam. The stronger beam is focused via lens into either a sapphire or calcium fluoride crystal to generate a while light continuum and then directed to the sample. Afterwards, it passes through a 10 nm band-pass filter set to the desired wavelength and then focused into another Si photodiode, serving as the signal. 26 Both the signal and reference photodiodes are connected to a Tektronix 2465A 350 MHz oscilloscope. The oscilloscope was used to measure the intensity of light from the signal beam passing through the unpumped sample (I0), then an iris and neutral-density filters are used to adjust the detected signal from the reference beam so that it matches I0 from the signal beam. Finally, both photodiodes are disconnected from the oscilloscope and connected to a Stanford Research: SR830 DSP digital lock-in amplifier which is synchronized to the chopper. A LabVIEW virtual instrument then records the data from the lock-in amplifier and this data was fit using Igor Pro software to the following exponential decay equation: 𝛥𝐴(𝑡) = 𝑦0 + 𝛥𝐴0𝑒− Where Δt is the time between when the pump and probe pulses reach the sample, and τ is the time 2.1 𝛥𝑡 𝜏 constant for relaxation. The presence of y0 considers any offset the relaxation curve may have from the x-axis despite the decay having reached its minimum. Offset may result from scattering of the pump beam interfering with the signal or from photodecomposition of the sample. VT-TA measurements were only recorded on WE and followed a protocol that has previously been described in detail.6 A Janis Research STVP-100-2 optical Dewar is placed at the sample region. The optical Dewar is connected to an International Cryogenics liquid nitrogen storage Dewar via a Janis Research transfer line. A continuous flow of liquid nitrogen from the storage Dewar is used to control the temperature of the sample with ranges from 235-295 K. The temperature of the sample was monitored and recorded by two sensors placed at the top and bottom of the cryostat chamber connected to a Lake Shore 335 temperature controller. The sample was held at the desired temperature for 15 minutes before data collection began to ensure thermal equilibrium between the sample and its surroundings. Regarding RR, a Coherent Mantis Ti:sapphire oscillator generates a ~300 mW 800 nm modelocked beam. This beam is directed into a Coherent Evolotion Nd:YLF pumped Coherent Legend Elite Ti:sapphire regenerative amplifier. The Legend Elite generates ~1.1 W 800 nm pulsed beam with a repetition rate of 1 kHz. This beam is split 70:30 into a pump beam and probe beam, respectively. Both beams are directed into separate Coherent OperA Solo OPAs to tune the wavelength of the beam within the visible spectrum. Both beams then travel through a double prism compressor in order to compress the excitation pulse, as discussed in Chapter 1.4. The prisms are placed on Thor Labs dovetail optical rails, so that the distance between each prism may be controlled. The pump beam is directed to a 225 mm Newport DL225 delay stage which is 27 controlled by Newport Controller Series software. This stage controls the delay between the pump and probe pulses. Then, the pump beam is passed through an optical chopper at a frequency of 469 Hz. Both beams are then directed through waveplates, so their polarizations are at magic angle. A neutral density filter was used to set the power of the pump to about 4 mW, and the probe beam was similarly set to about 1/10th of that power. The pump beam is then focused into the sample before hitting a beam block. The probe beam is split 90:10 into a signal beam and reference beam, as described above. The probe beam passes through a Scientific Measurements monochrometer. A Tektronix 2467B 400 MHz oscilloscope is used alongside an iris and neutral density filters to set the signal of the reference beam equal to the I0 of the signal beam passing through the unpumped sample. Both the signal and reference photodiodes are connected to a Stanford Research: SR810 DSP digital lock-in amplifier synchronized to the chopper. Data is recorded from the lock-in amplifier as described for the WE system. The data is fit using Igor Pro either to Equation 2.1, or the following double exponential decay equation if more than one process if being probed: 𝛥𝐴(𝑡) = 𝑦0 + 𝛥𝐴01𝑒 − 𝛥𝑡 𝜏1 + 𝛥𝐴02𝑒 − 𝛥𝑡 𝜏2 2.2 For both systems, the IRF was determined from the optical Kerr effect (OKE) observed from collecting TA data on neat acetonitrile with an analyzing polarizer set perpendicular to the probe polarization placed in front of the detector.10 The reference photodiode must be blocked, since the polarizer blocks all light from reaching the signal photodiode. The early portion of the data (up to just past the peak), is fit to a gaussian, Equation 2.3, and the full width half maximum Figure 2.1: An OKE dataset collected on HPLC grade acetonitrile. The pump and probe beams were both set to 490 nm. 28 (FWHM) extracted from the fit provides the IRF. A sample OKE dataset collected on RR is provided in Figure 2.1. 𝛥𝐴(𝑡) = 𝑦0 + 𝛥𝐴0𝑒−( 2 𝛥𝑡 𝑤𝑖𝑑𝑡ℎ ) 2.2 When measuring the IRF of the pump pulse on RR, the output from the probe OPA is blocked. A 90:10 beam splitter is placed in the pump beam path before the beam reaches the delay stage. The lower energy beam is then directed into the probe line to be used as the probe pulse. This set-up can also be used for single color experiments on RR, rather than relying on output from both OPAs. Samples were prepared in quartz optical cuvettes that were either 1 or 2 mm in pathlength. The 2 mm pathlength cells were primarily used for VT-TA measurements but could also be used to increase the signal to noise ratio (s/n) for low absorption complexes. Samples were prepared in HPLC grade acetonitrile solution for room temperature measurements. Similar to sample preparation for nanosecond TA spectroscopy, samples were prepared with concentrations yielding an absorbance range of 0.5-0.7 AU. The ground state absorption spectra of samples prepared for TA were collected using a Varian Cary 50 UV-Vis spectrophotometer before and after TA data was collected. Lastly, some points to consider regarding s/n for VT-TA measurements. The use of the cryostat at low temperatures introduces several hurdles that harm s/n. First to consider is the solvent being used at low temperatures. For VT-TA measurements, spectrophotometric grade acetonitrile that was stored in an argon atmosphere was used instead of the HPLC grade acetonitrile used at room temperature. This was done to minimize water contamination in the solvent. HPLC grade acetonitrile solutions were found to provide the same results as the spectrophotometric grade acetonitrile solutions at room temperature, but it was discovered that when brought to low temperatures, the small amounts of water present in the HPLC grade solvent would freeze into clusters and generate a large amount of pump scattering, which reduces the s/n and alters the observed lifetime. Using spectrophotometric grade solvent stored in an argon atmosphere greatly reduced scattering and improved s/n. Next to consider is the scattering caused by the cryostat glass. The cryostat chamber, illustrated in Figure 2.2, introduces much more glass into the beam path, 12 mm of Si-fused glass in total. This creates a great deal of pump scattering when the cryostat is installed. This scattering cannot be avoided, however, by using pump/probe wavelengths combinations that were at least 30 nm apart, the amount of scattered pump light that could pass through the wavelength separator is reduced. This further improves s/n while the cryostat is 29 installed. Furthermore, the glass windows reduce the power of the pump beam by about 0.6 mW. So increasing the power of the pump at the sample to compensate for this helps maintain good s/n. The glass also introduces a great deal of chirp, which as explained in Chapter 1.4, increases the duration of the pulse. Given the IRF of the instrument used for VT-TA is about 150 fs and the complexes studied with VT-TA have excited state lifetimes on the order of nanoseconds, this change in pulse duration is of no concern for the VT-TA data discussed in this dissertation. The last aspect to keep in mind is the pressure inside the liquid nitrogen storage Dewar. The temperature of the sample chamber is controlled by a continuous flow of liquid nitrogen from the storage Dewar. High pressures within the storage Dewar drive push liquid nitrogen from the Dewar to the cryostat via a transfer line. If the pressure inside the storage Dewar is too high (from my experience, exceeding 3 atm), the flow rate of liquid nitrogen is too high to control. This runs the risk of temperatures plummeting or high pressures within the sample chamber causing the sample to freeze. In either case, not only is the VT-TA data compromised and must be recollected, but the cuvette may also be damaged. Therefore, consistent monitoring of pressure inside the storage Dewar is extremely important. Figure 2.2: A schematic of a top-down view of the cryostat used in VT-TA measurements. A vacuum pump is used to create a “vacuum jacket” to prevent exterior heat from entering the sample chamber. A storage Dewar is used to provide a continuous flow of liquid nitrogen to lower the temperature inside the chamber. A temperature controller connects to two temperature probes, one at the top of the sample chamber and one at the bottom, to monitor the temperature. 30 REFERENCES 1. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Petersson, G. A.; Nakatsuji, H.; Li, X.; Caricato, M.; Marenich, A. V.; Bloino, J.; Janesko, B. G.; Gomperts, R.; Mennucci, B.; Hratchian, H. P.; Ortiz, J. V.; Izmaylov, A. F.; Sonnenberg, J. L.; Williams-Young, D.; Ding, F.; Lipparini, F.; Egidi, F.; Goings, J.; Peng, B.; Petrone, A.; Henderson, T.; Ranasinghe, D.; Zakrzewski, G.; Gao, J.; Rega, N.; Zheng, G.; Liang, W.; Hada, M.; Ehara, M.; Toyota, K.; Fukada, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Throssell, K.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J. J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Keith, T. A.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Millam, J. M.; Klene, M.; Adamo, C.; Cammi, R.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Farkas, O.; Foresman, J. B. and Fox, D. J. Gaussian 16, Rev B.01. Gaussian Inc.: Wallingford CT 2016. 2. Nance, J.; Bowman, D. N.; Mukherjee, S.; Kelley, C. T. and Jakubikova, E. Inorg. Chem. 2015, 54, 11259 − 11268. 3. Yarranton, J. T. and McCusker, J. K. J. Am. Chem. Soc. 2022, 144, 12488 − 12500. 4. Woodhouse, M. D. and McCusker, J. K. J. Am. Chem. Soc. 2020, 142, 16229 – 16233. 5. Preus, S. DecayFit - Fluorescence Decay Analysis. https://www.fluortools.com/software/decayfit-fluorescence-decay-analysis (accessed Apr 15, 2024). 6. Carey, M. C.; Adelman, S. L. and McCusker, J. K. Chem. Sci. 2019, 10, 134 – 144. 7. Juban, E. A. and McCusker, J. K. J. Am. Chem. Soc. 2005, 127, 6857 – 6865. 8. Brown, A. M.; McCusker, C. E.; Carey, M. C.; Blanco-Rodríguez, A. M.; Towrie, M.; Clark, I. P.; Vlček, A. and McCusker J. K. J. Phys. Chem. A. 2018, 122, 7941 – 7953. 9. Paulus, B.C.; Adelman, S.L.; Jamula, L. and McCusker J. K. Nature. 2020, 582, 214 – 218. 10. Albrecht, H-S.; Heist, P.; Kleinschmidt, J.; Lap, D. V.; and Schroder, T. Meas. Sci. Technol. 1993, 4, 492 – 495. 31 CHAPTER 3: REVIEW AND ANALYSIS OF NON-RADIATIVE DECAY THEORIES 3.1 Introduction to Non-Radiative Decay Theories The combination of transient absorption spectroscopy (TA) with variable temperature (VT) techniques, as described in the previous chapters, allows for the study of knr, the nonradiative rate constant, as a function of temperature. Doing so grants the ability to study excited state relaxation through three temperature dependent rate theories: Arrhenius theory, transition state theory, and Marcus theory. All three theories describe non-radiative relaxation as a mono-exponential decay function but do so in different ways. Arrhenius theory describes the temperature dependence of decay rates as a function of the activation energy, or energy barrier, of the transition between states. Transition state theory is similar to Arrhenius theory but can be used to further describe the formation of the transition state between electronic states by describing the energy barrier as a free energy cost.1 The benefits of doing so will be discussed later in this chapter. Lastly, Marcus theory is a model which describes nonradiative decay as a function of driving force, electronic coupling, and reorganization energy.2 The above theories are typically used to describe kinetic interactions between separate systems (i.e. chemical reactions or electron transfer), rather than the transition between electronic states. However, so long as excited state relaxation behaves classically, rather than quantum mechanically, the underlying principles are applicable.3 There is a limit on the temperature dependency that relaxation rates experience. As the temperature decreases, the ambient energy in the sample decreases, reducing the vibrational levels within the sample that are populated. Eventually, the sample will settle at the zero-point energy of the long-lived excited state and can go no lower. Without enough ambient energy to overcome the energy barrier, the only pathway available for relaxation is through quantum mechanical tunneling, which bypasses the barrier.4 At this point, there is no temperature dependence for excited state relaxation. This so called “classical limit” is where excited state relaxation can no longer be described using the above theories. Thankfully, this limit is easy to determine. Equation 3.2 provides the linear form of the Arrhenius equation. If experimental data is plotted to equation 3.2, the plot should be linear. If the slope of the plot goes to zero as temperature decreases, that indicates that the classical limit has been crossed, and if the slope is constant then the temperature is still above the classical limit. The data collected in this dissertation was done at temperatures above the classical limit, therefore allowing the data to be analyzed through these theories. 32 This section will begin with a brief examination of Arrhenius theory. The equation for Arrhenius theory is provided in equation 3.1: 𝑘𝑛𝑟 = 𝐴𝑒 − 𝐸𝑎 𝑘𝐵𝑇 3.1 Where knr is the nonradiative rate constant, A is the frequency factor (the rate constant of the process when there is no barrier), Ea is the activation energy or energy barrier, kB is the Boltzmann’s constant (0.695 cm-1/K), and T is temperature. Taking the natural log of equation 3.1 yields the following linear equation: ln(𝑘𝑛𝑟) = ln 𝐴 − 𝐸𝑎 𝑘𝐵 ∗ 1 𝑇 3.2 Since the ground state recovery (GSR) process can be accurately described through the exponential decay equation, the time constant (τ) for relaxation can be extracted by fitting the data to equation 3.3, and then τ can be inverted to provide knr. 𝛥𝐴(𝛥𝑡) = 𝛥𝐴0𝑒− 𝛥𝑡 𝜏 𝑘𝑛𝑟 = 1 𝜏 3.3 3.4 ΔA is the difference in absorbance between the excited and ground states, the signal of TA spectroscopic measurements, and Δt is the amount of time the probe pulse takes to reach the sample after it is excited by the pump pulse. Therefore, collecting knr at various temperatures allows an Arrhenius plot to be constructed which can then be fit to equation 3.2, providing Ea from the slope and A from the y-intercept of the plot. Next, transition state theory is described by the equation provided below:1 𝑘𝑛𝑟 = 𝜅𝑘𝐵𝑇 ℎ ∗ 𝑒 − ‡ ΔG 𝑘𝐵𝑇 3.5 Where κ is the transmission coefficient, which in most cases is assumed to be 1, h is Planck’s constant, and ΔG‡ is the free energy cost associated with the formation of the transition state. Upon the formation of the transition state, it is possible for the complex to relax back to the reactant state rather than continuing to the product state. This process dampens the rate of relaxation. By assuming a κ value of 1, this “back slide” to the reactant state is assumed to be negligible. Since transition state theory describes the energy barrier as free energy, which is dependent on changes in enthalpy and entropy, equation 3.5 can be rearranged to provide the following equation, dubbed the Eyring equation:1 33 𝑘𝑛𝑟 = 𝜅𝑘𝐵𝑇 ℎ ∗ 𝑒 ‡ ΔS 𝑘𝐵 ∗ 𝑒 − ‡ ΔH 𝑘𝐵𝑇 3.6 Where ΔH‡ and ΔS‡ are the enthalpy and entropy costs, respectively, associated with forming the transition state. Using the Eyring equation to analyze data allows for a more detailed examination of how the energy barrier is influenced by changes to the system being studied, namely through changing either ΔH‡ or ΔS‡. Lastly is Marcus theory, here the semi-classical Marcus equation for nonradiative decay was used (equation 3.7).4 The term semi-classical refers to the methods used to describe the electronic and nuclear components of the transition. At temperatures higher than the classical limit, the nuclear component can be described using classical terms, given as the exponential expression in equation 3.7. The pre-exponential component, due to the inclusion of the Hamiltonian dependent term Hab, treats the electronic part of the system quantum mechanically. The combination of quantum mechanical and classical treatments leads to the semi-classical Marcus equation: 𝑘𝑛𝑟 = 2𝜋 ℏ ∗ |𝐻𝑎𝑏|2 ∗ 1 √4𝜋𝜆𝑘𝐵𝑇 − (∆𝐺𝑜+𝜆)2 4𝜆𝑘𝐵𝑇 ∗ 𝑒 3.7 Where ħ is the reduced Planck’s constant (h/2π), Hab is the electronic coupling constant, λ is the reorganization energy, and ΔGo is the zero-point energy difference between the reactant and product states. λ represents the energy required to shift molecule from the geometry of the reactant state to the geometry of the product state. Furthermore, λ is split into two components, inner sphere (λi) and outer sphere (λo). λ = λ𝑖 + λ𝑜 3.8 λi stems from the structural changes within the molecule, while λo stems from the changes needed for the solvent shell surrounding the molecule. Hab is an off-diagonal Hamiltonian matrix element which describes the electronic communication between the reactant and product states. In the following chapters, which focus on the GSR from the longest-lived excited state, the reactant state (a) is 5T2 and the product state (b) is 1A1. Hab is obtained from the following matrix:2,5 𝐻𝑎𝑎 − 𝐸 | 𝐻𝑎𝑏 𝐻𝑎𝑏 𝐻𝑏𝑏 − 𝐸 | = 0 3.9 Where E is the energy of the system at a given point on the reaction coordinate, and Haa and Hbb are the diabatic energies of each state, described by the following expressions: 𝐻𝑎𝑎 = ⟨𝛹𝑎|Ĥ|𝛹𝑎⟩ 𝐻𝑏𝑏 = ⟨𝛹𝑏|Ĥ|𝛹𝑏⟩ 34 3.10 3.11 Where Ĥ is the Hamiltonian operator for the system and Ψ is the wave function of the state. As Hab describes the communication between states, it relies on Ψ for each state and is described by the following: 𝐻𝑎𝑏 = ⟨𝛹𝑎|Ĥ|𝛹𝑏⟩ 3.12 If Ψa and Ψb are orthogonal, meaning the wave functions have zero overlap, then there is no communication between states and Hab = 0. Following Equation 3.7, if Hab is zero, then so is knr; meaning that the process cannot occur. This scenario is described in the diabatic electron transfer model, pictured below in Figure 3.1. However, if Ψa and Ψb do overlap, then Hab is greater than Figure 3.1: Diabatic model showing the potential energy surfaces of the 5T2 and 1A1 electronic states. In this model, Hab is zero and non-radiative decay cannot occur. zero and there is communication between each state. This leads to the development of the non- adiabatic model where the potential energy surfaces of each state mix and are split into an upper well and a lower well. This splitting is dependent on the strength of Hab and is derived from the following equations:2 (𝐻𝑎𝑎 − 𝐸)(𝐻𝑏𝑏 − 𝐸) − 𝐻𝑎𝑏 𝐸2 − (𝐻𝑎𝑎 + 𝐻𝑏𝑏)𝐸 + (𝐻𝑎𝑎𝐻𝑏𝑏) − 𝐻𝑎𝑏 2 = 0 2 = 0 𝐸± = (𝐻𝑎𝑎+𝐻𝑏𝑏)±√(𝐻𝑎𝑎+𝐻𝑏𝑏)2−4(𝐻𝑎𝑎𝐻𝑏𝑏−𝐻𝑎𝑏 2) 2 3.13 3.14 3.15 Equation 3.13 is a representation of the matrix from equation 3.9, which is reordered to a quadratic equation in equation 3.14. Solving equation 3.14 for E yields equation 3.15. Equation 3.15 provides two values for E: E+ and E-. E+ represents the energy of the upper non-adiabatic well and E- represents the lower well. At the point on the reaction coordinate where the potential energy 35 surface of each state crosses (which is the transition state), Haa is equal to Hbb, and equation 3.16 collapses into the following: This means that at the transition state, the energy gap between the upper and lower non-adiabatic well is equal to 2Hab. The non-adiabatic model is illustrated in Figure 3.2. 𝐸± = 𝐻𝑎𝑎 ± |𝐻𝑎𝑏| 3.16 Figure 3.2: Non-adiabatic model of electron transfer. The energy gap between the upper and lower well at the diabatic crossing, shown with dashed grey lines, is equal to twice Hab. The energy gap caused by Hab pushes against the barrier, therefore a larger Hab will lead to a smaller energy barrier and faster transition. A larger Hab will also reduce the chance of a phenomenon known as “well hopping.” If the thermal energy of the system (kBT) is greater than 2Hab, the molecule can jump the 2Hab energy gap into the upper well, keeping the molecule trapped between states until it relaxes back to either the reactant or product state.2 This shows that knr is dependent on Hab in two different ways: Hab affects the energy barrier, and 2Hab affects the rate of “well hopping,” hence equation 3.7 shows that knr ∝ 2|Hab|2. If Hab is very large, the barrier can be reduced to effectively zero; this regime is called adiabatic. For the molecules studied in this dissertation, the experimentally predicted values for Hab are significantly less than both ΔG‡ determined through the Eyring equation and Ea determined through the Arrhenius equation. Therefore, the non-adiabatic model is best for understanding the 5T2 → 1A1 relaxation. The models pictured in Figures 3.1 and 3.2 are operating in what is called the Marcus normal region, which is characterized by having the two potential energy surfaces displaced along the nuclear coordinate. Within the Marcus normal region, the value for λ is greater than the absolute 36 value of ΔGo. However, there are two other regions that two states may occupy: the barrierless region and the inverted region. The relationship between the energy barrier, ΔGo, and λ is described in the following expression: 𝛥𝐺‡ ∝ (𝛥𝐺𝑜+𝜆)2 4𝜆 3.17 Equation 3.17 is derived from setting the exponential terms from equations 3.5 and 3.7 as equal. Since ΔGo has a negative value and λ has a positive value, when λ and |ΔGo| are equal, the numerator of equation 3.17 collapses to zero and the barrier diminishes. This is known as the barrierless region, where there is effectively no energy barrier between states. The inverted region occurs when λ is less than |ΔGo|, in this region the zero-point energy of the reactant state is nested within the potential energy surfaces of the product state, rather than being displaced from it. The three regions are illustrated in Figure 3.3. Equation 3.17 shows that there is a parabolic relationship between the energy barrier and either ΔGo or λ, while the other variable is held constant. It is worth noting that the relationship between ΔGo and the energy barrier while holding λ constant is symmetric regardless of which region the system is in. However, if ΔGo is held constant, the parabolic relationship between λ and the energy barrier changes as the system crosses the barrierless region. This change in behavior is the result of the 4λ term being present in the denominator of equation 3.17. Figure 3.3: Diabatic Marcus theory models displaying the Marcus inverted region (left), barrierless region (middle), and normal region (right). In this picture, ΔGo is held constant. The model changes regions as ΔQ is increased, as λ is dependent on the reaction coordinate. The previously discussed relationship between energy barrier, λ, and ΔGo highlights that there are two ways to shift between the Marcus regions. The first, as displayed in Figure 3.3, is to change ΔQ, the difference in reaction coordinate between the zero-point energy levels of each well. The expression describing λ is provided below:2 37 𝜆 = 1 2 𝑓(∆𝑄)2 3.18 Where f is the force constant, which influences the width of the potential energy surface of the reactant state along the kinetically relevant vibrational modes, and ΔQ is the difference in reaction coordinate between the zero-point energy of each state. Therefore, increasing ΔQ will also increase λ, and vice versa. The other method of changing between regions is to alter ΔGo. If λ is held constant, the potential energy surfaces will be in the normal region at low ΔGo, and increasing the value of ΔGo will push the surfaces towards the barrierless region and then eventually into the inverted region. This creates an inverted parabolic relationship between ΔGo and knr, where the rate increases as the system approaches the barrierless region and then decreases due to the barrier growing larger as the system moves further from the barrierless region. This creates opposite relationships between ΔGo and knr depending on which region the system is in; this is illustrated in Figure 3.4. Figure 3.4: A plot showing the ln(knr) vs -ΔGo, while holding ΔQ constant. At low ΔGo, the system is in the Marcus normal region (orange) and at high ΔGo, the system is in the Marcus inverted region (blue). When |ΔGo| = λ, the system is in the barrierless region, and knr is at its peak value. This behavior highlights the importance of understanding which Marcus region an electronic state transition is occurring in. Doing so will aid in developing a model that can accurately predict how excited state evolution of the ligand field states will change as adjustments are made to a molecule, especially those that will impact ligand field strength. In addition, having an accurate model to reference will aid in understanding unexpecting results. Using these theories 38 to analyze VT data uniquely allows the observation of Marcus normal or inverted region behavior within a system by being able to extract the energy barrier and predict the energy gap from experimental data. 3.2 Variable Temperature Analysis As mentioned previously, Arrhenius theory, transition state theory, and Marcus theory (equations 3.1, 3.6, and 3.7, respectively) are all temperature dependent. Therefore, by using VT techniques to measure knr as a function of temperature, each theory can be used to provide different information about the relaxation dynamics of the process being measured. The exponential component of each theory has a 1 𝑇⁄ component, meaning that if the natural log is taken on each side of the equation, as seen in equation 3.2, then a relationship will be established between the ln(𝑘𝑛𝑟) and 1 𝑇⁄ which can be fit to the linear equation: 𝑦 = 𝑚𝑥 + 𝑏 3.19 Where y is dependent on ln(𝑘𝑛𝑟), x represents 1 𝑇⁄ , b is equal to the natural log of the pre- exponential term, and m is equal to the natural log of the exponential term with the temperature component removed. This relationship is shown in the Arrhenius plot provided in Figure 3.5. Figure 3.5: Arrhenius plot constructed from VT-TA data collected for the GSR process of [Fe(terpy)2](PF6)2. This data was collected following the procedures outlined in Chapter 2 from temperature ranges of 240 – 295 K in 5 K steps. However, regarding transition state theory and Marcus theory, further action will be needed to properly fit the data. As shown in equations 3.6 and 3.7, both transition state and Marcus theory have temperature components in the pre-exponential terms. To address this, before the natural log of each side is taken the equation must be rearranged to place temperature on the y-axis. After 39 doing this, the linear forms of the Eyring equation and semi-classical Marcus equation are as follows: ln ( 𝑘𝑛𝑟 𝑇 ) = [ln ( 𝜅𝑘𝐵 ℎ ) + 𝛥𝑆‡ 𝑘𝐵 ] + (− 𝛥𝐻‡ 𝑘𝐵 ∗ 1 𝑇 ) ln(𝑘𝑛𝑟 ∗ √𝑇) = ln (2𝜋 ℏ ∗ |𝐻𝑎𝑏|2 ∗ 1 √4𝜋𝜆𝑘𝐵 ) + (− (∆𝐺𝑜+𝜆)2 4𝜆𝑘𝐵 ∗ 1 𝑇 ) 3.20 3.21 Now, “Arrhenius-type” plots can be constructed using a knr value alongside the temperature that value was collected at to construct the y-axis. This removes temperature from the y-intercept of the fit, and the results can now be accurately determined. Using the Eyring equation to treat the VT data gives ΔS‡ from the y-intercept and ΔH‡ from the slope. Analyzing the data through Marcus theory provides two λ-dependent ratios: |𝐻𝑎𝑏|4 𝜆⁄ is calculated from the y-intercept and (𝛥𝐺𝑜 + λ)2 λ⁄ . These ratios are derived from the following expressions: = |𝐻𝑎𝑏|4 𝜆 (𝛥𝐺𝑜+λ)2 𝜆 𝑘𝐵ℏ2𝑒2𝑏 𝜋 = 4𝑚𝑘𝑏 3.22 3.23 Where b and m are taken from fitting the “Arrhenius-type” plots, shown in Figure 3.6, to equation 3.19. Figure 3.6: Arrhenius-type plots constructed from VT-TA data collected for the GSR process of [Fe(terpy)2](PF6)2. Plot A uses the Eyring equation to analyze the data, while plot B uses the semi- classical Marcus theory equation. The ratios displayed in equations 3.22 and 3.23 highlight that relying solely on VT-TA data will only provide limited information from Marcus analysis of the data. Instead, one of the terms must be determined independently to determine Hab or ΔGo from λ. Depending on the type of electronic state transition, ΔGo may be determined using electrochemistry or emission spectroscopy. However, the GSR process of the molecules studied in this work occurs from the 40 non-emissive, metal centered ligand field excited state, 5T2. The lack of emission from the ΔS = 2 transition, as well as the inability of electrochemistry to directly probe the eg* orbitals (instead electrochemical reduction probes the energy of the ligand π* orbitals), means that ΔGo cannot be determined via the experimental methods at our disposal.3,6 Therefore, a different means of decoupling these variables is needed. A technique known as Nelsen’s four-point method uses calculations to predict λ for a given transition.7 This technique has been used in previous studies to provide an accurate prediction for λi.8-10 This method involves calculating the optimized geometry of the reactant and product states using density functional theory (DFT). Then, the energy of the reactant state within both geometries is calculated. The difference between these energies yields λi. Given the metal-centered nature of the transitions studied with this analysis, we believe that outer-sphere perturbations will be small, resulting in minimal influence from λo compared to λi, on λtotal. Therefore, the λ predicted by this method, should provide a reasonable approximation for the system, and allow Hab and ΔGo to be calculated from the experimentally determined ratios shown in equations 3.22 and 3.23. Figure 3.7: Schematic of how λ is calculated using DFT. The geometry of the molecule is optimized for both the reactant state (5T2) and the product state (1A1). Then the energy of the reactant state is determined at both geometries (shown as the green dots), the difference between these energies is equal to λ. The use of these theories to analyze VT-TA data, alongside DFT calculations, allows for the determination of the energy barrier, energy gap, and electronic coupling tied to the electronic 41 transition being studied. Chapters 4 and 5 of this work will use this analysis to study the excited state properties of the 5T2 states of several Fe(II)-based polypyridine complexes. 3.3 Determining the Energy Barrier Arrhenius theory and transition state theory both are used to describe the energy barrier, but through different means. Arrhenius theory describes the barrier through Ea, a term that is derived from the slope of an Arrhenius plot fit to equation 3.2. Transition state theory uses the linear form of the Eyring equation, shown in equation 3.20, to extract ΔH‡ from the slope of an “Arrhenius-type” plot shown in Figure 3.6A, and ΔS‡ from the y-intercept. The barrier, ΔG‡, is then calculated from these variables using the Gibbs free energy equation: ∆𝐺‡ = ∆𝐻‡ − 𝑇∆𝑆‡ 3.24 This means that as opposed to Arrhenius theory, which only considers the slope, transition state theory uses information from both the slope and y-intercept to determine the energy barrier. Our group has historically relied on Arrhenius theory to describe the excited state dynamics of Fe(II)- polypyridines,3 but this difference in method raises the question of which theory can more accurately describe the system and if so, why. This section is dedicated to answering those questions. For clarification, throughout this discussion Ea will be used to represent the energy barrier determined through Arrhenius theory analysis, and ΔG‡ will be used to represent the energy barrier determined through transition state theory analysis. Lastly, full data tables for each plot shown in this section will be provided in the supplemental information (SI) of this chapter. Starting with Arrhenius theory, equation 3.1 shows that knr has an inverse exponential relationship with Ea. Therefore, Ea should have a direct relationship with the inverse of knr, τ. Figure 3.8 plots GSR τ collected at room temperature (τ295) against GSR Ea across a series of [Fe(terpy)2](PF6)2 complexes substituted at the 4’-position (the motivations for this series will be elaborated upon in the following chapter). When this plot is fit to an exponential equation, it can be observed that several points rest well outside of the fit. Most notably is the 4’-hyrdoxy (OH) substituted complex. Arrhenius analysis predicts the 4’-OH substituted complex has the greatest energy barrier but has a drastically faster τ295 than what the fit would suggest. This discrepancy is further highlighted when compared to the 4’-diethanol amino (N(EtOH)2) substituted complex. Both have similar Ea values (935 ± 10 cm-1 for [Fe(4’-OH-terpy)2](PF6)2 and 900 ± 20 cm-1 for [Fe(4’-N(EtOH)2-terpy)2](PF6)2), but the GSR lifetime of the 4’-OH substituted complex is nearly half that of the 4’-N(EtOH)2 substituted complex (24.8 ± 0.4 ns and 55.3 ± 0.9 ns, respectively). 42 Figure 3.8: The τ295 value of GSR collected across a series of [Fe(4’-R-terpy)2](PF6)2 complexes plotted against the Ea determined for each complex via VT-TA analysis. Beside each point is the substituent on that complex. Doing the same analysis with transition state theory yields Figure 3.9. When analyzed through transition state theory, the experimental data follows the model much more closely than the Arrhenius treatment. For example, the 4’-OH substituted complex now has a barrier value that is better in line with the measured τ295. The difference between the plots in Figures 3.8 and 3.9 displays clear evidence that transition state theory much more accurately describes the energy barrier of these systems than Arrhenius theory. Understanding why transition state theory is more Figure 3.9: The τ295 value of GSR collected across a series of [Fe(4’-R-terpy)2](PF6)2 complexes plotted against the ΔG‡ determined for each complex via VT-TA analysis. Beside each point is the substituent on that complex. 43 accurate relies on determining the difference in how the barrier is calculated. Namely, the exclusion of ΔS‡ in Arrhenius theory means that Arrhenius theory is ignoring critical information about the barrier. As mentioned earlier, ΔS‡ is calculated from the y-intercept (therefore, the pre-exponential term) and ΔH‡ is determined from the slope (the exponential term). Directly comparing the Arrhenius terms to the transition state terms highlights the consequence of this distinction. Table 3.1 compares the terms derived from all three theories for the complexes discussed earlier. Looking at the values for Ea and ΔG‡ shows very different values across the series. Ea is consistently lower in energy than ΔG‡ by a significant margin. Comparing Ea to ΔH‡ instead shows that the two values are much closer in energy, and that they follow the same trends. This relationship is seen across the entire series, displayed in Figure 3.10. Table 3.1: The values for τ295, the Arrhenius terms (Ea and A), transition state terms (ΔH‡, ΔS‡, and ΔG‡), and Marcus terms (Hab and ΔGo) for GSR in a select few complexes studied. This information is provided for the entire series in Table S3.1 in the SI. The ΔG‡ used through this chapter were calculated with a temperature of 295 K. Low temperature treatment (T = 240 K), is available in the SI. The Marcus terms were calculated from the experimentally determined ratios in equations 3.22 and 3.23 using estimated values for λ determined via the Nelsen method as described above. Substituent τ295 (ns) Ea (cm-1) A (ns-1) ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) Hab (cm-1) ΔGo (cm-1) H 4.9 ± 0.2 650 ± 10 4.89 ± 0.15 465 ± 10 -5.58 ± 0.02 2110 ± 15 5.90 ± 0.08 -5770 ± 40 OH 24.8 ± 0.4 935 ± 10 3.81 ± 0.11 750 ± 10 -5.76 ± 0.02 2450 ± 10 5.20 ± 0.08 -4445 ± 20 N(EtOH)2 55.3 ± 0.9 900 ± 20 1.43 ± 0.15 715 ± 20 -6.44 ± 0.07 2615 ± 40 3.15 ± 0.16 -4060 ± 65 Figure 3.10: A: Ea plotted against ΔG‡. B: Ea plotted against ΔH‡. Plot B shows stronger linearity than plot A, establishing a more direct relationship between Ea and ΔH‡ than there is between Ea and ΔG‡. 44 The strongly direct linear relationship between Ea and ΔH‡, coupled with the fact that both terms are derived solely from the slope of their respective plots, suggests that Arrhenius theory only produces the enthalpy portion of the energy barrier by ignoring information from the y-axis, and therefore inadequately describes the energy barrier when ΔS‡ ≠ 0. Not only does the inclusion of ΔS‡ increase the energy value of the calculated barrier, but the relationship between τ295 and the barrier is far more accurate to an exponential fit when ΔS‡ is used. This means that ΔS‡ contains critical information that must be considered alongside enthalpy for an accurate description of the barrier. To discern what this critical information is, Marcus theory must be considered. Both ΔS‡ and Hab are derived from the pre-exponential components of their respective equations, meaning that they both contain information from the y-axis of their respective “Arrhenius-type” plots. Looking at Table 3.1 reveals that the 4’-OH substituted complex has much stronger electronic coupling than the 4’-N(EtOH)2 substituted complex. Recall from equation 3.7 that knr is directly proportional to the square of Hab. This means that increasing Hab decreases τ295, which is observed in Table 3.1. Furthermore, stronger Hab correlates with less negative ΔS‡ values, suggesting that there is a relationship between the two. Setting the pre-exponential components of the Eyring and Marcus equations proportional to one another and solving for ΔS‡ yields the following equations: 𝜅𝑘𝐵𝑇 ℎ ∗ 𝑒 ‡ ΔS 𝑘𝐵 ∝ 2𝜋 ℏ ∗ |𝐻𝑎𝑏|2 ∗ 1 √4𝜋𝜆𝑘𝐵𝑇 𝛥𝑆‡ ∝ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3) 3.25 3.26 Equation 3.26 shows that ΔS‡ should be linearly proportional to the natural log of |𝐻𝑎𝑏|4 𝜆⁄ , which is one of the ratios that can be experimentally determined through VT Marcus theory analysis. Taking the natural log of each ratio and plotting it against the ΔS‡ determined for each molecule in the series used in the previous plots provides Figure 3.11. The strong linearity of the plot in Figure 3.11 supports the validity of using equation 3.26 to describe the data. This equation shows that by including information from the y-axis through the inclusion of ΔS‡, transition state theory considers the impact that electronic coupling has on the energy barrier. Looking back at the non-adiabatic model of Marcus theory provided in Figure 3.2, the relationship between the ΔG‡ and Hab becomes clear. Since the splitting between the upper and lower non-adiabatic wells at the transition state is equal to 2Hab, then increasing the strength of Hab stabilizes the lower well while destabilizing the upper well. The crossing dashed lines in 45 Figure 3.11: Plot of 𝑙𝑛(|𝐻𝑎𝑏|4 𝜆⁄ ) vs ΔS‡ for GSR across a series of [Fe(4’-R-terpy)2](PF6)2 complexes. Figure 3.2 represent what the barrier would be in a diabatic model; if Hab = 0. In other words, increasing Hab reduces the energy needed to form the transition state, represented by ΔG‡. This is a relationship that Arrhenius theory cannot account for, but transition state theory does. It is worthwhile to address the other Marcus term of the ratio embedded in ΔS‡: λ. As seen in equation 3.18, λ contains information about the reaction coordinate. Specifically, ΔQ represents the difference in reaction coordinate between electronic states and f represents the width of the potential energy surface of the reactant state along that reaction coordinate. Both terms influence the energy barrier in different ways. Increasing ΔQ increases the barrier, while increasing f reduces the barrier. However, the influence λ has on ΔG‡ does not stem exclusively through ΔS‡. Just as the pre-exponential components of the Eyring and Marcus equations can be set equal, the same treatment can be done for the exponential components. Doing so yields the following: 𝛥𝐻‡ ∝ (𝛥𝐺𝑜+𝜆)2 4𝜆 3.27 However, the same treatment could be done using equation 3.5 which would set equation 3.27 equal to ΔG‡ rather than ΔH‡. The ratio (𝛥𝐺𝑜 + λ)2 λ⁄ can be determined experimentally using equation 3.23, so to determine which term, ΔG‡ or ΔH‡, more directly correlates to the exponential component of the Marcus equation, each can be plotted against this ratio. These plots are provided in Figure 3.12. 46 Figure 3.12: A: ΔG‡ plotted against (𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ . B: ΔH‡ plotted against (𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ . Plot B shows stronger linearity than plot A, supporting that (𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ is more directly related to ΔH‡ than ΔG‡. Figure 3.12 supports the use of equation 3.27, since the plot using ΔH‡ had near perfect linearity, while the plot using ΔG‡ had notable deviations from linearity. Since both ΔH‡ and ΔS‡ are dependent on λ in some way, both must be considered to understand the nature of the relationship between λ and the energy barrier. Having established Marcus theory descriptions for ΔH‡ and ΔS‡, equations 3.26 and 3.27 can be substituted into the Gibbs free energy equation to describe the energy barrier using Marcus theory terms: 𝛥𝐺‡ ∝ [ (𝛥𝐺𝑜+𝜆)2 4𝜆 ] − 𝑇 [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)] 3.28 This equation shows that the dependence of the energy barrier on λ can roughly be described by the following expression: 𝛥𝐺‡ ∝ λ − ln(𝜆) 3.29 Equation 3.29 makes clear that the bulk of the dependency ΔG‡ has on λ comes from the ΔH‡, which Ea should be able to account for. This highlights that the key difference between Arrhenius and transition state theory lies in how Hab impacts the energy barrier. Overall, Figures 3.8 and 3.9 clearly display that transition state theory is a much better model for determining the energy barrier than Arrhenius theory in these systems. This is because Ea more closely determines the enthalpy of activation, rather than the entire barrier, as evidenced by the models shown in Figure 3.10. Since Arrhenius theory only includes information from the slope of the plot, information from the pre-exponential component of the equation is left out of Ea. Instead, ΔG‡ is calculated using information from both the slope and the y-intercept. Equation 3.28 displays ΔG‡ described in Marcus theory terms and shows that by using information from both the 47 exponential component (derived from the slope) and the pre-exponential term (derived from the y-intercept), transition state theory can incorporate information about ΔGo, λ, and Hab, each of which influence the barrier in different ways. By leaving out the pre-exponential term, Arrhenius theory is incapable of accounting for how Hab influences the barrier. Furthermore, since Arrhenius theory does not include −𝑇∆𝑆‡ when calculating the energy barrier, Arrhenius theory under- estimates the energy values of the barrier. This analysis concludes that transition state theory, rather than Arrhenius theory, should be used to interpret the results of variable temperature experiments on transition metal-based chromophores. 3.4 Resolving Discrepancies between Transition State and Marcus Theory Finally, it is important to address the difference in how temperature is applied to the y-axis between transition state and Marcus theory. For transition state theory, the natural log of 𝑘𝑛𝑟 𝑇⁄ is used for the y-axis, but Marcus theory uses the natural log of 𝑘𝑛𝑟 ∗ √𝑇. This discrepancy brings a challenge when drawing direct comparisons between each theory. For example, when comparing the Eyring equation to the semi-classical Marcus theory equation, ΔH‡ should be equal to (𝛥𝐺𝑜 + λ)2 4λ⁄ , however this does not work out empirically. For example, the experimentally determined values for [Fe(terpy)2]2+ are as follows: ΔH‡ = 465 ± 10 cm-1 and (𝛥𝐺𝑜 + λ)2 4λ⁄ = 742.5 ± 7 cm-1. This is true for the rest of the series, shown in Table 3.2: Table 3.2: The exponential components from Eyring theory (ΔH‡) and Marcus theory ((𝛥𝐺𝑜 + 𝜆)2 4𝜆⁄ ) for GSR across a series of 4’-position substituted [Fe(4’-R-terpy)2]2+ complexes. The substituent is given for each complex. Substituent ΔH‡ (cm-1) (ΔGo+λ)2/4λ (cm-1) SO2Me 240 ± 20 515 ± 17 CN 330 ± 25 608 ± 21 H 465 ± 10 743 ± 7 Cl 495 ± 10 770 ± 10 Furan 505 ± 15 783 ± 14 SMe OMe 655 530 ± 20 ± 20 933 810 ± 7 ± 21 N(EtOH)2 715 ± 10 993 ± 20 OH 750 ± 10 1025 ± 7 Despite the discrepancy in values, Figure 3.12B displays clearly that two values have strong linear correlation. This is why equations comparing transition state theory parameters to Marcus theory parameters (such as equations 3.26 and 3.27) use the “proportional to” symbol rather than the “equal to” symbol. Though, taking the difference between each value yields the same result across all complexes in the series: 277 ± 3. This means that the offset in values created by the different treatment of temperature in the y-axis is a constant: 277 cm-1, which will be called βH. This suggests that the following adaptation of equation 3.27 is true: 48 𝛥𝐻‡ = (𝛥𝐺𝑜+𝜆)2 4𝜆 − 𝛽𝐻 3.30 To test the validity of equation 3.30, the Marcus analysis-derived VT-TA data from a series of [Fe(bpy)3]2+ complexes were run through equation 3.30, and the result for each complex is within error of ΔH‡ determined through Eyring analysis. These results are summarized in Table 3.3. Table 3.3: The exponential components from Eyring theory (ΔH‡) and Marcus theory ((𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ ) for GSR across a heterologous series of substituted [Fe(X,X’-diR-bpy)2]2+ complexes. The substituent is given for each complex. 5,5’- diOMe 4,4’- 4,4’- diOMe diMe 120 ± 10 125 ± 20 155 ± 20 160 ± 10 165 ± 30 310 ± 25 335 ± 35 1720 ± 30 4,4’-diCl 5,5’-diCl 1750 ± 120 2420 ± 130 5,5’- diMe 1715 ± 60 1565 ± 40 1600 ± 75 2320 ± 90 Substituent H 114 ± 10 123 ± 19 152 ± 15 152 ± 8 161 ± 30 303 ± 23 328 ± 33 ΔH‡ (cm-1) (ΔGo + λ)2/λ (cm-1) [(ΔGo + λ)2/4λ] – βH (cm-1) The agreement between transition state and Marcus theory descriptions of the enthalpic term support the use of βH as a corrective term between the different analytic methods. This only addresses the discrepancy in enthalpy, but the same analysis was done for the entropic descriptions from each theory as well. Equation 3.26 shows that ΔS‡ should be proportional to [ ln ( 𝑘𝐵 2 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)]. Within that expression is the experimentally determinable ratio |𝐻𝑎𝑏|4 𝜆⁄ alongside several constants, as well as T, for this analysis, T is assumed to be 295 K. Plugging these values within that expression provides the data summarized in Table 3.4. Table 3.4: The Eyring and Marcus theory descriptions for the entropy of activation for GSR across a series of 4’-position substituted [Fe(4’-R-terpy)2]2+ complexes. Marcus entropy values were calculated from experimental data using the following expression: [ 𝑙𝑛 ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)]. Substituent ΔS‡ (cm-1/K) Marcus Entropy (cm-1/K) SO2Me -5.64 ± 0.06 -4.71 ± 0.06 CN -5.42 ± 0.08 -4.48 ± 0.08 H -5.58 ± 0.02 -4.65 ± 0.02 Cl -5.64 ± 0.03 -4.71 ± 0.04 Furan -5.60 ± 0.05 -4.66 ± 0.05 OMe N(EtOH)2 OH -6.44 ± -5.99 ± 0.07 0.07 -5.52 ± -5.06 ± 0.08 0.08 -5.76 ± 0.02 -4.82 ± 0.02 𝑘𝐵 2 SMe -6.02 ± 0.07 -5.08 ± 0.08 Like the enthalpic term, the difference between entropy values is the same across the series: 0.93 ± 0.01 cm-1/K. Using this constant, dubbed βS, equation 3.26 is adjusted to the following at 295 K: 𝛥𝑆‡ = [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)] − 𝛽𝑆 3.31 Once again, this adapted equation was tested by using the Marcus analysis-derived VT-TA data from the same heterologous substituted [Fe(bpy)3]2+ series. And again, the result for each complex 49 fell within error of the Eyring analysis-determined ΔS‡ value, supporting the use of βS as a corrective term for calculating activation entropy using Marcus theory. These results are shown in Table 3.5: Table 3.5: The Eyring and Marcus theory descriptions for the entropy of activation for GSR across a heterologous series of substituted [Fe(X,X’-diR-bpy)2]2+ complexes. Adjusted Marcus entropy following expression: from experimental data using values were calculated the [𝑙𝑛 ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3) ∗ 𝑘𝐵 2 ] − 0.93. Substituent 5,5’-diMe H ΔS‡ (cm-1/K) |Hab|4/λ (cm-3) Adjusted Marcus Entropy (cm-1/K) -5.68 ± 0.03 0.080 ± 0.008 -5.68 ± 0.03 -5.66 ± 0.07 0.086 ± 0.017 -5.66 ± 0.07 5,5’- diOMe -5.47 ± 0.06 0.145 ± 0.025 -5.48 ± 0.06 4,4’-diMe 4,4’-diCl 5,5’-diCl -5.66 ± 0.02 0.082 ± 0.007 -5.67 ± 0.03 -5.85 ± 0.11 0.049 ± 0.017 -5.87 ± 0.13 -5.42 ± 0.08 0.168 ± 0.043 -5.44 ± 0.09 4,4’- diOMe -5.91 ± 0.06 0.029 ± 0.010 -6.06 ± 0.12 However, it is vital to note that since equation 3.26 has a temperature dependency, βS is only equal to 0.93 at 295 K. To determine the relationship between βS and temperature, equation 3.31 is rearranged to solve for βS. 𝛽𝑆 = [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)] − 𝛥𝑆‡ 3.32 Equation 3.32 shows that there is a linear relationship between βS and the natural log of T-3. To construct a model for determining βS, equation 3.32 was rearranged to resemble to a linear model: 𝛽𝑆 = 𝑘𝐵 2 ln(𝑇−3) + [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2𝑘𝐵 3 ) − 𝛥𝑆‡] 3.33 Where 𝑘𝐵 2⁄ is the slope of the model and the expression [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2𝑘𝐵 3 ) − 𝛥𝑆‡] represents the y-intercept. This y-intercept expression was solved using the experimentally determined |𝐻𝑎𝑏|4 𝜆⁄ Table 3.6: The pre-exponential components from Eyring theory (ΔS‡) and Marcus theory (|𝐻𝑎𝑏|4 𝜆⁄ ) for GSR across a series of 4’-position substituted [Fe(4’-R-terpy)2]2+ complexes. The 3) − 𝛥𝑆‡] from equation 3.33 using the Y-intercept is solution for the expression [ 𝑙𝑛 ( 𝑘𝐵 2 |𝐻𝑎𝑏|4 4𝜋𝜆𝜅2𝑘𝐵 listed variables for that complex. 50 Substituent SO2Me CN H Cl Furan SMe OMe N(EtOH)2 OH ΔS‡ (cm-1/K) -5.64 ± 0.06 -5.42 ± 0.08 -5.58 ± 0.02 -5.64 ± 0.03 -5.60 ± 0.05 -6.02 ± 0.07 -5.99 ± 0.07 -6.44 ± 0.07 -5.76 ± 0.02 |Hab|4/λ (cm-3) 0.092 ± 0.016 0.179 ± 0.040 0.108 ± 0.006 0.091 ± 0.010 0.105 ± 0.015 0.032 ± 0.007 0.034 ± 0.008 0.009 ± 0.002 0.065 ± 0.004 Y-intercept 6.865 6.877 6.861 6.862 6.871 6.878 6.869 6.858 6.865 Figure 3.13: Plot of the activation entropy corrective constant, βS against ln(T-3). Shown alongside 𝑙𝑛(𝑇−3) + (𝑌 − the temperature dependent constant are models fit to the expression [ 𝑘𝐵 2 𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡)] , where the slope is 𝑘𝐵 2⁄ = 0.3475 cm-1/K and the y-intercept is substituted for the values listed in the legend. ratio and ΔS‡ values for each complex in the series. Each result equaled 6.868 ± 0.010 cm-1/K. This data is summarized in Table 3.6. This provides a small window of possible values that the y-intercept for modeling βS could fall under. To determine which value would provide the most accurate model, equation 3.33 was solved substituting the expression [ 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2𝑘𝐵 3 ) − 𝛥𝑆‡] for several values ranging from 6.859 cm-1/K to 6.874 cm-1/K. These predictive models were plotted against experimentally determined values for βS determined at each temperature within the VT-TA data collection window (240 K to 295 K). This plot, seen in Figure 3.13, finds that a y-intercept of 6.862 cm-1/K provided the best fit for the experimental results. This means that βS can be determined at any temperature by using the following equation: 𝛽𝑆 = 𝑘𝐵 2 ln(𝑇−3) + 6.862 3.34 This establishes constants that act as corrective terms for the enthalpy and entropy of activation, βH and βS. These empirically determined constants allow ΔH‡ and ΔS‡ to be accurately calculated through Marcus theory analysis of VT kinetic data. This also means that ΔG‡ can now be accurately calculated from Marcus theory parameters by using the following equation: 𝛥𝐺‡ = [( (𝛥𝐺𝑜+𝜆)2 4𝜆 ) − 𝛽𝐻] − 𝑇 [( 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)) − 𝛽𝑆] 3.35 51 To verify the accuracy of these constants, ΔG‡ was calculated using only Marcus theory parameters following equations 3.34 and 3.35 across both series of complexes discussed in this chapter. When compared to the Eyring analysis-determined ΔG‡ values, the Marcus analysis- determined ΔG‡ display a strong degree of accuracy. The data for [Fe(4’-R-terpy)2]2+ series is found in Table 3.7, and the same data for the [Fe(bpy)3]2+ series is found in Figure 3.8. Table 3.7: The energy barrier of GSR across a series of 4’-position substituted [Fe(4’-R-terpy)2]2+ complexes. Eyring ΔG‡ was determined through Eyring analysis of the VT kinetic data collected for the series. Marcus ΔG‡ was calculated using equation 3.35 with Marcus theory parameters determined by analyzing the VT kinetic data through Marcus theory. This done was at temperatures of 295 K and 240 K, βS was calculated at each temperature using equation 3.34. Substituent SO2Me OMe N(EtOH)2 OH Furan SMe CN Cl H Eyring ΔG‡ (cm-1) T = 295 K Marcus ΔG‡ (cm-1) T = 295 K Eyring ΔG‡ (cm-1) T = 240 K Marcus ΔG‡ (cm-1) T = 240 K 1905 ± 35 1930 ± 50 2110 ± 15 2160 ± 20 2155 ± 30 2310 ± 40 2425 ± 40 2615 ± 40 2450 ± 10 1905 ± 35 1930 ± 45 2110 ± 15 2160 ± 20 2155 ± 30 2305 ± 45 2425 ± 45 2620 ± 45 2450 ± 15 1595 ± 35 1630 ± 45 1805 ± 15 1850 ± 20 1850 ± 25 1980 ± 40 2095 ± 35 2260 ± 35 2130 ± 10 1595 ± 30 1630 ± 40 1805 ± 10 1845 ± 20 1850 ± 25 1975 ± 40 2095 ± 40 2265 ± 40 2130 ± 10 Table 3.8: The energy barrier of GSR across a heterologous series of substituted [Fe(X,X’-diR- bpy)2]2+ complexes. Eyring ΔG‡ was determined through Eyring analysis of the VT kinetic data collected for the series. Marcus ΔG‡ was calculated using equation 3.35 with Marcus theory parameters determined by analyzing the VT kinetic data through Marcus theory. This done was at temperatures of 295 K and 240 K, βS was calculated at each temperature using equation 3.34. Substituent 5,5’-diMe 5,5’-diOMe 4,4’-diMe 5,5’-diCl 4,4’-diOMe 4,4’-diCl H Eyring ΔG‡ (cm-1) T = 295 K Marcus ΔG‡ (cm-1) T = 295 K Eyring ΔG‡ (cm-1) T = 240 K Marcus ΔG‡ (cm-1) T = 240 K 1790 ± 20 1795 ± 40 1770 ± 35 1830 ± 15 1890 ± 60 1905 ± 45 2080 ± 70 1790 ± 20 1795 ± 40 1770 ± 35 1830 ± 15 1895 ± 65 1905 ± 50 2115 ± 70 1475 ± 10 1485 ± 40 1450 ± 20 1510 ± 10 1540 ± 30 1590 ± 20 1720 ± 35 1480 ± 15 1485 ± 35 1465 ± 30 1515 ± 15 1570 ± 60 1610 ± 45 1780 ± 60 52 These results establish that by using the corrective constants, βH for the enthalpy barrier and βS for the entropy barrier, the discrepancy between how Eyring and Marcus analysis determine the y-axis has been sufficiently resolved. For further rigorous testing, a mathematic program was developed using Microsoft Excel where any value for ΔH‡ and ΔS‡ could be input, and the program would then use transition state theory to calculate knr across a range of temperatures. Then the program performs Marcus-analysis on the data set, calculating the Marcus ratios discussed throughout this chapter and plugged these values into equations 3.30, 3.31, and 3.35 to calculate ΔH‡, ΔS‡, and ΔG‡. For any given input, without fail, the program was able to return the input values within three hundredths of a wavenumber for enthalpy, and three thousandths of a wavenumber for entropy; showing that this model should work for other transition metal chromophore systems. For example, ΔS = 1 electronic transitions within transition metal complexes typically have a ΔS‡ value near -3 cm-1/K, such as those based on Co(III).10 This model was tested using various entropy values near -3, and each returned the given input. This held true when using absurdly large or small values for enthalpy and entropy. The only scenario where the model could not reproduce the input values is when the barrier is so large that the computer treats knr as zero, in other words it fails in situations where no transition could reasonably occur. On the topic of units, the units for βH and βS are in wavenumbers, since that is the energy unit used for kB and ℏ, if a different energy is used for those constants, βH and βS would need to be converted as well. This analysis shows that, despite the model being determined through empiric investigation of the data, it is robust enough to describe systems across a wide range of energy barriers. This is because the model is simply addressing how adding either ln(√T) or ln(1/T) to the y-axis changes both the slope and y-intercept of the resulting plot. The actual values for knr are not relevant to how βH and βS were determined. 3.5 Concluding Thoughts In this chapter I have reviewed the three different theories used in this work to interpret the results of VT-TA experiments. Furthermore, I have explained how the variable temperature analysis is done, including ways our thinking regarding these analyses have evolved since VT-TA analysis was last used in publication from our group. Finally, an in-depth analysis of the differences between how Arrhenius theory and transition state theory calculate the energy barrier between electronic states was done. This analysis displayed that Arrhenius theory could only describe the 53 enthalpy portion of the barrier, and that transition state theory was able to accurately model the experimental results. This should hold true for any system with a significant ΔS‡ value. For this reason, I suggest that transition state theory should be relied on over Arrhenius theory examine VT-TA data moving forward. Lastly, the discrepancy in parameters between transition state theory and Marcus theory caused by the difference in how the y-axis is determined has been addressed. By using two empirically determined constants, one for correcting the activation enthalpy (βH) and one for correcting the activation entropy (βS), it is possible to accurately determine ΔG‡ through Marcus theory analysis of VT-TA data alone. It would be worthwhile to further test these constants by analyzing the VT-TA data of other series of complexes, especially those other than Fe(II)-based polypyridines. It is important to note that βH is dependent on both the range of temperatures used, as well as the number of data points within the range. As a result, βH is only equal to 277 cm-1 at the temperature set used in this study: [240 K, 245 K, 250 K, 255 K, 260 K, 265 K, 270 K, 275 K, 280 K, 285 K, 290 K, 295 K]. If either the number of points or the temperature window is altered, βH will change slightly. If these changes are small, a value of 277 cm-1 will still result in accurate approximations for ΔH‡. However, the new value of βH could be solved for any given temperature set by using a mathematic program, like the Excel program described above. Further work will be needed to develop a model for determining βH, as was done for βS. Until then, the use of 277 cm-1 should suffice for temperature sets similar to what was used in this study. 54 1. Laidler, K. J. and King, M. C. J. Phys. Chem. 1983, 87, 2657 – 2664. REFERENCES 2. Sutin, N. Prog. Inorg. Chem. 1983, 30, 441 – 498. 3. Carey, M. C.; Adelman, S. L. and McCusker, J. K. Chem. Sci. 2019, 10, 134 – 144 . 4. Marcus, R. A. and Sutin, N. Biochim. Biophys. Acta. 1985, 811, 265 – 322. 5. Hsu, C.-P. Acc. Chem. Res. 2009, 42, 509 – 518. 6. Adelman, S. L. Searching for Kinetic Control of Excited-State Evolution in Fe(II) Polypyridyl Chromophores. Doctoral Thesis, Michigan State University. ProQuest Dissertations Publishing. 7. Nelsen, S. F.; Blackstock, S. C. and Kim, Y. J. Am. Chem. Soc. 1987, 109, 677 – 682. 8. López-Estrada, O.; Laguna, H. G.; Barrueta-Flores, C. and Amador-Bedolla, C. ACS Omega. 2018, 3, 2130 – 2140. 9. Griffin, P. J. and Olshanksy, L. J. Am. Chem. Soc. 2023, 145, 20158 − 20162. 10. Ghosh, A.; Yarranton, J. T. and McCusker, J. K. Nature Chem. 2024. 55 APPENDIX: SUPPLEMENTAL INFORMATION 3.SI.1 Additional VT-TA Data The information used to produce the plots seen in this chapter is summarized in the following tables. A greater analysis of these data sets will be thoroughly explained in Chapter 4. Table 3.S1: The GSR time constant, Arrhenius, transition state, and Marcus theory parameters for a series of 4’-substituted [Fe(terpy)2](PF6)2 complexes. Hab and ΔG‡ were calculated from the values provided in Table S3.2. Substituent τ295 (ns) Ea (cm-1) A (ns-1) ΔH‡ (cm-1) ΔS‡ (cm-1/K) ΔG‡ (cm-1) Hab (cm-1) ΔGo (cm-1) SO2Me 1.75 ± 0.05 425 ± 20 4.50 ± 0.40 240 ± 20 -5.64 ± 0.06 1905 ± 35 5.60 ± 0.24 -5920 ± 75 CN 2.02 ± 0.06 515 ± 25 6.25 ± 0.74 330 ± 25 -5.42 ± 0.08 1930 ± 50 6.70 ± 0.38 -5875 ± 90 H Cl 4.9 ± 0.2 650 ± 10 4.89 ± 0.15 465 ± 10 -5.58 ± 0.02 2110 ± 15 5.90 ± 0.08 -5770 ± 40 6.0 ± 0.1 680 ± 15 4.49 ± 0.23 495 ± 10 -5.64 ± 0.03 2160 ± 20 5.70 ± 0.15 -5725 ± 25 Furan 5.8 ± 0.2 690 ± 15 4.81 ± 0.35 505 ± 15 -5.60 ± 0.05 2155 ± 30 5.90 ± 0.21 -5305 ± 50 SMe 12.5 ± 0.6 720 ± 25 2.61 ± 0.30 530 ± 20 -6.02 ± 0.07 2310 ± 40 4.40 ± 0.25 -5055 ± 80 OMe 21.8 ± 0.2 840 ± 20 2.73 ± 0.30 655 ± 20 -5.99 ± 0.07 2425 ± 40 4.40 ± 0.25 -4875 ± 75 OH 24.8 ± 0.4 935 ± 10 3.81 ± 0.11 750 ± 10 -5.76 ± 0.02 2450 ± 10 5.20 ± 0.08 -4445 ± 20 N(EtOH)2 55.3 ± 0.9 900 ± 20 1.43 ± 0.15 715 ± 20 -6.44 ± 0.07 2615 ± 40 3.15 ± 0.16 -4060 ± 65 Table 3.S2: The experimentally determined Marcus ratios alongside λ, which was calculated using DFT. Substituent |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) λ (cm-1) SO2Me 0.092 ± 0.016 2065 ± 65 CN H Cl 0.179 ± 0.040 2430 ± 85 0.108 ± 0.006 2970 ± 25 0.091 ± 0.010 3080 ± 40 Furan 0.105 ± 0.015 3130 ± 55 SMe OMe OH 0.032 ± 0.007 3240 ± 85 0.034 ± 0.008 3730 ± 85 0.065 ± 0.004 4105 ± 25 N(EtOH)2 0.009 ± 0.002 3970 ± 80 10600 11100 11600 11800 11200 11000 11400 11200 10500 56 3.SI.2 Linear Transition State Theory Plots The plot in Figure 3.9 can be lineated by using the natural log of the rate constant instead. In this case, the plot represents the following equation: The slope of this plot should be equal to − 1 𝑘𝐵𝑇⁄ , which at T = 295 K is equal to -0.00488. This ln 𝑘𝑛𝑟 = ln 𝜅𝑘𝐵𝑇 ℎ + (𝛥𝐺‡ ∗ − 1 𝑘𝐵𝑇 ) 3.S1 value is within error of the slope of the fit shown below. Figure 3.S1: Plot of ln(k295) vs ΔG‡ across a series of [Fe(4’-R-terpy)2](PF6)2 complexes. This same plot can be constructed using knr values collected at other temperatures. Shown in Figure 3.S2 is a plot using the knr values measured at 240 K and ΔG‡ values calculated using 240 K for T in the Gibbs free energy equation. The slope also changes to match equation 3.S1, where − 1 𝑘𝐵𝑇⁄ at T = 240 K is equal to -0.00599. Again, this rests within error of the slope from the fit. Figure 3.S2: Plot of ln(k240) vs ΔG‡ across a series of [Fe(4’-R-terpy)2](PF6)2 complexes. 57 This analysis was performed on a different series of Fe(II)-polypyridines: a series of 4,4’ or 5,5’-position substituted [Fe(bpy)3]2+ complexes. The counter anion in this series is mostly (PF6)-, however for the dichloro substituted complexes, a (BF4)- anion had to be used. The model was able to fit the experimental data of this series to a similar degree of accuracy. Table 3.S3: The experimentally determined τ, ln(knr), and ΔG‡ for a series of [Fe(bpy)3]2+ complexes substituted at either the 4,4’ or 5,5’-positions. These data sets were used to construct the plots in Figures 3.S3. Substituent τ295 (ps) ln(knr) ΔG‡ (cm-1) H 1030 ± 15 20.69 ± 0.01 1795 ± 40 4,4’-diMe 1220 ± 120 20.53 ± 0.10 1830 ± 15 5,5’-diMe 1035 ± 20 20.69 ± 0.02 1790 ± 20 4,4’-diOMe 4360 ± 460 19.25 ± 0.11 2080 ± 70 5,5’-diOMe 945 ± 25 20.78 ± 0.02 1770 ± 35 4,4’-diCl 1615 ± 30 20.24 ± 0.02 1890 ± 60 5,5’-diCl 1760 ± 80 20.16 ± 0.05 1905 ± 45 Figure 3.S3: The exponential (A) and linear (B) models for transition state theory analysis on a series of [Fe(bpy)3]2+ complexes substituted at either the 4,4’ or 5,5’-positions. The slope from plot B again falls within error of − 1 𝑘𝐵𝑇⁄ at T = 295 K (-0.00488). 3.SI.3 Additional Thoughts Regarding βS As another method of verifying the accuracy of equation 3.33, the values for βS calculated from experimental data were once again plotted against the natural log of T-3. However, this time instead of overlapping calculated models with varying y-intercept values, the experimental data was simply fit to the linear equation using IgorPro. The resulting y-intercept was 6.8618 ± 58 0.000261 and the slope was 0.3749 ± 1.56E-05. This falls exactly in line with what is predicted in equation 3.33. The plot is shown in Figure 3.S4. Figure 3.S4: Plot of the activation entropy corrective constant, βS against ln(T-3). Each βS was calculated from equation 3.33 using ΔS‡ and |𝐻𝑎𝑏|4 𝜆⁄ determined via Eyring and Marcus analysis of the VT-TA data, respectively. Furthermore, βS has a very slight dependency on the transmission coefficient, κ. The calculations performed throughout this dissertation all assume that κ = 1. However, given the relatively low estimations for Hab, κ is likely significantly less than 1. This has an impact on the value of ΔS‡ calculated for a transition. The value of ΔS‡ is directly related to the -ln(κ), so ΔS‡ at any given κ value can be determined from the following equation: Where x is the chosen value of κ. The ΔS‡ presented throughout the remainder of this dissertation 𝛥𝑆‡ 𝜅=𝑥 = 𝛥𝑆‡ 𝜅=1 − ln(𝑥) 3.S2 were calculated with a κ = 1 (𝛥𝑆‡ 𝜅=1). Equation 3.S2 was carried out for several κ values across several complexes, and these adjusted 𝛥𝑆‡ 𝜅=𝑥 values were used to observe how βS changes with κ. These results are provided for [Fe(terpy)2]2+ in Figure 3.S5. It was found that 𝛽𝑆𝜅=1 = 0.936421 and 𝛽𝑆𝜅=0.1 = 0.93640. Meaning that with a κ range from 0.1 to 1, the impact on βS was on the scale of 10-5. Considering the value provided for βS in this chapter stopped at three significant figures, the impact of κ on βS is considered negligible. 59 Figure 3.S5: The calculated value for βS at T = 295 K with κ values ranging from 0.1 to 1. The trend was fit to the following equation: 𝑦 = 𝑦0 + 𝐴𝑒𝑥𝑝 ൜− [ 𝑙𝑛(𝑥 𝑥0⁄ ) 𝑤𝑖𝑑𝑡ℎ 2 ] ൠ. 60 CHAPTER 4: VARIABLE-TEMPERATURE TIME-RESOLVED SPECTROSCOPY OF A HOMOLOGOUS SERIES OF FE(II)-TERPYRIDYL COMPLEXES 4.1 Introduction As briefly discussed in the first chapter, a quenching study performed by Woodhouse and McCusker on [Fe(tren(py)3]2+ (where tren(py)3 is tris(2-pyridyl-methylimino-ethyl)amine) showed that electron acceptors were able to dampen the lifetime of the long-lived high spin excited state, 5T1.1 A Sterm-Volmer plot constructed in that study shows that increasing the concentration of the quencher further reduced the 5T2 lifetime. This clearly established that it was possible to use Fe(II)-based polypyridine complexes to act as photosensitizers from the ligand field (LF) excited states. This suggests that the mechanism for previously reported uses of Fe(II) polypyridyl-based chromophores in photocatalysis also involved electron transfer from the long-lived LF excited state.2-4. Recall, that one of the largest hinderances towards using Fe-based systems as opposed to those relying on rarer metals was the rapid deactivation of the charge separated metal-to-ligand charge transfer (MLCT) excited state.5,6 However, by inducing electron transfer from the LF excited states, this issue is not only bypassed, but new photocatalysis mechanisms become possible. Therefore, further exploration of this electron transfer mechanism is important. To further explore electron transfer from the LF excited states of Fe(II)-polypyridyl chromophores, additional quenching studies were conducted on [Fe(bpy)3]2+ (where bpy is 2,2’- bipyridine) and [Fe(terpy)2]2+ (where terpy is 2,2’:6’,2”-terpyridine). Early results showed that quencher concentrations had no impact on the lifetime of the 5T2 excited state, seen in Figure 4.1. Figure 4.1: Single wavelength kinetics two Fe(II)-polypyridine solutions with varying concentrations of quenchers in acetonitrile. A: [Fe(bpy)3]2+ excited and probed by 500 and 550 nm pulses, respectively. Added to the solution was varying concentrations of tetracyanoethylene (TCNE). B: [Fe(terpy)2]2+ excited and probed by 560 and 530 nm pulses, respectively. Added to the solution was varying concentrations of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ). 61 The difficulty of inducing electron transfer from the LF excited states in these complexes highlights the importance of characterizing the 5T2 excited state. Understanding the nature of the excited state will provide knowledge in how to manipulate the electronic structure to yield excited state properties that will best enable electron transfer. One of the most fundamental characteristics regarding excited state relaxation is which Marcus region the transition between electronic states occurs in. As discussed in Chapter 2, there are three regions within Marcus theory that could describe the relationship between electronic states: the normal region, the barrierless region, and the inverted.7 Understand which region two electronic states are in is vital for understanding how changes to the electronic structure will impact excited state dynamics. As previously discussed, the relationship between the driving force, or ΔGo (the free energy between states), and knr has opposite relationships in the normal an inverted regions (direct in the normal region and inverted in the inverted region). It has long been assumed that ground state recovery (GSR) process in Fe(II)-polypyridines has occurred in the Marcus normal region,6 but this has never been experimentally established. The work in this chapter seeks to provide experimental evidence discerning which Marcus region GSR occurs in for these complexes, as well as establishing a method for doing so for other systems. In Chapter 3, it was established that, similarly to knr, the relationship between driving force and the energy barrier (ΔG‡) also changes depending on which Marcus region the excited state relaxation occurs in. This means that the identity of the Marcus region can be determined by examining how knr and ΔG‡ change with ΔGo. Furthermore, the primary identifying feature of the Marcus regions is the relationship between ΔGo and reorganization energy (λ). If ΔGo is less than λ, then the process occurs in the normal region, and if ΔGo is greater than λ, the process occurs in the inverted region. However, since ΔGo and λ cannot be decoupled through experimentation alone, DFT predictions will need to be relied on to determine the relationship between these variables. Therefore, the goal is to design a series of complexes that will modulate the ligand field strength (since ΔGo between LF states is dependent on LF strength) and study the series through the variable temperature transient absorption (VT-TA) analysis outlined in previous chapters.8 A concern when determining the series to be studied is controlling the reaction coordinate of the 5T2 → 1A1 relaxation. As shown in Figure 3.3 changing the difference in reaction coordinate between states (ΔQ) influences λ, knr, and ΔG‡. Therefore, in order to draw meaningful conclusions from the series, it must be designed to allow changes to ΔG‡ while restricting changes to ΔQ. In 62 Figure 4.2: The Fe(II) bis-terpyridine complexes studied throughout this chapter. this effort, a series of 4’-position substituted [Fe(terpy)2]2+ molecules were selected. At the 4’- position of the terpyridine ligand, as shown in Figure 4.2, the substituent is positioned away from the metal center. Here, the substituent should introduce minimal steric interference, keeping changes to ΔQ small. Furthermore, at the 4’-position, substituents are able to strongly interact with the π-system of the pyridine rings, impacting the π-donating strength of the ligand.9,10 This means that by using substituents of varying electron donating (ED) or withdrawing (EW) strengths, ΔGo for the system can be modulated.11 This creates a series that when analyzed through VT-TA will Figure 4.3: Model of how the substituents impact the electronic structure of the complex. An EW substituent will pull π-density away from the binding N of the ligand, causing the ligand to act as a π-acceptor and increasing the LF strength. An ED group will push π-density onto the binding N of the ligand, causing the ligand to act as a π-donor and increasing the LF strength. In the Marcus normal region, increasing the LF will reduce ΔG‡ and vice versa. In the inverted region, the opposite would occur. For a more in-depth examination of how substituents at this position impact LF strength, refer to Chapter 1.3. 63 determine which Marcus region the GSR process for Fe(II)-based polypyridyl chromophores occurs in. The results of this study showcase experimental evidence supporting that this process does occur in the Marcus normal region, providing important context for developing a model to best understand how to manipulate the LF excited states for use in photosensitization. 4.2 Experimental Methods 4.2.1 Synthesis of [Fe(4’-R-terpy)2](PF6)2 Series The proto-substituted complex was prepared and characterized by fellow McCusker group member Yi-Jyun Lien and the remainder of the series was sent by collaborators from Dr. Gyorgi Vanko’s group. i. bis(2,2’:6’,2”-terpyridine) iron(II) hexafluorophosphate, [Fe(terpy)2](PF6)2. [Fe(terpy)2](PF6)2 was synthesized by Lien, Y.-J. using previously reported procedures.12,13 Characterization was done via 1H NMR using a Bruker 500 MHz spectrometer and electrospray-ionization mass spectrometry using a Waters Xevo G2- XS Quadrupole time-of-flight spectrometer (ESI-TOF). Characterization information is available in the supplemental information. ii. bis(4’-R-2,2’:6’,2”-terpyridine) iron(II) hexafluorophosphate, where R = cyano (CN), methylsulfonyl (SO2Me), 2-furyl (furan), chloro (Cl), methylthio (SMe), methoxy (OMe), hydroxy (OH), and ylimino-bisethanol (N(EtOH)2). All 4’-substituted derivatives of [Fe(terpy)2](PF6)2 were synthesized by collaborators from the Vanko group at EvoBlocks Ltd., Budapest following literature methods.11,13- 17. All samples were characterized via UV-Vis spectroscopy, NMR, and elemental analysis as described previously by the group.11 4.2.2 Synthesis of [Co(4’-R-terpy)2](PF6)3 Derivatives The Co(III)-based complexes and their ligands were prepared and characterized by fellow McCusker group members Yi-Jyun Lien and Bekah Bowers. All characterization was done via 1H and 13C NMR using a Bruker 500 MHz spectrometer and ESI-TOF using a Waters Xevo G2-XS spectrometer. All NMR and ESI-TOF data are available in the SI. i. 4'-(methylthio)-2,2':6',2''-terpyridine, 4’-SMe-terpy. 4’-SMe-terpy synthesis was done by Lien, Y.-J. following modified literature procedure.13 Potassium tert-butoxide (KOtBu) (5.56 g, 49.5 mmol) was suspended in anhydrous tetrahydrofuran (THF) (40 mL) and cooled to 0°C. 2-Acetylpyridine (2.77 64 mL, 24.8 mmol) was slowly added to the flask followed by carbon disulfide (1.50 mL, 24.8 mmol) and methyl iodide (3.10 mL, 49.5 mmol). The mixture was stirred at room temperature for 5.5 hours and formed the intermediate 3,3-bis(methylthio)-1-(pyridin- 2-yl)prop-2-en-1-one. Additional KOtBu (5.56 g, 49.5 mmol) and 2-acetylpyridine (2.77 mL, 24.8 mmol) in anhydrous THF (40 mL) was then carefully added to the flask by cannula transfer. After the solution was stirred for 14 hours at room temperature, ammonium acetate (19.29 g, 3.3 mol) and glacial acetic acid (30.37 mL) were added. THF was slowly distilled off over a 4-hour period right after ammonium acetate and glacial acetic acid were added. The mixture was then washed with iced water (2 L) and rinsed with small amount of DCM (5 mL) and large amount of hexane (1 L). The colored impurities were removed, affording gray powder (4.02 g, 58% yield). ii. 2,2'-([2,2':6',2''-terpyridin]-4'-ylimino)bis(ethanol), 4’-N(EtOH)2-terpy. 4’-N(EtOH)2-terpy synthesis was done by Lien, Y.-J. following modified literature procedure.18 A 250 mL single-neck flask was charged with manganese(II) chloride tetrahydrate (0.15 g, 1.19 mmol), diethanolamine (3.93 g, 37.33 mmol), 4'-chloro- 2,2':6',2''-terpyridine (0.20 g, 0.74 mmol) and MeOH (10 mL). The mixture was then refluxed for 14 hrs. Upon cooling to room temperature, the solvent was removed under reduced pressure and residue was stirred with 50 mL sodium hydroxide 1:1 MeCN/water solution (pH >12) for continuous 16 hrs under exposure to air. MeCN was further removed under reduced pressure and the remaining aqueous portion was extracted with DCM (3 × 100 mL). The combined organic layers were washed with brine, dried with Na2SO4, and the solvent was removed. Purification was performed by column chromatography on neutral aluminum oxide with 2% MeOH in DCM solution as eluent, affording white powder (0.15 g, 58% yield). iii. 4'-(2-furyl)-2,2':6',2''-terpyridine, 4’-furan-terpy. 4’-furan-terpy synthesis was done by Lien, Y.-J. following modified literature procedure.19 A 250 mL single-neck flask was charged with 2-furaldehyde (0.69 g, 7.73 mmol), 2-acetylpyridine (1.73 g, 14.20 mmol), KOH (1.20 g, 21.40 mmol), aqueous ammonia solution (20 mL, 25%) and EtOH (30 mL). The mixture solution was then brought to 60°C and stirred overnight. Upon cooling to room temperature, the mixture was added with copious quantities of iced water and white precipitate gradually formed 65 in the mixture. After the mixture was stirred with iced water for 30 minutes, the precipitate was collected by filtration and rinsed with ether, affording white powder (1.69 g, 79% yield). iv. 4’-methoxy-2,2':6',2''-terpyridine, 4’-OMe-terpy. 4’-OMe-terpy was synthesized by Bowers, B. following literature procedure.20 v. 4’-chloro-2,2':6',2''-terpyridine, 4’-Cl-terpy. 4’-Cl-terpy was purchased from Alfa Aesar. vi. bis(4’-R-2,2’:6’,2”-terpyridine) cobalt(III) hexafluorophosphate, where R = SMe, N(EtOH)2, furan, OMe, or Cl. All [Co(4’-R-terpy)2](PF6)3 complex synthesis was done by Bowers, B. following modified literature procedure.20 2.1 equivalence of 4’-R-terpy were added to a dry round bottom flask and dissolved in chloroform, and 1.0 equivalence of Co(OAc)2 • 4H2O was taken into methanol and added to the CHCl3 solution of 4’-R-terpy. Upon addition of Co(II) starting material, the reaction turned either dark red or orange and was stirred for approximately 3 hours at room temperature. [Co(4’-R-terpy)2](PF6)2 complexes were precipitated upon addition of ammonium hexafluorophosphate and isolated via vacuum filtration and washed with MeOH, H2O, and Et2O. The crude [Co(4’-R-terpy)2](PF6)2 complexes were then suspended in H2O (20-30 mL) and a saturated solution of bromine (2 drops neat Br2 in 2 mL of H2O) was added dropwise to the suspension. The reaction was shielded from light and allowed to stir for 18 hours. Excess ammonium hexafluorophosphate was added to the suspension and allowed to stir for an additional 30 minutes before the product was isolated via vacuum filtration and washed with large amounts of H2O and Et2O. The solid was then taken into MeCN and recrystallized twice via ether diffusion. [Co(4’-SMe-terpy)2](PF6)3 was further purified via column chromatography on neutral alumina and eluted with MeCN as the first spot. Yields are as follows: [Co(4’-SMe-terpy)2](PF6)3 – 0.537 g, 75%; [Co(4’- N(EtOH)2-terpy)2](PF6)3 – 0.246 g, 70%; [Co(4’-furan-terpy)2](PF6)3 – 0.151 g, 60%; [Co(4’-OMe-terpy)2](PF6)3 – 0.116 g, 79%; and [Co(4’-Cl-terpy)2](PF6)3 – 0.523 g, 90%. 66 4.2.3 Spectroscopic Methods The [Fe(4’-R-terpy)](PF6)2 complexes were examined using UV-Vis spectroscopy, density functional theory calculations (DFT), and VT-TA using the procedures outlined in Chapter 2, and the [Co(4’-R-terpy)2](PF6)3 complexes were examined with UV-Vis spectroscopy following the procedure outlined in Chapter 2.3. Regarding VT-TA data collection, some complexes were studied with an ultrafast laser system, while the rest were studied with the nanosecond laser system. Furthermore, the cooling method used on the nanosecond system was changed from a liquid nitrogen dewar to a water-based cooling system. Given the ~5 ns instrument response function (IRF) of the system and the similarities in sample chamber design between the two cooling systems, the change between them is not expected to significantly influence the data. The laser and cooling systems used for each complex, alongside the pump/probe wavelengths are provided in Table 4.1. Table 4.1: The laser system, temperature control method, and pump/probe combination used for each complex. The proto complex was studied on both laser systems, as the low temperature lifetimes were too long to accurately be measured on the 13 ns delay stage the ultrafast system used. Multiple trials were performed for each complex and for some, different pump/probe combinations were used to improve s/n. Substituent Laser System Temperature Pump Wavelength Probe Wavelength Control Method (nm) Janis Dewar 590/580 Janis Dewar Ultrafast Ultrafast (nm) 550 560/550 530 570 560 580 530 530 530 590 560 540 530 550 570 570 570 SO2Me CN H SMe OMe Ultrafast and Nanosecond Janis Dewar Nanosecond Janis Dewar Nanosecond Janis Dewar N(EtOH)2 Nanosecond Janis Dewar Cl OH Furan Nanosecond Nanosecond Nanosecond QNW TC-1 QNW TC-1 QNW TC-1 67 4.3 Spectroscopic Results 4.3.1 [Fe(terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(terpy)2](PF6)2 is provided in Figure 4.4, alongside the pump and probe wavelengths used throughout TA measurements. To avoid influence from pump scattering, a 30 nm distance between pump and probe wavelengths was required, to get the best s/n ratio with this limitation, pump/probe wavelengths on opposite ends of the central peak were chosen, maximizing the absorptivity at each wavelength. Figure 4.4: Ground state absorption spectrum for [Fe(terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Hallmarks of this feature for Fe(II)- based terpyridine complexes are the sharp peak, with a broad blue shoulder and a weaker red shoulder. These features are observed across the series. Initial VT-TA experiments on this complex were run using the ultrafast (UF) system, however with a τ of 4.9 ± 0.2 ns at room temperature, and 10.1 ± 0.8 ns at 240 K, the 13 ns delay stage used on the UF system was insufficient to provide accurate lifetimes for GSR. Therefore, data was collected on the nanosecond (NS) system as well. A comparison of VT-TA data is provided in Figure 4.5. The two data sets were used in conjunction, the IRF of the NS system was very close to the higher temperature τ values, so data from the UF system was used to check the accuracy of deconvolution. Two trials were run on the UF system, and five more were done on the NS system. A Grubb’s test was run for each data set at each temperature to check for outliers caused 68 by experimental error, any that were not deemed as significant outliers were averaged together. This data is summarized in Table 4.2. Figure 4.5: VT-TA data for [Fe(terpy)2]2+ collected on the UF system (A, trial 1) and the NS system (B, trial 6). Table 4.2: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. † These experiments were conducted on the UF laser system, all others were done on the NS system. 69 τ (ns) Trial 1† Trial 2† Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Average 295 K 4.91 ± 0.22 4.87 ± 0.32 4.83 ± 0.04 4.86 ± 0.04 4.76 ± 0.04 4.98 ± 0.04 5.03 ± 0.05 4.9 ± 0.2 290 K 4.94 ± 0.23 5.51 ± 0.42 5.22 ± 0.04 5.60 ± 0.03 5.15 ± 0.03 5.23 ± 0.05 4.97 ± 0.04 5.2 ± 0.3 285 K 5.21 ± 0.25 7.63 ± 0.98 5.46 ± 0.05 5.77 ± 0.02 5.45 ± 0.04 5.59 ± 0.04 5.16 ± 0.05 5.4 ± 0.2 280 K 5.67 ± 0.34 7.44 ± 1.09 5.83 ± 0.04 6.06 ± 0.02 5.98 ± 0.04 5.56 ± 0.04 5.60 ± 0.05 5.8 ± 0.2 275 K 7.17 ± 0.53 10.73 ± 1.64 6.00 ± 0.04 6.20 ± 0.02 6.09 ± 0.04 6.13 ± 0.04 5.95 ± 0.05 6.1 ± 0.1 270 K 7.35 ± 0.64 16.08 ± 3.93 6.49 ± 0.04 6.83 ± 0.02 6.57 ± 0.04 6.48 ± 0.04 6.51 ± 0.05 6.6 ± 0.1 265 K 7.35 ± 0.51 13.60 ± 3.39 7.07 ± 0.03 7.08 ± 0.02 7.06 ± 0.04 6.88 ± 0.03 6.84 ± 0.05 7.0 ± 0.1 260 K 8.60 ± 0.80 12.54 ± 2.00 7.47 ± 0.03 7.64 ± 0.02 7.47 ± 0.04 7.61 ± 0.04 7.53 ± 0.04 7.6 ± 0.1 255 K 9.95 ± 1.14 8.58 ± 0.87 7.98 ± 0.04 8.26 ± 0.02 8.09 ± 0.04 8.07 ± 0.05 7.75 ± 0.05 8.1 ± 0.3 250 K 9.93 ± 1.04 12.40 ± 1.82 8.64 ± 0.04 8.60 ± 0.02 8.60 ± 0.04 8.79 ± 0.05 8.50 ± 0.04 8.6 ± 0.1 245 K 11.02 ± 1.51 14.35 ± 3.02 9.23 ± 0.04 9.39 ± 0.02 9.34 ± 0.04 9.33 ± 0.03 9.41 ± 0.05 9.3 ± 0.1 240 K 12.03 ± 1.91 12.23 ± 3.59 9.91 ± 0.03 9.91 ± 0.02 10.32 ± 0.05 9.99 ± 0.05 9.76 ± 0.05 10.1 ± 0.8 The averaged data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots were shown in Figure 3.6. The information extracted from these plots is provided in Table 4.3. Table 4.3: The transition state and Marcus theory parameters for [Fe(terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots in Figure 3.6 following the procedure outlined in Chapter 3. 4.3.2 [Fe(4’-SMe-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-SMe-terpy)2](PF6)2 is provided in Figure 4.6, alongside the pump and probe wavelengths used throughout TA measurements. Pump and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Figure 4.6: Ground state absorption spectrum for [Fe(4’-SMe-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Four VT-TA trials were run on this complex, the first two trials were done in 10 K steps. This is the case for several complexes in the series, it was determined after a few data sets were collected that it was worthwhile to increase the number of temperatures measured by going in 5 K steps through the same temperature window (except the lowest temperature is now 240 K). This was done to increase the density of points in the “Arrhenius-type” plots. The VT-TA plot of trial three is shown in Figure 4.7. The kinetic data from each trial is summarized in Table 4.4. 70 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) H 465 ± 10 -5.58 ± 0.02 2110 ± 15 0.108 ± 0.006 2970 ± 25 Figure 4.7: VT-TA data for [Fe(4’-SMe-terpy)2]2+. Shown is the data from the third trial. Table 4.4: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.8 and 71 τ (ns) Trial 1 Trial 2 Trial 3 Trial 4 Average 295 K 12.47 ± 0.08 12.93 ± 0.08 13.02 ± 0.10 11.51 ± 0.12 12.5 ± 0.6 290 K ― ― 13.03 ± 0.11 13.63 ± 0.18 13.3 ± 0.3 285 K 14.14 ± 0.06 14.44 ± 0.07 14.16 ± 0.10 13.78 ± 0.16 14.1 ± 0.2 280 K ― ― 15.67 ± 0.12 15.55 ± 0.17 15.6 ± 0.1 275 K 15.89 ± 0.06 15.93 ± 0.07 16.30 ± 0.12 15.29 ± 0.16 15.9 ± 0.4 270 K ― ― 17.78 ± 0.12 17.98 ± 0.16 17.9 ± 0.2 265 K 19.15 ± 0.06 18.80 ± 0.07 18.45 ± 0.13 19.75 ± 0.14 19.0 ± 0.5 260 K ― ― 20.87 ± 0.11 19.57 ± 0.22 20.2 ± 0.7 255 K 22.17 ± 0.06 22.16 ± 0.07 23.24 ± 0.13 21.88 ± 0.20 22.4 ± 0.5 250 K ― ― 24.37 ± 0.11 24.31 ± 0.19 24.3 ± 0.1 245 K 26.30 ± 0.06 25.56 ± 0.07 26.73 ± 0.10 25.77 ± 0.18 26.1 ± 0.5 240 K ― ― 26.89 ± 0.12 26.68 ± 0.21 26.8 ± 0.2 analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.5. Figure 4.8: “Arrhenius-type” plots for [Fe(4’-SMe-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). Table 4.5: The transition state and Marcus theory parameters for [Fe(4’-SMe-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.8. 4.3.3 [Fe(4’-SO2Me-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-SO2Me-terpy)2](PF6)2 is provided in Figure 4.9, alongside the pump and probe wavelengths used throughout TA measurements. For the initial trials, a pump wavelength of 580 nm was used, that was changed to 590 nm for the last trial Figure 4.9: Ground state absorption spectrum for [Fe(4’-SO2Me-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. 72 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) SMe 530 ± 20 -6.02 ± 0.07 2310 ± 40 0.032 ± 0.007 3240 ± 85 to reduce scattering caused by the cryostat from reaching the detector through the wavelength separator. Collecting VT-TA data with good s/n on this and the 4’-CN substituted complex proved very challenging. The central peak of the MLCT absorption feature is sharper for these two complexes than the rest of the series, which meant that in order to keep pump/probe combinations Figure 4.10: VT-TA data for [Fe(4’-SO2Me-terpy)2]2+. Shown is the data from the first trial. Table 4.6: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. 73 τ (ps) Trial 1 Trial 2 Trial 3 Average 295 K 1780 ± 30 1810 ± 20 1700 ± 20 1760 ± 50 290 K 1820 ± 30 ― 1920 ± 30 1870 ± 60 285 K 1840 ± 30 ― 1980 ± 40 1910 ± 80 280 K 1890 ± 30 ― 2000 ± 40 1940 ± 60 275 K 1990 ± 30 ― 2100 ± 50 2050 ± 60 270 K 2140 ± 30 ― 2250 ± 50 2190 ± 70 265 K 2130 ± 40 ― 2220 ± 40 2180 ± 60 260 K 2250 ± 40 1810 ± 20 2270 ± 40 2260 ± 40 255 K 2460 ± 50 2370 ± 30 2380 ± 50 2400 ± 50 250 K 2540 ± 60 2560 ± 40 2590 ± 40 2570 ± 40 245 K 2600 ± 80 2770 ± 50 2700 ± 40 2740 ± 80 240 K 2690 ± 80 2900 ± 30 2760 ± 50 2830 ± 100 with a 30 nm difference, the pump and probe pulses would need to be located much further from the maximum of the peak than what occurs for the other complexes. When the first several trials were collected, the ultrafast laser system was operating with low power. The led to a very weak signal and very high error bars on the measured knr. After repairs were done and system power was improved, some trials with reasonable s/n were collected. Furthermore, it was of vital importance to use spectrometry grade solvent taken from an air free atmosphere for these complexes. The slightest water contamination in the solvent would lead to high scattering and weak signal at low temperatures, as was discussed in Chapter 2. Due to time constraints, the second trial only included a portion of the temperature points. The remaining temperatures were attempted to be collected the following day, however due to overnight laser drift, the system could not be properly aligned, and the attempt was abandoned. The VT-TA plot of trial 1 is shown in Figure 4.10 and the kinetic data from each trial is summarized in Table 4.6. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.11 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.7. Figure 4.11: “Arrhenius-type” plots for [Fe(4’-SO2Me-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). Table 4.7: The transition state and Marcus theory parameters for [Fe(4’-SO2Me-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.11. 74 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) SO2Me 240 ± 20 -5.64 ± 0.06 1905 ± 35 0.092 ± 0.016 2065 ± 65 4.3.4 [Fe(4’-OMe-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-SO2Me-terpy)2](PF6)2 is provided in Figure 4.12, alongside the pump and probe wavelengths used throughout TA measurements. Pump Figure 4.12: Ground state absorption spectrum for [Fe(4’-OMe-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Three trials were conducted on this complex, similarly to the 4’-SMe substituted complex, the first trial was collected in 10 K temperature steps before switching to 5 K steps for the latter two trials. The second trial is shown in Figure 4.13, and kinetic data is summarized in Table 4.8. Figure 4.13: VT-TA data for [Fe(4’-OMe-terpy)2]2+. Shown is the data from the second trial. 75 Table 4.8: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.14 and analyzed using transition state and Marcus theory. Observed in the plots in Figure 4.14 at the lowest temperature is a “flattening” of the slope, resulting in a “hockey stick” shape to the plot. This is observed for other complexes in the series but is very pronounced here. This feature could be emblematic of approaching the “classical limit” discussed in Chapter 3.1, but it is worth noting that 240 K is near the freezing point of acetonitrile (229 K). At this temperature, the near-frozen Figure 4.14: “Arrhenius-type” plots for [Fe(4’-OMe-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). 76 τ (ns) Trial 1 Trial 2 Trial 3 Average 295 K 21.70 ± 0.10 22.10 ± 0.06 21.74 ± 0.10 21.8 ± 0.2 290 K ― 23.87 ± 0.06 23.30 ± 0.11 23.6 ± 0.3 285 K 25.19 ± 0.13 25.77 ± 0.05 24.93 ± 0.09 25.3 ± 0.4 280 K ― 27.34 ± 0.07 27.22 ± 0.09 27.3 ± 0.1 275 K 30.42 ± 0.11 29.38 ± 0.08 29.67 ± 0.10 29.8 ± 0.4 270 K ― 33.24 ± 0.07 32.74 ± 0.13 33.0 ± 0.3 265 K 35.93 ± 0.12 35.53 ± 0.06 35.19 ± 0.12 35.6 ± 0.3 260 K ― 39.58 ± 0.08 39.21 ± 0.13 39.4 ± 0.2 255 K 43.23 ± 0.13 42.79 ± 0.08 42.50 ± 0.15 42.8 ± 0.3 250 K ― 47.13 ± 0.08 48.03 ± 0.15 47.6 ± 0.5 245 K 52.52 ± 0.18 50.35 ± 0.09 50.99 ± 0.16 51.3 ± 0.9 240 K ― 54.17 ± 0.08 52.73 ± 0.19 53.5 ± 0.7 solvent will influence the excited state dynamics, this phenomenon is most likely the culprit behind this feature. The information extracted from these plots is provided in Table 4.9. Table 4.9: The transition state and Marcus theory parameters for [Fe(4’-OMe-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.14. 4.3.5 [Fe(4’-N(EtOH)2-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-N(EtOH)2-terpy)2](PF6)2 is provided in Figure 4.15, alongside the pump and probe wavelengths used throughout TA measurements. Pump and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Figure 4.15: Ground state absorption spectrum for [Fe(4’-N(EtOH)2-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Four VT-TA trials were run on this complex. As with the 4’-SMe substituted complex, the first two trials were done in 10 K steps and the last two trials were done in 5 K steps through the same temperature window (except the lowest temperature is now 240 K). The VT-TA spectra from the fourth trial is shown in Figure 4.16. The kinetic data from each trial is provided in Table 4.10. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.17 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.11. 77 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) OMe 655 ± 20 -5.99 ± 0.07 2425 ± 40 0.034 ± 0.008 3730 ± 85 Figure 4.16: VT-TA data for [Fe(4’-N(EtOH)2-terpy)2]2+. Shown is the data from the fourth trial. Table 4.10: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. 78 τ (ns) Trial 1 Trial 2 Trial 3 Trial 4 Average 295 K 56.27 ± 0.09 53.91 ± 0.07 55.01 ± 0.10 55.96 ± 0.10 55.3 ± 0.9 290 K ― ― 59.70 ± 0.11 60.14 ± 0.10 59.9 ± 0.2 285 K 65.76 ± 0.09 64.64 ± 0.16 65.05 ± 0.19 66.03 ± 0.10 65.4 ± 0.6 280 K ― ― 71.11 ± 0.52 71.33 ± 0.11 71.2 ± 0.3 275 K 77.87 ± 0.09 79.32 ± 0.14 78.57 ± 0.18 77.34 ± 0.11 78.3 ± 0.7 270 K ― ― 85.42 ± 0.19 85.86 ± 0.14 85.6 ± 0.3 265 K 92.56 ± 0.10 95.28 ± 0.13 92.38 ± 0.18 94.73 ± 0.15 93.7 ± 1.3 260 K ― ― 103.09 ± 0.20 102.58 ± 0.14 102.8 ± 0.3 255 K 114.71 ± 0.10 115.46 ± 0.11 115.03 ± 0.17 114.37 ± 0.15 114.9 ± 0.4 250 K ― ― 126.50 ± 0.17 125.72 ± 0.15 126.1 ± 0.4 245 K 143.76 ± 0.10 139.86 ± 0.12 137.38 ± 0.15 136.14 ± 0.16 139.3 ± 2.9 240 K ― ― 146.45 ± 0.18 144.76 ± 0.26 145.6 ± 0.9 Figure 4.17: “Arrhenius-type” plots for [Fe(4’-N(EtOH)2-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). Table 4.11: The transition state and Marcus theory parameters for [Fe(4’-N(EtOH)2-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.17. 4.3.6 [Fe(4’-CN-terpy)](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-N(EtOH)2-terpy)2](PF6)2 is provided in Figure 4.15, alongside the pump and probe wavelengths used throughout TA measurements. Similarly to the 4’-SO2Me substituted complex, the last two trials collected on this complex were done with different pump/probe combinations from the first two, once again to Figure 4.18: Ground state absorption spectrum for [Fe(4’-CN-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. 79 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) N(EtOH)2 715 ± 20 -6.44 ± 0.07 2615 ± 40 0.009 ± 0.002 3970 ± 80 reduce scattering caused by the cryostat from reaching the detector. This scattering was preventing the TA spectra from relaxing back to the zero line, and the new pump/probe combination proved helpful in this regard. Unlike the SO2Me complex, the pump pulse was held constant while the probe pulse was shifted. Figure 4.19: VT-TA data for [Fe(4’-CN-terpy)2]2+. Shown is the data from the fourth trial. Table 4.12: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. 80 τ (ps) Trial 1 Trial 2 Trial 3 Trial 4 Average 295 K 1950 ± 20 2030 ± 100 2020 ± 30 2070 ± 30 2020 ± 60 290 K 2070 ± 10 2050 ± 50 2150 ± 40 2110 ± 30 2090 ± 50 285 K 2290 ± 20 2300 ± 60 2290 ± 50 2250 ± 50 2280 ± 40 280 K 2190 ± 20 2270 ± 50 2390 ± 60 2240 ± 50 2270 ± 80 275 K 2230 ± 20 2350 ± 50 2400 ± 50 2270 ± 50 2310 ± 80 270 K 2430 ± 20 2490 ± 60 2560 ± 60 2550 ± 50 2510 ± 70 265 K 2570 ± 20 2290 ± 20 2700 ± 70 2650 ± 60 2550 ± 170 260 K 2710 ± 20 2520 ± 20 2780 ± 60 2860 ± 60 2710 ± 130 255 K 2780 ± 30 2680 ± 20 3020 ± 70 3040 ± 60 2880 ± 160 250 K 2950 ± 30 2830 ± 30 3560 ± 110 3280 ± 70 3160 ± 290 245 K ― 3070 ± 30 3300 ± 80 3750 ± 90 3370 ± 290 240 K ― 3290 ± 30 3360 ± 120 4200 ± 120 3620 ± 420 The VT-TA data for this complex, like the 4’-SO2Me substituted complex, was collected on the ultrafast laser system, and faced similar issues. Reaching a suitable s/n ratio proved challenging, though this was mitigated once repairs to the system improved laser output power. A total of four trials were used, however the first trial is missing the lowest temperature points. This is because during data collection at 245 K, the sample had begun to freeze, which could cause damage to the cuvette. At first notice of this, data collection was halted, and the sample chamber was slowly brought to room temperature. The VT-TA data for trial four is provided in Figure 4.19 and the kinetic data across all trials is summarized in Table 4.12. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.20 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.13. Figure 4.20: “Arrhenius-type” plots for [Fe(4’-CN-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). Table 4.13: The transition state and Marcus theory parameters for [Fe(4’-CN-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.20. 81 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) CN 330 ± 25 -5.42 ± 0.08 1930 ± 50 0.179 ± 0.040 2430 ± 85 4.3.7 [Fe(4’-Cl-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-Cl-terpy)2](PF6)2 is provided in Figure 4.21, alongside the pump and probe wavelengths used throughout TA measurements. Pump and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Figure 4.21: Ground state absorption spectrum for [Fe(4’-Cl-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. This and the following compounds were sent by collaborators after data collection had finished on the previous complexes. By this point, the dewar on the NS laser system had been replaced with the Quantum Northwest water-based cooling system. The same temperature range was achievable with this system and the sample chamber was similar in design to the Janis dewar Figure 4.22: VT-TA data for [Fe(4’-Cl-terpy)2]2+. Shown is the data from the first trial. 82 Table 4.14: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. used previously. Therefore, the change in systems is not expected to cause significant differences in the VT-TA data collected. For this complex, four trials were conducted. The VT-TA data for the first trial is shown in Figure 4.22 and the kinetic data from each trial is summarized in Table 4.14. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.23 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.15. Figure 4.23: “Arrhenius-type” plots for [Fe(4’-Cl-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). 83 τ (ns) Trial 1 Trial 2 Trial 3 Trial 4 Average 295 K 6.11 ± 0.05 5.95 ± 0.04 5.70 ± 0.29 5.99 ± 0.04 6.0 ± 0.1 290 K 6.38 ± 0.04 6.48 ± 0.04 6.44 ± 0.32 6.41 ± 0.04 6.4 ± 0.1 285 K 6.81 ± 0.05 6.97 ± 0.04 6.45 ± 0.32 6.73 ± 0.04 6.8 ± 0.1 280 K 7.29 ± 0.05 7.56 ± 0.04 7.21 ± 0.36 7.29 ± 0.04 7.3 ± 0.2 275 K 7.78 ± 0.05 7.99 ± 0.05 7.22 ± 0.36 7.53 ± 0.05 7.6 ± 0.3 270 K 8.38 ± 0.05 8.19 ± 0.05 8.01 ± 0.40 8.23 ± 0.04 8.3 ± 0.1 265 K 8.86 ± 0.04 9.03 ± 0.05 8.10 ± 0.40 8.71 ± 0.04 8.9 ± 0.1 260 K 9.58 ± 0.04 9.62 ± 0.04 8.83 ± 0.44 9.51 ± 0.04 9.6 ± 0.1 255 K 10.30 ± 0.04 10.28 ± 0.04 9.67 ± 0.48 10.41 ± 0.05 10.3 ± 0.1 250 K 11.07 ± 0.03 11.10 ± 0.04 10.52 ± 0.53 11.12 ± 0.04 11.1 ± 0.1 245 K 12.24 ± 0.04 11.99 ± 0.04 11.37 ± 0.57 11.86 ± 0.04 12.0 ± 0.2 240 K 12.80 ± 0.05 12.64 ± 0.04 12.28 ± 0.05 12.88 ± 0.05 12.7 ± 0.2 Table 4.15: The transition state and Marcus theory parameters for [Fe(4’-Cl-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.23. 4.3.8 [Fe(4’-OH-terpy)2](PF6)2 Figure 4.24: Ground state absorption spectrum for [Fe(4’-OH-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. The UV-Vis spectrum of the MLCT feature for [Fe(4’-OH-terpy)2](PF6)2 is provided in Figure 4.24, alongside the pump and probe wavelengths used throughout TA measurements. Pump and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Figure 4.25: VT-TA data for [Fe(4’-OH-terpy)2]2+. Shown is the data from the first trial. 84 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) Cl 495 ± 10 -5.64 ± 0.03 2160 ± 20 0.091 ± 0.010 3080 ± 40 Three trials of VT-TA data were collected on this complex, the plot from the first trial is shown in Figure 4.25 and the kinetic data across all trials is given in Table 4.16. Table 4.16: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.26 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.17. Figure 4.26: “Arrhenius-type” plots for [Fe(4’-OH-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). 85 τ (ns) Trial 1 Trial 2 Trial 3 Average 295 K 24.65 ± 0.08 24.51 ± 0.10 25.33 ± 0.09 24.8 ± 0.4 290 K 27.01 ± 0.09 27.60 ± 0.10 27.26 ± 0.10 27.3 ± 0.3 285 K 28.83 ± 0.07 29.31 ± 0.09 29.58 ± 0.10 29.2 ± 0.3 280 K 30.81 ± 0.10 31.45 ± 0.10 32.62 ± 0.11 31.6 ± 0.8 275 K 33.66 ± 0.10 34.85 ± 0.11 35.29 ± 0.12 34.6 ± 0.7 270 K 37.02 ± 0.11 38.32 ± 0.11 38.72 ± 0.11 38.0 ± 0.7 265 K 41.30 ± 0.11 43.00 ± 0.12 42.13 ± 0.12 42.1 ± 0.7 260 K 45.05 ± 0.10 46.75 ± 0.11 47.04 ± 0.13 46.3 ± 0.9 255 K 50.44 ± 0.12 51.51 ± 0.13 52.39 ± 0.12 51.4 ± 0.8 250 K 55.29 ± 0.12 57.41 ± 0.14 57.63 ± 0.14 56.8 ± 1.1 245 K 63.28 ± 0.14 63.88 ± 0.16 63.31 ± 0.14 63.5 ± 0.3 240 K 68.67 ± 0.13 71.54 ± 0.16 71.14 ± 0.14 70.5 ± 1.3 Table 4.17: The transition state and Marcus theory parameters for [Fe(4’-Cl-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.23. 4.3.9 [Fe(4’-furan-terpy)2](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4’-OH-terpy)2](PF6)2 is provided in Figure 4.27, alongside the pump and probe wavelengths used throughout TA measurements. Pump and probe wavelengths were chosen for similar reasons to the [Fe(terpy)2](PF6)2 TA measurements. Figure 4.27: Ground state absorption spectrum for [Fe(4’-furan-terpy)2]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Figure 4.28: VT-TA data for [Fe(4’-furan-terpy)2]2+. Shown is the data from the third trial. 86 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) OH 750 ± 10 -5.76 ± 0.02 2450 ± 10 0.065 ± 0.004 4105 ± 25 Four trials of VT-TA data were collected on this complex, the plot from the third trial is shown in Figure 4.28 and the kinetic data across all trials is given in Table 4.18. Table 4.18: The GSR τ in ns across each temperature. The averaged data is boldened for ease of reading. The averaged data was used to construct “Arrhenius-type” plots shown in Figure 4.29 and analyzed using transition state and Marcus theory. The information extracted from these plots is provided in Table 4.19. Figure 4.29: “Arrhenius-type” plots for [Fe(4’-furan-terpy)2]2+ analyzed through transition state theory (A) and Marcus theory (B). 87 τ (ns) Trial 1 Trial 2 Trial 3 Trial 4 Average 295 K 6.09 ± 0.05 5.77 ± 0.05 5.68 ± 0.04 5.74 ± 0.05 5.8 ± 0.2 290 K 6.36 ± 0.04 7.19 ± 0.13 6.52 ± 0.05 6.36 ± 0.05 6.6 ± 0.3 285 K 6.91 ± 0.04 6.89 ± 0.06 6.64 ± 0.04 6.63 ± 0.04 6.8 ± 0.1 280 K 7.44 ± 0.05 7.39 ± 0.05 6.95 ± 0.05 6.83 ± 0.04 7.2 ± 0.3 275 K 7.79 ± 0.04 7.72 ± 0.04 7.70 ± 0.05 7.58 ± 0.05 7.7 ± 0.1 270 K 8.19 ± 0.05 8.47 ± 0.05 8.08 ± 0.05 7.95 ± 0.05 8.2 ± 0.2 265 K 8.94 ± 0.05 8.97 ± 0.05 8.92 ± 0.06 8.49 ± 0.04 8.8 ± 0.2 260 K 9.58 ± 0.04 9.61 ± 0.04 9.30 ± 0.06 9.39 ± 0.05 9.5 ± 0.1 255 K 10.49 ± 0.06 10.33 ± 0.05 10.18 ± 0.05 9.90 ± 0.04 10.2 ± 0.2 250 K 10.85 ± 0.04 11.09 ± 0.05 10.98 ± 0.05 10.78 ± 0.05 10.9 ± 0.1 245 K 12.19 ± 0.04 12.08 ± 0.04 11.83 ± 0.05 11.92 ± 0.05 12.0 ± 0.1 240 K 13.04 ± 0.04 12.96 ± 0.06 12.94 ± 0.05 12.97 ± 0.04 13.0 ± 0.1 Table 4.19: The transition state and Marcus theory parameters for [Fe(4’-Cl-terpy)2]2+. These values were calculated from the slopes and y-intercepts of the plots shown in Figure 4.23. 4.3.10 Reaction Coordinate Calculations As mentioned before, it is important to keep the 5T2 → 1A1 reaction coordinate relatively constant across the series, otherwise meaningful interpretations of the data above will be impossible. It is assumed that by placing substituents at the 4’-positions changes in ΔQ across the series are minimized, but it is important to check if that assumption is reasonable. By using the Nelsen method described in Chapter 3.2, it is possible to predict λ for a given complex. Since λ is directly influenced by the reaction coordinate, calculated λ across the series will aid in determining whether the assumption that reaction coordinate is relatively unchanging is reasonable or not. These calculations were performed on each complex in the series and the results are provided in Table 4.20. The parameters used to best simulate the most accurate results were determined by running time dependent DFT (TD-DFT) calculations to generate simulated ground state absorption spectra, and these were compared to an experimentally collected spectrum.21 The parameters that led to the best fit were selected for further calculations. More information regarding this procedure Table 4.20: DFT calculated λ across the series. Calculations were performed using the Nelsen method previously described. 88 Substituent λ (cm-1) H 11600 SMe 11000 SO2Me 10600 OMe 11400 N(EtOH)2 10500 CN 11100 Cl 11800 OH 11200 Furan 11200 Substituent ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) Furan 505 ± 15 -5.60 ± 0.05 2155 ± 30 0.105 ± 0.015 3130 ± 55 is available in the SI. The results of the DFT calculations reveal that λ does not change very much with substitution. The energy across all complexes is within 1000 cm-1 of 11000 cm-1 and the standard deviation of the series is about 3.5% of the energy value for the proto-substituted complex. Therefore, theoretical calculations support the validity of the assumption that ΔQ is not changing much across the series. 4.3.11 Determining Driving Force Order – Co(III) Derivative Ligand Field Analysis The utility of the series being studied is contingent on being able to identify how ΔGo changes with each substituent. More so than determine the values of ΔGo it is important to establish a trend of driving force strength. Doing so will allow relationships between ΔGo and knr or ΔG‡ to be determined from the experimental results. To this end, the absorption energies of the LF excited states can be useful in determining such trends. The free energy difference between the 5T2 excited state and 1A1 ground state are dependent on the energy cost of populating two eg* orbitals. This energy cost is directly related to the energy gap between the eg* orbitals and the t2g orbitals, in other words, the LF strength. Therefore, by determining the energy required to excite a molecule into a LF excited state, such as the 3T1 state, a ΔGo trend can be established. However, as a d-d transition, LF excited state absorptions have incredibly weak signals, this is especially true for those of ΔS > 0. This creates a problem when looking at the LF excited state absorptions of Fe(II)- based polypyridines because the region where the LF absorption features are present is dominated Figure 4.30: Molar Absorptivity in the UV-Vis region for [Fe(terpy)2]2+ alongside a Co(III)-based analog. The CT feature of the Co(III) based analog is heavily blue shifted compared to that of the Fe(II)-based complex. This opens the visible region, allowing the incredibly weak LF excited state signals to be seen with high concentration samples. Both samples shown in this figure were prepared in acetonitrile solution. 89 by the intense and broad charge transfer (CT) absorption feature. However, this is not the case for complexes with the same ligands bound to Co(III). In Co(III)-based complexes, the CT feature (LMCT in this case) is heavily blue-shifted, as displayed in Figure 4.30. This reveals the LF excited state absorption features, meaning that relying on Co(III)-based analogs for the complexes in the series allows this LF analysis to establish a trend in ΔGo with substituent changes for the Fe(II)- based complexes.22 Co(III)-analogs can be used to approximate the LF strength of Fe(II)-based complexes since Co(III) is isoelectronic with Fe(II). Therefore, complexes using either metal have access to the same electronic states, such as the 3T1 excited state shown in Figure 4.31. Once measured, the peak was deconvolved using a multi-peak fitting software, and the energy of the spin-forbidden peak was recorded. This was done for several [Co(4’-R-terpy)2]3+ complexes, and the results are summarized in Table 4.21. Spectra for the other complexes are available in the SI. Table 4.21: The 1A1 → 3T1 excitation energies of the [Co(4’- R-terpy)2]3+ complexes. These energies are used to establish trends in LF strength, and therefore ΔGo. †: Data has been previously reported.22 Figure 4.31: Molar Absorptivity of the 1A1 → 3T1 absorption feature for [Co(4’-SMe-terpy)2](PF6)3. The sample was prepared as a ~12 mM solution in acetonitrile in a 10 cm pathlength cell. Igor Pro’s multipeak fitting tool was used to deconvolve the gaussian of the peak, providing the energy needed for the excitation. Unfortunately, Co(III)-based analogs could be made for only a portion of the series. The 4’-CN and 4’-SO2Me substituted ligands were synthesized (procedure and characterization in the SI) but would not bind to the metal, and a 4’-OH substituted ligand could not be isolated. This limitation means that the rest of the series will need to be supplemented with further characterization of the driving force. 90 Substituent 3T1 Peak Energy (cm-1) Cl 15800 H† 15600 Furan 15500 SMe 14900 OMe 15000 N(EtOH)2 14200 4.3.12 Determining Driving Force Order – Marcus Theory Analysis Marcus theory analysis of the VT-TA data provides two ratios: |𝐻𝑎𝑏|4 𝜆⁄ from the y- intercept and (𝛥𝐺𝑜 + λ)2 λ⁄ from the slope. By using the λ values determined via DFT (Table 4.20), both Hab and ΔGo can be estimated for each complex. The estimated ΔGo values are given in Table 4.22. Table 4.22: Estimated ΔGo values determined via Marcus theory analysis of VT-TA data across the series. The ΔGo trend seen in Table 4.22 is in excellent agreement with the Co(III)-based analog 1A1 → 3T1 absorption energy values given in Table 4.21. This supports the accuracy of the rest of the trend, specifically regarding the complexes with no Co(III)-based counterpart. Therefore, through a combination of LF analysis and Marcus theory, a trend in driving force across the series is established. 4.4 Discussion 4.4.1 Experimental Evidence of Marcus Normal Region Behavior In order to establish which Marcus region the 5T2 → 1A1 relaxation occurs in, there are three relationships that must be observed: the relationship between ΔGo and λ, the relationship between ΔGo and knr, and the relationship between ΔGo and ΔG‡ (the energy barrier). The first two relationships are illustrated in Figure 3.4 (provided again below). If |ΔGo| < λ, the zero-point energy of the 5T2 excited state would sit outside the potential well of the 1A1 ground state, placing the transition in the normal region. However, if |ΔGo| > λ, then the zero-point energy of the 5T2 state would rest within the potential well of the 1A1 state, placing the transition in the inverted region. Tables 4.20 and 4.22 provided values for ΔGo and λ estimated via DFT calculations. Those results Table 4.23: The estimated values for λ and ΔGo across the series. The values for λ were determined via DFT, and ΔGo was calculated from the Marcus theory-derived ratio (𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ using those λ values. 91 Substituent SO2Me CN Cl H Furan SMe OMe OH N(EtOH)2 ΔGo (cm-1) -5920 ± 75 -5875 ± 90 -5770 ± 40 -5725 ± 25 -5305 ± 50 -5055 ± 80 -4875 ± 75 -4445 ± 20 -4060 ± 65 Substituent SO2Me CN Cl H Furan SMe OMe OH N(EtOH)2 λ (cm-1) 10600 11100 11800 11600 11200 11000 11400 11200 10500 |ΔGo| (cm-1) 5920 ± 75 5875 ± 90 5770 ± 40 5725 ± 25 5305 ± 50 5055 ± 80 4875 ± 75 4445 ± 20 4060 ± 65 are summarized again in Table 4.23, which shows clearly that |ΔGo| < λ across the series, instantly pointing to the Marcus normal region. However, it is worth noting that these values are dependent on DFT calculations, so to be certain of Marcus normal region identification, experimental data must be considered. Figure 3.4: A plot showing the ln(knr) vs -ΔGo, while holding ΔQ constant. At low ΔGo the system is in the Marcus normal region (orange), and at high ΔGo the system is in the Marcus inverted region (blue). When |ΔGo| = λ, the system is in the barrierless region, and knr is at its peak value. As seen in the plot shown in Figure 3.4, ΔGo and knr have opposite relationships depending on which Marcus region the electronic transition occurs in. This behavior is explained when looking at the semi-classical Marcus equation: 𝑘𝑛𝑟 = 2𝜋 ℏ ∗ |𝐻𝑎𝑏|2 ∗ 1 √4𝜋𝜆𝑘𝐵𝑇 − (∆𝐺𝑜+𝜆)2 4𝜆𝑘𝐵𝑇 ∗ 𝑒 4.1 This shows that knr is exponentially proportional to −(𝛥𝐺𝑜 + 𝜆)2, in other words as the absolute values of the sum of ΔGo and λ increases (|sum|), knr will decrease. In this series ΔQ, and therefore λ, are assumed to be held constant, so changes in that sum are entirely dependent on changes in ΔGo. If the transition is in the Marcus normal region (|ΔGo| < λ), increasing the driving force will result in a lower |sum|, and higher knr, since ΔGo and λ have opposite signs. In the inverted region (|ΔGo| > λ), increasing the driving force increases |sum|, thus reducing knr. Since a ΔGo order has been established across the series, seeing how knr changes across the series will point towards the Marcus region GSR occurs in. This trend is shown in Figure 4.32B, which shows that as the driving 92 Figure 4.32: A: The GSR TA spectra collected at 295 K across the series. B: The ln(knr) at room temperature plotted against the driving force. The substituent at the 4’-position of a given complex is provided next to the point correlating to that complex. force of the transition is increased, so does the ln(knr). This provides the first piece of experimental evidence supporting the notion that GSR occurs in the Marcus normal region. Lastly is the trend between ΔGo and ΔG‡. The relationship between the two is represented in the following expression which was derived in Chapter 3. 𝛥𝐺‡ ∝ (𝛥𝐺𝑜+𝜆)2 𝜆 4.2 Which means that in this series, which again is assumed to hold λ constant, ΔG‡ has the opposite relationship with ΔGo than knr does: they are inversely related in the normal region and directly related in the inverted region. As shown in Figure 4.33, the ΔG‡ has an inverse relationship with the driving force. Once again, this displays experimental evidence of Marcus normal region behavior in the GSR process of a series of Fe(II)-based polypyridyl chromophores. Figure 4.33: The ΔG‡ at room temperature plotted against the driving force. The substituent at the 4’-position of a given complex is provided next to the point correlating to that complex. 93 Notably, the trends shown in Figures 4.32 and 4.33 display opposite but equal patterns. This is consistent with the analysis conducted in Chapter 3, which established that the ln(knr) is directly proportional to ΔG‡ (see Figure S3.1). Both trends display some slight deviations, which can be understood by looking at the following expression which describes the energy barrier in Marcus parameters (derived in Chapter 3): 𝛥𝐺‡ = [( (𝛥𝐺𝑜+𝜆)2 4𝜆 ) − 𝛽𝐻] − 𝑇 [( 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)) − 𝛽𝑆] 4.3 This equation shows that ΔG‡ is reliant on three variables (when at constant temperature): ΔGo, λ, and Hab. The model shown in Figure 3.4 assumes that ΔGo is the only variable changing, however that is not the case in this system. DFT calculations show that λ does change slightly with substituent placement. Furthermore, the method used to predict λ only considers inner sphere reorganization. Complexes with bulky or polar substituents (such as N(EtOH)2 or Cl) likely have larger perturbations to λ than what is given through DFT calculations. Furthermore, Hab also changes across the series (this will be explored further in the next section). As a result, deviations from the expected trend for the series likely stem from the changes in λ and Hab across each complex. Overall, the trends in both knr and ΔG‡ are consistent with the Marcus normal region, and DFT calculations predict that |ΔGo| is less than λ, a hallmark of the Marcus normal region. By performing VT-TA analysis using transition state and Marcus theory on a homologous series of substituted [Fe(4’-R-terpy)2]2+ complexes, experimental evidence supporting a long-held belief regarding the GSR process of Fe(II)-based polypyridyl chromophores was produced. This method can be used to establish the Marcus region of other complexes and electronic state transitions, granted the time constant of the process is long enough for thermalization within the reactant state to occur. 4.4.2 Electronic Coupling and Reorganization Energy As mentioned above, Hab changes across the series. Plot A in Figure 4.34 displays that Hab trends upwards as the driving force of the transition increases. At initial glance, this may seem counter intuitive as increasing the energy gap between the two states should result in weaker coupling. However, it is important to keep in mind that GSR in these complexes is a ΔS = 2 transition, thus direct coupling between each state cannot occur. Instead, coupling must occur through a triplet state, in this case the 3T1 state. The complexes in this study are low-spin d6 with the lowest energy excited state being the 5T2 state. This places the LF strength of each complex in 94 Figure 4.34: A: Hab plotted against |ΔGo| across the series, each point has the substituent on the associated complex listed near it. B: Tanabe-Sugano diagram for d6 octahedral complexes. The area highlighted in orange denotes the ligand field strength region Fe(II)-polypyridyl complexes typically fall under. the series within the region highlighted in orange in the Tanabe-Sugano diagram shown in Figure 4.34B. Within this region, the slope of the 5T2 state energy is sharper than that of the 3T1 state, so increasing the ligand field strength (which is tied to ΔGo), reduces the energy gap between the quintet and triplet states faster than the energy between the singlet and triplet states increases. This results in the overall coupling between the 5T2 and 1A1 states increasing. However, there are significant deviations from linear behavior shown in Figure 3.34A, indicating that something other than LF strength is influencing electronic coupling between the systems. Since the deviations are so strong, it is unclear what relationship one should expect between coupling and the energy gap between the three states in which coupling occurs through (i.e. linear, logarithmic, etc.). In other words, Figure 4.34A is unreliable for establishing a pattern that will aid in identifying what factors may be leading to the deviations from the trend in Hab across the series. Instead, the linear form of the semi-classical Marcus equation can be used: ln(𝑘𝑛𝑟 ∗ √𝑇) = ln (2𝜋 ℏ ∗ |𝐻𝑎𝑏|2 ∗ 1 √4𝜋𝜆𝑘𝐵 ) + (− (∆𝐺𝑜+𝜆)2 4𝜆𝑘𝐵 ∗ 1 𝑇 ) 4.4 This equation reveals that there should be a linear relationship between ln(knr) and [−(𝛥𝐺𝑜 + λ)2 λ⁄ ]. Therefore, when plotted against each other at constant temperature, any deviations from linear behavior must be proportional to the magnitude and direction of deviations in expected ln(|𝐻𝑎𝑏|2 √𝜆⁄ ratio will have a faster rate than what the value for [−(𝛥𝐺𝑜 + λ)2 λ⁄ ] would predict, and vice versa, meaning that complexes with points above the slope could have disproportionately stronger ) values. Complexes with stronger than expected electronic coupling 95 electronic coupling, and complexes below the slope could have disproportionately weaker coupling. This plot is shown in Figure 3.35. Figure 4.35: The ln(knr) at 295 K plotted against [−(𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ ] for each complex in the series, each point has the substituent on the associated complex listed near it. Of course, since deviations are dependent on a ratio between two variables, it is important to understand how strongly λ influences the ln(|𝐻𝑎𝑏|2 √𝜆⁄ ) ratio. This can be determined by plotting this ratio against the ln(knr), producing a plot with the ln(|𝐻𝑎𝑏|2 √𝜆⁄ ) on the y-axis (found in the SI, Figure 4.S21). The pattern between Figure 4.S21 and Figure 4.35 match, the same complexes are above and below the fit in each plot. Importantly, the fit overlaps with the point correlating to the 4’-OMe substituted complex. Therefore, the y-intercept of the fit can be treated as having a λ similar to the 4’-OMe substituted complex, which DFT calculations predict as 11400 cm-1 (Table 4.19). Since the y-axis is dependent on ln(1 √𝜆⁄ ), the fit can be adjusted to represent the extremes of λ (10500 cm-1 for the 4’-N(EtOH)2 substituted complex and 11800 cm-1 for the 4’-Cl substituted complex) by subtracting Δln(1 √𝜆⁄ ) from the y-intercept of the fit. This produces a “λ window” of how the extremes of λ values influence both ln(|𝐻𝑎𝑏|2 √𝜆⁄ ) and ln(knr). The ln(knr) window was applied to the fit shown in Figure 4.35, providing Figure 4.36 which shows the window that changes in λ can account for. With the window in place, it becomes clear that deviations from expected behavior can mostly be attributed to deviations in λ across the series. The complexes with the lower λ values, such as the SO2Me, SMe, and N(EtOH)2 substituted complexes, can be found near the bottom of the “λ window.” Meanwhile, the complexes with the highest λ values, the proto and chloro substituted complexes, sit at the top of the window. However, there are still some 96 notable exceptions. The CN, furan, and OH substituted complexes all have relatively low λ values, but they all rest either at or above the top end of the window. This likely results from a discrepancy between the actual value for λ and the value that was used the given complex, which was estimated through DFT. The substituents listed, as well as N(EtOH)2 whose complex also sits outside the window, are expected to significant outer sphere reorganization components (λo). The Nelsen method used to calculate λ can only consider inner sphere reorganization energy (λi). The data presented in Figure 4.36 suggests that the Nelsen method is adequate for predicting λ for most of the complexes in the series, but either underestimates or overestimates (in the case of the N(EtOH)2 substituted complex) λ for the complexes sitting outside of the window. At least, with respect to λ for the other complexes in the series, this analysis provides no information regarding the accuracy of the DFT determined λ to reality. If λo could be taken into consideration, the “λ window” would likely be large enough to encompass the whole series. Figure 4.36: The ln(knr) at 295 K plotted against [−(𝛥𝐺𝑜 + 𝜆)2 𝜆⁄ ] for each complex in the series, each point has the substituent on the associated complex listed near it. The indigo and green lines provide the window of influence different λ values have on the fit, the indigo correlates to the fit adjusted with the maximum λ values in the series and the green line correlates to the minimum λ value. Now, looking back at Figure 3.34A with this context, the trends observed begin to fall make more sense. Both terms in that plot, are calculated from experimentally determined ratios by using the DFT determined λi. The complexes that deviate most significantly from the overall trend between Hab and ΔGo are the same complexes that would be expected to have significant λo values. 97 Suggesting that the most significant factor influencing Hab is indeed the strength of ΔGo, as discussed before. 4.5 Concluding Thoughts A homologous series of 4’-position substituted [Fe(terpy)2]2+ complexes was studied using VT-TA spectroscopy. The GSR kinetic data obtained was then analyzed through the lenses of transition state and Marcus theory, assisted by DFT calculations. A trend in driving force for the 5T2 → 1A1 transition was established using a combination of Co(III) derivative-based ligand field analysis and Marcus theory analysis of the VT kinetic data. DFT predictions found that |ΔGo| was less than λ across the series, but importantly experimental data found trends between driving force, non-radiative rate, and the energy barrier which all exhibit Marcus normal region behavior. This provides experimental evidence supporting a long-held belief regarding Fe(II)-polypyridyl complexes. Slight deviations from Marcus normal region behavior were observed across the series, but these were sufficiently accounted for by examination of how Hab and λ shift across the series, both of which the Marcus normal region model assumes to be constant. 4.6 Future Directions A significant portion of this project was the development of a method for identifying which Marcus region an excited state transition occurs in. This relies on the modulation of ΔGo within a molecular framework, primarily through placing substituents with varying degrees of π-interaction on the ligand. This method has been successfully employed on other systems, for example it has been used to identify that GSR in [Co(bpy)3]3+ complexes occurs within the Marcus inverted region.22 However, not all systems have the luxury of using substituents to modulate ΔGo without making significant changes to λ. In systems like this, it would be ideal to use a method that allows for Marcus region identification through the use of a single molecule. In that effort, this could be done not by changing ΔGo with a constant λ, but instead by changing λ with a constant ΔGo. If λ could be modified while holding ΔGo constant, tracking changes in activation enthalpy (ΔH‡ = [(𝛥𝐺𝑜 + λ)2 4λ⁄ ]) with trends in λ could identify which Marcus region a transition is occurring within. As illustrated by Figure 3.37, ΔH‡ will have opposite relationships with λ depending on which Marcus region the transition occurs in. Therefore, VT-TA measurements could be used in the same manner as throughout this chapter to discern the Marcus region. 98 Figure 4.37: A model of how ΔH‡ changes with λ. This model was calculated using a ΔGo of - 7500 cm-1 from the following equation: 𝛥𝐻‡ = [(𝛥𝐺𝑜 + 𝜆)2 4𝜆⁄ ]. To keep ΔGo the same while changing λ, the same complex should be used while altering the solvent used to prepare the sample. A previous study by our group found that the GSR lifetime of a series of [Fe(bpy)3]2+ changed systematically with an increase of the static dielectric constant of the solvent used (DS).24 This change in lifetime was largely attributed to shifts in λo caused by different DS values. The relationship between λo and DS is illustrated by the following model:7 𝜆𝑜 = (𝛥𝑒)2 8𝜋 ( 1 𝐷𝑂𝑝 − 1 𝐷𝑆 ) ∫(𝐷𝑅 − 𝐷𝑃)2𝑑𝑡 4.5 Where Δe is the charge being transferred, DOp is the optical dielectric constant, which is equal to the refractive index of the solvent squared, and the term ∫(𝐷𝑅 − 𝐷𝑃)2𝑑𝑡 refers to difference in dielectric displacement poles between the reactant and product state integrated across the time of the transition. Equation 4.5 shows a general model which may change depending on the nature of the system, but λo is always proportional to (1 𝐷𝑂𝑝 − 1 𝐷𝑆⁄ ). So, in solvents where DS is greater than DOp, such as alcohols, increasing DS will lead to a larger λo. So, plotting ΔH‡ against DS should ⁄ yield a positive slope in the Marcus normal region, and a smaller slope in the inverted region. It would also be interesting to test this method on transitions with a very small energy gap, such as GSR in spin-crossover complexes. The parabola in Figure 4.37 is asymmetric, the inverted region increases more sharply than the normal region as λ shifts away from the barrierless region. This asymmetricity is not very pronounced at high ΔGo values, but with |ΔGo| values less than 1500 cm-1, this behavior becomes extremely pronounced, illustrated in Figure 4.38. 99 Figure 4.38: A model of how ΔH‡ changes with λ. This model was calculated using a ΔGo of -500 cm-1 from the following equation: 𝛥𝐻‡ = [(𝛥𝐺𝑜 + 𝜆)2 4𝜆⁄ ]. This behavior could be exploited to drastically increase the energy barrier between states with a low energy gap within the inverted region by using solvents with weaker DS, which should lead to increased time constants for the transition. If this model is correct, then determining the Marcus region of a transition could be simplified, only relying on a single complex rather than a series of complexes. Furthermore, depending on the region and energy gap between states, the solvent used could drastically impact the rate of transition. To test this model, VT-TA data should be collected for the GSR of both [Fe(bpy)3]2+, which is a normal region process, and [Co(bpy)3]3+, which is an inverted region process, and analyzed via Marcus and transition state theory. The Fe(II) complex should see a positive trend between ΔH‡ and DS, while the Co(III) complex should see a negative trend. 100 REFERENCES 1. Woodhouse, M. D. and McCusker, J. K. J. Am. Chem. Soc. 2020, 142, 16229 – 16233. 2. Gualandi, A.; Marchini, M.; Mengozzi, L.; Natali, M.; Lucarini, M.; Ceroni, P. and Cozzi, P. G. ACS Catal. 2015, 5, 5927 − 5931. 3. Parisien-Collette, S.; Hernandez-Perez, A. C. and Collins, S. K. Org. Lett. 2016, 18, 4994 − 4997. 4. Xia, S.; Hu, K.; Lei, C. and Jin, J. Org. Lett. 2020, 22, 1385 − 1389. 5. Creutz, C.; Chou, M.; Netzel, T. L.; Okumura, M. and Sutin, N. J. Am. Chem. Soc. 1980, 102, 1309 − 1319. 6. McCusker, J. K. Science. 2019, 363, 484 – 488. 7. Sutin, N. Prog. Inorg. Chem. 1983, 30, 441 – 498. 8. Carey, M. C.; Adelman, S. L.; McCusker, J. K. Chem. Sci. 2019, 10, 134 − 144. 9. Ashley, D. C. and Jakubikova, E. Inorg. Chem. 2018, 57, 9907 – 9917. 10. Papp, M.; Keszthelyi, T.; Vancza, A.; Bajnóczi, É. G.; Kováts, É.; Németh, Z.; Bogdán, C.; Bazsó, G.; Rozgonyi, T.; and Vankó, G. Inorg. Chem. 2023, 62, 6397 – 6410. 11. Szemes, D. S.; Keszthelyi, T.; Papp, M.; Varga, L.; Vanko, G. Chem. Commun. 2020, 56, 11831 − 11834. 12. England, J.; Scarborough, C. C.; Weyhermüller, T.; Sproules, S. and Wieghardt, K. Eur. J. Inorg. Chem. 2012, 2012, 4605 – 4621. 13. Potts, K. T.; Usifer, D. A.; Guadalupe, A. and Abruna, H. D. J. Am. Chem. Soc. 1987, 109, 3961 – 3967. 14. Jameson, D. L. and Guise, L. E. Tetrahedron Lett. 1991, 32, 1999 – 2002. 15. Husson, J.; Dehaudt, J. and Guyard, L. Nat. Protoc. 2014, 9, 21 – 26. 16. Medlycott, E. A. and Hanan, G. S. Chem. Soc. Rev. 2005, 34, 133 – 142. 17. Pal, A. K. and Hanan, G. S. Chem. Soc. Rev. 2014, 43, 6184 – 6197. 18. Schlingloff, G.; Wieprecht, T.; Bachmann, F.; Dannacher, J.; Dubs, M.-J.; Hazenkamp, M.; Hänsler, G.; Schmidt, B.; Schneider, A. and Weingartner, P. WO2002088289 A2, Nov 7th, 2002. 101 19. Constable, E. C.; Dunphy, E. L.; Housecroft, C. E.; Neuburger, M.; Schaffner, S.; Schaper, F. and Batten, S. R. Datlon Trans. 2007, 38, 4323 – 4332. 20. Chambers, J.; Eaves, B.; Parker, D.; Claxton, R.; Ray, P. R. and Slatter, S. J. Inorganica Chim. Acta. 2006, 359, 2400 – 2406. 21. Yarranton, J. T. Synthesis and Characterization of Low-Spin d6 Polypyridyl Complexes: Manipulation of Excited-State Electronic Structure through Ligand Design. Doctoral Thesis, Michigan State University. ProQuest Dissertations Publishing. 22. Yarranton, J. T. and McCusker, J. K. J. Am. Chem. Soc. 2022, 144, 12488 − 12500. 23. Hsu, C.-P. Acc. Chem. Res. 2009, 42, 509 – 518. 24. Miller, J. N. and McCusker, J. K. Chem. Sci. 2020, 11, 5191 – 5204. 102 APPENDIX: SUPPLEMENTAL INFORMATION 4.SI.1 Adjusting Outer-Sphere Reorganization Energy Section 4.4.2 of this chapter points out that ignoring outer-sphere reorganization energy (λo) leads to some discrepancies in observed trends. Both Hab and ΔGo were calculated using the DFT determined inner-sphere reorganization energy (λi). As highlighted by Figure 4.34A, this led to deviations from the expected trend between both values, as Hab is expected to increase with ΔGo. This expected trend can be used to predict the value of λo, as when it is included in calculations, the trend between Hab and ΔGo should be corrected. The work in this section tests this theory. The first step is to estimate the total reorganization energy (λtotal) for each complex. An estimated λtotal of 14100 cm-1 for [Fe(terpy)2]2+ was taken from previously published literature.1 The DFT calculated λi was subtracted from this value to get a baseline value for λo (2500 cm-1). This was added to the λi of each complex. These λtotal values were used to calculate Hab and ΔGo from the experimentally determined ratios and produce the plot in Figure 4.S1. Figure 4.S1: Electronic coupling vs. Driving Force calculated using estimated λtotal values. Since the unsubstituted complex is used as the baseline, the fit was adjusted to be in line with the point representing [Fe(terpy)2]2+. Then, the λo for each substituted complex was adjusted and both Hab and ΔGo were recalculated using the adjusted λtotal. If the original point was below the fit, λo reduced, and it was increased if the point was above the fit. Updated Hab and ΔGo values 103 were plotted and this process was repeated until each point was in line with the slope of the adjusted fit. The linear plot is provided in Figure 4.S2. Figure 4.S2: Electronic coupling vs. Driving Force calculated using adjusted λtotal values. It was found that ΔGo changed much more than Hab with changes in λo. This led to a reordering of the trend in ligand field strength between complexes. This initially appears to create deviations from expected Marcus normal region behavior, however it is important to keep in mind that λ is no longer constant across the series. The updated λtotal, Hab, and ΔGo values are provided in table 4.S1: Table 4.S1: Values for λtotal, Hab, and |ΔGo| calculated after estimating λo. Each complex is listed in order of decreasing driving force. Substituent λtotal (cm-1) Hab (cm-1) |ΔGo| (cm-1) CN H Furan Cl SO2Me OH OMe SMe N(EtOH)2 14200 14100 14300 14000 12350 14900 13400 12600 12200 7.10 6.25 6.22 5.97 5.81 5.58 4.62 4.48 3.24 8350 7600 7600 7450 7300 7100 6350 6250 5250 Table 4.S1 displays that λtotal shifts significantly across the series, by nearly 3000 cm-1. The expected normal region trend shown in Figure 3.4 expects λ to be held constant. So, when plotting 104 ln(knr) against |ΔGo|, the values for λo should be kept in mind. This plot is provided in Figure 4.S3, which displays that complexes with similar λtotal values still engage in Marcus normal region behavior. Figure 4.S3: Plot comparing ln(knr) vs. driving force. Complexes are grouped by similar λtotal values, both the OH and OMe substituted complexes are omitted from this plot as each had λtotal values distinct from the two groups. This analysis was omitted from the main body of the chapter as it relies on many assumptions. First: that the relationship between Hab and ΔGo is linear. Second: that the slope from the initial fit is accurate to reality. Third: that 14100 cm-1 is a reasonable estimate for the λtotal of [Fe(terpy)2]2+. To eliminate reliance on the third assumption, this analysis was performed with different λtotal values used for [Fe(terpy)2]2+ and while it was found that values calculated for all parameters changed, the trends among each complex did not. As a final note, this analysis does not change the overall findings of this work. Regardless of how λ is incorporated, the series of [Fe(4’- X-terpy)2]2+ complexes still displays Marcus normal region behavior. 4.SI.2 Synthesis and Characterization Provided below is the NMR and ESI-TOF data used to characterize the complexes synthesized in the McCusker research lab. Characterization on the remaining complexes was conducted by the Vanko group. Characterization of [Fe(terpy)2](PF6)2: 1H NMR (500 MHz, acetone): δ 9.24 (d, J = 8.0 Hz, 4H), 8.87 (t, J = 8.1 Hz, 2H), 8.79 (d, J = 8.0 Hz, 4H), 8.04 (t, J = 7.8 Hz, 4H), 7.44 (d, J = 5.6 Hz, 4H), 7.25 (t, J = 6.6 Hz, 4H). HRMS (ESI-TOF) m/z: 261.0637 (calculated for [M-2(PF6)]2+: 261.0628). 105 Characterization of [Co(4’-SMe-terpy)2](PF6)3: 1H NMR (500 MHz, CD3CN): δ 8.73 (s, 4H), 8.56 (ddd, J = 7.9, 1.5, 0.6 Hz, 4H), 8.21 (td, J = 7.8, 1.4 Hz, 4H), 7.43 (ddd, J = 7.6, 5.8, 1.4 Hz, 4H), 7.35 (ddd, J = 5.8, 1.3, 0.5 Hz, 4H), 3.02 (s, 6H). 13C NMR (125 MHz, CD3CN): δ 166.23, 156.88, 154.94, 153.37, 143.94, 131.74, 127.84, 123.84, 118.34, 15.75. HRMS (ESI-TOF) m/z: 907.0263 (calculated for [M-PF6]1+: 907.0276). Characterization of [Co(4’-N(EtOH)2-terpy)2](PF6)3: 1H NMR (500 MHz, CD3CN): δ 8.56 (dt, J = 8.0, 0.9 Hz, 4H), 8.18 – 8.09 (m, 8H), 7.45 – 7.36 (m, 8H), 4.12 (t, J = 5.3 Hz, 8H), 4.06 (q, J = 5.4 Hz, 8H), 3.59 (t, J = 5.5 Hz, 3H). 13C NMR (125 MHz, CD3CN): δ 160.06, 158.38, 154.70, 152.88, 143.31, 130.72, 126.52, 118.34, 111.08, 60.03, 55.15. HRMS (ESI-TOF) m/z: 1021.1537 (calculated for [M-PF6]1+: 1021.1788). Characterization of [Co(4’-furan-terpy)2](PF6)3: 1H NMR (500 MHz, CD3CN): δ 9.18 (s, 4H), 8.69 (d, J = 7.9 Hz, 4H), 8.24 (td, J = 7.7, 2.1 Hz, 4H), 8.11 (d, J = 1.8 Hz, 2H), 7.95 (d, J = 3.7 Hz, 2H), 7.45 – 7.39 (m, 8H), 7.00 (dd, J = 3.8, 1.8 Hz, 2H). 13C NMR (125 MHz, CD3CN): δ 157.04, 157.03, 153.35, 150.07, 149.80, 146.46, 144.06, 131.82, 128.08, 121.76, 118.34, 115.57. HRMS (ESI-TOF) m/z: 947.0737 (calculated for [M-PF6]1+: 947.0733). Characterization of [Co(4’-OMe-terpy)2](PF6)3: 1H NMR (500 MHz, CD3CN): δ 8.58 (dd, J = 8.0, 1.4 Hz, 4H), 8.54 (s, 4H), 8.21 (td, J = 7.8, 1.4 Hz, 4H), 7.43 (ddd, J = 7.5, 5.8, 1.4 Hz, 4H), 7.35 (d, J = 1.2 Hz, 4H), 4.46 (s, 6H). 13C NMR (125 MHz, CD3CN): δ 174.37, 157.80, 157.15, 153.25, 143.96, 131.73, 127.82, 118.34, 115.47, 59.91. HRMS (ESI-TOF) m/z: 875.0711 (calculated for [M-PF6]1+: 875.0733). Characterization of [Co(4’-Cl-terpy)2](PF6)3: 1H NMR (500 MHz, CD3CN): δ 9.15 (s, 4H), 8.57 (dd, J, = 7.9, 1.0 Hz, 4H), 8.26 (td, J = 7.8, 1.3 Hz, 4H), 7.48 (ddd, J = 7.6, 5.9, 1.5 Hz, 4H), 7.38 (dd, J = 6.0, 1.2 Hz, 4H). 13C{1H} NMR (125 MHz, CD3CN): δ 157.70, 156.00, 155.30, 153.79, 144.34, 132.44, 129.24, 128.78, 118.34. 106 HRMS (ESI-TOF) m/z: 882.9749 (calculated for [M-PF6]1+: 882.9742). Unused Ligands As mentioned in section 4.3.11 both 4’-CN-2,2’:6’,2”-terpyridine and 4’-SO2Me- 2,2’:6’,2”-terpyridine were synthesized, but neither bound to the metal. This is likely because the strongly electronegative substituents reduced the σ-donating strength of the ligand too much. The synthesis for each ligand is provided below and they were characterized by the same methods listed for the other complexes. Synthesis and characterization for each ligand was carried out by Yi-Jyun Lien. Synthesis and Characterization of 4'-cyano-2,2':6',2''-terpyridine. 4’-CN-terpy was synthesized by modified literature procedures.2 4'-(methylsulfonyl)-2,2':6',2''-terpyridine (0.63 g, 2.02 mmol) and potassium cyanide (0.46 g, 7.06 mmol) were added to anhydrous DMF (10 mL) and heated to 110°C for overnight. Upon cooling to room temperature, the mixture was diluted with water (40 mL) and the aqueous portion was extracted with DCM (3 × 100 mL). The combined organic layers were washed with brine, dried with Na2SO4, and the solvent was removed. (0.42 g, 81% yield). 1H NMR (500 MHz, CDCl3) δ 8.72 (d, J = 4.8 Hz, 2H), 8.70 (s, 2H), 8.60 (d, J = 7.9 Hz, 2H), 7.90 (td, J = 7.7, 1.8 Hz, 2H), 7.40 (ddd, J = 7.5, 4.7, 1.2 Hz, 2H). Synthesis of 4'-(methylsulfonyl)-2,2':6',2''-terpyridine. 4’-SO2Me-terpy was synthesized by modified literature procedures.3 4'-(methylthio)-2,2':6',2''-terpyridine (0.24 g, 0.96 mmol) was first dissolved in DCM (5 mL) and cooled to 0°C. ~70% m-chloroperoxybenzoic acid (mCPBA) (0.90 g, 5.22 mmol) was slowly added to the flask. After 20 minutes the mixture was allowed to warm to room temperature and stirring continued for 6 hours. The resulting solution was then extracted with DCM (3 × 100 mL). The combined organic layers were washed with sat. NaHCO3 (aq.) once and brine twice, dried with Na2SO4, and the solvent was removed. Purification was performed by column chromatography on silica gel with DCM solution as eluent, affording white powder. In the chromatography purification process, despite some decomposition of the target ligand occurring, its presence was still detected in the H-NMR spectrum as compared to the literature. 107 4.SI.3 DFT Calculations Figure 4.S4: Near UV-Vis spectrum of [Fe(terpy)2](PF6)2 (in black), alongside TD-DFT simulated spectra for [Fe(terpy)2]2+ calculated using HF of 20% (in purple), 15% (in green), and 10% (in red). DFT calculations were conducted following the procedure outlined in Chapter 2.2. To find the Hartree-Fock (HF) percent that would yield the most accurate results, various HF% was used to optimize the geometry of the singlet state. Then, a time dependent DFT calculation (TD-DFT) was conducted to generate a simulated UV-Vis spectrum and compared to the experimental spectrum. The experimental spectrum has a unique feature where there is a very sharp peak with a broad shoulder. Since the simulated spectra provides the same width to all peaks, this unique shape Figure 4.S5: UV-Vis spectrum of [Fe(terpy)2](PF6)2 (in black), alongside TD-DFT simulated spectrum for [Fe(terpy)2]2+ (in red) and the individual states (in blue). 108 cannot be replicated by TD-DFT, instead the energy of the peaks and presence of excitations that could allow the shape were used to determine the most accurate spectrum. Figure 4.S4 provides three of the TD-DFT spectra alongside the experimental spectrum. It was found that using 10% HF yielded the spectrum closest in energy to the experimental. Figure 4.S5 shows the individual excitation states calculated through TD-DFT. These states closely match the features of the experimental spectrum. There is a single intense peak around 530 nm which accounts for the sharp MLCT feature. Near the intense peak are several weaker peaks which account for the broad shoulder near the sharp MLCT feature. Lastly, the red shoulder around 600 nm is predicted by TD- DFT. This process was repeated for each complex in the series, 10% HF provided the best fit for most complexes. However, for some 10% HF produced poor fits of the experimental data. Instead, 5% HF provided the best fit for the 4’-CN and 4’-SO2Me substituted complexes and 20% HF provided the best fit for the 4’-SMe and 4’-N(EtOH)2 substituted complexes. There was a concern that using a different HF% would yield incomparable values for λ. To check this, λ was calculated across 20%, 15%, and 10% for [Fe(terpy)2]2+, and it was found that λ only varied by 100 cm-1, less than 1% of the value for λ at 10% HF. This was reproduced within 200 cm-1 across other complexes in the series, with two exceptions. The complexes where 20% HF provided the best fit both had negative values for λ when using 10% HF, which should not be possible. Results for λ using other HF% were still within 200 cm-1 of the value at calculated with 20% HF. Provided below are the optimized singlet and quintet geometries, and relevant structural information for each complex in the series. The pbd file information for each complex in the series will be available at the end of the SI. [Fe(terpy)2]2+ DFT Geometries Figure 4.S6: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(terpy)2]2+. 109 Table 4.S2: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.93 2.18/2.19 2.03 2.24/2.26 179.8 166.7 162.1/162.2 148.2/149.1 [Fe(4’-SMe-terpy)2]2+ DFT Geometries Figure 4.S7: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-SMe-terpy)2]2+. Table 4.S3: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.92 2.15 2.03 2.24 179.9 178.7 160.4 147.3 [Fe(4’-SO2Me-terpy)2]2+ DFT Geometries Figure 4.S8: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-SO2Me-terpy)2]2+. Table 4.S4: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.95 2.18 2.07 2.28 179.9 178.5 161.8/161.9 150.1 110 Fe(4’-OMe-terpy)2]2+ DFT Geometries Figure 4.S9: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-OMe-terpy)2]2+. Table 4.S5: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.94 2.18 2.03 2.24/2.25 179.9 164.9 161.7 148.1/148.2 Fe(4’-N(EtOH)2-terpy)2]2+ DFT Geometries Figure 4.S10: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-N(EtOH)2-terpy)2]2+. Table 4.S6: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.92 2.13 2.04 2.24 179.8 179.7 159.8/159.9 147.3/147.4 111 Fe(4’-CN-terpy)2]2+ DFT Geometries Figure 4.S11: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-CN-terpy)2]2+. Table 4.S7: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.95 2.17 2.07 2.27/2.28 179.9 179.0 161.9 150.3 Fe(4’-Cl-terpy)2]2+ DFT Geometries Figure 4.S12: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-Cl-terpy)2]2+. Table 4.S8: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.94 2.20 2.05/2.06 2.26/2.27 179.9 174.6 161.3 147.7/147.9 112 Fe(4’-OH-terpy)2]2+ DFT Geometries Figure 4.S13: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-OH-terpy)2]2+. Table 4.S9: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.95 2.19 2.05/2.06 2.27 179.9 178.6 161.0 148.0 Fe(4’-furan-terpy)2]2+ DFT Geometries Figure 4.S14: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4’-furan-terpy)2]2+. Table 4.S10: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-Nax (Å) Fe-Neq (Å) Nax-Fe-Nax (deg) Neq-Fe-Neq (deg) 1.94 2.18 2.05/2.06 2.27 179.8 178.7 161.1/161.2 148.0/148.1 113 4.SI.4 Cobalt(III) Derivatives The UV-Vis spectra for the 3T1 absorption feature for all Co(III) analogs that were used is provided in Figure 4.S12. Given the transition is both spin and Laporte forbidden, the molar absorptivity is very weak, ranging from ~2.5 to ~13 M-1 cm-1. The deconvolution fit for each complex is also given in the following figures. Figure 4.S15: The UV-Vis spectra of the 3T1 absorption feature across several [Co(4’-R- terpy)2](PF6)3 complexes. Figure 4.S16: UV-Vis spectra for [Co(4’-furan-terpy)2](PF6)3 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 3T1 peak is at 15500 cm- 1. There is a warp in the baseline corrected peak near 735 nm (which is visible in Figure 4.S12). This is an artifact caused by removing the solvent features from the spectrum. This is done for each complex by measuring a solvent blank (typically acetonitrile), then subtracting the blank from the spectrum. This complex had poor solubility in acetonitrile so some water was added to the solution, which increased solubility. The acetonitrile:water ratio could not be perfectly recreated, which led to these artifacts appearing where solvent features are present in the non-corrected spectrum. 114 Figure 4.S17: UV-Vis spectra for [Co(4’-N(EtOH)2-terpy)2](PF6)3 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 3T1 peak is 14200 cm-1. Figure 4.S18: UV-Vis spectra for [Co(4’-OMe-terpy)2](PF6)3 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 3T1 peak is 15000 cm-1. Figure 4.S19: UV-Vis spectra for [Co(4’-Cl-terpy)2](PF6)3 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 3T1 peak is 15800 cm-1. 115 In addition to the 3T1 peaks, the 1T1 peaks were also measured, which are shown in Figure 4.S20. However, the proximity to the CT excitation features made deconvolving the 1T1 peaks nearly impossible for the 4’-furan and 4’-N(EtOH)2 complexes nearly impossible. On top of this, there is a 1T2 peak buried under the CT feature.4 Properly fitting this peak is an unclear practice, and the energy of the 1T1 peak would shift when the 1T2 peak was not included in the fit. Furthermore, the data set was largely in agreement with the trend seen in the 3T1 data, so for these reasons the 1T1 data was deemed unnecessary. Figure 4.S20: The UV-Vis spectra of the 1T1 absorption feature across several [Co(4’-R- terrpy)2](PF6)3 complexes. 4.SI.5 Reorganization Energy Variance impact on ln(knr) Figure 4.S21 plots ln(Hab 2/√λ) vs ln(knr). In order to see how greatly changes in λ impact each factor, the fit from the plot was adjusted by the boundaries of λ determined through DFT calculations (11800 cm-1 to 10500 cm-1). To do this, the fit was treated as being representative of λ = 11400 cm-1, since the fit crossed the point for the diOMe substituted complex (which has a λ of 11400 cm-1). The y-intercept value for ln(Hab 2/√λ) was used to calculate Hab with a λ of 11400 cm-1, this effectively represents what Hab would be at ln(knr) = 0 (or knr = 1). Then, holding that y- intercept Hab value constant, new ln(Hab 2/√λ) values were calculated substituting λ with 11800 and 10500 cm-1 to generate new y-intercept values for λ-adjusted fits (keeping the slope the same). The difference of x-intercept values of each fit represents the range of influence λ has on ln(knr) for the series. This range was used to make the adjusted fits present in Figure 4.36. 116 Furan H Cl CN SO2Me OH SMe OMe N(EtOH)2 2/√λ) vs ln(knr). The linear fit of the plot was adjusted to represent Figure 4.S21: Plot of ln(Hab different values for λ across the series. This was done to determine how greatly the λ variance across the series would influence both Hab and knr. 4.SI.6 DFT Optimized Geometry Files [Fe(terpy)2]2+: Table 4.S11: The optimized geometry file for the lowest energy singlet state of [Fe(terpy)2]2+. TITLE Fe-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.003 3.969 6.122 N HETATM 2 N 0 0.100 1.688 5.331 N HETATM 3 N 0 1.722 5.062 3.902 N HETATM 4 C 0 0.550 3.199 7.181 C HETATM 5 C 0 0.617 3.696 8.517 C HETATM 6 C 0 1.160 5.000 8.736 C HETATM 7 C 0 0.025 1.869 6.719 C HETATM 8 C 0 -0.349 0.504 4.794 C HETATM 9 C 0 -0.503 0.862 7.579 C HETATM 10 C 0 -0.887 -0.543 5.597 C HETATM 11 C 0 -0.966 -0.362 7.011 C HETATM 12 C 0 1.531 5.234 6.319 C HETATM 13 C 0 1.624 5.783 7.633 C HETATM 14 C 0 1.952 5.871 5.023 C HETATM 15 C 0 2.070 5.540 2.660 C HETATM 16 C 0 2.534 7.166 4.906 C HETATM 17 C 0 2.655 6.826 2.473 C HETATM 18 C 0 2.891 7.652 3.613 C HETATM 19 N 0 0.772 2.559 2.525 N HETATM 20 N 0 -0.960 3.944 3.839 N HETATM 21 N 0 2.698 2.348 4.225 N HETATM 22 C 0 -0.372 2.794 1.781 C HETATM 23 C 0 -0.487 2.287 0.451 C HETATM 24 C 0 0.603 1.539 -0.092 C HETATM 25 C 0 -1.381 3.601 2.547 C HETATM 26 C 0 -1.804 4.684 4.634 C 117 Table 4.S11 (cont’d) HETATM 27 C 0 -2.653 4.001 2.043 C HETATM 28 C 0 -3.089 5.114 4.192 C HETATM 29 C 0 -3.520 4.767 2.876 C HETATM 30 C 0 1.835 1.835 2.012 C HETATM 31 C 0 1.778 1.305 0.687 C HETATM 32 C 0 2.958 1.712 3.004 C HETATM 33 C 0 3.655 2.299 5.212 C HETATM 34 C 0 4.182 1.023 2.764 C HETATM 35 C 0 4.899 1.627 5.037 C HETATM 36 C 0 5.167 0.979 3.794 C HETATM 37 H 0 0.254 3.080 9.369 H HETATM 38 H 0 1.222 5.408 9.769 H HETATM 39 H 0 -0.266 0.407 3.689 H HETATM 40 H 0 -0.549 1.037 8.676 H HETATM 41 H 0 -1.236 -1.482 5.111 H HETATM 42 H 0 -1.383 -1.162 7.664 H HETATM 43 H 0 2.048 6.799 7.794 H HETATM 44 H 0 1.868 4.861 1.802 H HETATM 45 H 0 2.705 7.783 5.816 H HETATM 46 H 0 2.918 7.164 1.446 H HETATM 47 H 0 3.347 8.661 3.499 H HETATM 48 H 0 -1.406 2.473 -0.147 H HETATM 49 H 0 0.535 1.135 -1.127 H HETATM 50 H 0 -1.426 4.932 5.650 H HETATM 51 H 0 -2.955 3.713 1.011 H HETATM 52 H 0 -3.732 5.710 4.876 H HETATM 53 H 0 -4.518 5.089 2.502 H HETATM 54 H 0 2.631 0.724 0.273 H HETATM 55 H 0 3.402 2.817 6.163 H HETATM 56 H 0 4.357 0.528 1.783 H HETATM 57 H 0 5.638 1.618 5.869 H HETATM 58 H 0 6.130 0.446 3.627 H HETATM 59 Fe 0 0.889 3.262 4.324 Fe END CONECT 1 4 12 59 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 38 CONECT 7 2 4 9 CONECT 8 2 10 39 CONECT 9 7 11 40 CONECT 10 8 11 41 CONECT 11 9 10 42 CONECT 12 1 13 14 CONECT 13 12 6 43 CONECT 14 3 12 16 CONECT 15 3 17 44 CONECT 16 14 18 45 CONECT 17 15 18 46 CONECT 18 16 17 47 CONECT 19 22 30 59 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 48 CONECT 24 23 31 49 CONECT 25 22 20 27 CONECT 26 20 28 50 CONECT 27 25 29 51 CONECT 28 26 29 52 CONECT 29 27 28 53 CONECT 30 19 31 32 CONECT 31 24 30 54 CONECT 32 21 30 34 CONECT 33 21 35 55 CONECT 34 32 36 56 118 Table 4.S11 (cont’d) CONECT 35 33 36 57 CONECT 36 34 35 58 CONECT 37 5 CONECT 38 6 CONECT 39 8 CONECT 40 9 CONECT 41 10 CONECT 42 11 CONECT 43 13 CONECT 44 15 CONECT 45 16 CONECT 46 17 CONECT 47 18 CONECT 48 23 CONECT 49 24 CONECT 50 26 CONECT 51 27 CONECT 52 28 CONECT 53 29 CONECT 54 31 CONECT 55 33 CONECT 56 34 CONECT 57 35 CONECT 58 36 CONECT 59 1 19 Table 4.S12: The optimized geometry file for the lowest energy quintet state of [Fe(terpy)2]2+. TITLE Fe-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.947 3.979 6.359 N HETATM 2 N 0 0.050 1.522 5.834 N HETATM 3 N 0 1.741 5.077 4.053 N HETATM 4 C 0 0.510 3.286 7.470 C HETATM 5 C 0 0.553 3.870 8.771 C HETATM 6 C 0 1.062 5.199 8.905 C HETATM 7 C 0 0.001 1.887 7.171 C HETATM 8 C 0 -0.385 0.274 5.461 C HETATM 9 C 0 -0.492 0.994 8.170 C HETATM 10 C 0 -0.890 -0.672 6.400 C HETATM 11 C 0 -0.942 -0.302 7.777 C HETATM 12 C 0 1.447 5.258 6.477 C HETATM 13 C 0 1.517 5.905 7.754 C HETATM 14 C 0 1.901 5.880 5.175 C HETATM 15 C 0 2.136 5.568 2.832 C HETATM 16 C 0 2.462 7.193 5.089 C HETATM 17 C 0 2.703 6.864 2.665 C HETATM 18 C 0 2.868 7.691 3.817 C HETATM 19 N 0 0.809 2.574 2.254 N HETATM 20 N 0 -1.132 3.822 3.615 N HETATM 21 N 0 2.787 2.071 3.987 N HETATM 22 C 0 -0.299 2.893 1.501 C HETATM 23 C 0 -0.364 2.541 0.116 C HETATM 24 C 0 0.744 1.853 -0.463 C HETATM 25 C 0 -1.386 3.630 2.263 C HETATM 26 C 0 -2.061 4.487 4.377 C HETATM 27 C 0 -2.583 4.109 1.649 C HETATM 28 C 0 -3.284 4.986 3.839 C HETATM 29 C 0 -3.546 4.793 2.449 C HETATM 30 C 0 1.883 1.907 1.710 C HETATM 31 C 0 1.884 1.528 0.332 C HETATM 32 C 0 3.016 1.641 2.687 C HETATM 33 C 0 3.763 1.878 4.933 C HETATM 34 C 0 4.239 1.005 2.312 C HETATM 35 C 0 5.006 1.244 4.638 C HETATM 36 C 0 5.245 0.801 3.303 C HETATM 37 H 0 0.199 3.309 9.662 H HETATM 38 H 0 1.106 5.680 9.908 H 119 Table 4.S12 (cont’d) HETATM 39 H 0 -0.324 0.038 4.373 H HETATM 40 H 0 -0.524 1.301 9.238 H HETATM 41 H 0 -1.231 -1.673 6.052 H HETATM 42 H 0 -1.330 -1.014 8.541 H HETATM 43 H 0 1.917 6.937 7.849 H HETATM 44 H 0 1.988 4.889 1.962 H HETATM 45 H 0 2.582 7.817 6.001 H HETATM 46 H 0 3.006 7.210 1.651 H HETATM 47 H 0 3.307 8.711 3.729 H HETATM 48 H 0 -1.255 2.794 -0.499 H HETATM 49 H 0 0.719 1.568 -1.539 H HETATM 50 H 0 -1.804 4.619 5.453 H HETATM 51 H 0 -2.763 3.954 0.563 H HETATM 52 H 0 -4.006 5.515 4.500 H HETATM 53 H 0 -4.487 5.174 1.990 H HETATM 54 H 0 2.749 0.992 -0.115 H HETATM 55 H 0 3.530 2.250 5.958 H HETATM 56 H 0 4.406 0.672 1.264 H HETATM 57 H 0 5.764 1.108 5.442 H HETATM 58 H 0 6.205 0.306 3.032 H HETATM 59 Fe 0 0.818 3.045 4.389 Fe END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 38 CONECT 7 2 4 9 CONECT 8 2 10 39 CONECT 9 7 11 40 CONECT 10 8 11 41 CONECT 11 10 9 42 CONECT 12 1 13 14 CONECT 13 12 6 43 CONECT 14 3 12 16 CONECT 15 3 17 44 CONECT 16 14 18 45 CONECT 17 15 18 46 CONECT 18 16 17 47 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 48 CONECT 24 23 31 49 CONECT 25 20 22 27 CONECT 26 20 28 50 CONECT 27 25 29 51 CONECT 28 26 29 52 CONECT 29 27 28 53 CONECT 30 19 31 32 CONECT 31 24 30 54 CONECT 32 21 30 34 CONECT 33 21 35 55 CONECT 34 32 36 56 CONECT 35 33 36 57 CONECT 36 34 35 58 CONECT 37 5 CONECT 38 6 CONECT 39 8 CONECT 40 9 CONECT 41 10 CONECT 42 11 CONECT 43 13 CONECT 44 15 CONECT 45 16 CONECT 46 17 120 Table 4.S12 (cont’d) CONECT 47 18 CONECT 48 23 CONECT 49 24 CONECT 50 26 CONECT 51 27 CONECT 52 28 CONECT 53 29 CONECT 54 31 CONECT 55 33 CONECT 56 34 CONECT 57 35 CONECT 58 36 [Fe(4’-SMe-terpy)2]2+: Table 4.S13: The optimized geometry file for the lowest energy singlet state of [Fe(4’-SMe- terpy)2]2+. TITLE Fe-SMe-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.228 4.024 6.055 N HETATM 2 N 0 0.153 1.778 5.514 N HETATM 3 N 0 1.752 5.011 3.762 N HETATM 4 C 0 0.876 3.346 7.160 C HETATM 5 C 0 1.104 3.872 8.428 C HETATM 6 C 0 1.713 5.130 8.539 C HETATM 7 C 0 0.252 2.044 6.846 C HETATM 8 C 0 -0.398 0.619 5.130 C HETATM 9 C 0 -0.205 1.144 7.806 C HETATM 10 C 0 -0.874 -0.324 6.035 C HETATM 11 C 0 -0.775 -0.057 7.395 C HETATM 12 C 0 1.813 5.236 6.135 C HETATM 13 C 0 2.071 5.819 7.366 C HETATM 14 C 0 2.116 5.806 4.806 C HETATM 15 C 0 1.986 5.442 2.515 C HETATM 16 C 0 2.719 7.045 4.604 C HETATM 17 C 0 2.585 6.667 2.242 C HETATM 18 C 0 2.957 7.481 3.305 C HETATM 19 N 0 0.555 2.488 2.598 N HETATM 20 N 0 -0.974 3.990 3.980 N HETATM 21 N 0 2.637 2.244 4.048 N HETATM 22 C 0 -0.620 2.734 1.984 C HETATM 23 C 0 -0.896 2.195 0.737 C HETATM 24 C 0 0.069 1.382 0.113 C HETATM 25 C 0 -1.504 3.605 2.787 C HETATM 26 C 0 -1.701 4.781 4.780 C HETATM 27 C 0 -2.773 4.013 2.386 C HETATM 28 C 0 -2.975 5.226 4.442 C HETATM 29 C 0 -3.519 4.835 3.224 C HETATM 30 C 0 1.492 1.716 2.023 C HETATM 31 C 0 1.283 1.143 0.772 C HETATM 32 C 0 2.700 1.576 2.863 C HETATM 33 C 0 3.684 2.176 4.880 C HETATM 34 C 0 3.821 0.833 2.504 C HETATM 35 C 0 4.834 1.452 4.585 C HETATM 36 C 0 4.903 0.769 3.376 C HETATM 37 H 0 0.814 3.314 9.305 H HETATM 38 H 0 -0.459 0.443 4.065 H HETATM 39 H 0 -0.117 1.374 8.859 H HETATM 40 H 0 -1.310 -1.244 5.669 H HETATM 41 H 0 -1.136 -0.768 8.128 H HETATM 42 H 0 2.541 6.792 7.430 H HETATM 43 H 0 1.684 4.783 1.712 H HETATM 44 H 0 2.999 7.661 5.448 H HETATM 45 H 0 2.750 6.965 1.215 H HETATM 46 H 0 3.425 8.442 3.130 H HETATM 47 H 0 -1.838 2.393 0.244 H HETATM 48 H 0 -1.248 5.067 5.720 H 121 Table 4.S13 (cont’d) HETATM 49 H 0 -3.176 3.695 1.434 H HETATM 50 H 0 -3.518 5.864 5.126 H HETATM 51 H 0 -4.508 5.162 2.928 H HETATM 52 H 0 2.047 0.526 0.325 H HETATM 53 H 0 3.598 2.719 5.812 H HETATM 54 H 0 3.852 0.311 1.557 H HETATM 55 H 0 5.651 1.431 5.294 H HETATM 56 H 0 5.783 0.196 3.112 H HETATM 57 Fe 0 0.892 3.256 4.326 Fe HETATM 58 S 0 2.072 5.930 10.068 S HETATM 59 S 0 -0.350 0.727 -1.469 S HETATM 60 C 0 1.483 4.758 11.333 C HETATM 61 H 0 1.702 5.244 12.284 H HETATM 62 H 0 0.409 4.599 11.254 H HETATM 63 H 0 2.027 3.816 11.282 H HETATM 64 C 0 1.111 -0.254 -1.944 C HETATM 65 H 0 1.284 -1.068 -1.242 H HETATM 66 H 0 0.859 -0.671 -2.919 H HETATM 67 H 0 1.994 0.376 -2.043 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 30 21 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 122 Table 4.S13 (cont’d) CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 19 CONECT 58 6 60 CONECT 59 24 64 CONECT 60 58 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 59 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 Table 4.S14: The optimized geometry file for the lowest energy quintet state of [Fe(4’-SMe- terpy)2]2+. TITLE Fe-SMe-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.267 4.128 6.257 N HETATM 2 N 0 0.159 1.785 5.808 N HETATM 3 N 0 1.871 5.264 3.960 N HETATM 4 C 0 0.917 3.449 7.361 C HETATM 5 C 0 1.136 3.967 8.634 C HETATM 6 C 0 1.738 5.225 8.761 C HETATM 7 C 0 0.287 2.124 7.108 C HETATM 8 C 0 -0.396 0.610 5.493 C HETATM 9 C 0 -0.147 1.275 8.127 C HETATM 10 C 0 -0.854 -0.288 6.451 C HETATM 11 C 0 -0.725 0.054 7.792 C HETATM 12 C 0 1.847 5.340 6.358 C HETATM 13 C 0 2.097 5.918 7.593 C HETATM 14 C 0 2.187 5.983 5.058 C HETATM 15 C 0 2.144 5.768 2.751 C HETATM 16 C 0 2.790 7.237 4.958 C HETATM 17 C 0 2.744 7.009 2.571 C HETATM 18 C 0 3.071 7.755 3.698 C HETATM 19 N 0 0.518 2.391 2.394 N HETATM 20 N 0 -1.140 3.951 3.715 N HETATM 21 N 0 2.723 2.074 3.793 N HETATM 22 C 0 -0.655 2.628 1.775 C HETATM 23 C 0 -0.935 2.084 0.530 C HETATM 24 C 0 0.024 1.269 -0.095 C HETATM 25 C 0 -1.596 3.503 2.526 C HETATM 26 C 0 -1.922 4.746 4.454 C HETATM 27 C 0 -2.862 3.849 2.055 C HETATM 28 C 0 -3.194 5.138 4.051 C HETATM 29 C 0 -3.670 4.677 2.828 C HETATM 30 C 0 1.448 1.615 1.813 C HETATM 31 C 0 1.236 1.035 0.566 C HETATM 32 C 0 2.695 1.433 2.605 C HETATM 33 C 0 3.805 1.960 4.570 C HETATM 34 C 0 3.771 0.658 2.173 C HETATM 35 C 0 4.916 1.206 4.207 C HETATM 36 C 0 4.895 0.544 2.985 C HETATM 37 H 0 0.846 3.405 9.507 H HETATM 38 H 0 -0.478 0.383 4.436 H HETATM 39 H 0 -0.042 1.552 9.166 H HETATM 40 H 0 -1.297 -1.227 6.146 H HETATM 41 H 0 -1.067 -0.616 8.571 H HETATM 42 H 0 2.562 6.891 7.667 H HETATM 43 H 0 1.873 5.155 1.900 H 123 Table 4.S14 (cont’d) HETATM 44 H 0 3.039 7.806 5.842 H HETATM 45 H 0 2.945 7.374 1.572 H HETATM 46 H 0 3.539 8.727 3.603 H HETATM 47 H 0 -1.877 2.279 0.038 H HETATM 48 H 0 -1.512 5.080 5.401 H HETATM 49 H 0 -3.222 3.484 1.104 H HETATM 50 H 0 -3.788 5.784 4.684 H HETATM 51 H 0 -4.655 4.956 2.477 H HETATM 52 H 0 1.997 0.417 0.117 H HETATM 53 H 0 3.781 2.492 5.514 H HETATM 54 H 0 3.744 0.148 1.220 H HETATM 55 H 0 5.769 1.145 4.871 H HETATM 56 H 0 5.739 -0.054 2.664 H HETATM 57 Fe 0 0.907 3.277 4.315 Fe HETATM 58 S 0 2.086 6.015 10.293 S HETATM 59 S 0 -0.397 0.610 -1.670 S HETATM 60 C 0 1.496 4.834 11.550 C HETATM 61 H 0 1.710 5.316 12.504 H HETATM 62 H 0 0.422 4.672 11.466 H HETATM 63 H 0 2.042 3.894 11.495 H HETATM 64 C 0 1.063 -0.371 -2.149 C HETATM 65 H 0 1.240 -1.183 -1.446 H HETATM 66 H 0 0.807 -0.790 -3.122 H HETATM 67 H 0 1.945 0.260 -2.253 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 30 21 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 124 Table 4.S14 (cont’d) CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 60 CONECT 59 24 64 CONECT 60 58 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 59 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 [Fe(4’-SO2Me-terpy)2]2+: Table 4.S15: The optimized geometry file for the lowest energy singlet state of [Fe(4’-SO2Me- terpy)2]2+. TITLE Fe-SO2Me-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.977 3.969 6.145 N HETATM 2 N 0 0.062 1.639 5.403 N HETATM 3 N 0 1.589 5.103 3.873 N HETATM 4 C 0 0.578 3.182 7.244 C HETATM 5 C 0 0.681 3.693 8.592 C HETATM 6 C 0 1.212 5.021 8.763 C HETATM 7 C 0 0.052 1.824 6.812 C HETATM 8 C 0 -0.401 0.424 4.894 C HETATM 9 C 0 -0.416 0.795 7.712 C HETATM 10 C 0 -0.884 -0.639 5.739 C HETATM 11 C 0 -0.891 -0.453 7.170 C HETATM 12 C 0 1.494 5.268 6.324 C HETATM 13 C 0 1.625 5.833 7.647 C HETATM 14 C 0 1.856 5.925 5.002 C HETATM 15 C 0 1.880 5.611 2.605 C HETATM 16 C 0 2.417 7.249 4.865 C HETATM 17 C 0 2.439 6.923 2.403 C HETATM 18 C 0 2.714 7.756 3.549 C HETATM 19 N 0 0.611 2.534 2.529 N HETATM 20 N 0 -1.117 3.921 3.913 N HETATM 21 N 0 2.646 2.351 4.155 N HETATM 22 C 0 -0.583 2.756 1.816 C HETATM 23 C 0 -0.750 2.235 0.478 C HETATM 24 C 0 0.345 1.491 -0.089 C HETATM 25 C 0 -1.587 3.565 2.620 C HETATM 26 C 0 -1.960 4.668 4.737 C HETATM 27 C 0 -2.896 3.956 2.151 C HETATM 28 C 0 -3.277 5.087 4.328 C HETATM 29 C 0 -3.754 4.726 3.015 C HETATM 30 C 0 1.674 1.797 1.968 C HETATM 31 C 0 1.566 1.249 0.635 C HETATM 32 C 0 2.856 1.687 2.916 C HETATM 33 C 0 3.674 2.324 5.099 C HETATM 34 C 0 4.087 0.992 2.624 C HETATM 35 C 0 4.927 1.651 4.866 C HETATM 36 C 0 5.138 0.971 3.611 C HETATM 37 H 0 0.346 3.086 9.481 H 125 Table 4.S15 (cont’d) HETATM 38 H 0 -0.381 0.307 3.773 H HETATM 39 H 0 -0.407 0.968 8.826 H HETATM 40 H 0 -1.246 -1.598 5.266 H HETATM 41 H 0 -1.261 -1.269 7.857 H HETATM 42 H 0 2.022 6.876 7.808 H HETATM 43 H 0 1.653 4.938 1.730 H HETATM 44 H 0 2.619 7.878 5.779 H HETATM 45 H 0 2.652 7.276 1.352 H HETATM 46 H 0 3.153 8.788 3.422 H HETATM 47 H 0 -1.703 2.392 -0.104 H HETATM 48 H 0 -1.560 4.937 5.756 H HETATM 49 H 0 -3.240 3.660 1.118 H HETATM 50 H 0 -3.911 5.689 5.042 H HETATM 51 H 0 -4.781 5.039 2.666 H HETATM 52 H 0 2.401 0.645 0.177 H HETATM 53 H 0 3.479 2.863 6.070 H HETATM 54 H 0 4.224 0.470 1.633 H HETATM 55 H 0 5.719 1.666 5.670 H HETATM 56 H 0 6.107 0.432 3.400 H HETATM 57 Fe 0 0.795 3.253 4.337 Fe HETATM 58 S 0 1.331 5.755 10.529 S HETATM 59 S 0 0.155 0.767 -1.854 S HETATM 60 O 0 1.311 7.289 10.379 O HETATM 61 O 0 0.255 5.053 11.383 O HETATM 62 O 0 1.031 -0.500 -1.919 O HETATM 63 O 0 -1.357 0.671 -2.140 O HETATM 64 C 0 0.911 2.089 -2.989 C HETATM 65 H 0 0.320 3.048 -2.850 H HETATM 66 H 0 2.004 2.209 -2.706 H HETATM 67 H 0 0.795 1.676 -4.042 H HETATM 68 C 0 3.044 5.224 11.148 C HETATM 69 H 0 3.138 5.674 12.187 H HETATM 70 H 0 3.066 4.090 11.179 H HETATM 71 H 0 3.818 5.650 10.435 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 126 Table 4.S15 (cont’d) CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 19 1 CONECT 58 60 61 CONECT 59 62 63 CONECT 60 58 CONECT 61 58 CONECT 62 59 CONECT 63 59 Table 4.S16: The optimized geometry file for the lowest energy quintet state of [Fe(4’-SO2Me- terpy)2]2+. TITLE Fe-SO2Me-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.980 4.040 6.360 N HETATM 2 N 0 0.073 1.562 5.712 N HETATM 3 N 0 1.634 5.338 4.064 N HETATM 4 C 0 0.613 3.247 7.456 C HETATM 5 C 0 0.727 3.747 8.811 C HETATM 6 C 0 1.217 5.087 8.983 C HETATM 7 C 0 0.077 1.855 7.090 C HETATM 8 C 0 -0.428 0.333 5.289 C HETATM 9 C 0 -0.412 0.905 8.067 C HETATM 10 C 0 -0.938 -0.658 6.203 C HETATM 11 C 0 -0.924 -0.366 7.616 C HETATM 12 C 0 1.460 5.344 6.542 C HETATM 13 C 0 1.583 5.916 7.867 C HETATM 14 C 0 1.837 6.077 5.247 C HETATM 15 C 0 1.976 5.918 2.845 C HETATM 16 C 0 2.378 7.419 5.216 C HETATM 17 C 0 2.527 7.245 2.738 C HETATM 18 C 0 2.727 8.009 3.947 C HETATM 19 N 0 0.610 2.462 2.316 N HETATM 20 N 0 -1.250 3.915 3.660 N HETATM 21 N 0 2.763 2.120 3.940 N HETATM 22 C 0 -0.565 2.715 1.596 C HETATM 23 C 0 -0.740 2.207 0.251 C HETATM 24 C 0 0.330 1.427 -0.307 C HETATM 25 C 0 -1.620 3.533 2.355 C HETATM 26 C 0 -2.161 4.633 4.432 C HETATM 27 C 0 -2.910 3.884 1.801 C HETATM 28 C 0 -3.467 5.008 3.950 C HETATM 29 C 0 -3.843 4.631 2.609 C HETATM 30 C 0 1.646 1.698 1.762 C HETATM 31 C 0 1.532 1.144 0.428 C HETATM 32 C 0 2.872 1.522 2.669 C 127 Table 4.S16 (cont’d) HETATM 33 C 0 3.844 2.037 4.816 C HETATM 34 C 0 4.070 0.816 2.266 C HETATM 35 C 0 5.069 1.356 4.480 C HETATM 36 C 0 5.180 0.732 3.184 C HETATM 37 H 0 0.433 3.126 9.704 H HETATM 38 H 0 -0.410 0.146 4.175 H HETATM 39 H 0 -0.403 1.148 9.167 H HETATM 40 H 0 -1.332 -1.639 5.806 H HETATM 41 H 0 -1.310 -1.118 8.365 H HETATM 42 H 0 1.945 6.970 8.031 H HETATM 43 H 0 1.792 5.286 1.927 H HETATM 44 H 0 2.531 8.005 6.167 H HETATM 45 H 0 2.787 7.666 1.724 H HETATM 46 H 0 3.150 9.055 3.906 H HETATM 47 H 0 -1.680 2.395 -0.341 H HETATM 48 H 0 -1.818 4.915 5.470 H HETATM 49 H 0 -3.193 3.579 0.753 H HETATM 50 H 0 -4.166 5.590 4.619 H HETATM 51 H 0 -4.855 4.911 2.192 H HETATM 52 H 0 2.344 0.506 -0.024 H HETATM 53 H 0 3.707 2.532 5.822 H HETATM 54 H 0 4.147 0.340 1.247 H HETATM 55 H 0 5.913 1.318 5.229 H HETATM 56 H 0 6.121 0.185 2.885 H HETATM 57 Fe 0 0.812 3.229 4.345 Fe HETATM 58 S 0 1.350 5.808 10.757 S HETATM 59 S 0 0.132 0.706 -2.076 S HETATM 60 O 0 1.210 7.339 10.632 O HETATM 61 O 0 0.367 5.017 11.642 O HETATM 62 O 0 0.990 -0.574 -2.140 O HETATM 63 O 0 -1.381 0.630 -2.365 O HETATM 64 C 0 0.907 2.020 -3.205 C HETATM 65 H 0 0.326 2.985 -3.066 H HETATM 66 H 0 2.000 2.127 -2.917 H HETATM 67 H 0 0.791 1.612 -4.260 H HETATM 68 C 0 3.124 5.398 11.291 C HETATM 69 H 0 3.227 5.828 12.338 H HETATM 70 H 0 3.235 4.269 11.288 H HETATM 71 H 0 3.831 5.904 10.560 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 17 16 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 128 Table 4.S16 (cont’d) CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 60 61 CONECT 59 62 63 CONECT 60 58 CONECT 61 58 CONECT 62 59 CONECT 63 59 [Fe(4’-OMe-terpy)2]2+: Table 4.S17: The optimized geometry file for the lowest energy singlet state of [Fe(4’-OMe- terpy)2]2+. TITLE Fe-OMe-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.078 3.784 6.195 N HETATM 2 N 0 -0.009 1.584 5.410 N HETATM 3 N 0 1.757 4.880 3.971 N HETATM 4 C 0 0.628 3.023 7.266 C HETATM 5 C 0 0.788 3.474 8.600 C HETATM 6 C 0 1.432 4.743 8.826 C HETATM 7 C 0 -0.007 1.738 6.804 C HETATM 8 C 0 -0.563 0.446 4.873 C HETATM 9 C 0 -0.562 0.750 7.667 C HETATM 10 C 0 -1.135 -0.580 5.680 C HETATM 11 C 0 -1.135 -0.426 7.099 C HETATM 12 C 0 1.694 5.002 6.399 C HETATM 13 C 0 1.894 5.524 7.711 C HETATM 14 C 0 2.092 5.645 5.097 C HETATM 15 C 0 2.077 5.366 2.724 C HETATM 16 C 0 2.748 6.905 4.978 C HETATM 17 C 0 2.733 6.617 2.536 C HETATM 18 C 0 3.073 7.399 3.681 C HETATM 19 N 0 0.586 2.459 2.585 N HETATM 20 N 0 -0.992 3.924 3.999 N HETATM 21 N 0 2.574 2.103 4.177 N HETATM 22 C 0 -0.581 2.767 1.899 C HETATM 23 C 0 -0.802 2.290 0.583 C HETATM 24 C 0 0.212 1.472 -0.033 C HETATM 25 C 0 -1.499 3.626 2.727 C HETATM 26 C 0 -1.747 4.704 4.844 C 129 Table 4.S17 (cont’d) HETATM 27 C 0 -2.767 4.109 2.293 C HETATM 28 C 0 -3.023 5.219 4.473 C HETATM 29 C 0 -3.541 4.918 3.177 C HETATM 30 C 0 1.564 1.677 2.006 C HETATM 31 C 0 1.417 1.158 0.687 C HETATM 32 C 0 2.731 1.472 2.935 C HETATM 33 C 0 3.580 1.985 5.108 C HETATM 34 C 0 3.900 0.719 2.620 C HETATM 35 C 0 4.773 1.248 4.857 C HETATM 36 C 0 4.936 0.604 3.593 C HETATM 37 H 0 0.433 2.876 9.468 H HETATM 38 H 0 -0.539 0.370 3.764 H HETATM 39 H 0 -0.545 0.903 8.769 H HETATM 40 H 0 -1.570 -1.482 5.193 H HETATM 41 H 0 -1.574 -1.210 7.756 H HETATM 42 H 0 2.393 6.505 7.853 H HETATM 43 H 0 1.793 4.721 1.864 H HETATM 44 H 0 2.999 7.489 5.892 H HETATM 45 H 0 2.968 6.963 1.505 H HETATM 46 H 0 3.585 8.382 3.567 H HETATM 47 H 0 -1.731 2.529 0.022 H HETATM 48 H 0 -1.302 4.912 5.841 H HETATM 49 H 0 -3.139 3.854 1.276 H HETATM 50 H 0 -3.592 5.846 5.195 H HETATM 51 H 0 -4.535 5.306 2.858 H HETATM 52 H 0 2.216 0.531 0.240 H HETATM 53 H 0 3.409 2.503 6.078 H HETATM 54 H 0 3.993 0.230 1.625 H HETATM 55 H 0 5.556 1.185 5.645 H HETATM 56 H 0 5.856 0.020 3.366 H HETATM 57 Fe 0 0.833 3.123 4.390 Fe HETATM 58 O 0 1.555 5.119 10.148 O HETATM 59 O 0 -0.069 1.039 -1.313 O HETATM 60 C 0 2.211 6.422 10.440 C HETATM 61 H 0 2.195 6.508 11.555 H HETATM 62 H 0 3.270 6.416 10.060 H HETATM 63 H 0 1.629 7.263 9.971 H HETATM 64 C 0 0.943 0.187 -1.994 C HETATM 65 H 0 0.499 -0.039 -2.996 H HETATM 66 H 0 1.910 0.750 -2.109 H HETATM 67 H 0 1.105 -0.762 -1.411 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 130 Table 4.S17 (cont’d) CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 19 1 CONECT 58 6 60 CONECT 59 24 64 CONECT 60 58 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 59 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 Table 4.S18: The optimized geometry file for the lowest energy quintet state of [Fe(4’-OMe- terpy)2]2+. TITLE Fe-OMe-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.836 4.177 6.258 N HETATM 2 N 0 0.094 1.636 5.951 N HETATM 3 N 0 1.813 5.045 3.928 N HETATM 4 C 0 0.329 3.598 7.400 C HETATM 5 C 0 0.231 4.312 8.629 C HETATM 6 C 0 0.683 5.676 8.655 C HETATM 7 C 0 -0.106 2.151 7.223 C HETATM 8 C 0 -0.261 0.335 5.693 C HETATM 9 C 0 -0.678 1.363 8.268 C HETATM 10 C 0 -0.834 -0.515 6.684 C HETATM 11 C 0 -1.046 0.012 7.993 C HETATM 12 C 0 1.280 5.486 6.279 C HETATM 13 C 0 1.216 6.271 7.460 C HETATM 14 C 0 1.826 5.982 4.951 C HETATM 15 C 0 2.291 5.409 2.694 C HETATM 16 C 0 2.323 7.307 4.752 C HETATM 17 C 0 2.804 6.710 2.416 C HETATM 18 C 0 2.819 7.675 3.467 C HETATM 19 N 0 0.926 2.354 2.335 N HETATM 20 N 0 -1.037 3.638 3.603 N HETATM 21 N 0 2.903 1.957 4.082 N 131 Table 4.S18 (cont’d) HETATM 22 C 0 -0.179 2.602 1.555 C HETATM 23 C 0 -0.250 2.175 0.195 C HETATM 24 C 0 0.874 1.469 -0.352 C HETATM 25 C 0 -1.291 3.352 2.269 C HETATM 26 C 0 -1.989 4.310 4.329 C HETATM 27 C 0 -2.512 3.741 1.636 C HETATM 28 C 0 -3.232 4.729 3.770 C HETATM 29 C 0 -3.495 4.439 2.398 C HETATM 30 C 0 2.007 1.664 1.819 C HETATM 31 C 0 2.023 1.207 0.477 C HETATM 32 C 0 3.140 1.453 2.811 C HETATM 33 C 0 3.868 1.811 5.048 C HETATM 34 C 0 4.361 0.790 2.484 C HETATM 35 C 0 5.111 1.160 4.799 C HETATM 36 C 0 5.360 0.642 3.493 C HETATM 37 H 0 -0.181 3.825 9.536 H HETATM 38 H 0 -0.074 -0.025 4.655 H HETATM 39 H 0 -0.835 1.792 9.282 H HETATM 40 H 0 -1.105 -1.564 6.428 H HETATM 41 H 0 -1.493 -0.620 8.795 H HETATM 42 H 0 1.567 7.325 7.490 H HETATM 43 H 0 2.254 4.622 1.906 H HETATM 44 H 0 2.324 8.043 5.586 H HETATM 45 H 0 3.181 6.952 1.397 H HETATM 46 H 0 3.211 8.703 3.291 H HETATM 47 H 0 -1.151 2.382 -0.417 H HETATM 48 H 0 -1.733 4.517 5.393 H HETATM 49 H 0 -2.695 3.506 0.565 H HETATM 50 H 0 -3.970 5.271 4.403 H HETATM 51 H 0 -4.453 4.751 1.923 H HETATM 52 H 0 2.886 0.653 0.049 H HETATM 53 H 0 3.623 2.234 6.050 H HETATM 54 H 0 4.534 0.395 1.459 H HETATM 55 H 0 5.862 1.066 5.616 H HETATM 56 H 0 6.319 0.127 3.258 H HETATM 57 Fe 0 0.961 3.016 4.415 Fe HETATM 58 O 0 0.651 6.488 9.765 O HETATM 59 O 0 0.951 0.999 -1.642 O HETATM 60 C 0 -0.213 1.238 -2.540 C HETATM 61 H 0 0.079 0.775 -3.516 H HETATM 62 H 0 -1.130 0.737 -2.124 H HETATM 63 H 0 -0.386 2.343 -2.662 H HETATM 64 C 0 0.121 5.914 11.034 C HETATM 65 H 0 0.753 5.040 11.355 H HETATM 66 H 0 0.197 6.747 11.777 H HETATM 67 H 0 -0.950 5.601 10.899 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 132 Table 4.S18 (cont’d) CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 64 CONECT 59 24 60 CONECT 60 59 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 58 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 [Fe(4’-N(EtOH)2-terpy)2]2+: Table 4.S19: The optimized geometry file for the lowest energy singlet state of [Fe(4’-N(EtOH)2- terpy)2]2+. TITLE Fe-N(EtOH)2-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.281 4.154 5.996 N HETATM 2 N 0 0.843 1.760 5.250 N HETATM 3 N 0 1.247 5.492 3.832 N HETATM 4 C 0 1.256 3.296 7.032 C HETATM 5 C 0 1.450 3.724 8.334 C HETATM 6 C 0 1.697 5.098 8.597 C HETATM 7 C 0 1.003 1.904 6.595 C HETATM 8 C 0 0.612 0.536 4.758 C HETATM 9 C 0 0.932 0.812 7.457 C HETATM 10 C 0 0.528 -0.596 5.560 C HETATM 11 C 0 0.691 -0.454 6.934 C HETATM 12 C 0 1.494 5.465 6.207 C HETATM 13 C 0 1.699 5.973 7.478 C HETATM 14 C 0 1.477 6.241 4.946 C HETATM 15 C 0 1.213 6.104 2.641 C 133 Table 4.S19 (cont’d) HETATM 16 C 0 1.676 7.617 4.868 C HETATM 17 C 0 1.402 7.474 2.492 C HETATM 18 C 0 1.638 8.243 3.627 C HETATM 19 N 0 0.709 2.864 2.420 N HETATM 20 N 0 -1.033 3.676 4.088 N HETATM 21 N 0 2.914 3.094 3.670 N HETATM 22 C 0 -0.538 2.809 1.919 C HETATM 23 C 0 -0.785 2.354 0.635 C HETATM 24 C 0 0.296 1.940 -0.189 C HETATM 25 C 0 -1.550 3.281 2.892 C HETATM 26 C 0 -1.870 4.118 5.036 C HETATM 27 C 0 -2.919 3.329 2.640 C HETATM 28 C 0 -3.246 4.192 4.851 C HETATM 29 C 0 -3.779 3.790 3.631 C HETATM 30 C 0 1.758 2.467 1.676 C HETATM 31 C 0 1.595 1.999 0.384 C HETATM 32 C 0 3.039 2.600 2.406 C HETATM 33 C 0 4.019 3.253 4.410 C HETATM 34 C 0 4.283 2.261 1.878 C HETATM 35 C 0 5.290 2.936 3.945 C HETATM 36 C 0 5.423 2.431 2.656 C HETATM 37 H 0 1.411 3.006 9.136 H HETATM 38 H 0 0.491 0.465 3.685 H HETATM 39 H 0 1.064 0.945 8.522 H HETATM 40 H 0 0.339 -1.561 5.109 H HETATM 41 H 0 0.633 -1.314 7.590 H HETATM 42 H 0 1.854 7.032 7.604 H HETATM 43 H 0 1.028 5.474 1.781 H HETATM 44 H 0 1.859 8.195 5.764 H HETATM 45 H 0 1.365 7.916 1.506 H HETATM 46 H 0 1.791 9.312 3.550 H HETATM 47 H 0 -1.800 2.313 0.276 H HETATM 48 H 0 -1.420 4.422 5.972 H HETATM 49 H 0 -3.312 3.013 1.683 H HETATM 50 H 0 -3.876 4.557 5.652 H HETATM 51 H 0 -4.845 3.834 3.451 H HETATM 52 H 0 2.459 1.682 -0.175 H HETATM 53 H 0 3.876 3.649 5.407 H HETATM 54 H 0 4.363 1.871 0.873 H HETATM 55 H 0 6.149 3.086 4.586 H HETATM 56 H 0 6.397 2.172 2.260 H HETATM 57 Fe 0 0.994 3.506 4.209 Fe HETATM 58 N 0 1.930 5.555 9.865 N HETATM 59 N 0 0.097 1.515 -1.473 N HETATM 60 C 0 1.947 4.640 11.010 C HETATM 61 H 0 2.564 5.090 11.788 H HETATM 62 H 0 2.434 3.702 10.734 H HETATM 63 C 0 2.231 6.967 10.119 C HETATM 64 H 0 2.807 7.025 11.043 H HETATM 65 H 0 2.873 7.361 9.328 H HETATM 66 C 0 0.986 7.847 10.251 C HETATM 67 H 0 0.401 7.544 11.126 H HETATM 68 H 0 0.348 7.748 9.365 H HETATM 69 C 0 0.558 4.345 11.581 C HETATM 70 H 0 0.114 5.262 11.983 H HETATM 71 H 0 -0.106 3.960 10.799 H HETATM 72 C 0 1.220 1.095 -2.316 C HETATM 73 H 0 0.920 1.217 -3.357 H HETATM 74 H 0 2.076 1.756 -2.161 H HETATM 75 C 0 -1.241 1.508 -2.072 C HETATM 76 H 0 -1.124 1.599 -3.152 H HETATM 77 H 0 -1.801 2.387 -1.747 H HETATM 78 C 0 1.648 -0.357 -2.087 C HETATM 79 H 0 0.837 -1.037 -2.369 H HETATM 80 H 0 1.883 -0.527 -1.030 H HETATM 81 C 0 -2.046 0.245 -1.760 C HETATM 82 H 0 -1.554 -0.632 -2.195 H HETATM 83 H 0 -2.120 0.092 -0.678 H 134 Table 4.S19 (cont’d) HETATM 84 O 0 2.799 -0.575 -2.904 O HETATM 85 H 0 3.064 -1.498 -2.825 H HETATM 86 O 0 -3.340 0.437 -2.335 O HETATM 87 H 0 -3.860 -0.363 -2.201 H HETATM 88 O 0 0.741 3.376 12.614 O HETATM 89 H 0 -0.109 3.205 13.032 H HETATM 90 O 0 1.452 9.190 10.393 O HETATM 91 H 0 0.694 9.769 10.524 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 19 1 CONECT 58 6 60 63 CONECT 59 24 72 75 135 Table 4.S19 (cont’d) CONECT 60 58 61 62 69 CONECT 61 60 CONECT 62 60 CONECT 63 58 64 65 66 CONECT 64 63 CONECT 65 63 CONECT 66 63 67 68 90 CONECT 67 66 CONECT 68 66 CONECT 69 60 70 71 88 CONECT 70 69 CONECT 71 69 CONECT 72 59 73 74 78 CONECT 73 72 CONECT 74 72 CONECT 75 59 76 77 81 CONECT 76 75 CONECT 77 75 CONECT 78 72 79 80 84 CONECT 79 78 CONECT 80 78 CONECT 81 75 82 83 86 CONECT 82 81 CONECT 83 81 CONECT 84 78 85 CONECT 85 84 CONECT 86 81 87 CONECT 87 86 CONECT 88 69 89 CONECT 89 88 CONECT 90 66 91 CONECT 91 90 Table 4.S20: The optimized geometry file for the lowest energy quintet state of [Fe(4’-N(EtOH)2- terpy)2]2+. TITLE Fe-N(EtOH)2-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.318 4.231 6.190 N HETATM 2 N 0 0.879 1.724 5.548 N HETATM 3 N 0 1.316 5.727 4.033 N HETATM 4 C 0 1.292 3.380 7.234 C HETATM 5 C 0 1.476 3.807 8.537 C HETATM 6 C 0 1.721 5.178 8.809 C HETATM 7 C 0 1.048 1.956 6.867 C HETATM 8 C 0 0.655 0.474 5.128 C HETATM 9 C 0 0.995 0.917 7.798 C HETATM 10 C 0 0.587 -0.611 5.994 C HETATM 11 C 0 0.762 -0.381 7.355 C HETATM 12 C 0 1.530 5.542 6.416 C HETATM 13 C 0 1.726 6.048 7.688 C HETATM 14 C 0 1.532 6.389 5.189 C HETATM 15 C 0 1.301 6.409 2.882 C HETATM 16 C 0 1.740 7.769 5.206 C HETATM 17 C 0 1.499 7.784 2.819 C HETATM 18 C 0 1.722 8.473 4.006 C HETATM 19 N 0 0.686 2.798 2.220 N HETATM 20 N 0 -1.203 3.621 3.845 N HETATM 21 N 0 3.027 2.985 3.391 N HETATM 22 C 0 -0.559 2.736 1.710 C HETATM 23 C 0 -0.810 2.272 0.432 C HETATM 24 C 0 0.263 1.851 -0.396 C HETATM 25 C 0 -1.631 3.206 2.634 C HETATM 26 C 0 -2.096 4.063 4.737 C HETATM 27 C 0 -2.985 3.228 2.298 C HETATM 28 C 0 -3.460 4.115 4.476 C HETATM 29 C 0 -3.909 3.687 3.230 C HETATM 30 C 0 1.726 2.392 1.465 C 136 Table 4.S20 (cont’d) HETATM 31 C 0 1.560 1.915 0.178 C HETATM 32 C 0 3.055 2.503 2.131 C HETATM 33 C 0 4.174 3.121 4.066 C HETATM 34 C 0 4.257 2.145 1.520 C HETATM 35 C 0 5.410 2.785 3.524 C HETATM 36 C 0 5.447 2.287 2.226 C HETATM 37 H 0 1.432 3.090 9.339 H HETATM 38 H 0 0.527 0.341 4.060 H HETATM 39 H 0 1.135 1.106 8.853 H HETATM 40 H 0 0.404 -1.605 5.606 H HETATM 41 H 0 0.719 -1.198 8.065 H HETATM 42 H 0 1.880 7.106 7.818 H HETATM 43 H 0 1.124 5.831 1.983 H HETATM 44 H 0 1.915 8.295 6.134 H HETATM 45 H 0 1.477 8.291 1.864 H HETATM 46 H 0 1.882 9.544 4.003 H HETATM 47 H 0 -1.824 2.228 0.075 H HETATM 48 H 0 -1.700 4.387 5.693 H HETATM 49 H 0 -3.323 2.897 1.326 H HETATM 50 H 0 -4.145 4.480 5.230 H HETATM 51 H 0 -4.963 3.711 2.984 H HETATM 52 H 0 2.421 1.593 -0.382 H HETATM 53 H 0 4.097 3.512 5.074 H HETATM 54 H 0 4.276 1.761 0.509 H HETATM 55 H 0 6.311 2.912 4.109 H HETATM 56 H 0 6.388 2.014 1.765 H HETATM 57 Fe 0 1.007 3.514 4.204 Fe HETATM 58 N 0 1.945 5.632 10.074 N HETATM 59 N 0 0.062 1.419 -1.673 N HETATM 60 C 0 1.959 4.715 11.220 C HETATM 61 H 0 2.574 5.165 12.000 H HETATM 62 H 0 2.447 3.778 10.944 H HETATM 63 C 0 2.236 7.046 10.335 C HETATM 64 H 0 2.808 7.104 11.261 H HETATM 65 H 0 2.880 7.445 9.548 H HETATM 66 C 0 0.985 7.917 10.463 C HETATM 67 H 0 0.396 7.607 11.333 H HETATM 68 H 0 0.354 7.818 9.572 H HETATM 69 C 0 0.568 4.420 11.786 C HETATM 70 H 0 0.122 5.337 12.186 H HETATM 71 H 0 -0.093 4.034 11.001 H HETATM 72 C 0 1.182 0.991 -2.517 C HETATM 73 H 0 0.879 1.107 -3.558 H HETATM 74 H 0 2.038 1.654 -2.368 H HETATM 75 C 0 -1.279 1.404 -2.269 C HETATM 76 H 0 -1.164 1.488 -3.349 H HETATM 77 H 0 -1.840 2.283 -1.947 H HETATM 78 C 0 1.610 -0.459 -2.279 C HETATM 79 H 0 0.798 -1.141 -2.551 H HETATM 80 H 0 1.851 -0.621 -1.222 H HETATM 81 C 0 -2.079 0.139 -1.947 C HETATM 82 H 0 -1.584 -0.739 -2.375 H HETATM 83 H 0 -2.151 -0.006 -0.863 H HETATM 84 O 0 2.757 -0.683 -3.101 O HETATM 85 H 0 3.021 -1.605 -3.018 H HETATM 86 O 0 -3.373 0.323 -2.522 O HETATM 87 H 0 -3.890 -0.478 -2.383 H HETATM 88 O 0 0.750 3.451 12.819 O HETATM 89 H 0 -0.101 3.281 13.236 H HETATM 90 O 0 1.443 9.261 10.614 O HETATM 91 H 0 0.680 9.835 10.745 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 137 Table 4.S20 (cont’d) CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 60 63 CONECT 59 24 72 75 CONECT 60 58 61 62 69 CONECT 61 60 CONECT 62 60 CONECT 63 58 64 65 66 CONECT 64 63 CONECT 65 63 CONECT 66 63 67 68 90 CONECT 67 66 CONECT 68 66 CONECT 69 60 70 71 88 CONECT 70 69 CONECT 71 69 CONECT 72 59 73 74 78 CONECT 73 72 CONECT 74 72 CONECT 75 59 76 77 81 138 Table 4.S20 (cont’d) CONECT 76 75 CONECT 77 75 CONECT 78 72 79 80 84 CONECT 79 78 CONECT 80 78 CONECT 81 75 82 83 86 CONECT 82 81 CONECT 83 81 CONECT 84 78 85 CONECT 85 84 CONECT 86 81 87 CONECT 87 86 CONECT 88 69 89 CONECT 89 88 CONECT 90 66 91 CONECT 91 90 [Fe(4’-CN-terpy)2]2+: Table 4.S21: The optimized geometry file for the lowest energy singlet state of [Fe(4’-CN- terpy)2]2+. TITLE Fe-CN-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.003 3.972 6.137 N HETATM 2 N 0 0.085 1.656 5.355 N HETATM 3 N 0 1.732 5.099 3.899 N HETATM 4 C 0 0.543 3.191 7.218 C HETATM 5 C 0 0.613 3.688 8.568 C HETATM 6 C 0 1.166 5.014 8.791 C HETATM 7 C 0 0.009 1.844 6.762 C HETATM 8 C 0 -0.374 0.448 4.825 C HETATM 9 C 0 -0.528 0.829 7.637 C HETATM 10 C 0 -0.919 -0.603 5.646 C HETATM 11 C 0 -1.000 -0.412 7.074 C HETATM 12 C 0 1.543 5.258 6.345 C HETATM 13 C 0 1.638 5.812 7.671 C HETATM 14 C 0 1.970 5.912 5.041 C HETATM 15 C 0 2.083 5.604 2.646 C HETATM 16 C 0 2.562 7.225 4.930 C HETATM 17 C 0 2.674 6.906 2.470 C HETATM 18 C 0 2.919 7.730 3.628 C HETATM 19 N 0 0.771 2.550 2.510 N HETATM 20 N 0 -0.994 3.954 3.823 N HETATM 21 N 0 2.734 2.333 4.217 N HETATM 22 C 0 -0.391 2.786 1.747 C HETATM 23 C 0 -0.511 2.270 0.407 C HETATM 24 C 0 0.597 1.504 -0.141 C HETATM 25 C 0 -1.416 3.608 2.510 C HETATM 26 C 0 -1.860 4.709 4.617 C HETATM 27 C 0 -2.700 4.020 1.991 C HETATM 28 C 0 -3.154 5.146 4.157 C HETATM 29 C 0 -3.582 4.798 2.824 C HETATM 30 C 0 1.850 1.809 1.985 C HETATM 31 C 0 1.792 1.268 0.650 C HETATM 32 C 0 2.992 1.683 2.979 C HETATM 33 C 0 3.718 2.285 5.205 C HETATM 34 C 0 4.231 0.984 2.732 C HETATM 35 C 0 4.975 1.605 5.019 C HETATM 36 C 0 5.237 0.943 3.764 C HETATM 37 H 0 0.249 3.071 9.437 H HETATM 38 H 0 -0.298 0.327 3.708 H HETATM 39 H 0 -0.578 1.006 8.750 H HETATM 40 H 0 -1.274 -1.556 5.158 H HETATM 41 H 0 -1.423 -1.217 7.743 H HETATM 42 H 0 2.066 6.839 7.847 H HETATM 43 H 0 1.881 4.938 1.759 H HETATM 44 H 0 2.742 7.848 5.853 H 139 Table 4.S21 (cont’d) HETATM 45 H 0 2.935 7.258 1.430 H HETATM 46 H 0 3.383 8.754 3.523 H HETATM 47 H 0 -1.437 2.450 -0.208 H HETATM 48 H 0 -1.497 4.967 5.652 H HETATM 49 H 0 -3.007 3.735 0.944 H HETATM 50 H 0 -3.808 5.753 4.848 H HETATM 51 H 0 -4.589 5.128 2.435 H HETATM 52 H 0 2.649 0.675 0.223 H HETATM 53 H 0 3.484 2.811 6.174 H HETATM 54 H 0 4.410 0.477 1.740 H HETATM 55 H 0 5.730 1.601 5.857 H HETATM 56 H 0 6.212 0.401 3.587 H HETATM 57 Fe 0 0.888 3.260 4.324 Fe HETATM 58 C 0 1.249 5.554 10.166 C HETATM 59 C 0 0.506 0.962 -1.516 C HETATM 60 N 0 1.316 5.992 11.282 N HETATM 61 N 0 0.432 0.523 -2.631 N END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 140 Table 4.S21 (cont’d) CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 19 CONECT 58 6 60 CONECT 59 24 61 CONECT 60 58 CONECT 61 59 Table 4.S22: The optimized geometry file for the lowest energy quintet state of [Fe(4’-CN- terpy)2]2+. TITLE Fe-CN-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.012 4.062 6.339 N HETATM 2 N 0 0.013 1.639 5.642 N HETATM 3 N 0 1.871 5.309 4.088 N HETATM 4 C 0 0.528 3.300 7.413 C HETATM 5 C 0 0.579 3.801 8.767 C HETATM 6 C 0 1.162 5.110 8.995 C HETATM 7 C 0 -0.027 1.923 7.021 C HETATM 8 C 0 -0.443 0.399 5.198 C HETATM 9 C 0 -0.540 0.964 7.976 C HETATM 10 C 0 -0.959 -0.607 6.092 C HETATM 11 C 0 -1.011 -0.315 7.504 C HETATM 12 C 0 1.572 5.330 6.553 C HETATM 13 C 0 1.671 5.889 7.881 C HETATM 14 C 0 2.036 6.048 5.276 C HETATM 15 C 0 2.236 5.895 2.877 C HETATM 16 C 0 2.584 7.387 5.264 C HETATM 17 C 0 2.782 7.226 2.788 C HETATM 18 C 0 2.963 7.981 4.005 C HETATM 19 N 0 0.764 2.464 2.307 N HETATM 20 N 0 -1.156 3.888 3.594 N HETATM 21 N 0 2.892 2.204 3.971 N HETATM 22 C 0 -0.403 2.679 1.559 C HETATM 23 C 0 -0.530 2.171 0.213 C HETATM 24 C 0 0.587 1.441 -0.356 C HETATM 25 C 0 -1.487 3.494 2.282 C HETATM 26 C 0 -2.069 4.654 4.315 C HETATM 27 C 0 -2.751 3.857 1.680 C HETATM 28 C 0 -3.343 5.060 3.777 C HETATM 29 C 0 -3.689 4.649 2.438 C HETATM 30 C 0 1.844 1.752 1.766 C HETATM 31 C 0 1.793 1.226 0.422 C HETATM 32 C 0 3.039 1.581 2.716 C HETATM 33 C 0 3.917 2.077 4.907 C HETATM 34 C 0 4.227 0.827 2.380 C HETATM 35 C 0 5.126 1.338 4.644 C HETATM 36 C 0 5.283 0.706 3.356 C HETATM 37 H 0 0.186 3.201 9.634 H HETATM 38 H 0 -0.395 0.223 4.084 H HETATM 39 H 0 -0.572 1.202 9.077 H HETATM 40 H 0 -1.317 -1.595 5.679 H HETATM 41 H 0 -1.415 -1.074 8.236 H HETATM 42 H 0 2.126 6.902 8.065 H HETATM 43 H 0 2.083 5.263 1.954 H HETATM 44 H 0 2.712 7.971 6.220 H HETATM 45 H 0 3.063 7.652 1.781 H HETATM 46 H 0 3.392 9.025 3.978 H HETATM 47 H 0 -1.464 2.333 -0.396 H HETATM 48 H 0 -1.759 4.941 5.363 H HETATM 49 H 0 -3.009 3.533 0.632 H HETATM 50 H 0 -4.046 5.681 4.405 H HETATM 51 H 0 -4.680 4.942 1.982 H 141 Table 4.S22 (cont’d) HETATM 52 H 0 2.659 0.662 -0.025 H HETATM 53 H 0 3.754 2.596 5.897 H HETATM 54 H 0 4.335 0.332 1.374 H HETATM 55 H 0 5.925 1.268 5.438 H HETATM 56 H 0 6.217 0.120 3.110 H HETATM 57 Fe 0 0.908 3.263 4.323 Fe HETATM 58 C 0 1.239 5.653 10.371 C HETATM 59 C 0 0.497 0.911 -1.737 C HETATM 60 N 0 1.300 6.094 11.486 N HETATM 61 N 0 0.424 0.482 -2.855 N END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 60 142 Table 4.S22(cont’d) CONECT 59 24 61 CONECT 60 58 CONECT 61 59 [Fe(4’-Cl-terpy)2]2+: Table 4.S23: The optimized geometry file for the lowest energy singlet state of [Fe(4’-Cl- terpy)2]2+. TITLE Fe-Cl-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.003 3.971 6.130 N HETATM 2 N 0 0.092 1.673 5.356 N HETATM 3 N 0 1.732 5.090 3.912 N HETATM 4 C 0 0.548 3.204 7.191 C HETATM 5 C 0 0.611 3.695 8.528 C HETATM 6 C 0 1.156 4.996 8.732 C HETATM 7 C 0 0.020 1.867 6.741 C HETATM 8 C 0 -0.361 0.480 4.841 C HETATM 9 C 0 -0.506 0.870 7.614 C HETATM 10 C 0 -0.897 -0.556 5.657 C HETATM 11 C 0 -0.973 -0.359 7.066 C HETATM 12 C 0 1.533 5.235 6.332 C HETATM 13 C 0 1.627 5.787 7.644 C HETATM 14 C 0 1.960 5.885 5.042 C HETATM 15 C 0 2.085 5.592 2.680 C HETATM 16 C 0 2.543 7.182 4.943 C HETATM 17 C 0 2.670 6.879 2.512 C HETATM 18 C 0 2.903 7.688 3.661 C HETATM 19 N 0 0.772 2.554 2.516 N HETATM 20 N 0 -0.981 3.947 3.819 N HETATM 21 N 0 2.720 2.335 4.208 N HETATM 22 C 0 -0.372 2.786 1.770 C HETATM 23 C 0 -0.494 2.278 0.442 C HETATM 24 C 0 0.598 1.530 -0.084 C HETATM 25 C 0 -1.390 3.598 2.525 C HETATM 26 C 0 -1.841 4.690 4.596 C HETATM 27 C 0 -2.658 3.993 2.006 C HETATM 28 C 0 -3.120 5.114 4.138 C HETATM 29 C 0 -3.537 4.761 2.822 C HETATM 30 C 0 1.832 1.826 2.001 C HETATM 31 C 0 1.778 1.291 0.680 C HETATM 32 C 0 2.965 1.698 2.984 C HETATM 33 C 0 3.695 2.280 5.178 C HETATM 34 C 0 4.185 1.005 2.730 C HETATM 35 C 0 4.933 1.605 4.987 C HETATM 36 C 0 5.183 0.956 3.744 C HETATM 37 H 0 0.249 3.092 9.387 H HETATM 38 H 0 -0.289 0.355 3.739 H HETATM 39 H 0 -0.552 1.050 8.710 H HETATM 40 H 0 -1.248 -1.499 5.182 H HETATM 41 H 0 -1.387 -1.149 7.731 H HETATM 42 H 0 2.050 6.798 7.821 H HETATM 43 H 0 1.892 4.936 1.803 H HETATM 44 H 0 2.715 7.792 5.856 H HETATM 45 H 0 2.934 7.230 1.490 H HETATM 46 H 0 3.360 8.698 3.565 H HETATM 47 H 0 -1.406 2.455 -0.166 H HETATM 48 H 0 -1.486 4.950 5.617 H HETATM 49 H 0 -2.957 3.704 0.975 H HETATM 50 H 0 -3.771 5.713 4.813 H HETATM 51 H 0 -4.530 5.078 2.434 H HETATM 52 H 0 2.620 0.707 0.253 H HETATM 53 H 0 3.469 2.795 6.136 H HETATM 54 H 0 4.353 0.508 1.750 H HETATM 55 H 0 5.682 1.595 5.809 H HETATM 56 H 0 6.141 0.419 3.562 H HETATM 57 Fe 0 0.890 3.262 4.323 Fe 143 Table 4.S23 (cont’d) HETATM 58 Cl 0 1.255 5.652 10.397 Cl HETATM 59 Cl 0 0.486 0.873 -1.748 Cl END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 19 CONECT 58 6 CONECT 59 24 Table 4.S24: The optimized geometry file for the lowest energy quintet state of [Fe(4’-Cl- terpy)2]2+. TITLE Fe-Cl-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 144 Table 4.S24 (cont’d) HETATM 1 N 0 0.993 4.030 6.376 N HETATM 2 N 0 0.053 1.601 5.751 N HETATM 3 N 0 1.788 5.245 4.116 N HETATM 4 C 0 0.553 3.290 7.455 C HETATM 5 C 0 0.610 3.807 8.783 C HETATM 6 C 0 1.140 5.118 8.958 C HETATM 7 C 0 0.016 1.910 7.102 C HETATM 8 C 0 -0.411 0.377 5.335 C HETATM 9 C 0 -0.492 0.987 8.066 C HETATM 10 C 0 -0.932 -0.595 6.237 C HETATM 11 C 0 -0.971 -0.281 7.625 C HETATM 12 C 0 1.509 5.297 6.552 C HETATM 13 C 0 1.599 5.884 7.853 C HETATM 14 C 0 1.958 5.988 5.277 C HETATM 15 C 0 2.173 5.803 2.920 C HETATM 16 C 0 2.520 7.303 5.254 C HETATM 17 C 0 2.739 7.105 2.818 C HETATM 18 C 0 2.915 7.868 4.008 C HETATM 19 N 0 0.785 2.520 2.256 N HETATM 20 N 0 -1.131 3.893 3.550 N HETATM 21 N 0 2.841 2.140 3.946 N HETATM 22 C 0 -0.346 2.774 1.511 C HETATM 23 C 0 -0.462 2.310 0.165 C HETATM 24 C 0 0.630 1.574 -0.375 C HETATM 25 C 0 -1.425 3.569 2.234 C HETATM 26 C 0 -2.055 4.604 4.276 C HETATM 27 C 0 -2.654 3.960 1.621 C HETATM 28 C 0 -3.305 5.025 3.737 C HETATM 29 C 0 -3.607 4.697 2.384 C HETATM 30 C 0 1.837 1.801 1.732 C HETATM 31 C 0 1.797 1.305 0.392 C HETATM 32 C 0 3.008 1.592 2.682 C HETATM 33 C 0 3.853 2.000 4.864 C HETATM 34 C 0 4.197 0.889 2.321 C HETATM 35 C 0 5.066 1.311 4.579 C HETATM 36 C 0 5.239 0.745 3.283 C HETATM 37 H 0 0.259 3.223 9.658 H HETATM 38 H 0 -0.359 0.179 4.240 H HETATM 39 H 0 -0.518 1.244 9.146 H HETATM 40 H 0 -1.295 -1.572 5.849 H HETATM 41 H 0 -1.371 -1.013 8.363 H HETATM 42 H 0 2.011 6.903 8.009 H HETATM 43 H 0 2.020 5.175 2.013 H HETATM 44 H 0 2.651 7.884 6.192 H HETATM 45 H 0 3.033 7.504 1.821 H HETATM 46 H 0 3.354 8.891 3.972 H HETATM 47 H 0 -1.365 2.506 -0.450 H HETATM 48 H 0 -1.776 4.841 5.328 H HETATM 49 H 0 -2.875 3.700 0.564 H HETATM 50 H 0 -4.018 5.597 4.370 H HETATM 51 H 0 -4.570 5.010 1.923 H HETATM 52 H 0 2.637 0.729 -0.047 H HETATM 53 H 0 3.677 2.460 5.864 H HETATM 54 H 0 4.320 0.455 1.305 H HETATM 55 H 0 5.853 1.226 5.361 H HETATM 56 H 0 6.173 0.199 3.019 H HETATM 57 Fe 0 0.868 3.180 4.353 Fe HETATM 58 Cl 0 1.231 5.812 10.606 Cl HETATM 59 Cl 0 0.531 0.969 -2.058 Cl END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 145 Table 4.S24 (cont’d) CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 CONECT 59 24 [Fe(4’-OH-terpy)2]2+: Table 4.S25: The optimized geometry file for the lowest energy singlet state of [Fe(4’-OH- terpy)2]2+. TITLE Fe-OH-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.020 3.940 6.145 N HETATM 2 N 0 0.106 1.645 5.372 N HETATM 3 N 0 1.745 5.059 3.928 N HETATM 4 C 0 0.571 3.175 7.210 C HETATM 5 C 0 0.641 3.660 8.544 C HETATM 6 C 0 1.188 4.966 8.767 C HETATM 7 C 0 0.039 1.839 6.758 C HETATM 8 C 0 -0.353 0.453 4.857 C HETATM 9 C 0 -0.488 0.840 7.630 C HETATM 10 C 0 -0.889 -0.583 5.673 C 146 Table 4.S25 (cont’d) HETATM 11 C 0 -0.959 -0.387 7.082 C HETATM 12 C 0 1.549 5.203 6.352 C HETATM 13 C 0 1.649 5.753 7.661 C HETATM 14 C 0 1.972 5.854 5.059 C HETATM 15 C 0 2.096 5.563 2.695 C HETATM 16 C 0 2.552 7.153 4.958 C HETATM 17 C 0 2.678 6.851 2.526 C HETATM 18 C 0 2.911 7.660 3.676 C HETATM 19 N 0 0.783 2.520 2.525 N HETATM 20 N 0 -0.966 3.914 3.826 N HETATM 21 N 0 2.732 2.301 4.214 N HETATM 22 C 0 -0.358 2.748 1.776 C HETATM 23 C 0 -0.483 2.238 0.453 C HETATM 24 C 0 0.604 1.478 -0.090 C HETATM 25 C 0 -1.375 3.563 2.532 C HETATM 26 C 0 -1.826 4.659 4.602 C HETATM 27 C 0 -2.644 3.960 2.014 C HETATM 28 C 0 -3.105 5.084 4.145 C HETATM 29 C 0 -3.521 4.730 2.829 C HETATM 30 C 0 1.840 1.787 2.006 C HETATM 31 C 0 1.784 1.248 0.692 C HETATM 32 C 0 2.975 1.662 2.991 C HETATM 33 C 0 3.709 2.248 5.183 C HETATM 34 C 0 4.194 0.967 2.738 C HETATM 35 C 0 4.947 1.571 4.993 C HETATM 36 C 0 5.194 0.920 3.750 C HETATM 37 H 0 0.286 3.060 9.409 H HETATM 38 H 0 -0.284 0.333 3.755 H HETATM 39 H 0 -0.529 1.020 8.726 H HETATM 40 H 0 -1.244 -1.524 5.198 H HETATM 41 H 0 -1.374 -1.177 7.748 H HETATM 42 H 0 2.073 6.768 7.829 H HETATM 43 H 0 1.902 4.906 1.820 H HETATM 44 H 0 2.723 7.764 5.871 H HETATM 45 H 0 2.941 7.203 1.504 H HETATM 46 H 0 3.364 8.672 3.580 H HETATM 47 H 0 -1.400 2.421 -0.150 H HETATM 48 H 0 -1.469 4.919 5.622 H HETATM 49 H 0 -2.943 3.670 0.983 H HETATM 50 H 0 -3.755 5.684 4.820 H HETATM 51 H 0 -4.515 5.048 2.441 H HETATM 52 H 0 2.621 0.658 0.260 H HETATM 53 H 0 3.483 2.767 6.140 H HETATM 54 H 0 4.360 0.468 1.758 H HETATM 55 H 0 5.697 1.563 5.814 H HETATM 56 H 0 6.152 0.381 3.569 H HETATM 57 Fe 0 0.903 3.230 4.335 Fe HETATM 58 O 0 1.240 5.403 10.083 O HETATM 59 H 0 1.642 6.313 10.132 H HETATM 60 O 0 0.586 0.932 -1.365 O HETATM 61 H 0 -0.278 1.137 -1.818 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 147 Table 4.S25 (cont’d) CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 60 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 19 CONECT 58 6 59 CONECT 59 58 CONECT 60 24 61 CONECT 61 60 Table 4.S26: The optimized geometry file for the lowest energy quintet state of [Fe(4’-OH- terpy)2]2+. TITLE Fe-OH-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.983 4.101 6.337 N HETATM 2 N 0 0.055 1.670 5.692 N HETATM 3 N 0 1.797 5.335 4.100 N HETATM 4 C 0 0.532 3.357 7.409 C HETATM 5 C 0 0.573 3.867 8.738 C HETATM 6 C 0 1.098 5.182 8.946 C HETATM 7 C 0 0.001 1.978 7.043 C HETATM 8 C 0 -0.402 0.444 5.274 C HETATM 9 C 0 -0.519 1.053 8.001 C HETATM 10 C 0 -0.932 -0.529 6.168 C HETATM 11 C 0 -0.991 -0.216 7.557 C HETATM 12 C 0 1.494 5.371 6.534 C HETATM 13 C 0 1.567 5.949 7.831 C HETATM 14 C 0 1.955 6.072 5.264 C HETATM 15 C 0 2.186 5.893 2.906 C HETATM 16 C 0 2.513 7.388 5.247 C HETATM 17 C 0 2.750 7.198 2.810 C 148 Table 4.S26 (cont’d) HETATM 18 C 0 2.915 7.957 4.005 C HETATM 19 N 0 0.790 2.456 2.291 N HETATM 20 N 0 -1.118 3.887 3.514 N HETATM 21 N 0 2.857 2.164 3.974 N HETATM 22 C 0 -0.335 2.684 1.524 C HETATM 23 C 0 -0.446 2.178 0.197 C HETATM 24 C 0 0.643 1.415 -0.333 C HETATM 25 C 0 -1.418 3.500 2.216 C HETATM 26 C 0 -2.043 4.622 4.215 C HETATM 27 C 0 -2.659 3.850 1.599 C HETATM 28 C 0 -3.300 5.009 3.667 C HETATM 29 C 0 -3.610 4.615 2.334 C HETATM 30 C 0 1.845 1.722 1.781 C HETATM 31 C 0 1.809 1.183 0.467 C HETATM 32 C 0 3.017 1.551 2.739 C HETATM 33 C 0 3.868 2.061 4.898 C HETATM 34 C 0 4.200 0.818 2.415 C HETATM 35 C 0 5.077 1.351 4.647 C HETATM 36 C 0 5.242 0.717 3.381 C HETATM 37 H 0 0.211 3.268 9.602 H HETATM 38 H 0 -0.339 0.244 4.180 H HETATM 39 H 0 -0.560 1.311 9.081 H HETATM 40 H 0 -1.289 -1.507 5.776 H HETATM 41 H 0 -1.399 -0.948 8.289 H HETATM 42 H 0 1.972 6.968 8.006 H HETATM 43 H 0 2.038 5.264 1.999 H HETATM 44 H 0 2.635 7.968 6.188 H HETATM 45 H 0 3.049 7.601 1.817 H HETATM 46 H 0 3.352 8.980 3.974 H HETATM 47 H 0 -1.350 2.365 -0.423 H HETATM 48 H 0 -1.759 4.905 5.254 H HETATM 49 H 0 -2.888 3.534 0.558 H HETATM 50 H 0 -4.012 5.604 4.280 H HETATM 51 H 0 -4.581 4.896 1.867 H HETATM 52 H 0 2.647 0.591 0.043 H HETATM 53 H 0 3.695 2.570 5.874 H HETATM 54 H 0 4.313 0.328 1.424 H HETATM 55 H 0 5.864 1.299 5.431 H HETATM 56 H 0 6.170 0.150 3.145 H HETATM 57 Fe 0 0.906 3.262 4.320 Fe HETATM 58 O 0 1.185 5.775 10.192 O HETATM 59 H 0 0.835 5.161 10.895 H HETATM 60 O 0 0.640 0.872 -1.604 O HETATM 61 H 0 -0.217 1.077 -2.070 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 149 Table 4.S26 (cont’d) CONECT 24 23 31 60 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 59 CONECT 59 58 CONECT 60 24 61 CONECT 61 60 [Fe(4’-furan-terpy)2]2+: Table 4.S27: The optimized geometry file for the lowest energy singlet state of [Fe(4’-furan- terpy)2]2+. TITLE Fe-furan-terpy Singlet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.000 3.958 6.134 N HETATM 2 N 0 0.084 1.666 5.360 N HETATM 3 N 0 1.725 5.083 3.921 N HETATM 4 C 0 0.550 3.194 7.198 C HETATM 5 C 0 0.616 3.683 8.532 C HETATM 6 C 0 1.160 4.994 8.779 C HETATM 7 C 0 0.018 1.858 6.746 C HETATM 8 C 0 -0.373 0.475 4.842 C HETATM 9 C 0 -0.507 0.857 7.616 C HETATM 10 C 0 -0.909 -0.562 5.656 C HETATM 11 C 0 -0.978 -0.369 7.066 C HETATM 12 C 0 1.529 5.223 6.345 C HETATM 13 C 0 1.624 5.767 7.653 C HETATM 14 C 0 1.954 5.876 5.054 C HETATM 15 C 0 2.077 5.586 2.689 C HETATM 16 C 0 2.536 7.174 4.952 C HETATM 17 C 0 2.661 6.873 2.520 C HETATM 18 C 0 2.895 7.681 3.671 C HETATM 19 N 0 0.759 2.555 2.519 N HETATM 20 N 0 -0.989 3.942 3.826 N HETATM 21 N 0 2.709 2.324 4.202 N HETATM 22 C 0 -0.381 2.789 1.769 C HETATM 23 C 0 -0.501 2.292 0.442 C 150 Table 4.S27 (cont’d) HETATM 24 C 0 0.584 1.536 -0.132 C HETATM 25 C 0 -1.399 3.598 2.531 C HETATM 26 C 0 -1.847 4.682 4.608 C HETATM 27 C 0 -2.669 3.996 2.017 C HETATM 28 C 0 -3.128 5.108 4.155 C HETATM 29 C 0 -3.546 4.760 2.838 C HETATM 30 C 0 1.816 1.827 1.992 C HETATM 31 C 0 1.759 1.306 0.672 C HETATM 32 C 0 2.951 1.692 2.976 C HETATM 33 C 0 3.684 2.263 5.173 C HETATM 34 C 0 4.170 0.997 2.720 C HETATM 35 C 0 4.921 1.586 4.980 C HETATM 36 C 0 5.168 0.941 3.734 C HETATM 37 H 0 0.252 3.066 9.380 H HETATM 38 H 0 -0.304 0.357 3.739 H HETATM 39 H 0 -0.548 1.034 8.713 H HETATM 40 H 0 -1.263 -1.503 5.179 H HETATM 41 H 0 -1.391 -1.161 7.730 H HETATM 42 H 0 2.051 6.781 7.807 H HETATM 43 H 0 1.882 4.929 1.813 H HETATM 44 H 0 2.709 7.784 5.866 H HETATM 45 H 0 2.924 7.225 1.499 H HETATM 46 H 0 3.350 8.692 3.575 H HETATM 47 H 0 -1.419 2.483 -0.152 H HETATM 48 H 0 -1.489 4.937 5.629 H HETATM 49 H 0 -2.969 3.711 0.985 H HETATM 50 H 0 -3.777 5.704 4.834 H HETATM 51 H 0 -4.541 5.078 2.453 H HETATM 52 H 0 2.615 0.727 0.263 H HETATM 53 H 0 3.458 2.777 6.132 H HETATM 54 H 0 4.337 0.502 1.738 H HETATM 55 H 0 5.670 1.571 5.802 H HETATM 56 H 0 6.125 0.402 3.550 H HETATM 57 Fe 0 0.882 3.254 4.327 Fe HETATM 58 C 0 1.244 5.540 10.153 C HETATM 59 C 0 1.698 6.755 10.694 C HETATM 60 O 0 0.782 4.708 11.199 O HETATM 61 C 0 1.508 6.673 12.138 C HETATM 62 H 0 2.117 7.613 10.132 H HETATM 63 C 0 0.953 5.418 12.393 C HETATM 64 H 0 1.753 7.448 12.892 H HETATM 65 H 0 0.633 4.890 13.311 H HETATM 66 C 0 0.498 1.008 -1.513 C HETATM 67 C 0 1.364 0.256 -2.327 C HETATM 68 O 0 -0.690 1.286 -2.229 O HETATM 69 C 0 0.682 0.061 -3.603 C HETATM 70 H 0 2.371 -0.117 -2.055 H HETATM 71 C 0 -0.553 0.701 -3.493 C HETATM 72 H 0 1.062 -0.486 -4.489 H HETATM 73 H 0 -1.413 0.835 -4.176 H END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 151 Table 4.S27 (cont’d) CONECT 18 17 16 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 66 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 19 1 CONECT 58 6 59 60 CONECT 59 58 61 62 CONECT 60 58 63 CONECT 61 59 63 64 CONECT 62 59 CONECT 63 60 61 65 CONECT 64 61 CONECT 65 63 CONECT 66 24 67 68 CONECT 67 66 69 70 CONECT 68 66 71 CONECT 69 67 71 72 CONECT 70 67 CONECT 71 68 69 73 CONECT 72 69 CONECT 73 71 Table 4.S28: The optimized geometry file for the lowest energy quintet state of [Fe(4’-furan- terpy)2]2+. TITLE Fe-furan-terpy Quintet SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.995 4.118 6.333 N HETATM 2 N 0 0.058 1.711 5.616 N HETATM 3 N 0 1.802 5.420 4.134 N HETATM 4 C 0 0.540 3.345 7.383 C HETATM 5 C 0 0.576 3.819 8.724 C HETATM 6 C 0 1.098 5.133 8.995 C 152 Table 4.S28 (cont’d) HETATM 7 C 0 0.009 1.977 6.977 C HETATM 8 C 0 -0.401 0.499 5.161 C HETATM 9 C 0 -0.508 1.022 7.906 C HETATM 10 C 0 -0.928 -0.502 6.026 C HETATM 11 C 0 -0.982 -0.232 7.424 C HETATM 12 C 0 1.504 5.381 6.570 C HETATM 13 C 0 1.569 5.919 7.884 C HETATM 14 C 0 1.968 6.119 5.321 C HETATM 15 C 0 2.190 6.013 2.957 C HETATM 16 C 0 2.534 7.431 5.343 C HETATM 17 C 0 2.761 7.316 2.899 C HETATM 18 C 0 2.936 8.036 4.117 C HETATM 19 N 0 0.801 2.594 2.243 N HETATM 20 N 0 -1.108 3.989 3.507 N HETATM 21 N 0 2.868 2.252 3.914 N HETATM 22 C 0 -0.324 2.844 1.482 C HETATM 23 C 0 -0.430 2.381 0.141 C HETATM 24 C 0 0.657 1.637 -0.438 C HETATM 25 C 0 -1.407 3.639 2.198 C HETATM 26 C 0 -2.034 4.705 4.227 C HETATM 27 C 0 -2.647 4.009 1.589 C HETATM 28 C 0 -3.289 5.108 3.690 C HETATM 29 C 0 -3.598 4.753 2.344 C HETATM 30 C 0 1.856 1.878 1.709 C HETATM 31 C 0 1.819 1.388 0.376 C HETATM 32 C 0 3.027 1.676 2.662 C HETATM 33 C 0 3.877 2.122 4.837 C HETATM 34 C 0 4.210 0.952 2.319 C HETATM 35 C 0 5.085 1.419 4.566 C HETATM 36 C 0 5.251 0.822 3.283 C HETATM 37 H 0 0.207 3.192 9.562 H HETATM 38 H 0 -0.343 0.333 4.061 H HETATM 39 H 0 -0.545 1.247 8.994 H HETATM 40 H 0 -1.287 -1.467 5.604 H HETATM 41 H 0 -1.387 -0.987 8.135 H HETATM 42 H 0 1.977 6.936 8.057 H HETATM 43 H 0 2.034 5.413 2.031 H HETATM 44 H 0 2.666 7.981 6.300 H HETATM 45 H 0 3.058 7.749 1.918 H HETATM 46 H 0 3.379 9.057 4.116 H HETATM 47 H 0 -1.337 2.585 -0.465 H HETATM 48 H 0 -1.751 4.959 5.275 H HETATM 49 H 0 -2.874 3.724 0.539 H HETATM 50 H 0 -4.002 5.687 4.318 H HETATM 51 H 0 -4.568 5.050 1.885 H HETATM 52 H 0 2.678 0.820 -0.039 H HETATM 53 H 0 3.702 2.602 5.827 H HETATM 54 H 0 4.326 0.490 1.315 H HETATM 55 H 0 5.871 1.344 5.350 H HETATM 56 H 0 6.179 0.260 3.032 H HETATM 57 Fe 0 0.915 3.339 4.293 Fe HETATM 58 C 0 1.149 5.667 10.373 C HETATM 59 C 0 1.583 6.883 10.931 C HETATM 60 O 0 0.672 4.825 11.405 O HETATM 61 C 0 1.365 6.791 12.369 C HETATM 62 H 0 2.008 7.747 10.382 H HETATM 63 C 0 0.814 5.530 12.604 C HETATM 64 H 0 1.589 7.562 13.133 H HETATM 65 H 0 0.479 4.995 13.513 H HETATM 66 C 0 0.589 1.139 -1.829 C HETATM 67 C 0 1.460 0.394 -2.644 C HETATM 68 O 0 -0.585 1.442 -2.560 O HETATM 69 C 0 0.800 0.229 -3.932 C HETATM 70 H 0 2.459 0.004 -2.361 H HETATM 71 C 0 -0.432 0.880 -3.830 C HETATM 72 H 0 1.188 -0.305 -4.822 H HETATM 73 H 0 -1.279 1.035 -4.526 H END 153 Table 4.S28 (cont’d) CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 17 16 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 66 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 59 60 CONECT 59 58 61 62 CONECT 60 58 63 CONECT 61 59 63 64 CONECT 62 59 CONECT 63 60 61 65 CONECT 64 61 CONECT 65 63 CONECT 66 24 67 68 CONECT 67 66 69 70 CONECT 68 66 71 CONECT 69 67 71 72 154 Table 4.S28 (cont’d) CONECT 70 67 CONECT 71 68 69 73 CONECT 72 69 CONECT 73 71 155 REFERENCES 1. Carey, M. C.; Adelman, S. L. and McCusker, J. K. Chem. Sci. 2019, 10, 134 – 144 . 2. Stippich, K.; Weiss, D.; Guether, A.; Görls, H. and Beckert, R. J. Sulfur Chem. 2009, 30, 109 − 118. 3. Machan, C. W.; Adelhardt, M.; Sarjeant, A. A.; Stern, C. L.; Sutter, J.; Meyer, K. and Mirkin, C. A. J. Am. Chem. Soc. 2012, 134, 16921 – 16924. 4. Yarranton, J. T. and McCusker, J. K. J. Am. Chem. Soc. 2022, 144, 12488 − 12500. 156 CHAPTER 5: LEVERAGING RESONANCE AND INDUCTION TO CONTROL THE ELECTRONIC STRUCTURE OF FE(II)-BIPYRIDYL COMPLEXES 5.1 Introduction With the work in the previous chapter having established that ground state recovery (GSR) in Fe(II)-based polypyridines occurs in the Marcus normal region, this presents an interesting challenge for improving the utility of these complexes as photosensitizers. Recall from the previous chapter that the same electron acceptor (2,3-dichloro-5,6-dicyano-1,4-benzoquinone, DDQ) that Woodhouse and McCusker used to quench the 5T2 excited state lifetime of [Fe(tren(py)3]2+ (where tren(py)3 is tris(2-pyridyl-methylimino-ethyl)amine) had no impact on the 5T2 lifetime of [Fe(bpy)3]2+ (where bpy is 2,2’-bipyridine).1 Since DDQ could not quench the GSR lifetime, no electron transfer (ET) was occurring from the 5T2 excited state of [Fe(bpy)3]2+. For an ET involving a photo-induced excited complex to occur, three conditions must be satisfied: there must be electronic coupling between the reactant and product, the transfer must be energetically favorable, and the excited state of the reactant complex must live long enough for the transfer to occur.2,3 Regarding the first point, given that ET is observed from the 5T2 state of [Fe(tren(py)3]2+, it is reasonable to assume that similar electronic coupling strength will occur between DDQ and the 5T2 state of other Fe(II)-polypyridines. This points towards the free energy difference and excited state lifetime as possible culprits. As stated before, the energy of the 5T2 state is dependent on the energy of the eg* orbitals, as both eg* orbitals are populated with an electron in this state. Therefore, destabilizing the eg* orbitals will increase the energy of the 5T2 state, leading to a more energetically favorable ET to an acceptor like DDQ. The next point to consider is the excited state lifetime. For [Fe(tren(py)3]2+, the 5T2 excited state has a lifetime around 55 ns, long enough for ET to occur in solution dynamics (the rate of molecular diffusion in solvent occurs on a ns timescale).1 However, for [Fe(bpy)3]2+ the 5T2 excited state lifetime is around 1 ns (assuming acetonitrile as the solvent for both complexes). This is at the low end for lifetimes that could reasonably engage in solution-based ET reactions and reducing that lifetime even further would further decrease the likelihood of ET. This presents the challenge caused by GSR occurring in the Marcus normal region: increasing the energy gap between the 5T2 and the ground state (1A1) by destabilizing the eg* orbitals will reduce the GSR lifetime. By solving the energy problem, the lifetime problem worsens, and vice versa. Therefore, to improve the viability of [Fe(bpy)3]2+ complexes as photosensitizers, a system that 157 destabilizes both the eg* orbitals and the t2g orbitals, doing so will keep the ligand field (LF) energy gap low and GSR lifetime long. This is illustrated in Figure 5.1. y g r e n E Figure 5.1: Schematics of the LF orbitals of Fe(II)-polypyridyl complexes alongside the π* orbital of an electron acceptor. In schematic A, the energy of the eg* orbitals is less than the energy of the acceptor π* orbital, so ET to the acceptor will not happen. In schematic B, the eg* orbitals have been destabilized so that they are higher in energy than the acceptor π* orbitals. However, the increased LF strength leads to a faster excited GSR, making ET possible, but unlikely as GSR will out-compete it. In schematic C, the t2g orbitals have also been destabilized which reduces LF strength. In this scenario, ET will become more likely as it may have a rate that is competitive with GSR. Accomplishing this task relies on manipulating both the σ and π-donating strength of the ligand. Recall from Chapter 1.3 that the 3𝑑𝑧2 and 3𝑑𝑥2−𝑦2 orbitals of the Fe(II) complex share eg symmetry with the σ orbitals of the binding nitrogen of the ligand. It is the strength of the bond between these orbitals that impacts the eg-orbital splitting, so a stronger σ bond, which is caused by strong σ-donation from the ligand, will destabilize the eg* orbitals. Conversely, the 3𝑑𝑥𝑦, 3𝑑𝑥𝑧, and 3𝑑𝑦𝑧 orbitals share t2g splitting with the π orbitals of the binding nitrogen of the ligand. Figure 1.7 illustrates that a π-donating ligand will destabilize the t2g orbitals while a π-accepting ligand will stabilize the t2g orbitals (more accurately, with π-donating ligands the t2g* orbitals are stabilized below the eg* orbitals and become the lower ligand field orbitals in place of the t2g orbitals, but for simplicity, I will refer to the lower ligand field orbitals as the t2g orbitals regardless of the π-donating/accepting nature of the ligand throughout this chapter). As previously established, the π-donating strength of the ligand can be tuned with substituents through the resonance structures shown in Figure 1.8. Similarly, substituents can also tune the σ-donating strength of the ligand. An electron withdrawing (EW) substituent will make the ligand a poorer σ- donor while an electron donating (ED) substituent will do the opposite. 158 In 2019, Carey et al. conducted variable temperature transient absorption spectroscopic (VT-TA) measurements on a series of 4,4’-substituted [Fe(bpy)3]2+ complexes, including a dimethyl (diMe) substituted complex.4 Since the Me substituent should act as a σ-donor, it was expected that the substituent should increase LF strength and reduce the 5T2 lifetime, but the opposite was found; the 5T2 excited state lifetime of [Fe(4,4’-diMe-bpy)3](PF6)2 was measured around 1.3 ns. Furthermore, electrochemical studies found that the highest occupied molecular orbital (HOMO), which is the t2g orbital in low-spin d6 complexes like Fe(II)-polypyridines, was destabilized compared to the unsubstituted complex. This means that the diMe-substituted ligand acted as a stronger π-donor, better resembling scheme C as shown in Figure 5.1, as opposed to scheme A. In this work, to further explore how substituents can be used to manipulate the LF structure, the series was expanded to include substituents of varying π-donating strengths: methoxy (OMe) and chloro. Additionally, the placement of the substituent was altered from 4,4’ to 5,5’. In the 5,5’ position, as displayed in Figure 5.2, the resonance structure of the ligand weakens the ability of the substituent to impact the π-donating strength of the ligand by only modulating one of the binding nitrogens rather than both, as is seen in the 4,4’-substituted ligand. This is because the substituent at this position can only interact with the nitrogen on the opposite ring through resonance, while the resonance structure enabling one susbtituent interaction is active, the opposite substituent is cut off from the other accesible resonance pathway. These complexes will be studied Figure 5.2: Schematic of resonance structures available to bipyridine ligands substituted with OMe at the 4,4’ (top) and 5,5’-positions (bottom). 159 via VT-TA, cyclic voltammetry, and acid titration on the unbound ligands to explore how the substituents and susbtituent placement impacts the electronic structure and excited state dynamics in [Fe(bpy)3]2+ complexes. 5.2 Experimental Methods 5.2.1 Synthesis of [Fe(X,X’-diR-bpy)3]2+ Series, where X = 4, 5 and R = H, Me, OMe, Cl The proto-substituted complex and the 4,4’-diMe substituted complex were both synthesized by former group member Dr. Chris Tichnell. The 4,4’-diOMe substituted complex was synthesized by myself following procedures developed by Dr. Tichnell. The 5,5’-diMe and 5,5’- diOMe substituted complexes were both synthesized by former group member Dr. Sara Adelman and current group member Bekah Bowers. Finally, both diCl substituted complexes were synthesized by Bekah Bowers. All complexes were characterized via 1H NMR, data is in the SI. iii. 2,2’-bipyridine, bpy; 4,4’-dimethyl-bipyridine, 4,4’-diMe-bpy; 4,4’-dimethoxy- bipyridine, 4,4’-diOMe-bpy; and 5,5’-dimethyl-bipyridine, 5,5’-diMe-bpy. These ligands used throughout the procedures below were purchased from Alfa Aesar. iv. 4,4’-dichloro-bipyridine, 4,4’-diCl-bpy. 4,4’-diCl-bpy was synthesized by Bowers, B. following literature procedure.5 v. 5,5’-dimethoxy-bipyridine, 5,5’-diOMe-bpy. 5,5’-diOMe-bpy synthesis was done by Adelman, S and Bowers, B. Nickel chloride heahydrate (0.63 g, 2.65 mmol) and triphenylphosphine (1.38 g, 5.26 mmol) were added to a dry Schlenk flask and dissolved in 13 mL of dry DMF taken from an argon atmospheric environment. The solution was degassed for 30 minutes using nitrogen flow. Activated zinc (0.235 g, 3.59 mmol) was then added and the solution was degassed for an additional hour. 2-bromo-5-methoxy-pyridine (0.5 g, 2.66 mmol) purchased from Alfa Aesar was then added. The solution was stirred under nitrogen atmosphere at 80 oC for three hours while monitoring product formation via TLC with 2% methanol in DCM. The solution was then poured into a 1:1 mixture of water and ammonium hydroxide. Chloroform was then added to this mixture to extract the product from the solution and was then decanted. This was repeated twice. Excess sodium sulfate was then added to the chloroform solution to remove any water, and the chloroform was then evaporated, leaving the powdered product (0.115 g, 40% yield). 160 vi. 5,5’-dichloro-bipyridine, 5,5’-diCl-bpy. 5,5’-diCl-bpy was synthesized by Bowers, B. following modified literature procedure.6 2-bromo-5-chloropyridine (2.102 g, 10.4 mmol), Pd(OAc)2 (0.114g, 0.52 mmol), K2CO3 (2.163 g, 15.6 mmol), and TBABr (3.349 g, 10.4 mmol) were added to a flask under nitrogen. A solution mixture containing 24 mL of DMF and 3.2 mL of isopropanol was degassed with nitrogen before adding to the flask. The reaction was heated to 100°C for 20 hours, cooled and filtered over celite. The product was extracted into DCM and washed with H2O 3 times (100 mL). The DCM solution was dried over MgSO4 before removing the organic solvent. The product was recrystallized from MeOH (yield: 0.561 g, 24%). vii. tris(2,2’-bipyridine) iron(II) hexafluorophosphate, [Fe(bpy)3](PF6)2. [Fe(bpy)3](PF6)2 was synthesized by Tichnell, C. following literature procedure.7 viii. tris(4,4’-dimethyl-2,2’-bipyridine) iron(II) hexafluorophosphate, [Fe(4,4’diMe- bpy)3](PF6)2. [Fe(4,4’-diMe-bpy)3](PF6)2 was synthesized by Tichnell, C. following literature procedure.8 ix. tris(4,4’-dimethoxy-2,2’-bipyridine) iron(II) hexafluorophosphate, [Fe(4,4’diOMe-bpy)3](PF6)2. [Fe(4,4’-diOMe-bpy)3](PF6)2 was synthesized by myself following a procedure adapted from literature by Tichnell, C.8 The original procedure was done under nitrogen, the version provided below was tested with small scale under atmospheric conditions. After synthesis proved successful, the procedure was scaled up using the amounts provided below. 4,4’-diOMe-bpy (0.26 g, 1.2 mmol) was dissolved in 100 mL of methanol. Ammonium iron(II) sulfate (0.157 g, 0.4 mmol) dissolved in a 10 mL aqueous solution was then added to the methanol solution, which was then stirred for one hour. Ammonium hexafluorophosphate (0.13 g, 0.8 mmol) dissolved in a 10 mL aqueous solution was then added, which was stirred for an addition 20 minutes. A red solid was filtered from the solution using an medium fritted glass filter, which was then washed with water and then ethanol. The remaining solid was scraped into a glass vial and dried in a desiccator overnight (0.302 g, 76% yield). 161 x. tris(4,4’-dichloro-2,2’-bipyridine) iron(II) tetrafluoroborate, [Fe(4,4’diCl- bpy)3](BF4)2. [Fe(4,4’-diCl-bpy)3](PF6)2 was synthesized by Bowers, B. The complex synthesis was originally attempted using FeCl2 and metathesizing to the PF6 salt; however, upon addition of NH4PF6, the ligand dissociated from the Fe-center. The final complex was synthesized by dissolving Fe(BF4)2 · 6H2O (0.143 g, 0.43 mmol) in 5 mL of acetone under N2 and adding 3.1 equivalence of 4,4’-diCl-bpy (0.299 g, 1.33 mmol) in CHCl3. The mixture was stirred at room temperature, overnight under N2 before adding diethyl ether to precipitate the final complex for collection via vacuum filtration. Purification was done by recrystallization from MeCN via ether diffusion 2 times (yield: 0.332 g, 84%). xi. tris(5,5’-dimethyl-2,2’-bipyridine) iron(II) hexafluorophosphate, [Fe(5,5’diMe- bpy)3](PF6)2. [Fe(5,5’-diMe-bpy)3](PF6)2 was initially synthesized by Adelman, S. and additional sample was synthesized later by Bowers, B. via the following procedure FeCl2 (0.070 g, 0.56 mmol) was dissolved in degassed H2O and added dropwise via cannula transfer to a solution of 5,5’-diMe-bpy (0.32 g, 1.73 mmol) in MeOH under N2. The dark red solution was stirred for 1 hour at room temperature after which an excess of NH4PF6 was added to precipitate the final complex. The product was collected via vacuum filtration and recrystallized via ether diffusion into a MeCN solution. xii. tris(5,5’-dimethoxy-2,2’-bipyridine) iron(II) hexafluorophosphate, [Fe(5,5’diOMe-bpy)3](PF6)2. [Fe(5,5’-diOMe-bpy)3](PF6)2 was synthesized by Adelman, S. and additional sample was synthesized later by Bowers, B. Under an argon atmosphere, 5,5’-diOMe-bpy (0.043 g, 0.199 mmol) was dissolved in acetonitrile. Iron(II) tetrafluoroborate bishydrate (0.017 g, 0.064 mmol) dissolved in acetonitrile was added to the solution dropwise while stirring. The solution was then stirred for 8 hours. 10 equivalents of ammonium hexafluorophosphate (0.1 g, 0.613 mmol) was then added to the solution. The product was precipitated from solution by adding diethyl ether. The solution was then removed from argon atmosphere and the precipitate was filtered out using a fine 162 fritted glass filter. Powdered sample was recrystallized from diethyl ether diffusion in acetonitrile. (yield unreported). xiii. tris(5,5’-dichloro-2,2’-bipyridine) iron(II) tetrafluoroborate, [Fe(5,5’diCl- bpy)3](BF4)2. [Fe(5,5’-diCl-bpy)3](PF6)2 was synthesized by Bowers, B. using the same procedure for [Fe(4,4’-diCl-bpy)3](BF4)2. 5.2.2 Spectroscopic Methods The [Fe(X,X’-diR-bpy)](PF6)2 complexes were examined using UV-Vis spectroscopy, density functional theory calculations (DFT), and VT-TA using procedures outlined in Chapter 2. 5.2.3 Electrochemical Methods Electrochemical data were collected using a CH Instruments Model CHI620D electrochemical workstation. A standard three-electrode setup was employed to obtain Fe(II/III) oxidation and bpy/bpy·- reduction potentials using cyclic voltammetry (CV) in acetonitrile solution with a 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6) supporting electrolyte, a platinum working electrode, a silver wire reference electrode, and a platinum wire counter electrode. TBAPF6 was purchased from Oakwood Chemical Company and recrystallized from ethanol before use. All potentials presented are reference to the ferrocene/ferrocenium (Fc/Fc+) potential couple. Ferrocene was purchased from Sigma-Aldrich and recrystallized from pentane before use. CV spectra with Fc/Fc+ potential included is available in the supplemental information. Once the potentials were adjusted vs. Fc/Fc+, the E1/2 was recorded for both the oxidation and reduction potentials. Most electrochemical data was collected under argon atmosphere, but for those that were collected in air were degassed for 5 minutes before data collection and a low pressure nitrogen flow was applied to the solution surface to prevent oxygen from interfering with data collection. 5.2.4 Acid Titration The pKb was determined for both diMe substituted ligands, both diCl substituted ligands, and the 5,5’-diOMe substituted ligand. This was done by dissolving a small amount of the ligand in 150 mL of acetone (for the diMe substituted ligands, acetonitrile had to be used instead because the equivalence point was too difficult to determine). The pH was measured using a Fisher Scientific Accumet Basic pH meter. A 0.1 M nitric acid solution, prepared from a 70% nitric acid stock purchased from Sigma-Aldrich, was then added dropwise using a Gilson p1000 pipettor. 163 With each drop added, the pH of the solution was again measured and recorded. Once completed, the pH was plotted against the volume of acid added, as seen in Figure 5.3. The pH of the solution half-way to the equivalence point (called the mid point) provides the pKb of the ligand. Figure 5.3: Titration curve of a 4,4’-diMe-bpy solution in acetonitrile titrated with 0.1 M nitric acid. pKb = 3.16. 5.3 Experimental Results 5.3.1 [Fe(bpy)3](PF6)2 The ultrafast spectroscopic data for this complex was already collected and reported by Dr. Monica Carey.4 However, since the time of publication, VT-TA data analysis has been expanded to include transition state theory, as well as an altered approach to calculate Marcus parameters that corrects for the temperature dependency of the pre-exponential component of the Marcus equation (see Chapter 3 for further explanation). Therefore, Carey’s original data has been reanalyzed under these new methods. Additionally, since publication the Nelsen method has been employed for calculating inner-sphere reorganization energy (λi) using DFT. While the reported GSR time constant (τ) and electrochemical potentials (E1/2) are unchanged; new energy barrier (ΔG‡), electronic coupling constant (Hab), and driving force (ΔGo) values are used throughout this work. Furthermore, new parameters were determined using transition state theory: the enthalpy barrier (ΔH‡) and the entropy barrier (ΔS‡) related to GSR. These terms are summarized in Table 5.1, and the transition state and Marcus theory “Arrhenius-type” plots are shown in Figure 5.4, provided below. The VT-TA data for this complex, and the 4,4’-diMe substituted complex were 164 measured in 5 K steps. The remainder of the VT-TA data in this work was recorded in 10 K steps to save time for the lengthy VT-TA experiments. Table 5.1: The GSR lifetime at 295 K, electrochemical potentials, transition state, and Marcus theory parameters for [Fe(bpy)3](PF6)2. †: These values were reported in literature.4 Substituent H τ295 (ps)† 1050 ± 20 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+)† 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+)† ΔH‡ (cm-1) ΔS‡ (cm-1 * K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) 0.68 -1.74 125 ± 20 -5.66 ± 0.07 1795 ± 40 0.086 ± 0.017 1600 ± 75 Figure 5.4: “Arrhenius-type” plots for [Fe(bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). 5.3.2 [Fe(4,4’-diMe-bpy)3](PF6)2 Like the unsubstituted complex, the ultrafast spectroscopic data for this complex was already collected and reported by Dr. Monica Carey.4 Again, this complex was re-analyzed using the techniques outlined in Chapter 3. The electrochemical and spectroscopic data for this complex is summarized in Table 5.2. The “Arrhenius-type” plots for this complex are shown in Figure 5.5. Figure 5.5: “Arrhenius-type” plots for [Fe(4,4’-diMe-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). 165 The pKb for the ligand of this complex was also measured. The titration curve is provided in Figure 5.3 and the pKb was measured to be 3.16 in acetonitrile. Table 5.2: The GSR lifetime at 295 K, electrochemical potentials, transition state, and Marcus theory parameters for [Fe(4,4’-diMe-bpy)3](PF6)2. †: These values were reported in literature.4 Substituent 4,4’-diMe τ295 (ps)† 1320 ± 20 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+)† 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+)† ΔH‡ (cm-1) ΔS‡ (cm-1/K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) 0.52 -1.84 160 ± 10 -5.66 ± 0.02 1830 ± 15 0.082 ± 0.007 1720 ± 30 5.3.3 [Fe(5,5’-diMe-bpy)3](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(5,5’-diMe-bpy)3](PF6)2 is provided in Figure 5.6, alongside the pump and probe wavelengths used throughout TA measurements. Typically, it is preferred to use the lower energy pulse for the pump, however since the sapphire crystal used to generate white light in the probe line produces little power at bluer wavelengths, it was decided to use 505 nm as the pump instead, which led to good signal to noise (s/n). Figure 5.6: Ground state absorption spectrum for [Fe(5,5’-diMe-bpy)3]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Four VT-TA trials were conducted on this complex. The VT-TA plot of trial four is shown in Figure 5.7. The kinetic data from each trial is summarized in Table 5.3. Just as the VT-TA data was treated in the previous chapter, kinetic data from each trial was run through Grubb’s test to check for outliers and then averaged for each temperature. The second trial was not completed, but the data was collected was still used to produce the averaged kinetic values. 166 Figure 5.7: VT-TA data for [Fe(5,5’-diMe-bpy)3]2+. Shown is the data from the third trial. Table 5.3: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. τ (ps) 295 K 285 K 275 K 265 K 255 K 245 K 235 K Trial 3 1020 ± 20 1050 ± 25 1155 ± 25 1225 ± 20 1300 ± 20 1390 ± 25 1475± 20 Trial 4 1035 ± 10 1075 ± 20 1165 ± 15 1190 ± 20 1300 ± 20 1415 ± 15 1500 ± 15 Trial 1 1050 ± 25 1090 ± 20 1155 ± 20 1200 ± 20 1245 ± 20 1395 ± 20 1455 ± 25 Average 1035 ± 20 1070 ± 25 1160 ± 20 1205 ± 20 1280 ± 30 1390 ± 25 1480 ± 20 Trial 2 1035 ± 10 ― ― ― ― 1360 ± 10 1465 ± 10 The averaged kinetic data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots are shown in Figure 5.8. The cyclic voltammogram Figure 5.8: “Arrhenius-type” plots for [Fe(5,5’-diMe-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). 167 for this complex is shown in Figure 5.9. The information extracted from these plots is provided in Table 5.4. Figure 5.9: Cyclic voltammogram for [Fe(5,5’-diMe-bpy)3]2+. The positive potential is attributed to the Fe2+ oxidation, while the three negative potentials are attributed to the reduction of the bipyridine ligands into radicals. This assignment is consistent with previously reported CV data for [Fe(bpy)3]2+ complexes.4 Table 5.4: The transition state, Marcus theory, and electrochemical parameters for [Fe(5,5’-diMe- bpy)3](PF6)2. Substituent ΔH‡ (cm-1) ΔS‡ (cm-1/K) 5,5’-diMe 120 ± 10 -5.68 ± 0.03 ΔG‡ (cm-1) 1790 ± 20 |Hab|4/λ (cm-3) 0.080 ± 0.008 (ΔGo + λ)2/λ (cm-1) 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+) 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+) 1565 ± 40 0.59 -1.88 The pKb for the ligand of this complex was also measured. The titration curve is provided in Figure 5.10 and the pKb was measured to be 1.92 in acetonitrile. Figure 5.10: Titration curve of a 5,5’-diMe-bpy solution in acetonitrile titrated with 0.1 M nitric acid. pKb = 1.92. 168 5.3.4 [Fe(4,4’-diOMe-bpy)3](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4,4’-diOMe-bpy)3](PF6)2 is provided in Figure 5.11, alongside the pump and probe wavelengths used throughout TA measurements. Figure 5.11: Ground state absorption spectrum for [Fe(4,4’-diOMe-bpy)3]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Seven VT-TA trials were conducted on this complex. Like the rest of the complexes in this series, the first four were collected on the ultrafast laser system. However, the GSR lifetime for this complex, especially at low temperatures, was too long to accurately measure on the ultrafast system alone. Therefore, VT data was collected on the nanosecond (NS) laser system as well, since it is more suited to measuring the GSR lifetimes at low temperatures. The GSR lifetime being near the delay stage limit of the ultrafast system and the instrument response function of the NS system, has left the kinetic data for this complex with much larger error bars than the rest of the series. The VT-TA plot of trial two and trial five is shown in Figure 5.7. The kinetic data from each trial is summarized in Table 5.5. Figure 5.12: VT-TA data for [Fe(4,4’-diOMe-bpy)3]2+. Shown is the data from the second trial (A) collected on the ultrafast laser system and the fifth trial (B) collected on the NS laser system. 169 Table 5.5: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. †: These experiments were conducted on the NS laser system; all others were done on the ultrafast system. τ (ps) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5† Trial 6† Trial 7† Average 295 K 3600 ± 220 4560 ± 310 4095 ± 240 4180 ± 270 4650 ± 325 4635 ± 300 4780 ± 430 4360 ± 460 285 K 3965 ± 205 4665 ± 320 3725 ± 340 4395 ± 465 4825 ± 270 4745 ± 195 4350 ± 270 4380 ± 455 275 K 4655 ± 290 5335 ± 370 4155 ± 310 4405 ± 405 4915 ± 155 5655 ± 240 4820 ± 150 4850 ± 535 265 K 4320 ± 295 5990 ± 450 5010 ± 380 6880 ± 885 5745 ± 135 5340 ± 170 5510 ± 165 5540 ± 820 255 K 5045 ± 370 6775 ± 630 6580 ± 670 5930 ± 810 6580 ± 135 5990 ± 115 6650 ± 140 6220 ± 695 245 K 6175 ± 540 6755 ± 760 6005 ± 560 5990 ± 890 7015 ± 115 7305 ± 130 6660 ± 125 6560 ± 650 235 K 6725 ± 695 8155 ± 1220 9155± 145 11600 ± 3340 8390 ± 105 8170 ± 105 8015 ± 105 8110 ± 840 The averaged kinetic data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots were shown in Figure 5.13. The cyclic voltammogram for this complex is shown in Figure 5.14. The information extracted from these plots is provided in Table 5.6. Figure 5.13: “Arrhenius-type” plots for [Fe(4,4’-diOMe-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). Table 5.6: The transition state, Marcus theory, and electrochemical parameters for [Fe(4,4’- diOMe-bpy)3](PF6)2. Substituent ΔH‡ (cm-1) ΔS‡ (cm-1/K) 4,4’- diOMe 335 ± 35 -5.91 ± 0.13 ΔG‡ (cm-1) 2080 ± 70 |Hab|4/λ (cm-3) 0.029 ± 0.010 (ΔGo + λ)2/λ (cm-1) 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+) 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+) 2420 ± 130 0.37 -1.89 170 Figure 5.14: Cyclic voltammogram for [Fe(4,4’-diOMe-bpy)3]2+. 5.3.5 [Fe(5,5’-diOMe-bpy)3](PF6)2 The UV-Vis spectrum of the MLCT feature for [Fe(4,4’-diOMe-bpy)3](PF6)2 is provided in Figure 5.11, alongside the pump and probe wavelengths used throughout TA measurements. Collecting VT-TA data for this complex proved to be a challenge. Several trials were conducted with various pump/probe wavelength combinations in an attempt to improve s/n. The sapphire crystal to generate white light along the probe line proved to be a limiting factor, as very little power could be generated at blue wavelengths, yet the MLCT absorption feature is very blue shifted for this complex compared to the rest of the series. As a result, probe wavelengths that the sapphire crystal could generate at decent power ( e.g. 530 nm) led to very low s/n and kinetic data had high variance and large error bars. Eventually, a CaF2 crystal was set in place of the sapphire Figure 5.15: Ground state absorption spectrum for [Fe(5,5’-diOMe-bpy)3]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. 171 crystal, which was capable of generating decent power with a 460 nm probe. This probe wavelength provided better s/n and greatly improved the error bars on kinetic data. After the new pump/probe wavelength was used, two VT-TA trials were conducted. Shown in Figure 5.16 is the VT-TA data from the first trial. The kinetic data from each trial is summarized in Table 5.7. Figure 5.16: VT-TA data for [Fe(5,5’-diOMe-bpy)3]2+. Shown is the data from the first trial. Table 5.7: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. τ (ps) 295 K 285 K 275 K 265 K 255 K 245 K 235 K Trial 1 Trial 2 935 ± 30 950 ± 40 955 ± 20 975 ± 30 Average 945 ± 25 960 ± 30 1000 ± 35 1030 ± 30 1015 ± 30 1120 ± 40 1115 ± 35 1115 ± 30 1240 ± 50 1170 ± 40 1205 ± 50 1265 ± 40 1255 ± 40 1260 ± 30 1375 ± 35 1430 ± 45 1400 ± 40 The averaged kinetic data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots were shown in Figure 5.17. The cyclic voltammogram for this complex is shown in Figure 5.18. The information extracted from these plots is provided in Table 5.8. The ligand reduction peaks on CV collected under air show less reversibility, this is likely caused by the environment of data collection. However, the metal 172 oxidation peak remains clear and reversible, and the bulk of discussion surrounding CV data focuses on this information. Figure 5.17: “Arrhenius-type” plots for [Fe(5,5’-diOMe-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). Figure 5.18: Cyclic voltammogram for [Fe(5,5’-diOMe-bpy)3]2+. This and other CV data collected in air had the currents reversed, so the y-axis of the plot is reversed to better resemble the spectra collected under argon atmosphere, this has no impact on how E1/2 values are calculated. Table 5.8: The transition state, Marcus theory, and electrochemical parameters for [Fe(5,5’- diOMe-bpy)3](PF6)2. Substituent ΔH‡ (cm-1) ΔS‡ (cm-1/K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+) 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+) 5,5’-diOMe 155 ± 20 -5.47 ± 0.06 1770 ± 35 0.145 ± 0.010 1715 ± 60 0.66 -1.85 The pKb for the ligand of this complex was also measured. The titration curve is provided in Figure 5.19 and the pKb was measured to be 5.84 in acetone. 173 Figure 5.19: Titration curve of a 5,5’-diOMe-bpy solution in acetone titrated with 0.1 M nitric acid. pKb = 5.84. 5.3.6 Fe(4,4’-diCl-bpy)3](BF4)2 The UV-Vis spectrum of the MLCT feature for [Fe(4,4’-diCl-bpy)3](BF4)2 is provided in Figure 5.20, alongside the pump and probe wavelengths used throughout TA measurements. Figure 5.20: Ground state absorption spectrum for [Fe(4,4’-diCl-bpy)3]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. Two VT-TA trials were initially conducted, after this, two more trials were run focusing on temperatures where the τ had high error from the previous trials (though room temperature kinetic data was collected for each trial regardless). This was done to save time, VT-TA experiments require a significant amount of time, and multiple trials are often necessary to narrow error bars for the averaged data. Focusing on specific temperatures rather than recollecting data at all temperatures significantly sped up the process, and the last two trials were able to be collected 174 within one day. Shown in Figure 5.21 is the VT-TA data from the first trial. The kinetic data from each trial is summarized in Table 5.9. Figure 5.21: VT-TA data for [Fe(4,4’-diCl-bpy)3]2+. Shown is the data from the first trial. Table 5.9: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. τ (ps) Trial 1 Trial 2 295 K 1615 ± 35 1615 ± 40 285 K 1640 ± 40 1830 ± 45 Trial 3 1625 ± 30 2075 ± 50 Trial 4 Average 1605 ± 10 1615 ± 25 1610 ± 110 1790 ± 195 275 K 2000 ± 65 1915 ± 50 ― ― 1955 ± 65 265 K 1850 ± 50 2010 ± 50 1935 45 ± 20 2050 ± 240 1960 ± 130 255 K 2220 ± 60 2360 ± 55 2060 ± 270 245 K 2340 ± 70 2365 ± 110 2350 ± 60 235 K 2400 ± 90 2515 ± 70 ― ― ― ― 2215 ± 180 2355 ± 65 2505 ± 70 The averaged kinetic data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots were shown in Figure 5.22. The cyclic voltammogram for this complex is shown in Figure 5.23. The information extracted from these plots is provided in Table 5.10. Table 5.10: The transition state, Marcus theory, and electrochemical parameters for [Fe(4,4’-diCl- bpy)3](BF4)2. Substituent ΔH‡ (cm-1) ΔS‡ (cm-1/K) ΔG‡ (cm-1) |Hab|4/λ (cm-3) (ΔGo + λ)2/λ (cm-1) 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+) 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+) 4,4’-diCl 165 ± 30 -5.85 ± 0.11 1890 ± 60 0.049 ± 0.017 1750 ± 120 0.91 -1.43 175 Figure 5.22: “Arrhenius-type” plots for [Fe(4,4’-diCl-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). Figure 5.23: Cyclic voltammogram for [Fe(4,4’-diCl-bpy)3]2+. The pKb for the ligand of this complex was also measured. The titration curve is provided in Figure 5.24 and the pKb was measured to be 5.08 in acetone. Figure 5.24: Titration curve of a 4,4’-diCl-bpy solution in acetone titrated with 0.1 M nitric acid. pKb = 5.08. 176 5.3.7 Fe(5,5’-diCl-bpy)3](BF4)2 Figure 5.25: Ground state absorption spectrum for [Fe(5,5’-diCl-bpy)3]2+ solution within the visible region. This region is dominated by the MLCT excitation feature. The UV-Vis spectrum of the MLCT feature for [Fe(5,5’-diCl-bpy)3](BF4)2 is provided in Figure 5.25, alongside the pump and probe wavelengths used throughout TA measurements. Two VT-TA trials were initially conducted, after this, two more trials were run focusing on temperatures where the τ had high error from the previous trials. That said, collecting data with good s/n for this complex proved challenging for similar reasons as the 5,5’-diOMe substituted complex, so more focus was put on reduced the error bars on the “Arrhenius-type” plot fitted parameters (slope and y-intercept) than the kinetic data. Shown in Figure 5.26 is the VT-TA data from the first trial. The kinetic data from each trial is summarized in Table 5.11. Figure 5.26: VT-TA data for [Fe(5,5’-diCl-bpy)3]2+. Shown is the data from the first trial. 177 Table 5.11: The GSR τ in ps across each temperature. The averaged data is boldened for ease of reading. τ (ps) 295 K 285 K 275 K 265 K 255 K 245 K 235 K Trial 1 Trial 2 Trial 3 1715 ± 70 1790 ± 135 1785 ± 50 Trial 4 ― Average 1765 ± 85 2015 ± 140 2075 ± 120 2020 ± 110 1925 ± 140 2010 ± 120 2145 ± 115 2445 ± 155 2180 ± 215 1890 ± 140 2165 ± 235 2220 ± 110 2435 ± 190 2625 ± 200 2530 ± 160 ― ― 2750 ± 180 2675 ± 265 3175 ± 235 3600 ± 320 3020 ± 225 ― ― ― ― ― 2330 ± 165 2575 ± 155 2865 ± 290 3310 ± 370 The averaged kinetic data was used to construct “Arrhenius-type” plots and analyzed using transition state and Marcus theory. These plots were shown in Figure 5.27. The cyclic Figure 5.27: “Arrhenius-type” plots for [Fe(5,5’-diCl-bpy)3]2+ analyzed through transition state theory (A) and Marcus theory (B). Figure 5.28: Cyclic voltammogram for [Fe(5,5’-diCl-bpy)3]2+. The peak near -0.3 V vs Fc/Fc+ is likely caused by oxygen. 178 voltammogram for this complex is shown in Figure 5.28. The information extracted from these plots is provided in Table 5.12. Table 5.12: The transition state, Marcus theory, and electrochemical parameters for [Fe(4,4’- diCl-bpy)3](BF4)2. ΔH‡ (cm-1) (ΔGo + λ)2/λ (cm-1) 𝑂𝑥 𝐸1/2 (V vs Fc/Fc+) ΔS‡ (cm-1/K) |Hab|4/λ (cm-3) ΔG‡ (cm-1) Substituent 𝑅𝑒𝑑 𝐸1/2 (V vs Fc/Fc+) 5,5’-diCl 310 ± 25 -5.42 ± 0.08 1905 ± 45 0.168 ± 0.043 2320 ± 90 1.05 -1.41 The pKb for the ligand of this complex was also measured. The titration curve is provided in Figure 5.29 and the pKb was measured to be 5.79 in acetone. Figure 5.29: Titration curve of a 5,5’-diCl-bpy solution in acetone titrated with 0.1 M nitric acid. pKb = 5.79. 5.3.8 Determining Driving Force Order As with the previous series, the 1A1 → 3T1 excitation energies were recorded to provide insight on how the energy gap between the t2g and eg* orbitals (the LF orbitals) changes across the series. Unlike the 4’-substituted [Fe(terpy)2]2+ series, the 1A1 → 3T1 absorption feature was visible across all [Fe(bpy)3]2+ complexes. This means there was no need to rely on Co(III)-based derivatives. The UV-Vis spectra of the 1A1 → 3T1 absorption feature across the series is provided in Figure 5.30. Each spectrum was deconvolved using Igor Pro’s multipeak fitting tool. The deconvolution fit for [Fe(4,4’-diCl-bpy)3](BF4)2 is provided in Figure 5.31 and the energies of each 3T1 absorption across the series are summarized in Table 5.13. 179 Figure 5.30: UV-Vis spectra of the 3T1 absorption feature across a series of substituted [Fe(bpy)3]2+ complexes. The dips observed in the spectra are artifacts resulting from the removal of solvent features via subtracting a solvent blank. of Table 5.13: The 3T1 excitation [Fe(bpy)3]2+ the energies complexes. These energies are used to establish trends in LF strength. 3T1 Absorption (cm-1) Substituent H 12440 5,5’-diMe 12440 4,4’-diCl 12330 5,5’-diCl 12240 4,4’-diMe 12200 5,5’-diOMe 12140 4,4’-diOMe 12110 Figure 5.31: Molar Absorptivity of the 1A1 → 3T1 absorption feature for [Fe(4,4’-diCl-bpy)2](BF4)2. The sample was prepared as a ~5 mM solution in acetonitrile in a 10 cm pathlength cell. Igor Pro’s multipeak fitting tool was used to deconvolve the gaussian of the peak, providing the energy needed for the excitation. 5.3.9 Determining Inner-Sphere Reorganization Energy The inner-sphere reorganization energy (λi) was once again estimated using DFT calculations by following the Nelsen method described in Chapter 3.9 It was found that λi changed very little across the series. Standard deviation of λi values across the series was 140 cm-1, which is about 1.5% of the energy value for the proto-substituted complex. This means that any perturbations in λ can largely be attributed to changes in outer-sphere reorganization energy (λo). 180 As will be explored in detail in a future section, changes in λo across the series will have systematic and noticeable impacts the excited state dynamics for many complexes. The λi calculated for each complex is provided in Table 5.14. Table 5.14: DFT calculated λi across the series. Calculations were performed using the Nelsen method previously described. Substituent H 4,4’- diMe 5,5’- diMe 4,4’- diOMe 5,5’- diOMe 4,4’- diCl 5,5’- diCl Inner-Sphere Reorganization Energy (cm-1) 5.4 Discussion 9210 9095 9130 8830 9245 9220 9100 5.4.1 Controlling the Frontier Orbitals through Substitution The orbitals that define the ligand field in octahedral complexes, eg* and t2g, are influenced by the σ- and π-bonding strength, respectively, of the ligand. If the ligand is a stronger σ-donor, the σ-bond between ligand and metal will be stronger which will lead to greater splitting between the eg and eg* orbitals, destabilizing the eg* orbitals. Likewise, a weaker σ-donating ligand will lead to a more stable eg* orbital. The t2g orbitals are dependent on the π-donating strength of the ligand, a strong π-donating ligand will destabilize the t2g orbitals and a π-accepting ligand will do the opposite. These principles are explored in depth in Chapter 1.3 and illustrated in Figure 5.32. y g r e n E Figure 5.32: Schematic displaying how ligand σ- and π-bonding strengths impact the energy levels of the eg* and t2g orbitals, respectively. Whether each orbital set is stabilized or destabilized directly impacts the LF strength (represented by Δo) of the complex. The strength of ligand σ-donation is impacted by substituents through inductive effects. An electron donating (ED) substituent pushes electron density into the σ-bond network of the pyridine 181 ring, pushing higher electron density onto the nitrogen bound to the metal center. With higher electron density in the σ-orbitals of the nitrogen (the p orbitals that are involved in the formation of σ-bonds), the ligand becomes a stronger σ-donor. Inversely, an electron withdrawing (EW) substituent pulls electron density through the σ-bonds away from the σ-orbitals of the nitrogen, causing the ligand to become a worse σ-donor. The position of the substituent also affects how strongly the substituent can impact σ-donating strength. The further the substituent is from the nitrogen, the weaker the influence it will have on σ-donating strength. These substituent effects are illustrated in Figure 5.33. Figure 5.33: Schematic depicting how inductive effects influence the σ-donating strength of the ligand. A and B depict ED substituents which push electron density onto the nitrogen through induction, while C and D depict EW substituents which pull electron density away from the nitrogen through induction. Ligands with substituents placed at the 4,4’-positions have the substituent placed further from the nitrogen than ligands substituted at the 5,5’-positions, which result in weaker σ-donating impact compared to the 5,5’-substituted ligands. While the schematics here focus on σ-bonds, the same principles apply to how induction can influence the strength of π- donation by acting through the π-system of the ligand. The relative σ-donating strength of a ligand can be determined through measuring the basicity (represented through pKb) of the ligand. A lower pKb indicates a ligand more strongly donates the lone pair on the nitrogen, which directly correlates with the σ-donating strength of the ligand. Therefore, a low pKb value indicates that the ligand is a strong σ-donor and a high pKb indicates that the ligand is a weak σ-donor. The pKb of each ligand is summarized in Table 5.15: 182 Table 5.15: The pKb values of ligands used in the series. Different solvents needed to be used for ease of reading the equivalence point, and as such direct comparisons between values should not be drawn. However, trends in values can still be used to interpret how substituents impact σ- donating strength. †: Measured in acetonitrile. ‡: Measured at 3.33 in dilute dioxane, 4.33 – 4.44 in water, and 4.12 in dilute ethanol.10,11 α: Measured in acetone. Substituent 5,5’-diMe† 4,4’-diMe† H‡ 4,4’-diClα 5,5’-diClα 5,5’-diOMeα pKb 1.92 3.16 3.33 – 4.44 5.08 5.79 5.84 As predicted, the diMe (ED substituents) substituted ligands have lower pKb values while the ligands substituted with diCl (EW substituents) have higher pKb values, meaning diMe-substituted bpy is a much stronger σ-donor. Furthermore, the 5,5’-substituted ligands have more extreme pKb value changes than the 4,4’-substituted ligands. The 5,5’-diOMe substituted ligand has a similar pKb to the 5,5’-diCl substituted ligand, which displays that each substituent has roughly equivalent impact on σ-donating strength. Both chlorine and oxygen have similar electronegativity values,12,13 so it is unsurprising that each should have similar inductive strengths. This means that each substituent/placement combination has a predictable impact on the eg* energy. These energetic perturbations are illustrated in Figure 5.34. y g r e n E Figure 5.34: Schematic displaying how different substituents impact the eg* orbital energy. The ED substituent (diMe) destabilizes the eg* orbitals while the EW substituents stabilize the eg* orbitals. These effects are strengthened at with substituents placed at the 5,5’-positions compared to the 4,4’-positions. The next point to discuss is how the t2g orbitals are influenced. As stated before, the t2g orbital energy is directly tied to the π-donating/accepting strength of the ligand. The inductive effects illustrated in 5.33 are applicable to the π-system as well as the σ-bonds, however resonance 183 effects also play a role in determining the π-donating/accepting ability of a ligand. As displayed in Figure 5.2, a substituent with lone pairs within its π-orbitals (the orbitals involved in forming π- bonds) can enable resonance structures that push electron density into the π-orbitals of the nitrogen on the ligand, making the ligand a stronger π-donor. Figure 5.2 further shows that if the substituent is placed at the 5,5’-positions, the impact of resonance effects is weakened compared to substitution at the 4,4’-positions. This is in stark contrast to inductive effects, which are strengthened with substituents at the 5,5’-positions. Since the complexes in the series are all low spin, the relative impacts of substituents on the t2g orbitals can be directly measured through the Fe2+/Fe3+ oxidation potentials measured via cyclic voltammetry (CV). As the t2g orbital is destabilized, a weaker potential (closer to zero) will be required to oxidize the Fe(II) center to Fe(III). Additionally, the relative electron density within the ligand can also be observed through CV, by measuring the bpy/bpy·- reduction potential. If there is higher electron density within the π-system of the bpy ligand, a stronger potential (further from zero) will be required to reduce the ligand into the bpy radical, and a higher electron density within the π-system of the ligand corresponds to stronger π-donating ability from the ligand. The electrochemical potentials across the series are summarized in Table 5.16: Table 5.16: Electrochemical data for the Fe2+/Fe3+ oxidation potentials (𝐸1/2 reduction potentials (𝐸1/2 𝑅𝑒𝑑) across the series. 𝑂𝑥 ) and bpy/bpy·- Substituent 𝑂𝑥 (V vs Fc/Fc+) 𝐸1/2 𝑅𝑒𝑑 (V vs Fc/Fc+) 𝐸1/2 4,4’-diOMe 4,4’-diMe 5,5’-diMe 5,5’-diOMe H 4,4’-diCl 5,5’-diCl 0.37 -1.89 0.52 -1.84 0.59 -1.88 0.66 -1.85 0.68 0.91 -1.74 -1.43 1.05 -1.41 The bpy/bpy·- reduction potentials reveal that while the diMe and diOMe substituents increase π- donation, diCl substitution reduces π-donation. Having no lone pairs to donate for resonance, the diMe substituents must increase π-density within the ligand through induction alone. However, the diCl substituents are electronegative and can donate lone pairs for resonances structures, which have competing effects on π-donation. The reduction potentials reveal that the electronegativity of chlorine is strong enough to overcome the resonance effects, causing the diCl-substituted ligands to act as net π-acceptors. It is unclear based on available data whether the ligand acts as a π- acceptor or as a weaker π-donor, both scenarios result in the stabilization of the t2g orbitals so this distinction is not important for the analysis in this chapter, out of simplicity the ligands will be referred to as net π-acceptors. Conversely, the diOMe-substituted ligands acted as net π-donors, 184 Figure 5.35: DFT calculated geometry for the singlet state of [Fe(4,4’-diOMe-bpy)3]2+, the hydrogens have been hidden to improve clarity. Shown within the inset are p-orbitals representing the π-system of one of the 4,4’-diOMe-bpy ligands. The methyl groups being able to rotate about the oxygen hinder the electron withdrawing inductive effect the oxygen might otherwise produce for the ligand. with ligand π-density similar to the diMe substituted complex. Similarly to the diCl substituents, diOMe has competing resonance and induction effects, however in this case the π-donating resonance is stronger than the π-withdrawing induction, despite oxygen being more electronegative than chlorine. This is because the methyl group in the methoxy substituent can rotate freely around the oxygen, momentarily overlapping with the oxygen π-orbitals, illustrated in Figure 5.35. Being electron donating itself, the methyl group interferes with the electron withdrawing strength of the oxygen. Therefore, the methyl group weakens the inductive strength of the methoxy substituent enough for the resonance effects to dominate, leading the diOMe- substituted ligand to act as a net π-donor. As the chlorine substituent is mono-atomic, there is nothing to interfere with the induced electron withdrawing of the diCl substituents, so they remain net π-accepting. The Fe(II) oxidation potentials follow the same trend as the ligand reduction potentials, where net π-donating ligands destabilize the t2g orbitals (being easier to oxidize) and the net π- accepting ligands stabilize the t2g orbitals (being more difficult to oxidize). Furthermore, as 185 predicted by the resonance structures shown in Figure 5.2, π-interactions are weakened when the substituents are placed at the 5,5’-positions: for example, t2g orbitals are more destabilized with π- y g r e n E Figure 5.36: Schematic displaying how different substituents impact the t2g orbital energy. Net π- donors destabilize the t2g orbitals while net π-acceptors stabilize the t2g orbitals. donating substituents at 4,4’-positions. A schematic representing the relative t2g energy levels based on the Fe2+/Fe3+ oxidation potentials is shown in Figure 5.36. The data observed for t2g and eg* orbital energy shifts can be used alongside the 1A1 → 3T1 absorption energies to gain an understanding of how LF strength (and therefore the GSR ΔGo) changes across the series. Figure 5.37 displays the energetic changes of the LF orbitals across the series in order of decreasing LF strength (consistent with the 1A1 → 3T1 absorption energies provided in Table 5.13). 5.4.2 Long-Lived Excited State Dynamics The excited state dynamics of the diMe-substituted complexes are directly impacted by the position of the substituents. As stated before, the t2g and eg* orbitals are all destabilized with the addition the diMe substituents, the key difference is how strongly each orbital is destabilized. Since y g r e n E Figure 5.37: Schematics of the LF orbitals across the series ordered from strongest LF to weakest. The substituents correlating to each model are listed below. 186 substituent π-interactions are stronger at the 4,4’-positions, the t2g orbitals are more destabilized in the 4,4’-diMe substituted complex, and as substituent σ-interactions are strengthened at the 5,5’- positions, the eg* orbitals are more destabilized in the 5,5’-diMe substituted complex. The net result is a reduced LF strength in the 4,4’-diMe substituted complex, but in the 5,5’-diMe substituted complex, the eg* and t2g orbitals seem to have been destabilized to the same magnitude. The 5,5’-diMe substituted complex has a LF strength equivalent to that of the unsubstituted complex. The excited state kinetic data reflects this interpretation of the data. With a lower LF strength, the 5T2 → 1A1 relaxation will have a lower ΔGo in the 4,4’-diMe substituted complex compared to the unsubstituted and 5,5’-diMe substituted complexes. Decreasing ΔGo increases the enthalpy of activation (ΔH‡) within the Marcus normal region, as seen in the following equation which was derived in Chapter 3: 𝛥𝐺‡ = [( (𝛥𝐺𝑜+𝜆)2 4𝜆 ) − 𝛽𝐻] − 𝑇 [( 𝑘𝐵 2 ln ( 4𝜋3|𝐻𝑎𝑏|4 𝜆𝜅2(𝑘𝐵𝑇)3)) − 𝛽𝑆] 5.1 ‡ ΔH ‡ ΔS Increasing ΔH‡ leads to a higher energy barrier (ΔG‡) for the transition, which increases the lifetime of the excited state. Meanwhile, since the 5,5’-susbtituted complex has a LF strength on par with the unsubstituted complex, both the barrier and lifetime should match those of the unsubstituted complex. This is exactly what was observed in the spectroscopic results for each complex, summarized in Table 5.17: Table 5.17: Summary of the 5T2 excited state lifetimes at room temperature (τ295) and the transition state theory parameters for the diMe-substituted complexes in the series (at 295 K). Substituent τ295 (ps) ΔG‡ (cm-1) ΔH‡ (cm-1) ΔS‡ (cm-1 * K) H 1050 ± 20 1795 ± 40 125 ± 20 -5.66 ± 0.07 4,4’-diMe 1320 ± 20 1830 ± 15 160 ± 10 -5.66 ± 0.02 5,5’-diMe 1035 ± 20 1790 ± 20 120 ± 10 -5.68 ± 0.03 for Figure 5.38: Schematics of the LF diMe-substituted orbitals complexes, estimated via pKb, CV, and UV-Vis data. The substituents correlating to each model are listed below. the Unlike the diMe-substituted complexes, which has little influence on the λ of the complex, the excited state dynamics of the diOMe and diCl-substituted complexes are heavily influenced by perturbations in λ, specifically λo. Therefore, before looking into the energy barrier and lifetimes for these complexes, it is helpful to establish how λ changes with diOMe and diCl substitution. 187 This can be done by looking at the experimentally determined |Hab|4/λ ratio in combination with the relative LF strength determined through analysis of the 1A1 → 3T1 excitation energies. These data sets are provided in Table 5.18. Additionally, ΔH‡ and the entropy of activation (ΔS‡) is included in Table 5.18, as both factors are dependent on λ, as seen in equation 5.1. Understanding how these parameters change with λ will illuminate how ΔG‡ and τ295 change across the series. Table 5.18 shows that both diOMe-substituted complexes should have very similar LF strengths, the diCl-substituted complexes do as well but the energy gap between these complexes is notably greater than the energy gap between the diOMe-substituted complexes. It was established in the previous chapter that Hab is directly related to LF strength, so GSR in the diOMe-substituted complexes should have comparable Hab values, however the |Hab|4/λ ratios are wildly different between each complex. The 5,5’-substituted complex has a much higher ratio value, even higher than the unsubstituted complex, which should have the strongest electronic coupling based on LF strength. While the 4,4’-susbtituted complex has a very low ratio value. The same trend can be observed in the diCl-substituted complexes. If Hab can be treated as relatively constant, then a lower λ value will result in a higher ratio value, and vice versa. Therefore, since the 5,5’-substituted complexes have higher ratio values, it can be reasonably assumed that placing the substituents at the 5,5’-positions reduce λ with respect to the same substituents being placed at the 4,4’-positions. Table 5.14 shows that λi changes very little across the series, so these noticeable shifts in λ likely stem from outer-sphere interactions. This could explain the position dependency on λ, and why these effects are muted with the diMe substituents. All VT-TA data was collected on solutions prepared in acetonitrile, and both the diOMe and diCl substituents have electronegative atoms that the polar solvent molecules are able to interact with. At the 4,4’-positions these substituents strongly interact with the ligand π-system which allows these polar solvent molecules to interact with the ligand π-system more strongly than when the substituents are at the 5,5’-positions. Notably, the |Hab|4/λ ratio value is higher than expected for the 4,4’-diMe substituted complex, having a notably lower LF strength, the ratio should also be reduced compared to the unsubstituted complex. This means that λ is decreased and non-polar substituents have the opposite effect on polar solvent molecules, likely because the presence of the substituent introduces steric hinderance, albeit very weakly, as the solvent attempts to interact with the pyridine rings. This effect would likely be increased with a bulkier alkyl substituent, such as tert-butyl groups. 188 Table 5.18: Spectroscopic data across the series. An increase in |Hab|4/λ value is largely attributed to a decrease in λ, and vice versa. This change in λ directly causes observable changes in both ΔH‡ and ΔS‡. Substituent H 5,5’-diMe 4,4’-diCl 5,5’-diCl 4,4’-diMe 5,5’-diOMe 4,4’-diOMe 3T1 Absorption (cm-1) |Hab|4/λ (cm-3) 12440 12440 12330 12240 12200 12140 12110 0.086 ± 0.017 0.080 ± 0.008 0.049 ± 0.017 0.168 ± 0.043 0.082 ± 0.007 0.145 ± 0.025 0.029 ± 0.010 ΔH‡ (cm-1) 125 ± 20 120 ± 10 165 ± 30 310 ± 25 160 ± 10 155 ± 20 335 ± 35 ΔS‡ (cm-1/K) -5.66 ± 0.07 -5.68 ± 0.02 -5.85 ± 0.11 -5.42 ± 0.08 -5.66 ± 0.03 -5.47 ± 0.06 -5.91 ± 0.13 As stated before, these changes in λ cause observable changes in both ΔH‡ and ΔS‡. Shown in equation 5.1, ΔH‡ is dependent on both ΔGo and λ. If ΔGo is relatively unchanged (which the UV-Vis data for the 1A1 → 3T1 excitation feature suggests is the case for the diOMe complexes) then increasing λ will increase ΔH‡. We see this with the ΔH‡ values for the diOMe complexes. The 4,4’-diOMe complex, which has higher λ than the 5,5’-diOMe substituted complex, has a ΔH‡ value of 335 ± 35 cm-1 while the 5,5’-diOMe substituted complex has a value of only 155 ± 20 cm-1. However, the inverse is observed with the diCl complexes, the 5,5’-diCl substituted complex has the higher ΔH‡. Keeping in mind that the LF strength gap between the diCl substituted complexes is larger than the LF strength gap between the diOMe complexes and that the 5,5’- substituted complex has the weaker LF strength between the diCl-substituted complexes. Additionally, it is worth noting that the energy differences between the 1A1 → 3T1 excitation energies do not directly scale to ΔGo differences for the 5T2 → 1A1 relaxation, even though the trends should remain the same. Since formation of the 3T1 state only involves only one electron, while relaxing from the 5T2 state involves two electrons, energy differences between ΔGo across the series are likely greater than the energy differences for the 1A1 → 3T1 excitation. Under this context, the ΔH‡ of the diCl complexes are consistent with Marcus normal region behavior. The impact of λ on ΔS‡ is clearly observable. Equation 5.1 shows that ΔS‡ is dependent on − ln(𝜆). This means that increasing λ will decrease ΔS‡ and decreasing λ will increase ΔS‡ (in this case, more negative and less negative, respectively). This is exactly what is observed in the data, the 4,4’-position diOMe and diCl substituted complexes have ΔS‡ values lower than the unsubstituted complex, while the 5,5’-position counterparts have higher ΔS‡ values. 189 These changes in ΔH‡ and ΔS‡ values directly inform how ΔG‡ and τ295 change within these complexes. For example, the low LF strength and high λ in the 4,4’-diOMe substituted complex leads to a higher ΔH‡ and more negative ΔS‡. This combination leads to the greatest ΔG‡ and therefore longest τ295 across the series at 2080 ± 70 cm-1 and 4360 ± 460 ps, respectively. This information for the diOMe and diCl-substituted complexes is summarized in Table 5.19. In contrast to the 4,4’-diOMe substituted complex, the 5,5’-substituted complex has a ΔH‡ value that is slightly higher than the unsubstituted complex but a weaker activation entropic cost (less negative ΔS‡), this results in a ΔG‡ value that is within error of the unsubstituted complex at 1770 ± 70 cm-1. The reduced energy barrier results in a faster relaxation, with a GSR lifetime of 945 ± 25 ps at 295 K for the 5,5’-diOMe substituted complex. Table 5.19: Summary of the 5T2 excited state lifetimes at room temperature (τ295) and the transition state theory parameters for the diOMe and diCl-substituted complexes in the series (at 295 K). Substituent τ 295 (ps) ΔG‡ (cm-1) ΔH‡ (cm-1) ΔS‡ (cm-1 * K) H 1030 ± 15 1795 ± 40 125 ± 20 -5.66 ± 0.07 4,4’-diOMe 4360 ± 460 2080 ± 70 335 ± 35 -5.91 ± 0.13 4,4’-diCl 1615 ± 30 1890 ± 60 165 ± 30 -5.85 ± 0.11 5,5’-diOMe 945 ± 25 1770 ± 35 155 ± 20 -5.47 ± 0.06 5,5’-diCl 1760 ± 80 1905 ± 45 310 ± 25 -5.42 ± 0.08 The excited state dynamics of the diCl substituted complexes can be similarly explained. Despite having a ΔH‡ similar in energy to the 4,4’-diMe substituted complex, the 4,4’-diCl substituted complex has a greater λo (based on the previously described inferences from |Hab|4/λ ratios, the exact λo values are not known), which results in a more negative ΔS‡ and therefore greater ΔG‡ and longer τ295, at 1890 ± 60 cm-1 and 1615 ± 30 ps respectively. On the other hand, the 5,5’-diCl complex has a ΔH‡ within error of the 4,4’-diOMe complex, but similarly to the 5,5’- diOMe substituted complex, has a lower λo which results in a weaker activation entropic cost. The net result is a ΔG‡ value that rests between the 4,4’-diCl and 4,4’-diOMe substituted complexes at 1905 ± 45 cm-1, resulting in a GSR lifetime of 1760 ± 80 ps, which has observed values between the former complexes. Overall, the long-lived excited state dynamics within the series of substituted [Fe(bpy)3]2+ complexes are consistent with the expected trends in LF strength determined through pKb, CV, and UV-Vis analysis. However, a key factor in understanding the excited state dynamics is λ, which 190 plays a key role in both ΔH‡ and ΔS‡ depending on the position of the substituent. By examining trends in the experimentally determined |Hab|4/λ ratio, it was found that placing polar substituents at the 4,4’-position increases λo in a polar solvent but placing the same substituents at the 5,5’- positions reduces λo. Additionally, if the substituent is non-polar, placing the substituent at the 4,4’- position reduces λo in polar solvents, while the same substituent at the 5,5’ position had no observable impact on λo. Complexes with higher λ values had increased enthalpic and entropic activation costs, leading to a longer lived 5T2 excited state. The diOMe-substituted complexes display clearly the impact that λo has on the excited state dynamics. Despite having similar LF strengths, each diOMe-substituted complex has dramatically different excited state dynamics. The 4,4’-diOMe substituted complex has the greatest ΔG‡ and longest 5T2 lifetime across the series while the 5,5’-substituted complex has the lowest ΔG‡ and shortest 5T2 lifetime. 5.5 Conclusions Through the use of VT-TA spectroscopy combined with pKb analysis, cyclic voltammetry, and UV-Vis spectroscopy; the excited state dynamics and ligand field orbital energies were examined across a heterologous series of substituted [Fe(X,X’-diR-bpy)3]2+ complexes (where X is 4 or 5 and R is Me, OMe, or Cl). Through pKb analysis, it was found that placing electron donating substituents on the bpy ligand destabilized the eg* orbitals (which a controlled by σ- interactions) while electron withdrawing substituents stabilize eg* orbitals; each acting through induction effects which were strengthened when substituents were moved to the 5,5’-positions compared to the 4,4’-positions. Through cyclic voltammetry, it was found that induction effects on the π-system operated similarly to σ-induction, with the key difference being that π-interactions were weakened at the 5,5’-positions and strengthened at the 4,4’-positions (opposite to what was seen for σ-interactions). However, in addition to inductive effects, t2g orbital energy levels were also dependent on resonance effects. The diOMe and diCl substituents have competing inductive and resonance effects. Oxygen and chlorine are strongly electronegative, so inductive effects reduce π-density in the orbitals, however they also have lone pairs to donate for the formation of resonance structures which push π-density onto the ligand. It was found that for the diOMe- substituted complexes, the resonance effects overpower the inductive effects (because the methyl group on the substituent interferes with the electronegativity of the oxygen), so the substituent acts as a net π-donor despite the same substituent weakening σ-donation. On the other hand, it was found that the inductive effects of the diCl-substituted complexes overpowered the resonance 191 effects, causing the substituent to act as a net π-acceptor. This resulted in the t2g orbitals of the diMe and diOMe-substituted complexes being destabilized, with the 4,4’-subtituted complexes having those orbitals destabilized more than the 5,5’-subtituted complexes, and t2g orbitals of the diCl-substituted complexes were stabilized. In addition to the substituents and their positioning influencing ligand field orbital energies, and therefore the ΔGo of the 5T2 excited state relaxation, substituent position also impacts outer- sphere reorganization effects. It was found that placing electronegative substituents at the 4,4’- positions greatly increased outer-sphere reorganization energy in polar solvents, while placing the same substituents at the 5,5’-positions reduced outer-sphere reorganization energy. Alternatively, placing a non-polar substituent at the 4,4’-position reduced outer-sphere effects in the same solvent. These perturbations on outer-sphere interactions caused directly observable changes in the long-lived excited state dynamics. These impacts are highlighted in the diOMe complexes, which have similar ΔGo and, as a result, Hab values, so any changes to excited state dynamics stem from changes in reorganization energy. The 4,4’-diOMe complex VT-TA data show that by increasing outer-sphere reorganization energy, the energy barrier and excited state lifetime both increase. The same data for the 5,5’-diOMe complex displays that reducing outer-sphere reorganization energy leads to a reduction in energy barrier and excited state lifetime. These findings exhibit how substituent positioning can be used to manipulate the ligand field orbital energy levels and excited state dynamics within polypyridyl-based chromophores. 5.6 Future Directions With the goal of improving the efficacy that [Fe(bpy)3]2+ complexes engage in electron transfer from the long-lived 5T2 states, the pKb and cyclic voltammetry data provided a clear direction for substitution. Figure 5.1 shows that to improve the efficiency of electron transfer from a ligand field excited state, it is important to destabilize both the eg*and t2g orbitals. Destabilizing the eg* orbitals should make electron transfer to an acceptor more energetically favorable, while destabilizing the t2g orbitals should keep the ligand field strength low which results in a longer- lived excited state. The results of pKb analysis show that having electron donating substituents, such as diMe, placed at the 5,5’-positions leads to the strongest destabilization of the eg* orbitals. Meanwhile, CV data showed that strong resonance effects from the 4,4’-position led to the greatest destabilization of the t2g orbitals. In the series being studied, the diOMe substituent led to the greatest t2g destabilization, however it was found that π-inductive effects from electronegative 192 substituents competed with π-resonance effects. Therefore, using a less electronegative substituent, such as methylthiol (SMe), should destabilize the t2g orbitals further. Figure 5.39: A schematic of 4,4’-diSme;5,5’-diMe-bpy showcasing the σ-inductive effects and π- resonance effects that the substituents would enable. This should increase both the σ- and π- donating strength of the ligand. This would result in a complex using this ligand having destabilized t2g and eg* orbitals. Ideally, both effects should be combined, to yield the best results. Therefore, using a ligand with methyl substituents at the 5,5’-positions and methylthio substituents at the 4,4’-positions, like what is pictured in Figure 5.39, would be a viable next step. An Fe(II)-based complex using this ligand should be studied through the same methods outlined in this chapter and should be tested through quenching studies. Additionally, by having a polarizable atom placed at the 4,4’-positions, the outer-sphere reorganization energy for 5T2 relaxation should increase, which further increases the excited state lifetime. The electron withdrawing σ-inductive effects from the sulfur may dampen the effectiveness of the methyl groups as σ-donors. The electronegativity of sulfur may be addressed by using a structure similar to a formyl substituent, where a carbon atom sits between the oxygen and ligand. The results from pKb analysis on the ligands shows that adding additional distance between the substituent and the nitrogen weaken σ-induction, however without an additional methyl group on the electronegative atom to counter the electronegativity of the π-orbitals, electron withdrawing π- induction may overpower electron donating π-resonance, as was seen with the diCl-substituted complexes. Studying diformyl-substituted complexes and comparing the results to the diOMe substituted complexes should illuminate how such a strategy would impact the ligand field orbital energies. 193 1. Woodhouse, M. D. and McCusker, J. K. J. Am. Chem. Soc. 2020, 142, 16229 – 16233. REFERENCES 2. Sutin, N. Prog. Inorg. Chem. 1983, 30, 441 – 498. 3. Ferrere, S. and Gregg, B. A. J. Am. Chem. Soc. 1998, 120, 843 – 844. 4. Carey, M. C.; Adelman, S. L.; McCusker, J. K. Chem. Sci. 2019, 10, 134 – 144. 5. Ladouceur, S.; Swanick, K. N.; Gallagher-Duval, S.; Ding, Z. and Zysman-Colman, E. Eur. JIC. 2013, 30, 5329 – 5343. 6. Chadwick, N.; Kumar, D. K.; Ivaturi, A.; Grew, B. A.; Upadhyaya, H. M.; Yellowlees, L. J. and Robertson, N. Eur. JIC. 2015, 29, 4878 – 4884. 7. Creutz, C.; Chou, M.; Netzel, T. L.; Okumura, M. and Sutin, N. J. Am. Chem. Soc. 1980, 102, 1309 – 1319. 8. Hong, Y.-R. and Gorman, C. B. J. Org. Chem. 2003, 68, 9019 – 9025. 9. Nelsen, S. F.; Blackstock, S. C. and Kim, Y. J. Am. Chem. Soc. 1987, 109, 677 – 682. 10. Mullay, J. Structure and Bonding. 1987, 66, 1 – 25. 11. Sanderson, R. T. J. Am. Chem. Soc. 1983, 105, 2259 – 2261. 12. T. R. Harkins and H. Freiser. J. Am. Chem. Soc. 1955, 77, 1374 – 1376. 13. P. Krumholz. J. Am. Chem. Soc. 1951, 73, 3487 – 3492. 194 APPENDIX: SUPPLEMENTAL INFORMATION 5.SI.1 Synthetic Characterization All complexes were characterized via 1H NMR using a Bruker 500 MHz spectrometer. The NMR data provided below was collected by myself and Bekah Bowers. Characterization of [Fe(bpy)3](PF6)2: 1H NMR (500 MHz, Acetone-d6) δ 8.85 (dt, J = 8.1, 1.1 Hz, 1H), 8.28 (td, J = 7.8, 1.5 Hz, 1H), 7.77 – 7.72 (m, 1H), 7.59 (ddd, J = 7.3, 5.6, 1.3 Hz, 1H). Characterization of [Fe(4,4’-diMe-bpy)3](PF6)2: 1H NMR (500 MHz, Acetone-d6) δ 8.70 (dt, J = 1.8, 0.8 Hz, 1H), 7.51 (d, J = 5.8 Hz, 1H), 7.40 (ddd, J = 5.8, 1.8, 0.8 Hz, 1H), 2.58 (s, 3H), 2.09 (s, 2H). Characterization of [Fe(5,5’-diMe-bpy)3](PF6)2: 1H NMR (500 MHz, acetone-d6) δ 8.63 (d, J = 8.3 Hz, 2H), 8.04 (dd, J = 8.2, 1.9 Hz, 2H), 7.45 (d, J = 1.9 Hz, 2H), 2.19 (s, 6H). Characterization of [Fe(4,4’-diOMe-bpy)3](PF6)2: 1H NMR (500 MHz, Acetonitrile-d3) δ 8.07 (d, J = 2.7 Hz, 1H), 7.22 (d, J = 6.5 Hz, 1H), 6.98 (dd, J = 6.6, 2.7 Hz, 1H), 4.01 (s, 3H). Characterization of [Fe(5,5’-diOMe-bpy)3](PF6)2: 1H NMR (500 MHz, Acetonitrile-d3) δ 8.30 (d, J = 9.0 Hz, 1H), 7.64 (dd, J = 9.0, 2.7 Hz, 1H), 6.90 (d, J = 2.7 Hz, 1H), 3.75 (s, 3H). Characterization of [Fe(4,4’-diCl-bpy)3](BF4)2: 1H NMR (500 MHz, CD3CN): δ 8.61 (d, J = 2.2 Hz, 2H), 7.48 (dd, J = 6.2, 2.2 Hz, 2H), 7.32 (d, J = 6.2 Hz, 2H). Characterization of [Fe(5,5’-diCl-bpy)3](BF4)2: 1H NMR (500 MHz, CD3CN): δ 8.46 (d, J = 8.7 Hz, 2H), 8.18 (dd, J = 8.7, 2.2 Hz, 2H), 7.25 (d, J = 2.1 Hz, 2H). 5.SI.2 UV-Vis Data The molar absorptivity data for each complex in the series is summarized in Figure 5.S1. Below Figure 5.S1 are the individual UV-Vis spectra for 1A1 → 3T1 absorption features alongside the multi-peak fits for each complex. Figure 5.31 displays this information for [Fe(4,4’-diCl- bpy)3](BF4)2. 195 Figure 5.S1: The UV-Vis spectra of the MLCT absorption feature across the series of [Fe(X,X’- R-bpy)3]2+ complexes. Figure 5.S2: UV-Vis spectra for [Fe(bpy)3](PF6)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12440 cm-1. Figure 5.S3: UV-Vis spectra for [Fe(4,4’-diMe-(bpy)3](PF6)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12200 cm-1. 196 Figure 5.S4: UV-Vis spectra for [Fe(5,5’-diMe-(bpy)3](PF6)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12440 cm-1. As mentioned in the main body of the chapter, the dip observed in the spectrum is an artifact resulting from the removal of solvent features via subtracting a solvent blank. This feature is observed in several spectra, but it is particularly pronounced for complexes with low molar absorptivity for the 1A1 → 3T1 peak. This artifact causes some issues with the multipeak fitting tool, notably is over-estimates the height of the 1A1 → 3T1 peak. In these cases, I was looking for accuracy of the overall fit, which in this case had good agreement with the data. Figure 5.S5: UV-Vis spectra for [Fe(4,4’-diOMe-(bpy)3](PF6)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12110 cm-1. 197 Figure 5.S6: UV-Vis spectra for [Fe(5,5’-diOMe-(bpy)3](PF6)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12140 cm-1. Figure 5.S7: UV-Vis spectra for [Fe(5,5’-diCl-(bpy)3](BF4)2 alongside the deconvolution determined using the multipeak fitting tool of Igor Pro. The deconvolved 1A1 → 3T1 peak is at 12250 cm-1. 5.SI.3 DFT Calculations Similarly to the DFT calculations done on the [Fe(4’-R-terpy)2]2+ series discussed in Chapter 4, a modified B3LYP functional group was used to perform DFT calculations on the series studied in this chapter. The Hartree-Fock percent (HF%) was modified from 20% to yield the best results, and in order to determine which HF% should be used, TD-DFT calculations were conducted and the simulated spectrum was compared to an experimentally measured one for [Fe(bpy)3]2+. Through this examination, it was that both 15% HF and 10% HF yielded the same features as the experimental spectrum, with 15% percent having stronger energetic overlap of 198 major peaks with the experimentally determined spectrum. However, as will be made clear in the next chapter, 10% Hartree-Fock was used in order to be consistent with the settings used for [Fe(terpy)2]2+ DFT calculations. Figure 5.S8 provides a comparison of the TD-DFT determined spectra calculated with 10%, 15%, and 20% HF with the experimentally measured spectrum shown in Figure 5.S1. Figure 5.S8: UV-Vis spectrum of [Fe(bpy)3](PF6)2 (in black), alongside TD-DFT simulated spectra for [Fe(bpy)3]2+ calculated using HF of 20% (in purple), 15% (in green), and 10% (in red). Figure 5.S9 shows the individual states calculated through TD-DFT alongside both spectra. The states show that within the MLCT absorption feature, there are two states of roughly equivalent molar absorptivity. This matches what is seen in the experimental spectrum, however in the TD-DFT spectrum these states are closer in energy, causing the Gaussian of each state to Figure 5.S9: UV-Vis spectrum of [Fe(bpy)3](PF6)2 (in black), alongside TD-DFT simulated spectrum for [Fe(bpy)3]2+ (in red) and the individual states (in blue). 199 merge into one feature, where the experimental spectrum displays two peaks forming a wide MLCT absorption feature. Provided below are the optimized singlet and quintet geometries, and relevant structural information for each complex in the series. The pbd file information for each complex in the series will be available at the end of the SI. [Fe(bpy)3]2+ DFT Geometries Figure 5.S10: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(bpy)3]2+. Table 5.S1: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.25/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.5 167.5/169.9 95.3 74.8/95.8 [Fe(4,4’-diMe-bpy)3]2+ DFT Geometries Figure 5.S11: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4,4’-diMe-bpy)3]2+. 200 Table 5.S2: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.04/2.05 2.24/2.26 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.4/174.5 167.2/167.4/170 80.7/88.9/95.3 74.8/91.4/95 [Fe(5,5’-diMe-bpy)3]2+ DFT Geometries Figure 5.S12: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(5,5’-diMe-bpy)3]2+. Table 5.S3: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.25/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.7/174.8 167.4/167.7/170.2 80.9/88.8/95.3 75/91.3/95 [Fe(4,4’-diOMe-bpy)3]2+ DFT Geometries Figure 5.S13: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4,4’-diOMe-bpy)3]2+. 201 Table 5.S4: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.24/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.4 166.9/167.3/170.3 80.7/89/95 74.8/91.3/96 [Fe(5,5’-diOMe-bpy)3]2+ DFT Geometries Figure 5.S14: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(5,5’-diOMe-bpy)3]2+. Table 5.S5: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.25/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.9 167.4/167.8/170.4 81.2/88.7/95.2 75/91.2/96 [Fe(4,4’-diCl-bpy)3]2+ DFT Geometries Figure 5.S15: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(4,4’-diCl-bpy)3]2+. 202 Table 5.S6: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.25/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.6/174.7 166.5/168.4/170.7 80.8/88.7/95.4 74.6/92/97 [Fe(5,5’-diCl-bpy)3]2+ DFT Geometries Figure 5.S16: The DFT optimized geometries of the singlet state (left) and quintet state (right) for [Fe(5,5’-diCl-bpy)3]2+. Table 5.S7: Relevant structural parameters determined via DFT calculations. State Singlet Quintet Fe-N (Å) 2.05 2.25/2.27 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 174.8/174.9 167.6/168.4/169.4 80.8/88.5/95.5 74.6/91/97 Notably, within each complex it is observed that the high spin (quintet) state engages in Jahn-Teller distortion, elongation along the axial Fe-N bonds. 5.SI.4 DFT Optimized Geometry Files [Fe(bpy)3]2+: Table 5.S8: The optimized geometry file for the lowest energy singlet state of [Fe(bpy)3]2+. TITLE Fe-bpy Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.562 0.764 -1.076 N HETATM 2 N 0 1.550 -0.786 1.077 N HETATM 3 N 0 -0.107 -1.756 -1.042 N HETATM 4 N 0 -1.471 -0.928 1.079 N HETATM 5 N 0 -1.457 0.949 -1.081 N HETATM 6 N 0 -0.082 1.759 1.041 N HETATM 7 C 0 1.474 1.598 -2.172 C HETATM 8 H 0 0.450 1.833 -2.531 H HETATM 9 C 0 2.614 2.134 -2.830 C HETATM 10 H 0 2.476 2.799 -3.711 H HETATM 11 C 0 3.911 1.803 -2.340 C HETATM 12 C 0 4.018 0.945 -1.210 C HETATM 13 H 0 5.019 0.676 -0.809 H HETATM 14 C 0 2.832 0.436 -0.597 C HETATM 15 C 0 2.824 -0.481 0.597 C HETATM 16 C 0 4.002 -1.009 1.211 C HETATM 17 H 0 5.007 -0.758 0.808 H HETATM 18 C 0 3.881 -1.864 2.341 C 203 Table 5.S8 (cont’d) HETATM 19 C 0 2.579 -2.171 2.833 C HETATM 20 H 0 2.430 -2.832 3.715 H HETATM 21 C 0 1.447 -1.617 2.175 C HETATM 22 H 0 0.420 -1.833 2.535 H HETATM 23 C 0 0.669 -2.116 -2.125 C HETATM 24 H 0 1.387 -1.354 -2.493 H HETATM 25 C 0 0.568 -3.384 -2.761 C HETATM 26 H 0 1.220 -3.612 -3.633 H HETATM 27 C 0 -0.370 -4.333 -2.261 C HETATM 28 C 0 -1.175 -3.976 -1.144 C HETATM 29 H 0 -1.910 -4.701 -0.733 H HETATM 30 C 0 -1.028 -2.683 -0.553 C HETATM 31 C 0 -1.832 -2.196 0.622 C HETATM 32 C 0 -2.882 -2.944 1.241 C HETATM 33 H 0 -3.158 -3.949 0.856 H HETATM 34 C 0 -3.577 -2.390 2.351 C HETATM 35 C 0 -3.205 -1.096 2.818 C HETATM 36 H 0 -3.715 -0.619 3.683 H HETATM 37 C 0 -2.155 -0.402 2.157 C HETATM 38 H 0 -1.838 0.605 2.499 H HETATM 39 C 0 -2.148 0.433 -2.158 C HETATM 40 H 0 -1.845 -0.579 -2.500 H HETATM 41 C 0 -3.189 1.141 -2.820 C HETATM 42 H 0 -3.705 0.670 -3.685 H HETATM 43 C 0 -3.544 2.439 -2.353 C HETATM 44 C 0 -2.842 2.984 -1.242 C HETATM 45 H 0 -3.105 3.993 -0.858 H HETATM 46 C 0 -1.802 2.222 -0.624 C HETATM 47 C 0 -0.991 2.699 0.551 C HETATM 48 C 0 -1.119 3.995 1.140 C HETATM 49 H 0 -1.844 4.730 0.728 H HETATM 50 C 0 -0.308 4.342 2.256 C HETATM 51 C 0 0.616 3.380 2.757 C HETATM 52 H 0 1.271 3.600 3.629 H HETATM 53 C 0 0.699 2.110 2.123 C HETATM 54 H 0 1.406 1.338 2.492 H HETATM 55 Fe 0 -0.001 0.001 -0.000 Fe HETATM 56 H 0 -4.397 -2.961 2.843 H HETATM 57 H 0 -0.474 -5.336 -2.731 H HETATM 58 H 0 -4.356 3.021 -2.844 H HETATM 59 H 0 -0.398 5.347 2.725 H HETATM 60 H 0 4.790 -2.283 2.829 H HETATM 61 H 0 4.827 2.206 -2.827 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 204 Table 5.S8 (cont’d) CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 57 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 56 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 4 35 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 58 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 5 44 47 CONECT 47 46 6 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 59 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 34 CONECT 57 27 CONECT 58 43 CONECT 59 50 CONECT 60 18 CONECT 61 11 Table 5.S9: The optimized geometry file for the lowest energy quintet state of [Fe(bpy)3]2+. TITLE Fe-bpy Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.769 0.755 -1.130 N HETATM 2 N 0 1.763 -0.780 1.129 N HETATM 3 N 0 -0.230 -1.964 -1.112 N HETATM 4 N 0 -1.558 -1.170 1.150 N HETATM 5 N 0 -1.535 1.187 -1.148 N HETATM 6 N 0 -0.202 1.960 1.117 N HETATM 7 C 0 1.704 1.554 -2.250 C HETATM 8 H 0 0.683 1.790 -2.624 H HETATM 9 C 0 2.857 2.060 -2.910 C HETATM 10 H 0 2.744 2.700 -3.813 H HETATM 11 C 0 4.140 1.729 -2.385 C HETATM 12 C 0 4.220 0.905 -1.227 C HETATM 13 H 0 5.215 0.645 -0.806 H HETATM 14 C 0 3.017 0.425 -0.616 C HETATM 15 C 0 3.013 -0.461 0.616 C HETATM 16 C 0 4.212 -0.953 1.227 C HETATM 17 H 0 5.209 -0.701 0.806 H HETATM 18 C 0 4.125 -1.777 2.383 C HETATM 19 C 0 2.839 -2.097 2.908 C HETATM 20 H 0 2.719 -2.737 3.811 H HETATM 21 C 0 1.691 -1.580 2.249 C HETATM 22 H 0 0.667 -1.807 2.622 H HETATM 23 C 0 0.519 -2.332 -2.206 C HETATM 24 H 0 1.222 -1.564 -2.598 H HETATM 25 C 0 0.419 -3.612 -2.819 C HETATM 26 H 0 1.046 -3.854 -3.706 H HETATM 27 C 0 -0.488 -4.561 -2.265 C HETATM 28 C 0 -1.266 -4.192 -1.130 C 205 Table 5.S9 (cont’d) HETATM 29 H 0 -1.967 -4.928 -0.681 H HETATM 30 C 0 -1.124 -2.880 -0.576 C HETATM 31 C 0 -1.928 -2.400 0.624 C HETATM 32 C 0 -3.011 -3.147 1.186 C HETATM 33 H 0 -3.316 -4.123 0.752 H HETATM 34 C 0 -3.714 -2.624 2.308 C HETATM 35 C 0 -3.326 -1.363 2.847 C HETATM 36 H 0 -3.845 -0.916 3.723 H HETATM 37 C 0 -2.245 -0.672 2.233 C HETATM 38 H 0 -1.903 0.316 2.615 H HETATM 39 C 0 -2.229 0.698 -2.231 C HETATM 40 H 0 -1.906 -0.298 -2.608 H HETATM 41 C 0 -3.294 1.408 -2.850 C HETATM 42 H 0 -3.820 0.969 -3.726 H HETATM 43 C 0 -3.660 2.679 -2.318 C HETATM 44 C 0 -2.950 3.192 -1.195 C HETATM 45 H 0 -3.237 4.175 -0.765 H HETATM 46 C 0 -1.883 2.426 -0.626 C HETATM 47 C 0 -1.076 2.893 0.576 C HETATM 48 C 0 -1.193 4.209 1.125 C HETATM 49 H 0 -1.877 4.958 0.672 H HETATM 50 C 0 -0.414 4.564 2.263 C HETATM 51 C 0 0.471 3.599 2.825 C HETATM 52 H 0 1.097 3.829 3.715 H HETATM 53 C 0 0.548 2.316 2.215 C HETATM 54 H 0 1.234 1.534 2.612 H HETATM 55 Fe 0 -0.016 -0.003 0.002 Fe HETATM 56 H 0 -4.559 -3.196 2.753 H HETATM 57 H 0 -0.590 -5.578 -2.705 H HETATM 58 H 0 -4.492 3.266 -2.767 H HETATM 59 H 0 -0.497 5.585 2.701 H HETATM 60 H 0 5.051 -2.164 2.865 H HETATM 61 H 0 5.070 2.107 -2.867 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 57 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 30 4 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 56 206 Table 5.S9 (cont’d) CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 58 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 46 6 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 59 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 34 CONECT 57 27 CONECT 58 43 CONECT 59 50 CONECT 60 18 CONECT 61 11 [Fe(4,4’-diMe-bpy)3]2+: Table 5.S10: The optimized geometry file for the lowest energy singlet state of [Fe(4,4’-diMe- bpy)3]2+. TITLE Fe-4-dmb Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.563 0.761 -1.073 N HETATM 2 N 0 1.551 -0.783 1.076 N HETATM 3 N 0 -0.112 -1.753 -1.045 N HETATM 4 N 0 -1.468 -0.934 1.076 N HETATM 5 N 0 -1.454 0.956 -1.078 N HETATM 6 N 0 -0.088 1.756 1.045 N HETATM 7 C 0 1.485 1.597 -2.170 C HETATM 8 H 0 0.463 1.838 -2.532 H HETATM 9 C 0 2.626 2.126 -2.825 C HETATM 10 H 0 2.483 2.794 -3.705 H HETATM 11 C 0 3.939 1.804 -2.348 C HETATM 12 C 0 4.019 0.942 -1.213 C HETATM 13 H 0 5.016 0.668 -0.804 H HETATM 14 C 0 2.833 0.435 -0.598 C HETATM 15 C 0 2.825 -0.480 0.599 C HETATM 16 C 0 4.003 -1.007 1.213 C HETATM 17 H 0 5.004 -0.752 0.803 H HETATM 18 C 0 3.909 -1.866 2.350 C HETATM 19 C 0 2.591 -2.163 2.829 C HETATM 20 H 0 2.437 -2.826 3.711 H HETATM 21 C 0 1.459 -1.614 2.175 C HETATM 22 H 0 0.433 -1.836 2.539 H HETATM 23 C 0 0.657 -2.117 -2.133 C HETATM 24 H 0 1.376 -1.357 -2.504 H HETATM 25 C 0 0.548 -3.379 -2.771 C HETATM 26 H 0 1.199 -3.600 -3.647 H HETATM 27 C 0 -0.388 -4.350 -2.284 C HETATM 28 C 0 -1.181 -3.972 -1.158 C HETATM 29 H 0 -1.920 -4.692 -0.743 H HETATM 30 C 0 -1.030 -2.683 -0.561 C HETATM 31 C 0 -1.831 -2.200 0.621 C HETATM 32 C 0 -2.877 -2.952 1.240 C HETATM 33 H 0 -3.145 -3.957 0.848 H HETATM 34 C 0 -3.590 -2.422 2.358 C 207 Table 5.S10 (cont’d) HETATM 35 C 0 -3.202 -1.119 2.814 C HETATM 36 H 0 -3.712 -0.639 3.680 H HETATM 37 C 0 -2.158 -0.420 2.157 C HETATM 38 H 0 -1.848 0.589 2.504 H HETATM 39 C 0 -2.150 0.450 -2.159 C HETATM 40 H 0 -1.853 -0.562 -2.505 H HETATM 41 C 0 -3.185 1.163 -2.816 C HETATM 42 H 0 -3.700 0.690 -3.682 H HETATM 43 C 0 -3.555 2.471 -2.360 C HETATM 44 C 0 -2.836 2.992 -1.242 C HETATM 45 H 0 -3.091 4.001 -0.850 H HETATM 46 C 0 -1.800 2.227 -0.622 C HETATM 47 C 0 -0.993 2.699 0.559 C HETATM 48 C 0 -1.126 3.991 1.155 C HETATM 49 H 0 -1.855 4.721 0.739 H HETATM 50 C 0 -0.329 4.359 2.281 C HETATM 51 C 0 0.593 3.375 2.770 C HETATM 52 H 0 1.246 3.587 3.647 H HETATM 53 C 0 0.684 2.111 2.134 C HETATM 54 H 0 1.392 1.341 2.506 H HETATM 55 Fe 0 -0.001 0.001 0.000 Fe HETATM 56 C 0 -4.677 3.271 -3.050 C HETATM 57 H 0 -5.638 2.688 -3.047 H HETATM 58 H 0 -4.854 4.254 -2.540 H HETATM 59 H 0 -4.414 3.471 -4.126 H HETATM 60 C 0 -0.441 5.749 2.935 C HETATM 61 H 0 -1.310 6.327 2.525 H HETATM 62 H 0 -0.562 5.657 4.049 H HETATM 63 H 0 0.497 6.344 2.751 H HETATM 64 C 0 5.201 2.357 -3.038 C HETATM 65 H 0 5.126 3.470 -3.173 H HETATM 66 H 0 6.126 2.128 -2.448 H HETATM 67 H 0 5.316 1.905 -4.063 H HETATM 68 C 0 5.161 -2.442 3.038 C HETATM 69 H 0 5.072 -3.557 3.159 H HETATM 70 H 0 6.092 -2.218 2.454 H HETATM 71 H 0 5.278 -2.005 4.069 H HETATM 72 C 0 -4.722 -3.207 3.047 C HETATM 73 H 0 -5.677 -2.614 3.038 H HETATM 74 H 0 -4.908 -4.190 2.540 H HETATM 75 H 0 -4.465 -3.405 4.124 H HETATM 76 C 0 -0.523 -5.735 -2.944 C HETATM 77 H 0 -1.362 -6.326 -2.492 H HETATM 78 H 0 -0.709 -5.633 -4.048 H HETATM 79 H 0 0.429 -6.323 -2.821 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 64 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 68 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 208 Table 5.S10 (cont’d) CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 76 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 72 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 56 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 60 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 43 57 58 59 CONECT 57 56 CONECT 58 56 CONECT 59 56 CONECT 60 50 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 11 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 CONECT 68 18 69 70 71 CONECT 69 68 CONECT 70 68 CONECT 71 68 CONECT 72 34 73 74 75 CONECT 73 72 CONECT 74 72 CONECT 75 72 CONECT 76 27 77 78 79 CONECT 77 76 CONECT 78 76 CONECT 79 76 Table 5.S11: The optimized geometry file for the lowest energy quintet state of [Fe(4,4’-diMe- bpy)3]2+. TITLE Fe-4-dmb Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.759 0.754 -1.125 N HETATM 2 N 0 1.751 -0.779 1.128 N HETATM 3 N 0 -0.235 -1.953 -1.126 N HETATM 4 N 0 -1.555 -1.182 1.142 N HETATM 5 N 0 -1.535 1.198 -1.138 N HETATM 6 N 0 -0.209 1.951 1.132 N 209 Table 5.S11 (cont’d) HETATM 7 C 0 1.705 1.558 -2.244 C HETATM 8 H 0 0.686 1.802 -2.620 H HETATM 9 C 0 2.858 2.061 -2.898 C HETATM 10 H 0 2.740 2.707 -3.798 H HETATM 11 C 0 4.158 1.735 -2.387 C HETATM 12 C 0 4.210 0.904 -1.226 C HETATM 13 H 0 5.201 0.638 -0.798 H HETATM 14 C 0 3.007 0.424 -0.616 C HETATM 15 C 0 3.003 -0.464 0.617 C HETATM 16 C 0 4.200 -0.959 1.225 C HETATM 17 H 0 5.194 -0.705 0.795 H HETATM 18 C 0 4.139 -1.790 2.385 C HETATM 19 C 0 2.837 -2.100 2.898 C HETATM 20 H 0 2.712 -2.744 3.798 H HETATM 21 C 0 1.689 -1.582 2.246 C HETATM 22 H 0 0.667 -1.814 2.623 H HETATM 23 C 0 0.512 -2.325 -2.221 C HETATM 24 H 0 1.209 -1.553 -2.620 H HETATM 25 C 0 0.421 -3.604 -2.828 C HETATM 26 H 0 1.050 -3.836 -3.718 H HETATM 27 C 0 -0.473 -4.582 -2.281 C HETATM 28 C 0 -1.245 -4.194 -1.142 C HETATM 29 H 0 -1.941 -4.930 -0.684 H HETATM 30 C 0 -1.118 -2.880 -0.592 C HETATM 31 C 0 -1.927 -2.406 0.608 C HETATM 32 C 0 -3.017 -3.153 1.156 C HETATM 33 H 0 -3.317 -4.123 0.701 H HETATM 34 C 0 -3.744 -2.658 2.283 C HETATM 35 C 0 -3.338 -1.394 2.822 C HETATM 36 H 0 -3.861 -0.948 3.698 H HETATM 37 C 0 -2.255 -0.698 2.225 C HETATM 38 H 0 -1.920 0.287 2.621 H HETATM 39 C 0 -2.242 0.724 -2.222 C HETATM 40 H 0 -1.924 -0.268 -2.614 H HETATM 41 C 0 -3.311 1.437 -2.823 C HETATM 42 H 0 -3.840 0.998 -3.699 H HETATM 43 C 0 -3.696 2.710 -2.287 C HETATM 44 C 0 -2.963 3.195 -1.160 C HETATM 45 H 0 -3.248 4.169 -0.707 H HETATM 46 C 0 -1.888 2.430 -0.607 C HETATM 47 C 0 -1.074 2.892 0.594 C HETATM 48 C 0 -1.178 4.209 1.141 C HETATM 49 H 0 -1.859 4.957 0.679 H HETATM 50 C 0 -0.403 4.586 2.282 C HETATM 51 C 0 0.471 3.593 2.833 C HETATM 52 H 0 1.101 3.814 3.725 H HETATM 53 C 0 0.541 2.311 2.228 C HETATM 54 H 0 1.222 1.527 2.632 H HETATM 55 Fe 0 -0.025 -0.003 0.003 Fe HETATM 56 C 0 -4.849 3.521 -2.909 C HETATM 57 H 0 -5.764 2.881 -3.034 H HETATM 58 H 0 -5.114 4.411 -2.280 H HETATM 59 H 0 -4.558 3.888 -3.933 H HETATM 60 C 0 -0.485 6.002 2.881 C HETATM 61 H 0 -1.363 6.570 2.476 H HETATM 62 H 0 -0.568 5.959 4.001 H HETATM 63 H 0 0.449 6.582 2.638 H HETATM 64 C 0 5.439 2.253 -3.069 C HETATM 65 H 0 5.370 3.359 -3.261 H HETATM 66 H 0 6.348 2.053 -2.443 H HETATM 67 H 0 5.580 1.750 -4.066 H HETATM 68 C 0 5.415 -2.327 3.063 C HETATM 69 H 0 5.340 -3.436 3.233 H HETATM 70 H 0 6.328 -2.119 2.447 H HETATM 71 H 0 5.553 -1.845 4.071 H HETATM 72 C 0 -4.915 -3.448 2.898 C HETATM 73 H 0 -5.829 -2.799 2.986 H HETATM 74 H 0 -5.172 -4.351 2.284 H 210 Table 5.S11 (cont’d) HETATM 75 H 0 -4.651 -3.791 3.937 H HETATM 76 C 0 -0.579 -5.996 -2.885 C HETATM 77 H 0 -1.460 -6.554 -2.472 H HETATM 78 H 0 -0.673 -5.947 -4.004 H HETATM 79 H 0 0.350 -6.588 -2.655 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 64 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 68 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 76 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 72 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 56 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 60 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 43 57 58 59 CONECT 57 56 CONECT 58 56 CONECT 59 56 CONECT 60 50 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 211 Table 5.S11 (cont’d) CONECT 64 11 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 CONECT 68 18 69 70 71 CONECT 69 68 CONECT 70 68 CONECT 71 68 CONECT 72 34 73 74 75 CONECT 73 72 CONECT 74 72 CONECT 75 72 CONECT 76 27 77 78 79 CONECT 77 76 CONECT 78 76 CONECT 79 76 [Fe(5,5’-diMe-bpy)3]2+: Table 5.S12: The optimized geometry file for the lowest energy singlet state of [Fe(5,5’-diMe- bpy)3]2+. TITLE Fe-5-dmb Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.560 0.763 -1.079 N HETATM 2 N 0 1.547 -0.786 1.080 N HETATM 3 N 0 -0.109 -1.758 -1.041 N HETATM 4 N 0 -1.473 -0.926 1.082 N HETATM 5 N 0 -1.459 0.948 -1.084 N HETATM 6 N 0 -0.084 1.761 1.039 N HETATM 7 C 0 1.471 1.594 -2.178 C HETATM 8 H 0 0.443 1.823 -2.533 H HETATM 9 C 0 2.602 2.141 -2.855 C HETATM 10 C 0 3.896 1.798 -2.346 C HETATM 11 C 0 4.007 0.945 -1.217 C HETATM 12 H 0 5.011 0.681 -0.818 H HETATM 13 C 0 2.825 0.435 -0.596 C HETATM 14 C 0 2.817 -0.478 0.598 C HETATM 15 C 0 3.991 -1.007 1.220 C HETATM 16 H 0 4.999 -0.759 0.821 H HETATM 17 C 0 3.866 -1.857 2.349 C HETATM 18 C 0 2.566 -2.178 2.859 C HETATM 19 C 0 1.444 -1.614 2.180 C HETATM 20 H 0 0.413 -1.827 2.535 H HETATM 21 C 0 0.665 -2.117 -2.126 C HETATM 22 H 0 1.380 -1.349 -2.490 H HETATM 23 C 0 0.579 -3.382 -2.781 C HETATM 24 C 0 -0.369 -4.323 -2.262 C HETATM 25 C 0 -1.173 -3.971 -1.146 C HETATM 26 H 0 -1.907 -4.700 -0.739 H HETATM 27 C 0 -1.029 -2.680 -0.548 C HETATM 28 C 0 -1.827 -2.195 0.628 C HETATM 29 C 0 -2.871 -2.941 1.261 C HETATM 30 H 0 -3.148 -3.950 0.886 H HETATM 31 C 0 -3.558 -2.384 2.371 C HETATM 32 C 0 -3.203 -1.081 2.849 C HETATM 33 C 0 -2.154 -0.401 2.162 C HETATM 34 H 0 -1.835 0.611 2.494 H HETATM 35 C 0 -2.147 0.431 -2.164 C HETATM 36 H 0 -1.842 -0.585 -2.496 H HETATM 37 C 0 -3.188 1.125 -2.850 C HETATM 38 C 0 -3.525 2.433 -2.372 C HETATM 39 C 0 -2.830 2.980 -1.263 C HETATM 40 H 0 -3.093 3.993 -0.888 H HETATM 41 C 0 -1.796 2.221 -0.631 C HETATM 42 C 0 -0.990 2.696 0.545 C HETATM 43 C 0 -1.113 3.991 1.139 C HETATM 44 H 0 -1.835 4.730 0.730 H 212 Table 5.S12 (cont’d) HETATM 45 C 0 -0.305 4.332 2.255 C HETATM 46 C 0 0.629 3.378 2.777 C HETATM 47 C 0 0.695 2.110 2.124 C HETATM 48 H 0 1.398 1.331 2.490 H HETATM 49 Fe 0 -0.004 0.000 -0.000 Fe HETATM 50 H 0 4.779 -2.272 2.836 H HETATM 51 H 0 4.816 2.200 -2.831 H HETATM 52 H 0 -0.397 5.340 2.721 H HETATM 53 H 0 -4.331 3.022 -2.869 H HETATM 54 H 0 -0.478 -5.328 -2.731 H HETATM 55 H 0 -4.372 -2.962 2.868 H HETATM 56 C 0 -3.913 0.482 -4.049 C HETATM 57 H 0 -3.452 -0.503 -4.327 H HETATM 58 H 0 -4.998 0.305 -3.808 H HETATM 59 H 0 -3.876 1.157 -4.947 H HETATM 60 C 0 1.530 3.695 3.987 C HETATM 61 H 0 2.171 2.817 4.262 H HETATM 62 H 0 2.204 4.567 3.762 H HETATM 63 H 0 0.909 3.975 4.882 H HETATM 64 C 0 2.425 3.059 -4.081 C HETATM 65 H 0 2.916 2.610 -4.988 H HETATM 66 H 0 1.340 3.227 -4.315 H HETATM 67 H 0 2.906 4.059 -3.901 H HETATM 68 C 0 2.374 -3.092 4.085 C HETATM 69 H 0 2.870 -2.649 4.993 H HETATM 70 H 0 1.287 -3.244 4.318 H HETATM 71 H 0 2.841 -4.100 3.907 H HETATM 72 C 0 -3.919 -0.428 4.048 C HETATM 73 H 0 -3.446 0.551 4.325 H HETATM 74 H 0 -5.002 -0.240 3.809 H HETATM 75 H 0 -3.888 -1.104 4.947 H HETATM 76 C 0 1.476 -3.711 -3.992 C HETATM 77 H 0 2.131 -2.843 -4.265 H HETATM 78 H 0 2.136 -4.595 -3.769 H HETATM 79 H 0 0.851 -3.979 -4.888 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 64 CONECT 10 9 11 51 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 50 CONECT 18 17 19 68 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 76 CONECT 24 23 25 54 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 27 4 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 55 CONECT 32 31 33 72 213 Table 5.S12 (cont’d) CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 56 CONECT 38 37 39 53 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 52 CONECT 46 45 47 60 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 17 CONECT 51 10 CONECT 52 45 CONECT 53 38 CONECT 54 24 CONECT 55 31 CONECT 56 37 57 58 59 CONECT 57 56 CONECT 58 56 CONECT 59 56 CONECT 60 46 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 9 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 CONECT 68 18 69 70 71 CONECT 69 68 CONECT 70 68 CONECT 71 68 CONECT 72 32 73 74 75 CONECT 73 72 CONECT 74 72 CONECT 75 72 CONECT 76 23 77 78 79 CONECT 77 76 CONECT 78 76 CONECT 79 76 Table 5.S13: The optimized geometry file for the lowest energy quintet state of [Fe(5,5’-diMe- bpy)3]2+. TITLE Fe-5-dmb Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.762 0.752 -1.136 N HETATM 2 N 0 1.755 -0.776 1.135 N HETATM 3 N 0 -0.229 -1.962 -1.115 N HETATM 4 N 0 -1.558 -1.169 1.153 N HETATM 5 N 0 -1.535 1.188 -1.151 N HETATM 6 N 0 -0.200 1.960 1.119 N HETATM 7 C 0 1.698 1.544 -2.262 C HETATM 8 H 0 0.673 1.776 -2.632 H HETATM 9 C 0 2.842 2.057 -2.944 C HETATM 10 C 0 4.122 1.715 -2.397 C HETATM 11 C 0 4.205 0.900 -1.238 C HETATM 12 H 0 5.202 0.645 -0.818 H HETATM 13 C 0 3.005 0.421 -0.618 C HETATM 14 C 0 3.001 -0.456 0.617 C HETATM 15 C 0 4.197 -0.948 1.236 C HETATM 16 H 0 5.196 -0.701 0.817 H 214 Table 5.S13 (cont’d) HETATM 17 C 0 4.106 -1.763 2.395 C HETATM 18 C 0 2.823 -2.094 2.941 C HETATM 19 C 0 1.684 -1.569 2.260 C HETATM 20 H 0 0.657 -1.792 2.629 H HETATM 21 C 0 0.515 -2.327 -2.213 C HETATM 22 H 0 1.214 -1.551 -2.602 H HETATM 23 C 0 0.431 -3.604 -2.847 C HETATM 24 C 0 -0.482 -4.546 -2.272 C HETATM 25 C 0 -1.255 -4.186 -1.135 C HETATM 26 H 0 -1.953 -4.927 -0.689 H HETATM 27 C 0 -1.119 -2.876 -0.573 C HETATM 28 C 0 -1.915 -2.401 0.630 C HETATM 29 C 0 -2.988 -3.152 1.209 C HETATM 30 H 0 -3.289 -4.135 0.786 H HETATM 31 C 0 -3.684 -2.630 2.332 C HETATM 32 C 0 -3.318 -1.358 2.881 C HETATM 33 C 0 -2.242 -0.673 2.239 C HETATM 34 H 0 -1.903 0.321 2.611 H HETATM 35 C 0 -2.226 0.703 -2.238 C HETATM 36 H 0 -1.906 -0.299 -2.605 H HETATM 37 C 0 -3.286 1.406 -2.885 C HETATM 38 C 0 -3.628 2.687 -2.342 C HETATM 39 C 0 -2.925 3.199 -1.218 C HETATM 40 H 0 -3.208 4.189 -0.799 H HETATM 41 C 0 -1.870 2.429 -0.632 C HETATM 42 C 0 -1.070 2.890 0.573 C HETATM 43 C 0 -1.182 4.204 1.131 C HETATM 44 H 0 -1.863 4.958 0.681 H HETATM 45 C 0 -0.408 4.550 2.271 C HETATM 46 C 0 0.483 3.591 2.853 C HETATM 47 C 0 0.545 2.312 2.221 C HETATM 48 H 0 1.227 1.523 2.615 H HETATM 49 Fe 0 -0.020 -0.002 0.002 Fe HETATM 50 H 0 5.036 -2.145 2.876 H HETATM 51 H 0 5.055 2.088 -2.879 H HETATM 52 H 0 -0.494 5.572 2.708 H HETATM 53 H 0 -4.450 3.287 -2.798 H HETATM 54 H 0 -0.587 -5.565 -2.711 H HETATM 55 H 0 -4.519 -3.214 2.784 H HETATM 56 C 0 -4.020 0.809 -4.102 C HETATM 57 H 0 -3.575 -0.176 -4.403 H HETATM 58 H 0 -5.109 0.647 -3.870 H HETATM 59 H 0 -3.966 1.506 -4.983 H HETATM 60 C 0 1.341 3.915 4.092 C HETATM 61 H 0 1.977 3.040 4.389 H HETATM 62 H 0 2.018 4.791 3.891 H HETATM 63 H 0 0.690 4.190 4.967 H HETATM 64 C 0 2.698 2.937 -4.201 C HETATM 65 H 0 3.219 2.465 -5.080 H HETATM 66 H 0 1.620 3.089 -4.473 H HETATM 67 H 0 3.166 3.946 -4.037 H HETATM 68 C 0 2.671 -2.974 4.197 C HETATM 69 H 0 3.194 -2.507 5.077 H HETATM 70 H 0 1.591 -3.120 4.467 H HETATM 71 H 0 3.132 -3.987 4.033 H HETATM 72 C 0 -4.044 -0.749 4.097 C HETATM 73 H 0 -3.573 0.220 4.411 H HETATM 74 H 0 -5.126 -0.553 3.857 H HETATM 75 H 0 -4.019 -1.455 4.972 H HETATM 76 C 0 1.289 -3.943 -4.082 C HETATM 77 H 0 1.938 -3.077 -4.379 H HETATM 78 H 0 1.953 -4.827 -3.876 H HETATM 79 H 0 0.637 -4.213 -4.959 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 215 Table 5.S13 (cont’d) CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 64 CONECT 10 9 11 51 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 50 CONECT 18 17 19 68 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 76 CONECT 24 23 25 54 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 27 4 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 55 CONECT 32 31 33 72 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 56 CONECT 38 37 39 53 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 52 CONECT 46 45 47 60 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 17 CONECT 51 10 CONECT 52 45 CONECT 53 38 CONECT 54 24 CONECT 55 31 CONECT 56 37 57 58 59 CONECT 57 56 CONECT 58 56 CONECT 59 56 CONECT 60 46 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 9 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 CONECT 68 18 69 70 71 CONECT 69 68 CONECT 70 68 CONECT 71 68 CONECT 72 32 73 74 75 CONECT 73 72 216 Table 5.S13 (cont’d) CONECT 74 72 CONECT 75 72 CONECT 76 23 77 78 79 CONECT 77 76 CONECT 78 76 CONECT 79 76 [Fe(4,4’-diOMe-bpy)3]2+: Table 5.S14: The optimized geometry file for the lowest energy singlet state of [Fe(4,4’-diOMe- bpy)3]2+. TITLE Fe-4,4'-OMe-bpy Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.615 0.748 -1.096 N HETATM 2 N 0 1.609 -0.812 1.047 N HETATM 3 N 0 -0.074 -1.762 -1.077 N HETATM 4 N 0 -1.417 -0.954 1.063 N HETATM 5 N 0 -1.407 0.957 -1.081 N HETATM 6 N 0 -0.026 1.737 1.045 N HETATM 7 C 0 1.539 1.591 -2.183 C HETATM 8 H 0 0.517 1.842 -2.538 H HETATM 9 C 0 2.669 2.129 -2.854 C HETATM 10 C 0 3.974 1.782 -2.373 C HETATM 11 C 0 4.073 0.908 -1.241 C HETATM 12 H 0 5.082 0.639 -0.861 H HETATM 13 C 0 2.891 0.411 -0.632 C HETATM 14 C 0 2.887 -0.516 0.560 C HETATM 15 C 0 4.063 -1.049 1.150 C HETATM 16 H 0 5.073 -0.813 0.752 H HETATM 17 C 0 3.956 -1.918 2.286 C HETATM 18 C 0 2.650 -2.220 2.792 C HETATM 19 C 0 1.525 -1.648 2.139 C HETATM 20 H 0 0.502 -1.865 2.513 H HETATM 21 C 0 0.688 -2.127 -2.165 C HETATM 22 H 0 1.410 -1.370 -2.536 H HETATM 23 C 0 0.589 -3.384 -2.819 C HETATM 24 C 0 -0.353 -4.340 -2.317 C HETATM 25 C 0 -1.154 -3.977 -1.184 C HETATM 26 H 0 -1.883 -4.715 -0.787 H HETATM 27 C 0 -0.996 -2.697 -0.594 C HETATM 28 C 0 -1.794 -2.217 0.595 C HETATM 29 C 0 -2.838 -2.966 1.197 C HETATM 30 H 0 -3.135 -3.966 0.812 H HETATM 31 C 0 -3.539 -2.426 2.325 C HETATM 32 C 0 -3.154 -1.134 2.811 C HETATM 33 C 0 -2.101 -0.450 2.147 C HETATM 34 H 0 -1.784 0.552 2.506 H HETATM 35 C 0 -2.113 0.467 -2.158 C HETATM 36 H 0 -1.820 -0.541 -2.519 H HETATM 37 C 0 -3.158 1.172 -2.810 C HETATM 38 C 0 -3.512 2.472 -2.321 C HETATM 39 C 0 -2.788 2.998 -1.201 C HETATM 40 H 0 -3.061 4.003 -0.813 H HETATM 41 C 0 -1.753 2.228 -0.609 C HETATM 42 C 0 -0.933 2.691 0.571 C HETATM 43 C 0 -1.057 3.975 1.162 C HETATM 44 H 0 -1.774 4.729 0.771 H HETATM 45 C 0 -0.238 4.322 2.286 C HETATM 46 C 0 0.689 3.346 2.779 C HETATM 47 C 0 0.754 2.087 2.126 C HETATM 48 H 0 1.463 1.314 2.490 H HETATM 49 Fe 0 0.051 -0.014 -0.016 Fe HETATM 50 H 0 2.509 2.799 -3.725 H HETATM 51 H 0 2.484 -2.881 3.668 H HETATM 52 H 0 -3.648 -0.649 3.679 H HETATM 53 H 0 1.243 -3.589 -3.692 H HETATM 54 H 0 -3.670 0.697 -3.674 H 217 Table 5.S14 (cont’d) HETATM 55 H 0 1.356 3.538 3.646 H HETATM 56 O 0 -4.505 3.282 -2.837 O HETATM 57 O 0 -0.419 5.590 2.802 O HETATM 58 O 0 -4.543 -3.216 2.851 O HETATM 59 O 0 -0.567 -5.604 -2.832 O HETATM 60 O 0 5.147 -2.395 2.798 O HETATM 61 O 0 5.170 2.220 -2.908 O HETATM 62 C 0 5.095 -3.306 3.971 C HETATM 63 H 0 6.164 -3.547 4.191 H HETATM 64 H 0 4.627 -2.790 4.851 H HETATM 65 H 0 4.532 -4.244 3.716 H HETATM 66 C 0 5.126 3.131 -4.081 C HETATM 67 H 0 6.197 3.337 -4.321 H HETATM 68 H 0 4.625 2.629 -4.952 H HETATM 69 H 0 4.598 4.087 -3.818 H HETATM 70 C 0 0.238 -6.034 -4.004 C HETATM 71 H 0 -0.113 -7.072 -4.225 H HETATM 72 H 0 0.042 -5.365 -4.884 H HETATM 73 H 0 1.332 -6.043 -3.749 H HETATM 74 C 0 -5.308 -2.701 4.016 C HETATM 75 H 0 -5.841 -1.750 3.749 H HETATM 76 H 0 -6.049 -3.505 4.247 H HETATM 77 H 0 -4.627 -2.541 4.895 H HETATM 78 C 0 -5.294 2.781 -3.993 C HETATM 79 H 0 -5.841 1.840 -3.718 H HETATM 80 H 0 -6.022 3.599 -4.215 H HETATM 81 H 0 -4.627 2.607 -4.879 H HETATM 82 C 0 0.406 6.003 3.967 C HETATM 83 H 0 0.079 7.048 4.191 H HETATM 84 H 0 0.204 5.337 4.848 H HETATM 85 H 0 1.498 5.989 3.701 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 50 CONECT 10 9 11 61 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 60 CONECT 18 17 19 51 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 53 CONECT 24 23 25 59 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 3 25 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 58 CONECT 32 31 33 52 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 218 Table 5.S14 (cont’d) CONECT 37 35 38 54 CONECT 38 37 39 56 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 5 39 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 57 CONECT 46 45 47 55 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 9 CONECT 51 18 CONECT 52 32 CONECT 53 23 CONECT 54 37 CONECT 55 46 CONECT 56 38 78 CONECT 57 45 82 CONECT 58 31 74 CONECT 59 24 70 CONECT 60 17 62 CONECT 61 10 66 CONECT 62 60 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 61 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 56 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 57 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 Table 5.S15: The optimized geometry file for the lowest energy quintet state of [Fe(4,4’-diOMe- bpy)3]2+. TITLE Fe-4,4'-OMe-bpy Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.804 0.747 -1.143 N HETATM 2 N 0 1.804 -0.811 1.094 N HETATM 3 N 0 -0.199 -1.956 -1.164 N HETATM 4 N 0 -1.508 -1.204 1.118 N HETATM 5 N 0 -1.486 1.204 -1.131 N HETATM 6 N 0 -0.145 1.926 1.141 N HETATM 7 C 0 1.752 1.565 -2.247 C HETATM 8 H 0 0.733 1.819 -2.616 H HETATM 9 C 0 2.894 2.081 -2.915 C HETATM 10 C 0 4.185 1.731 -2.401 C HETATM 11 C 0 4.257 0.881 -1.248 C HETATM 12 H 0 5.261 0.618 -0.850 H HETATM 13 C 0 3.060 0.405 -0.646 C HETATM 14 C 0 3.060 -0.501 0.576 C 219 Table 5.S15 (cont’d) HETATM 15 C 0 4.254 -1.009 1.157 C HETATM 16 H 0 5.257 -0.773 0.742 H HETATM 17 C 0 4.179 -1.860 2.310 C HETATM 18 C 0 2.888 -2.175 2.847 C HETATM 19 C 0 1.749 -1.628 2.199 C HETATM 20 H 0 0.730 -1.856 2.585 H HETATM 21 C 0 0.553 -2.329 -2.252 C HETATM 22 H 0 1.249 -1.557 -2.650 H HETATM 23 C 0 0.486 -3.607 -2.869 C HETATM 24 C 0 -0.412 -4.575 -2.310 C HETATM 25 C 0 -1.202 -4.202 -1.172 C HETATM 26 H 0 -1.884 -4.959 -0.729 H HETATM 27 C 0 -1.081 -2.892 -0.633 C HETATM 28 C 0 -1.896 -2.422 0.566 C HETATM 29 C 0 -2.990 -3.162 1.087 C HETATM 30 H 0 -3.319 -4.123 0.636 H HETATM 31 C 0 -3.714 -2.657 2.220 C HETATM 32 C 0 -3.309 -1.407 2.794 C HETATM 33 C 0 -2.211 -0.731 2.201 C HETATM 34 H 0 -1.868 0.245 2.614 H HETATM 35 C 0 -2.208 0.746 -2.208 C HETATM 36 H 0 -1.892 -0.238 -2.621 H HETATM 37 C 0 -3.295 1.447 -2.793 C HETATM 38 C 0 -3.666 2.707 -2.218 C HETATM 39 C 0 -2.922 3.197 -1.092 C HETATM 40 H 0 -3.224 4.165 -0.639 H HETATM 41 C 0 -1.841 2.432 -0.578 C HETATM 42 C 0 -1.006 2.884 0.614 C HETATM 43 C 0 -1.090 4.197 1.150 C HETATM 44 H 0 -1.755 4.972 0.711 H HETATM 45 C 0 -0.283 4.552 2.284 C HETATM 46 C 0 0.592 3.563 2.840 C HETATM 47 C 0 0.622 2.282 2.224 C HETATM 48 H 0 1.301 1.493 2.620 H HETATM 49 Fe 0 0.019 -0.017 -0.012 Fe HETATM 50 H 0 2.757 2.734 -3.802 H HETATM 51 H 0 2.750 -2.825 3.736 H HETATM 52 H 0 -3.820 -0.955 3.670 H HETATM 53 H 0 1.128 -3.823 -3.749 H HETATM 54 H 0 -3.823 1.006 -3.664 H HETATM 55 H 0 1.245 3.763 3.715 H HETATM 56 O 0 -4.699 3.510 -2.652 O HETATM 57 O 0 -0.424 5.847 2.742 O HETATM 58 O 0 -4.761 -3.438 2.662 O HETATM 59 O 0 -0.588 -5.865 -2.770 O HETATM 60 O 0 5.387 -2.309 2.808 O HETATM 61 O 0 5.395 2.145 -2.923 O HETATM 62 C 0 5.371 -3.202 3.995 C HETATM 63 H 0 6.448 -3.422 4.200 H HETATM 64 H 0 4.911 -2.679 4.876 H HETATM 65 H 0 4.820 -4.153 3.767 H HETATM 66 C 0 5.383 3.030 -4.117 C HETATM 67 H 0 6.461 3.219 -4.342 H HETATM 68 H 0 4.892 2.513 -4.985 H HETATM 69 H 0 4.862 3.997 -3.886 H HETATM 70 C 0 0.213 -6.314 -3.939 C HETATM 71 H 0 -0.108 -7.370 -4.115 H HETATM 72 H 0 -0.017 -5.685 -4.840 H HETATM 73 H 0 1.310 -6.279 -3.700 H HETATM 74 C 0 -5.554 -2.965 3.828 C HETATM 75 H 0 -6.050 -1.985 3.596 H HETATM 76 H 0 -6.324 -3.758 3.988 H HETATM 77 H 0 -4.900 -2.874 4.737 H HETATM 78 C 0 -5.513 3.053 -3.810 C HETATM 79 H 0 -6.026 2.083 -3.571 H HETATM 80 H 0 -6.269 3.861 -3.962 H HETATM 81 H 0 -4.871 2.947 -4.725 H HETATM 82 C 0 0.394 6.277 3.906 C 220 Table 5.S15 (cont’d) HETATM 83 H 0 0.101 7.341 4.082 H HETATM 84 H 0 0.152 5.655 4.809 H HETATM 85 H 0 1.489 6.213 3.663 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 50 CONECT 10 9 11 61 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 60 CONECT 18 17 19 51 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 53 CONECT 24 23 25 59 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 3 25 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 58 CONECT 32 31 33 52 CONECT 33 4 32 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 54 CONECT 38 37 39 56 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 5 39 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 57 CONECT 46 45 47 55 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 9 CONECT 51 18 CONECT 52 32 CONECT 53 23 CONECT 54 37 CONECT 55 46 CONECT 56 38 78 CONECT 57 45 82 CONECT 58 31 74 CONECT 59 24 70 CONECT 60 17 62 CONECT 61 10 66 CONECT 62 60 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 221 Table 5.S15 (cont’d) CONECT 66 61 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 56 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 57 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 [Fe(5,5’-diOMe-bpy)3]2+: Table 5.S16: The optimized geometry file for the lowest energy singlet state of [Fe(5,5’-diOMe- bpy)3]2+. TITLE Fe-5,5'-OMe-bpy Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.560 0.765 -1.080 N HETATM 2 N 0 1.545 -0.791 1.080 N HETATM 3 N 0 -0.104 -1.754 -1.048 N HETATM 4 N 0 -1.472 -0.924 1.080 N HETATM 5 N 0 -1.455 0.947 -1.086 N HETATM 6 N 0 -0.077 1.757 1.044 N HETATM 7 C 0 1.463 1.591 -2.169 C HETATM 8 H 0 0.447 1.839 -2.541 H HETATM 9 C 0 2.609 2.133 -2.841 C HETATM 10 C 0 3.911 1.799 -2.351 C HETATM 11 C 0 4.005 0.942 -1.220 C HETATM 12 H 0 5.010 0.678 -0.825 H HETATM 13 C 0 2.828 0.431 -0.594 C HETATM 14 C 0 2.819 -0.480 0.596 C HETATM 15 C 0 3.986 -1.012 1.223 C HETATM 16 H 0 4.996 -0.766 0.829 H HETATM 17 C 0 3.876 -1.866 2.355 C HETATM 18 C 0 2.567 -2.175 2.844 C HETATM 19 C 0 1.432 -1.614 2.170 C HETATM 20 H 0 0.411 -1.843 2.542 H HETATM 21 C 0 0.667 -2.100 -2.126 C HETATM 22 H 0 1.390 -1.348 -2.508 H HETATM 23 C 0 0.570 -3.374 -2.778 C HETATM 24 C 0 -0.369 -4.330 -2.276 C HETATM 25 C 0 -1.165 -3.966 -1.155 C HETATM 26 H 0 -1.897 -4.699 -0.751 H HETATM 27 C 0 -1.027 -2.680 -0.551 C HETATM 28 C 0 -1.824 -2.197 0.623 C HETATM 29 C 0 -2.867 -2.936 1.259 C HETATM 30 H 0 -3.145 -3.947 0.888 H HETATM 31 C 0 -3.568 -2.392 2.370 C HETATM 32 C 0 -3.201 -1.087 2.828 C HETATM 33 C 0 -2.146 -0.392 2.149 C HETATM 34 H 0 -1.848 0.618 2.499 H HETATM 35 C 0 -2.136 0.427 -2.155 C HETATM 36 H 0 -1.853 -0.588 -2.505 H HETATM 37 C 0 -3.180 1.138 -2.835 C HETATM 38 C 0 -3.529 2.447 -2.376 C HETATM 39 C 0 -2.821 2.981 -1.264 C HETATM 40 H 0 -3.084 3.995 -0.892 H 222 Table 5.S16 (cont’d) HETATM 41 C 0 -1.789 2.226 -0.628 C HETATM 42 C 0 -0.986 2.697 0.547 C HETATM 43 C 0 -1.103 3.985 1.150 C HETATM 44 H 0 -1.823 4.729 0.745 H HETATM 45 C 0 -0.303 4.338 2.271 C HETATM 46 C 0 0.621 3.367 2.774 C HETATM 47 C 0 0.699 2.091 2.123 C HETATM 48 H 0 1.409 1.328 2.506 H HETATM 49 Fe 0 -0.001 0.000 -0.001 Fe HETATM 50 H 0 -0.405 5.344 2.729 H HETATM 51 H 0 -4.330 3.046 -2.860 H HETATM 52 H 0 4.836 2.189 -2.828 H HETATM 53 H 0 4.793 -2.272 2.832 H HETATM 54 H 0 -4.378 -2.979 2.854 H HETATM 55 H 0 -0.487 -5.335 -2.734 H HETATM 56 O 0 -3.772 -0.401 3.890 O HETATM 57 O 0 1.422 -3.552 -3.858 O HETATM 58 O 0 -3.760 0.460 -3.897 O HETATM 59 O 0 1.475 3.532 3.855 O HETATM 60 O 0 2.326 2.948 -3.928 O HETATM 61 O 0 2.269 -2.984 3.932 O HETATM 62 C 0 -4.866 -1.065 4.640 C HETATM 63 H 0 -4.500 -2.016 5.114 H HETATM 64 H 0 -5.158 -0.331 5.431 H HETATM 65 H 0 -5.741 -1.272 3.965 H HETATM 66 C 0 1.364 -4.846 -4.583 C HETATM 67 H 0 0.345 -5.003 -5.030 H HETATM 68 H 0 2.127 -4.754 -5.394 H HETATM 69 H 0 1.630 -5.696 -3.899 H HETATM 70 C 0 1.434 4.826 4.581 C HETATM 71 H 0 0.417 4.997 5.026 H HETATM 72 H 0 2.195 4.722 5.393 H HETATM 73 H 0 1.713 5.673 3.898 H HETATM 74 C 0 -4.844 1.141 -4.648 C HETATM 75 H 0 -4.464 2.087 -5.120 H HETATM 76 H 0 -5.146 0.412 -5.439 H HETATM 77 H 0 -5.716 1.360 -3.974 H HETATM 78 C 0 3.468 3.537 -4.670 C HETATM 79 H 0 4.114 2.729 -5.108 H HETATM 80 H 0 2.998 4.137 -5.488 H HETATM 81 H 0 4.073 4.206 -4.001 H HETATM 82 C 0 3.398 -3.592 4.677 C HETATM 83 H 0 4.059 -2.795 5.114 H HETATM 84 H 0 2.916 -4.181 5.496 H HETATM 85 H 0 3.992 -4.274 4.011 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 60 CONECT 10 9 11 52 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 53 CONECT 18 17 19 61 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 223 Table 5.S16 (cont’d) CONECT 23 21 24 57 CONECT 24 23 25 55 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 54 CONECT 32 31 33 56 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 58 CONECT 38 37 39 51 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 50 CONECT 46 45 47 59 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 45 CONECT 51 38 CONECT 52 10 CONECT 53 17 CONECT 54 31 CONECT 55 24 CONECT 56 32 62 CONECT 57 23 66 CONECT 58 37 74 CONECT 59 46 70 CONECT 60 9 78 CONECT 61 18 82 CONECT 62 56 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 57 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 60 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 61 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 Table 5.S17: The optimized geometry file for the lowest energy quintet state of [Fe(5,5’-diOMe- bpy)3]2+. TITLE Fe-5,5'-OMe-bpy Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 224 Table 5.S17 (cont’d) HETATM 1 N 0 1.762 0.755 -1.136 N HETATM 2 N 0 1.755 -0.784 1.135 N HETATM 3 N 0 -0.225 -1.959 -1.122 N HETATM 4 N 0 -1.557 -1.172 1.150 N HETATM 5 N 0 -1.531 1.191 -1.148 N HETATM 6 N 0 -0.194 1.954 1.128 N HETATM 7 C 0 1.689 1.545 -2.250 C HETATM 8 H 0 0.675 1.793 -2.638 H HETATM 9 C 0 2.848 2.054 -2.925 C HETATM 10 C 0 4.136 1.722 -2.399 C HETATM 11 C 0 4.203 0.901 -1.239 C HETATM 12 H 0 5.201 0.646 -0.823 H HETATM 13 C 0 3.009 0.419 -0.615 C HETATM 14 C 0 3.005 -0.459 0.613 C HETATM 15 C 0 4.194 -0.955 1.236 C HETATM 16 H 0 5.195 -0.709 0.820 H HETATM 17 C 0 4.120 -1.776 2.395 C HETATM 18 C 0 2.828 -2.097 2.921 C HETATM 19 C 0 1.675 -1.575 2.248 C HETATM 20 H 0 0.659 -1.814 2.635 H HETATM 21 C 0 0.518 -2.308 -2.213 C HETATM 22 H 0 1.227 -1.550 -2.615 H HETATM 23 C 0 0.420 -3.592 -2.848 C HETATM 24 C 0 -0.489 -4.549 -2.296 C HETATM 25 C 0 -1.257 -4.178 -1.156 C HETATM 26 H 0 -1.955 -4.923 -0.717 H HETATM 27 C 0 -1.123 -2.875 -0.582 C HETATM 28 C 0 -1.915 -2.407 0.619 C HETATM 29 C 0 -2.989 -3.151 1.200 C HETATM 30 H 0 -3.294 -4.133 0.776 H HETATM 31 C 0 -3.698 -2.643 2.324 C HETATM 32 C 0 -3.316 -1.371 2.859 C HETATM 33 C 0 -2.233 -0.673 2.228 C HETATM 34 H 0 -1.911 0.318 2.621 H HETATM 35 C 0 -2.212 0.704 -2.228 C HETATM 36 H 0 -1.911 -0.295 -2.616 H HETATM 37 C 0 -3.275 1.425 -2.867 C HETATM 38 C 0 -3.630 2.707 -2.339 C HETATM 39 C 0 -2.916 3.203 -1.212 C HETATM 40 H 0 -3.200 4.193 -0.794 H HETATM 41 C 0 -1.863 2.436 -0.622 C HETATM 42 C 0 -1.068 2.889 0.582 C HETATM 43 C 0 -1.178 4.195 1.154 C HETATM 44 H 0 -1.857 4.954 0.711 H HETATM 45 C 0 -0.411 4.550 2.299 C HETATM 46 C 0 0.473 3.574 2.859 C HETATM 47 C 0 0.549 2.288 2.225 C HETATM 48 H 0 1.239 1.515 2.633 H HETATM 49 Fe 0 -0.019 -0.004 0.003 Fe HETATM 50 H 0 -0.507 5.569 2.732 H HETATM 51 H 0 -4.447 3.317 -2.783 H HETATM 52 H 0 5.074 2.087 -2.869 H HETATM 53 H 0 5.054 -2.152 2.864 H HETATM 54 H 0 -4.531 -3.234 2.761 H HETATM 55 H 0 -0.604 -5.566 -2.729 H HETATM 56 O 0 -3.895 -0.728 3.943 O HETATM 57 O 0 1.239 -3.781 -3.952 O HETATM 58 O 0 -3.863 0.792 -3.952 O HETATM 59 O 0 1.288 3.745 3.969 O HETATM 60 O 0 2.595 2.837 -4.043 O HETATM 61 O 0 2.568 -2.879 4.038 O HETATM 62 C 0 -5.019 -1.406 4.634 C HETATM 63 H 0 -4.684 -2.387 5.066 H HETATM 64 H 0 -5.313 -0.707 5.455 H HETATM 65 H 0 -5.882 -1.559 3.931 H HETATM 66 C 0 1.182 -5.092 -4.643 C HETATM 67 H 0 0.152 -5.281 -5.049 H HETATM 68 H 0 1.915 -5.006 -5.482 H 225 Table 5.S17 (cont’d) HETATM 69 H 0 1.488 -5.920 -3.948 H HETATM 70 C 0 1.251 5.056 4.662 C HETATM 71 H 0 0.221 5.266 5.059 H HETATM 72 H 0 1.974 4.953 5.509 H HETATM 73 H 0 1.582 5.878 3.972 H HETATM 74 C 0 -4.964 1.496 -4.655 C HETATM 75 H 0 -4.600 2.467 -5.087 H HETATM 76 H 0 -5.269 0.802 -5.477 H HETATM 77 H 0 -5.829 1.672 -3.960 H HETATM 78 C 0 3.757 3.385 -4.786 C HETATM 79 H 0 4.398 2.553 -5.186 H HETATM 80 H 0 3.310 3.964 -5.631 H HETATM 81 H 0 4.362 4.067 -4.129 H HETATM 82 C 0 3.725 -3.441 4.779 C HETATM 83 H 0 4.375 -2.617 5.181 H HETATM 84 H 0 3.272 -4.017 5.623 H HETATM 85 H 0 4.322 -4.128 4.120 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 60 CONECT 10 9 11 52 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 53 CONECT 18 17 19 61 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 57 CONECT 24 23 25 55 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 54 CONECT 32 31 33 56 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 58 CONECT 38 37 39 51 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 50 CONECT 46 45 47 59 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 45 CONECT 51 38 226 Table 5.S17 (cont’d) CONECT 52 10 CONECT 53 17 CONECT 54 31 CONECT 55 24 CONECT 56 32 62 CONECT 57 23 66 CONECT 58 37 74 CONECT 59 46 70 CONECT 60 9 78 CONECT 61 18 82 CONECT 62 56 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 57 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 60 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 61 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 [Fe(4,4’-diCl-bpy)3]2+: Table 5.S18: The optimized geometry file for the lowest energy singlet state of [Fe(4,4’-diCl- bpy)3]2+. TITLE Fe-4,4'-Cl-bpy Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.566 0.764 -1.075 N HETATM 2 N 0 1.553 -0.785 1.078 N HETATM 3 N 0 -0.107 -1.758 -1.042 N HETATM 4 N 0 -1.472 -0.931 1.078 N HETATM 5 N 0 -1.459 0.953 -1.080 N HETATM 6 N 0 -0.084 1.763 1.041 N HETATM 7 C 0 1.481 1.602 -2.168 C HETATM 8 H 0 0.461 1.840 -2.534 H HETATM 9 C 0 2.616 2.145 -2.826 C HETATM 10 H 0 2.486 2.813 -3.705 H HETATM 11 C 0 3.903 1.804 -2.323 C HETATM 12 C 0 4.025 0.944 -1.198 C HETATM 13 H 0 5.029 0.681 -0.803 H HETATM 14 C 0 2.835 0.437 -0.596 C HETATM 15 C 0 2.827 -0.485 0.595 C HETATM 16 C 0 4.008 -1.015 1.196 C HETATM 17 H 0 5.016 -0.774 0.798 H HETATM 18 C 0 3.871 -1.869 2.323 C HETATM 19 C 0 2.579 -2.181 2.832 C HETATM 20 H 0 2.438 -2.843 3.713 H HETATM 21 C 0 1.454 -1.617 2.174 C HETATM 22 H 0 0.430 -1.834 2.544 H HETATM 23 C 0 0.669 -2.122 -2.122 C HETATM 24 H 0 1.386 -1.363 -2.499 H HETATM 25 C 0 0.579 -3.388 -2.758 C HETATM 26 H 0 1.229 -3.624 -3.628 H 227 Table 5.S18 (cont’d) HETATM 27 C 0 -0.358 -4.326 -2.242 C HETATM 28 C 0 -1.173 -3.983 -1.130 C HETATM 29 H 0 -1.901 -4.716 -0.723 H HETATM 30 C 0 -1.025 -2.686 -0.552 C HETATM 31 C 0 -1.835 -2.197 0.621 C HETATM 32 C 0 -2.887 -2.949 1.225 C HETATM 33 H 0 -3.171 -3.953 0.844 H HETATM 34 C 0 -3.577 -2.381 2.329 C HETATM 35 C 0 -3.217 -1.092 2.812 C HETATM 36 H 0 -3.736 -0.621 3.674 H HETATM 37 C 0 -2.162 -0.408 2.152 C HETATM 38 H 0 -1.850 0.598 2.503 H HETATM 39 C 0 -2.155 0.438 -2.154 C HETATM 40 H 0 -1.856 -0.571 -2.505 H HETATM 41 C 0 -3.202 1.135 -2.813 C HETATM 42 H 0 -3.727 0.671 -3.675 H HETATM 43 C 0 -3.547 2.429 -2.330 C HETATM 44 C 0 -2.849 2.988 -1.226 C HETATM 45 H 0 -3.121 3.995 -0.845 H HETATM 46 C 0 -1.806 2.224 -0.623 C HETATM 47 C 0 -0.990 2.703 0.550 C HETATM 48 C 0 -1.119 4.003 1.125 C HETATM 49 H 0 -1.837 4.745 0.718 H HETATM 50 C 0 -0.299 4.337 2.236 C HETATM 51 C 0 0.626 3.388 2.753 C HETATM 52 H 0 1.280 3.616 3.622 H HETATM 53 C 0 0.698 2.118 2.120 C HETATM 54 H 0 1.404 1.351 2.498 H HETATM 55 Fe 0 -0.001 0.001 0.000 Fe HETATM 56 Cl 0 -0.431 5.958 2.982 Cl HETATM 57 Cl 0 -4.863 3.356 -3.111 Cl HETATM 58 Cl 0 -4.905 -3.292 3.111 Cl HETATM 59 Cl 0 -0.515 -5.943 -2.992 Cl HETATM 60 Cl 0 5.331 -2.554 3.100 Cl HETATM 61 Cl 0 5.374 2.460 -3.102 Cl END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 59 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 228 Table 5.S18 (cont’d) CONECT 33 32 CONECT 34 32 35 58 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 57 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 56 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 50 CONECT 57 43 CONECT 58 34 CONECT 59 27 CONECT 60 18 CONECT 61 11 Table 5.S19: The optimized geometry file for the lowest energy quintet state of [Fe(4,4’-diCl- bpy)3]2+. TITLE Fe-4,4'-Cl-bpy Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.838 0.734 -1.117 N HETATM 2 N 0 1.810 -0.804 1.143 N HETATM 3 N 0 -0.222 -1.937 -1.084 N HETATM 4 N 0 -1.568 -1.105 1.144 N HETATM 5 N 0 -1.567 1.140 -1.169 N HETATM 6 N 0 -0.258 1.954 1.096 N HETATM 7 C 0 1.780 1.580 -2.201 C HETATM 8 H 0 0.767 1.797 -2.605 H HETATM 9 C 0 2.930 2.161 -2.801 C HETATM 10 H 0 2.832 2.836 -3.678 H HETATM 11 C 0 4.194 1.843 -2.231 C HETATM 12 C 0 4.284 0.973 -1.109 C HETATM 13 H 0 5.274 0.747 -0.663 H HETATM 14 C 0 3.077 0.426 -0.575 C HETATM 15 C 0 3.062 -0.514 0.623 C HETATM 16 C 0 4.256 -1.070 1.179 C HETATM 17 H 0 5.255 -0.860 0.744 H HETATM 18 C 0 4.140 -1.924 2.310 C HETATM 19 C 0 2.863 -2.217 2.865 C HETATM 20 H 0 2.744 -2.880 3.749 H HETATM 21 C 0 1.729 -1.635 2.235 C HETATM 22 H 0 0.707 -1.843 2.623 H HETATM 23 C 0 0.514 -2.318 -2.184 C HETATM 24 H 0 1.250 -1.576 -2.565 H HETATM 25 C 0 0.371 -3.576 -2.827 C HETATM 26 H 0 0.989 -3.832 -3.714 H HETATM 27 C 0 -0.585 -4.481 -2.287 C HETATM 28 C 0 -1.358 -4.119 -1.150 C HETATM 29 H 0 -2.095 -4.835 -0.733 H HETATM 30 C 0 -1.154 -2.828 -0.569 C HETATM 31 C 0 -1.931 -2.348 0.645 C HETATM 32 C 0 -2.975 -3.121 1.245 C HETATM 33 H 0 -3.274 -4.110 0.840 H HETATM 34 C 0 -3.640 -2.593 2.384 C 229 Table 5.S19 (cont’d) HETATM 35 C 0 -3.273 -1.324 2.911 C HETATM 36 H 0 -3.775 -0.888 3.801 H HETATM 37 C 0 -2.231 -0.620 2.249 C HETATM 38 H 0 -1.909 0.378 2.619 H HETATM 39 C 0 -2.250 0.646 -2.255 C HETATM 40 H 0 -1.917 -0.344 -2.639 H HETATM 41 C 0 -3.326 1.330 -2.885 C HETATM 42 H 0 -3.842 0.888 -3.764 H HETATM 43 C 0 -3.712 2.584 -2.336 C HETATM 44 C 0 -3.026 3.122 -1.213 C HETATM 45 H 0 -3.344 4.095 -0.782 H HETATM 46 C 0 -1.943 2.373 -0.657 C HETATM 47 C 0 -1.132 2.870 0.531 C HETATM 48 C 0 -1.244 4.204 1.032 C HETATM 49 H 0 -1.924 4.946 0.567 H HETATM 50 C 0 -0.441 4.578 2.145 C HETATM 51 C 0 0.453 3.644 2.740 C HETATM 52 H 0 1.090 3.910 3.611 H HETATM 53 C 0 0.508 2.342 2.171 C HETATM 54 H 0 1.192 1.573 2.593 H HETATM 55 Fe 0 0.023 0.013 0.002 Fe HETATM 56 Cl 0 -0.554 6.242 2.794 Cl HETATM 57 Cl 0 -5.075 3.495 -3.058 Cl HETATM 58 Cl 0 -4.953 -3.537 3.157 Cl HETATM 59 Cl 0 -0.819 -6.090 -3.038 Cl HETATM 60 Cl 0 5.616 -2.638 3.032 Cl HETATM 61 Cl 0 5.688 2.544 -2.923 Cl END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 59 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 58 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 230 Table 5.S19 (cont’d) CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 57 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 56 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 50 CONECT 57 43 CONECT 58 34 CONECT 59 27 CONECT 60 18 CONECT 61 11 [Fe(5,5’-diCl-bpy)3]2+: Table 5.S20: The optimized geometry file for the lowest energy singlet state of [Fe(5,5’-diCl- bpy)3]2+. TITLE Fe-5,5-Cl'-bpy Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.566 0.767 -1.075 N HETATM 2 N 0 1.553 -0.789 1.076 N HETATM 3 N 0 -0.105 -1.760 -1.040 N HETATM 4 N 0 -1.475 -0.929 1.078 N HETATM 5 N 0 -1.461 0.950 -1.080 N HETATM 6 N 0 -0.081 1.764 1.039 N HETATM 7 C 0 1.466 1.605 -2.162 C HETATM 8 H 0 0.449 1.844 -2.532 H HETATM 9 C 0 2.618 2.143 -2.802 C HETATM 10 C 0 3.919 1.824 -2.323 C HETATM 11 C 0 4.018 0.958 -1.202 C HETATM 12 H 0 5.022 0.694 -0.805 H HETATM 13 C 0 2.834 0.438 -0.593 C HETATM 14 C 0 2.826 -0.484 0.592 C HETATM 15 C 0 4.002 -1.023 1.201 C HETATM 16 H 0 5.009 -0.779 0.803 H HETATM 17 C 0 3.889 -1.886 2.324 C HETATM 18 C 0 2.583 -2.180 2.805 C HETATM 19 C 0 1.440 -1.623 2.165 C HETATM 20 H 0 0.419 -1.844 2.537 H HETATM 21 C 0 0.679 -2.114 -2.115 C HETATM 22 H 0 1.399 -1.360 -2.493 H HETATM 23 C 0 0.573 -3.391 -2.732 C HETATM 24 C 0 -0.356 -4.350 -2.243 C HETATM 25 C 0 -1.164 -3.982 -1.134 C HETATM 26 H 0 -1.894 -4.713 -0.726 H HETATM 27 C 0 -1.028 -2.686 -0.549 C HETATM 28 C 0 -1.837 -2.196 0.618 C HETATM 29 C 0 -2.896 -2.935 1.229 C HETATM 30 H 0 -3.183 -3.938 0.847 H HETATM 31 C 0 -3.600 -2.385 2.334 C HETATM 32 C 0 -3.211 -1.095 2.792 C HETATM 33 C 0 -2.155 -0.393 2.148 C HETATM 34 H 0 -1.844 0.611 2.502 H HETATM 35 C 0 -2.149 0.424 -2.150 C HETATM 36 H 0 -1.853 -0.585 -2.504 H HETATM 37 C 0 -3.195 1.140 -2.794 C HETATM 38 C 0 -3.567 2.435 -2.336 C HETATM 39 C 0 -2.855 2.975 -1.231 C HETATM 40 H 0 -3.128 3.983 -0.849 H 231 Table 5.S20 (cont’d) HETATM 41 C 0 -1.806 2.222 -0.620 C HETATM 42 C 0 -0.991 2.701 0.547 C HETATM 43 C 0 -1.109 4.000 1.131 C HETATM 44 H 0 -1.830 4.741 0.723 H HETATM 45 C 0 -0.296 4.358 2.239 C HETATM 46 C 0 0.621 3.387 2.729 C HETATM 47 C 0 0.709 2.108 2.113 C HETATM 48 H 0 1.419 1.344 2.492 H HETATM 49 Fe 0 -0.000 0.001 -0.000 Fe HETATM 50 Cl 0 -4.039 -0.329 4.183 Cl HETATM 51 Cl 0 1.628 -3.768 -4.129 Cl HETATM 52 Cl 0 2.346 -3.255 4.217 Cl HETATM 53 Cl 0 2.399 3.225 -4.212 Cl HETATM 54 Cl 0 1.681 3.751 4.125 Cl HETATM 55 Cl 0 -4.033 0.385 -4.184 Cl HETATM 56 H 0 -0.373 5.364 2.706 H HETATM 57 H 0 -4.386 3.006 -2.825 H HETATM 58 H 0 4.829 2.238 -2.809 H HETATM 59 H 0 4.792 -2.316 2.809 H HETATM 60 H 0 -4.428 -2.944 2.823 H HETATM 61 H 0 -0.447 -5.355 -2.710 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 53 CONECT 10 9 11 58 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 59 CONECT 18 17 19 52 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 51 CONECT 24 23 25 61 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 60 CONECT 32 31 33 50 CONECT 33 4 32 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 55 CONECT 38 37 39 57 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 41 6 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 56 CONECT 46 45 47 54 232 Table 5.S20 (cont’d) CONECT 47 6 46 48 CONECT 48 47 CONECT 50 32 CONECT 51 23 CONECT 52 18 CONECT 53 9 CONECT 54 46 CONECT 55 37 CONECT 56 45 CONECT 57 38 CONECT 58 10 CONECT 59 17 CONECT 60 31 CONECT 61 24 Table 5.S21: The optimized geometry file for the lowest energy quintet state of [Fe(5,5’-diCl- bpy)3]2+. TITLE Fe-5,5-Cl'-bpy Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.809 0.746 -1.112 N HETATM 2 N 0 1.787 -0.805 1.135 N HETATM 3 N 0 -0.215 -1.954 -1.097 N HETATM 4 N 0 -1.561 -1.142 1.142 N HETATM 5 N 0 -1.554 1.174 -1.167 N HETATM 6 N 0 -0.240 1.965 1.115 N HETATM 7 C 0 1.735 1.565 -2.214 C HETATM 8 H 0 0.722 1.802 -2.604 H HETATM 9 C 0 2.902 2.089 -2.837 C HETATM 10 C 0 4.186 1.767 -2.316 C HETATM 11 C 0 4.254 0.919 -1.178 C HETATM 12 H 0 5.250 0.664 -0.758 H HETATM 13 C 0 3.052 0.415 -0.587 C HETATM 14 C 0 3.040 -0.491 0.627 C HETATM 15 C 0 4.233 -1.005 1.230 C HETATM 16 H 0 5.235 -0.765 0.817 H HETATM 17 C 0 4.144 -1.844 2.373 C HETATM 18 C 0 2.850 -2.145 2.881 C HETATM 19 C 0 1.696 -1.614 2.240 C HETATM 20 H 0 0.677 -1.843 2.620 H HETATM 21 C 0 0.529 -2.314 -2.194 C HETATM 22 H 0 1.258 -1.569 -2.580 H HETATM 23 C 0 0.382 -3.584 -2.819 C HETATM 24 C 0 -0.551 -4.526 -2.301 C HETATM 25 C 0 -1.316 -4.148 -1.165 C HETATM 26 H 0 -2.043 -4.874 -0.741 H HETATM 27 C 0 -1.138 -2.854 -0.580 C HETATM 28 C 0 -1.923 -2.380 0.626 C HETATM 29 C 0 -2.984 -3.137 1.218 C HETATM 30 H 0 -3.287 -4.122 0.803 H HETATM 31 C 0 -3.676 -2.627 2.350 C HETATM 32 C 0 -3.281 -1.360 2.863 C HETATM 33 C 0 -2.225 -0.643 2.237 C HETATM 34 H 0 -1.899 0.350 2.618 H HETATM 35 C 0 -2.235 0.672 -2.247 C HETATM 36 H 0 -1.905 -0.314 -2.641 H HETATM 37 C 0 -3.315 1.379 -2.846 C HETATM 38 C 0 -3.719 2.635 -2.314 C HETATM 39 C 0 -3.010 3.148 -1.193 C HETATM 40 H 0 -3.323 4.120 -0.756 H HETATM 41 C 0 -1.921 2.403 -0.637 C HETATM 42 C 0 -1.111 2.887 0.552 C HETATM 43 C 0 -1.211 4.217 1.071 C HETATM 44 H 0 -1.888 4.965 0.607 H HETATM 45 C 0 -0.422 4.601 2.190 C HETATM 46 C 0 0.454 3.634 2.757 C HETATM 47 C 0 0.526 2.329 2.194 C HETATM 48 H 0 1.208 1.558 2.614 H 233 Table 5.S21 (cont’d) HETATM 49 Fe 0 0.001 0.010 0.001 Fe HETATM 50 Cl 0 -4.097 -0.651 4.291 Cl HETATM 51 Cl 0 1.386 -3.978 -4.251 Cl HETATM 52 Cl 0 2.649 -3.192 4.323 Cl HETATM 53 Cl 0 2.730 3.148 -4.272 Cl HETATM 54 Cl 0 1.478 4.038 4.171 Cl HETATM 55 Cl 0 -4.151 0.669 -4.262 Cl HETATM 56 H 0 -0.486 5.631 2.606 H HETATM 57 H 0 -4.568 3.200 -2.758 H HETATM 58 H 0 5.112 2.167 -2.785 H HETATM 59 H 0 5.060 -2.253 2.854 H HETATM 60 H 0 -4.503 -3.203 2.820 H HETATM 61 H 0 -0.677 -5.527 -2.768 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 53 CONECT 10 9 11 58 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 59 CONECT 18 17 19 52 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 51 CONECT 24 23 25 61 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 27 4 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 60 CONECT 32 31 33 50 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 55 CONECT 38 37 39 57 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 41 6 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 56 CONECT 46 45 47 54 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 32 CONECT 51 23 CONECT 52 18 CONECT 53 9 CONECT 54 46 CONECT 55 37 234 Table 5.S21 (cont’d) CONECT 56 45 CONECT 57 38 CONECT 58 10 CONECT 59 17 CONECT 60 31 CONECT 61 24 235 CHAPTER 6: USING VIBRATIONAL COHERENCE TO DEFINE THE REACTION COORDINATE OF MLCT DEACTIVATION WITHIN FE(II)-POLYPYRIDINE BASED CHROMOPHORES 6.1 Introduction As mentioned in Chapter 1.2 of this work, a key factor limiting the utilization of Fe(II)- based chromophores in electron transfer processes is the incredibly short lifetime of the metal-to- ligand charge transfer (MLCT) excited state within this class of compounds.1-5 Many efforts to address this short lifetime target the electronic structure in an attempt to replicate that of Ru(II)- based complexes by destabilizing the ligand field (LF) excited states.6-10 However, in addition to altering the electronic structure, changing the molecular structure can also impact the excited state kinetics of a complex. This method, which we refer to as the kinetic approach, seeks to design ligands that restrict key normal modes, or geometric vibrations, that a complex engages in during excited state relaxation. It has been observed that Fe(II)-based complexes undergo significant structural changes as they relax from the MLCT excited state to the long-lived 5T2 excited state. For example, the Fe-N bonds in both [Fe(bpy)3]2+ (where bpy is 2,2’-bipyridine) and [Fe(terpy)2]2+ (where terpy is 2,2’:6’,2”-terpyridine) were found to lengthen by 0.2 Å.11-13 Additionally, theoretical work pioneered by Jakubikova and coworkers, suggests that ligand twisting and rocking motions are relevant to excited state relaxation in these complexes.14,15 Therefore, designing ligands to restrict these geometric changes appears to be a promising step towards increasing the MLCT excited state lifetime, however the presence of these geometric changes does not inherently mean that they are tied to the reaction coordinate of MLCT deactivation. A phenomena that can be used to identify the reaction coordinate of excited state relaxation is excited state vibrational coherences, which takes advantage of the quantum mechanical wave packet formation when an ultrafast laser pulse excites a chromophore. Quantum mechanical coherence can be understood by looking at another photophysical phenomena utilized throughout ultrafast spectroscopy: wave interference. When two waves with different periods overlap, there are two types of interference: constructive and destructive. Constructive interference occurs where the waves are in phase, in other words when the sign of the amplitude is the same, and the intensity of the waves are added together. Destructive interference is the opposite, when the waves are out of phase, the intensity of the waves is reduced. This is illustrated in Figure 6.1A, where two waves of different wavelengths are added together 236 Figure 6.1: Plots displaying wave interference, with the combined waves shown above each plot in black. Plot A shows the summation of two waves. The combined wave has more intense amplitude at points where the two waves have amplitudes of the same sign, resulting in constructive interference. The combined amplitude is near zero at points where the two wave have amplitudes of equal intensity but opposite sign, resulting in destructive interference. Plot B shows the summation of several waves. Constructive interference at certain points results in localized peaks of intense amplitude in the combined wave, while destructive interference results in lower intensity amplitude throughout the rest of the wave. providing a new wave shown in black. This can be done with multiple waves, as shown in Figure 6.1B, to produce localized high intensity peaks amidst lower intensity waves, this is the principle behind the production of modelocked laser pulses used throughout ultrafast spectroscopy; where several waves with slightly different frequencies interfere with each other to produce brief spikes of intense light amidst near zero amplitude destructive interference. The same principles of interference are applied to quantum mechanical wavefunctions within electronic states of a molecule. When a laser pulse is spectrally broad enough to span multiple vibrational levels within an excited state, which is the case for the prism-compressed femtosecond pulses generated by RR, it is possible to impulsively excite the sample into multiple vibrational levels, illustrated in Figure 6.2A.16 At this point, the wave function of the sample becomes a superposition of states, represented by the following equation: 𝛹(𝑅, 𝑡) = ∑ 𝑐𝑖𝜓𝑖(𝑅)𝑒𝑥𝑝 ( 𝑖 −𝑖𝐸𝑖𝑡 ℏ ) 6.1 Where Ψ(R,t) is the time-dependent net wavefunction, t is time, and ℏ is the reduced Planck’s constant. Furthermore, ci is the population coefficient, ψi(R) is the wavefunction, and Ei is the energy of the ith vibrational level.16 Much like the combined wave seen in Figure 6.1B, constructive interference of each ψi(R) results in a localized intense feature within Ψ(R,t) known as a “wave packet.” The formation of this wave packet is illustrated in Figure 6.2B. Since the 237 Figure 6.2: Schematics showing the generation of a wave packet. Schematic A displays a complex excited by an energetically broad laser pulse into multiple vibrational levels within an excited state, in this case an MLCT state. Schematic B displays a figure generated by a MATLAB program written by Dr. Bryan Paulus which creates a coherent superposition of five oscillating wavefunctions, correlating to different vibrational levels within an excited state, resulting in the formation of a wave packet.17 wavefunction at each vibrational level oscillates as a function of time, Ψ(R,t) does as well, resulting in the wave packet oscillating back and forth across the reaction coordinate of the excited state potential energy surface, a process which will be called vibrational coherence throughout this work. This vibrational coherence has an impact on the photophysical properties of the chromophore, of particular importance is the impact on the absorption profile.18 The migration of the wave packet causes oscillations in the λmax of the absorption feature from the excited state in which the wave packet has formed. During a single-wavelength transient absorption (TA) experiment, this manifests in oscillations in the change in absorbance (ΔA) signal observed as a function of time. This is illustrated in Figure 6.3. These oscillations in ΔA can be isolated from the TA signal as residuals of the excited state exponential decay. Fast Fourier Transformation (FFT) of these residual oscillations provides the frequency of these oscillations, which can easily be converted to wavenumbers (cm-1), shown in Figure 6.3C. This provides observable frequencies to the normal modes that the molecule engages in as the wave packet oscillates across the nuclear coordinate.19 Additionally, Density Functional Theory (DFT) frequency calculations can be used to predict what normal modes correspond to observed frequencies. Several normal modes may be observed from coherent measurements, however some normal modes, called spectator modes, engage in molecular motions within an electronic state that 238 Figure 6.3: Plot A displays a model of an excited state absorption feature oscillating as a function of time, a result of the propagation of a coherent wave packet across the excited state reaction coordinate. This oscillating absorption feature causes the Apumped, and there ΔA, to change significantly over time. Plot B shows the 490 nm pumped ultrafast TA spectrum of [Fe(terpy)2](PF6)2 displaying these oscillatory features, which are shown isolated above as residuals to the exponential decay fit shown in black. Plot C shows the FFT spectrum of those residuals, this provides the frequencies of oscillations found in plot B which have been converted from THz to frequency in cm-1. The most prominent oscillation observed has a frequency of 98.1 cm-1. do not coincide with the reaction coordinate of the transition between states. A promoting mode, on the other hand, does oscillate along this reaction coordinate, and facilitates excited state relaxation. Identifying promoting and spectator modes from the observed normal modes is crucial to designing molecular structures that slow excited state relaxation as only restricting promoting modes will assist in this effort. Previous studies have demonstrated that vibrational coherences can be conserved as excited complexes relax to lower energy electronic states.20,21 DFT frequency calculations done at specific spin and electronic states can predict the frequencies of normal modes within those excited states. This means that normal modes present in two states with frequencies that match those observed from vibrational coherent oscillations are more likely to stem from oscillations along the reaction coordinate of the transition between electronic states. For example, 239 a normal mode with a frequency near 98 cm-1 was observed in both MLCT and 3T DFT frequency calculations for [Fe(terpy)2]2+, as seen in Figure 6.3, the same frequency was observed in vibrational coherent oscillations upon exciting [Fe(terpy)2]2+ into the MLCT excited state. Since the MLCT excited state relaxes into the 3T excited state,22,23 the same normal mode being present in both states at a frequency that matches one observed in coherent data indicates that the 98 cm-1 normal mode is a strong candidate for a promoting mode for the MLCT → 3T transition, rather than a spectator mode. As an additional means of analyzing the coherent data in search of promoting and spectator modes, Duschinsky displacement vector calculations were conducted. A Duschinsky displacement vector (K) describes the distortion of each normal mode needed to shift a molecule from one geometry to another. In other words, when calculated from the geometries of two distinct electronic states, K represents a shift of the normal mode along the reaction coordinate of the transition between the selected electronic states, as depicted in Figure 6.4. The equation used to calculate K is the following:24 𝐾 = [𝑀𝐿𝑖𝑛𝑖𝑡𝑖𝑎𝑙] † 𝛥𝑅 6.2 Where M is a diagonal matrix containing the atomic masses, Lintitial is a matrix that diagonalizes the mass weighted Cartesian Hessian of the initial state geometry, and ΔR is the Cartesian displacement vector between the initial and product state geometries. As the non-adiabatic Marcus Figure 6.4: A non-adiabatic model for the transition between electronic states. In this model, several vibrational levels are populated within the initial state, resulting in a vibrationally coherent wave packet. As the system relaxes to the final state, the wave packet is conserved and continues to oscillate along the reaction coordinate, but has been displaced along ΔQ, the change in reaction coordinate between the initial and final states. 240 theory model provided in Figure 6.4 displays, a transition between electronic states involves a change in the reaction coordinate (ΔQ) between each state involved.25 Therefore, a vibrationally coherent normal mode that is preserved during an electronic state transition should also be displaced along the reaction coordinate during the transition. Since K is a representation of normal mode displacement along the reaction coordinate, Duschinsky calculations can provide insight on which normal modes are likely to be promoting modes and which are likely to be spectator modes (i.e.: not oscillating along the reaction coordinate of a transition) by determining which normal modes displace the most during an electronic state transition. In the effort of defining the reaction coordinate of excited state relaxation within Fe(II)- based polypyridines, ultrafast fast TA spectroscopy was used to collect vibrational coherence data on a series of tris-bidentate and bis-tridentate Fe(II)-based polypyridyl complexes. In addition, DFT calculations were conducted to estimate the geometry and normal mode frequencies of the 1MLCT, 3T1, 5T2, and 1A1 states for each complex studied. DFT calculated normal mode frequencies were matched to those observed in coherence data and the geometric data was used to conduct Duschinsky calculations for transitions of initial excitation and along the subsequent excited state relaxation pathway: 1A1 → 1MLCT → 3T1 → 5T2. Of course, it has been observed that Fe(II)-based complexes undergo an intersystem crossing from the 1MLCT state to the 3MLCT state before relaxing to the 3T1 state,1 however we were unable to optimize the triplet charge- transfer excited state with the resources available, therefore the singlet charge transfer state was used to approximate the MLCT deactivation transition. We believe this is a reasonable approximation as significant structural changes are not expected between these two electronic states.26 The combination of these techniques will illuminate reaction coordinate for MLCT deactivation, providing information to guide ligand modification designed to prolong the MLCT excited state. The methodology of using coherence data to inform ligand design has been shown to be effective in a previous study by McCusker and coworkers focusing on an Fe(II)-based complex using a polypyridine “cage” ligand.19 The cage ligand was designed to prevent the Fe-N bond elongation that accompanies the population of the 5T2 excited state. Analysis of vibrational coherence data collected on the cage complex revealed that the complex was engaging in a twisting motion enabled by an opening in the ends of the cage ligand, therefore the ligand was modified to include Cu1+ ions in each end cap of the cage, restricting these twisting motions. Ultrafast TA 241 spectroscopy revealed that the MLCT lifetime of the modified cage complex was 20 times longer than that of the parent cage complex.19 This provides a clear example of geometric modifications informed by normal mode analysis of vibrational coherence data influencing the excited state dynamics of the chromophore. While this study focused on a specific ligand structure, the work presented in this chapter will look at several ligand structures with various modifications in order to create an excited state relaxation road map that can be used to inform synthetic structural design aimed at lengthening MLCT state lifetime. The ligand structures studied in this work were chosen to examine how the difference in dentate models, substituents, and increasing ligand rigidity impacts promoting modes. Complexes using bpy and terpy were studied to represent tris-bidentate and bis-tridentate complexes, respectively. In addition, chloro (Cl) and methoxy (OMe) substituted complexes were studied to observe the impact of both mono and multi-atomic substituents, respectively. Each substituent is placed at the 4’-position on terpy and both the 4,4’ and 5,5’-positions on bpy, allowing for the impact of difference in substituent position to be observed. Lastly, two ligands which increase the rigidity of the ligand for each structure type were utilized. For the tris-bidentate model, a complex using 1,10-phenanthroline ([Fe(phen)3 2+) was studied, since the phen ligand provides additional stability along the backbone of the bidentate ligand. For the bis-tridentate model, a dipyridoacridine based complex ([Fe(DPA)2 2+) was studied. Similarly to phen, DPA provides stability along the backbone of the tridentate ligand.27 The molecules studied in this chapter are illustrated in Figure 6.5. Figure 6.5: The molecules studied via vibrational coherence measurements, DFT calculations, and Duschinsky analysis throughout this chapter. A: [Fe(bpy)3]2+. B: [Fe(phen)3]2+. C: [Fe(terpy)2]2+. D: [Fe(DPA)2]2+. In addition to these, complex A is also studied with diCl and diOMe substituents placed at the 4,4’ and 5,5’-positions, and complex B is also studied with Cl and OMe substituents at the 4’-position. 242 6.2 Experimental Methods 6.2.1 Synthesis of Fe(II)-Based Polypyridines The synthetic procedures for [Fe(bpy)3](PF6)2, [Fe(4,4’-diOMe)3](PF6)2, [Fe(4,4’- diCl)3](BF4)2, [Fe(5,5’-diOMe)3](PF6)2, [Fe(5,5’-diCl)3](BF4)2, [Fe(terpy)2](PF6), [Fe(4’-OMe- terpy)2](PF6)2, and [Fe(4’-Cl-terpy)2](PF6)2 are provided in the previous chapters of this dissertation. The synthetics procedures for terpy-based complexes are found in Chapter 4.2.1 and the synthetic procedures for bpy-based complexes are found in Chapter 5.2.1. The 4’-substituted terpy-based complexes were provided by collaborators from Dr. Gyorgi Vanko’s research group, the remainder of the complexes were synthesized by fellow McCusker group members: Dr. Chris Tichnell, Dr. Sara Adelman, Yi-Jyun Lien, and Bekah Bowers. xiv. tris(1,10’-phenanthroline) iron(II) hexafluorophosphate, [Fe(phen)3](PF6)2. [Fe(phen)3](PF6)2 was synthesized by Bowers, B. following literature procedure using 1,10-phenantrholine ligand purchased from Sigma-Aldrich.28 xv. bis(dipyridoacridine) iron(II) tetrafluoroborate, [Fe(DPA)2](BF2)2. [Fe(DPA)2](PF6)2 was synthesized by Lien, Y.-J. The DPA ligand was synthesized following literature procedure.27 Ligand DPA (180 mg, 0.64 mmol) was dissolved in mixed solution of DCM and MeOH (1:1 12 mL). Fe(BF4)2•6H2O (105 mg, 0.31 mmol) was added to the solution. The resulting solution was stirred for 12 hours, and diethyl ether was added to facilitate precipitation of the complex. The crude precipitate was collected via filtration and washed with small portions of diethyl ether. Purification was performed by recrystallization from slowly diethyl ether diffusion into MeCN solution, affording crystallized complex [Fe(DPA)2](BF4)2 (0.21 g, 84%).29 6.2.2 Spectroscopic Methods Vibrational coherence data for all complexes was collected using the ~40 fs pulse ultrafast laser system called RR using the procedures outlined in Chapter 2. Optical Kerr Effect (OKE) measurements were collected before each use to ensure the pump pulse had an instrument response between 40 and 60 fs. Solvents blanks using a cuvette filled with neat spectroscopic grade acetonitrile were collected to check for coherent frequencies stemming from the solvent and cuvette. The pump pulse wavelength for each complex was selected based on UV-Vis data measured following the procedure provided in Chapter 2, additionally UV-Vis measurements were 243 collected before and after vibrational coherence measurements to check for photodecomposition. Probe pulse wavelengths were selected based on the absorbance features of the radical ligands of each complex; 600 nm probe pulses were used for complexes using bpy, phen, and terpy, chosen based on previously reported spectroelectrochemical data.30,31 A probe wavelength of 700 nm was chosen for [Fe(DPA)2]2+ based on spectroelectrochemical data collected by Lien, Y.-J. for a paper in preparation.29 DFT calculations for all complexes were performed using the procedures outlined in Chapter 2. In addition to a lowest energy singlet and quintet, calculations were performed for a lowest energy triplet state using the same procedure for each complex to simulate the 3T1 excited state. Calculations for a singlet charge transfer state were done by performing TD-DFT calculations solving for a single “root” state, rather than for multiple excited states. The results of ground state TD-DFT calculations were used to ensure that the state used for charge transfer calculations were a charge transfer transition by looking at the orbital densities of the molecular orbitals (MO) involved in the state, if the excited state was a charge transfer transition the initial MO would have orbital density centered on the metal and the final MO would have orbital density centered on the ligands. These calculations represent the 1MLCT excited state. 6.2.3 Duschinsky Vector Calculations Duschinsky displacement vectors were calculated using a MATLAB script developed by former group member Dr. Bryan Paulus.17 This program calculates K from equation 6.2 using the geometric data of each state determined via DFT calculation as well as the Hessian data for the initial state extracted from the formatted checkpoint file for initial state DFT frequency calculations.24 Additionally, it is important to reorient the geometries of the initial and final states so that there is no linear or angular momentum between each geometry before K can be calculated. The MATLAB script accomplishes this by following a procedure developed by Chang and coworkers.32 In addition to calculating K, the script also provides the inner-sphere reorganization energy associated with each normal mode by calculating the reduced mass and force constants for each normal mode. This script is available in the Supplemental Information (SI) of this chapter. 6.2.4 Analysis of Normal Modes Candidates for promoting modes were identified with the use of vibrational coherence data, DFT results, and Duschinsky displacement vectors. First, the frequencies of vibrational coherences observed in ultrafast TA data would need to be determined. This is accomplished by first fitting 244 the TA data to a multiexponential decay function and isolating the residuals of the fit using IgorPro, a multiexponential decay is used to ensure the decay feature fit is as accurate as possible so the residuals only reflect the vibrational coherence oscillations. The solvent blank was used as a reference to determine where the fit should begin, this way the time-zero solvent feature does not interfere with the vibrational coherence residuals. The IgorPro Fourier Transforms tool with output type set to “exponential squared” was used to Fast Fourier transform the residuals of the multiexponential fit. The x-axis of the resulting FFT spectrum was then converted from THz to cm-1. This process is illustrated by plots B and C shown in Figure 6.3. The frequencies of prominent peaks were recorded. This process was done across several trials (a minimum of three, though more may have been recorded depending on the s/n quality of the TA data) for each complex. Frequencies of similar value (within 10 cm-1) that appeared in multiple trials were averaged together; these averaged frequency values will be referred to as the “observed” frequency values. Lastly, these observed values were compared to those seen in solvent blank coherence data to ensure that they did not stem from coherences in the solvent or cuvette. This process is shown in Table 6.1 using vibrational coherence data for [Fe(4,4’-diCl-bpy)3](BF4)2 in acetonitrile. Next, Duschinsky analysis results were examined. For each electronic state transition, the normal modes were ordered by the strength of the absolute value of K, normal modes with low |K| values (less than 0.1) were disregarded. The initial state DFT determined frequencies of normal modes with sufficient |K| values were compared to the observed frequency values. Any normal modes with frequency values that matched observed values (within ~5-10 cm-1) were highlighted as potential promoting modes, which will be referred to as candidate modes. DFT frequency calculations were used to examine the vibrational motions of each candidate mode in the initial state, then the final state vibrations were examined to look for matching normal modes. This is important to do because normal modes often have different frequency values between states, so a candidate mode with a frequency that matches an observed frequency value in one state, may have a frequency with no observed match in another state. Candidate modes without matching frequency values in both states of a transition are ruled out as spectator modes; only normal modes with DFT frequency values in both states of a transition that match observed frequencies are considered as promoting modes. This process is shown in Table 6.2 using Duschinsky analysis data for the 1MLCT → 3T1 transition in [Fe(4,4’-diCl-bpy)3](BF4)2. 245 Table 6.1: The frequencies of vibrational coherence oscillations across four trials of ultrafast TA data for [Fe(4,4’-diCl-bpy)3](BF4)2 in acetonitrile collected with a pump wavelength of 550 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the four trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(4,4’-diCl-bpy)3](BF4)2 Coherent Oscillation Frequency (cm-1) Trial 1 Trial 2 Trial 3 15.162 31.174 45.076 Trial 4 8.0962 32.385 48.577 72.866 Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequency Values (cm-1) 13.671 ––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––→ 31.7795 46.8265 68.353 93.523 90.972 90.152 –––––––––––––––––––––––––––––––––––––––––––––––––→ 91.549 113.35 109.37 124.7 171.46 135.23 165.28 137.64 –––––→ 136.71 161.92 ––––––––––––––––––––––––––––––––––––→ 136.435 163.6 181.94 178.12 ––––––––––––––––––––––––––––––––––––→ 180.03 195.33 202.41 205.06 218.22 212.27 218.6 ––––––––––––––––––––––––––––––––––––→ 216.3633 240.41 232.4 264.98 272.92 303.24 270.46 –––––––––––––––––––→ 273.41 299.56 ––––––––––––––––––––––––––––––––––––→ 271.69 301.4 315.53 314.43 327.33 333.56 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 330.445 340.04 364.33 388.62 –––––→ –––––→ 360.61 390.66 465.79 358.5 363.89 405.26 576.72 454.86 500.35 361.8325 389.64 369.11 396.45 437.46 478.47 505.81 246 Table 6.2: Summary of the results of Duschinsky analysis on the 1MLCT → 3T1 relaxation of [Fe(4,4’-diCl-bpy)3](BF4)2. The normal modes are ordered by the strength of |K|, with no normal modes with |K| below 0.1 shown. The DFT determined frequencies for each normal mode at each state involved in the transition are provided on the right side of the table. The frequency of each normal mode was compared to the observed frequency values found in Table 6.1: values that match observed frequency values are highlighted in green, values that match frequency values observed in the solvent blank are highlighted in red, and those with no match are highlighted in yellow. Only normal modes with frequency values that match observed frequency values at each state are considered promoting modes. 1CT → 3MC Duschinsky Analysis Data 1MLCT Normal Modes Mode # |K| Frequency (cm-1) Mode # DFT Frequency (cm-1) 12 85.13 ––– –→ Mode # 3T1 Normal Modes DFT Frequency (cm-1) 77.33 11 12 4 21 3 2 20 14 9 1 15 5 8 6 18 51 17 7 24 13 53 34 35 41 16 39 26 57 19 10 42 31 45 29 101 105 2.92816 2.440268 2.151617 2.078108 2.036818 1.760414 0.965841 0.965809 0.791177 0.719581 0.429086 0.42218 0.40647 0.383113 0.366332 0.33885 0.291482 0.276664 0.245348 0.198468 0.188195 0.184467 0.184179 0.175854 0.171582 0.162372 0.16117 0.151404 0.150668 0.149362 0.129955 0.128124 0.124426 0.108982 0.103372 84.75846 21.28301 148.4854 20.95434 19.14324 139.1833 95.01957 78.19101 17.50892 96.51265 25.70481 75.35608 31.83571 119.5673 441.4502 116.7274 73.66465 191.8412 91.64303 447.3027 296.8722 298.888 363.7416 104.9304 331.8482 217.5565 510.6304 138.9031 79.11323 365.0638 266.6318 405.5878 237.8201 964.7284 980.3945 139.57 95.41 ––– –→ ––– –→ 21 15 137.37 92.98 96.91 25.77 ––– –→ ––– –→ 17 6 98.21 26.30 31.91 ––– –→ 7 30.38 91.76 ––– –→ 14 87.09 296.98 299.29 364.11 332.16 218.12 ––– –→ ––– –→ ––– –→ ––– –→ ––– –→ 37 36 42 40 28 296.38 295.18 355.59 343.17 217.91 139.27 ––– –→ 20 123.21 365.53 266.82 ––– –→ ––– –→ 41 33 348.91 258.18 20 14 15 5 6 13 34 35 41 39 26 19 42 31 247 This procedure was carried out for each complex and the transitions listed in the introduction of this chapter: 1A1 → 1MLCT → 3T1 → 5T2. Section 6.3 provides a list of likely promoting modes for each transition in each complex. As is seen in Table 6.2, a transition may have multiple candidate promoting modes, and it is possible that some of these modes are still spectator modes rather than promoting. A factor to consider when examining these modes is the frequency rate each mode was observed across the multiple trials. For example, Table 6.1 shows that some frequency values were observed in 75% of trials while others appeared in 50% of trials. To assist in finding the most relevant modes, the rate of appearance for each mode is also listed. 6.2.5 Normal Mode Description Nomenclature It is important to take the time to establish what each description of normal modes used throughout this chapter means. First to address are the vibrations that move the ligand with respect to the metal center (MC) while keeping the shape of the ligand intact. These motions are labeled Figure 6.6: Schematics illustrating the “Ligand” class of vibrational motions using bpy as an example, though these motions may be found in all ligands studied in this chapter. Ligand breathing has the ligand moving away and towards the MC. Ligand rocking and flapping both involved the ligand waving back and forth: in rocking the ligand stays in the plane of the page, while in flapping the ligand moves above and below the plane of the page. Ligand wagging involves one end of the ligand moving above the plane of the page while the other end goes below, and then reverses. Ligand shifting has the ligand maintain its orientation as it shifts back and forth within the plane of the page. 248 with the “Ligand” precursor. There are five “Ligand” vibrational motions that were observed: breathing, rocking, flapping, wagging, and shifting modes. Ligand breathing motions involve the ligand moving away and towards the MC. Ligand rocking motions involve the ligand rotating back and forth lengthwise around the MC, much like a waving hand. Ligand flapping motions involve the ligand face moving back and forth, much like the wings of a bird flapping. Ligand wagging motions involve the ligand twisting about the MC with one pyridine ring moving in one direction as the other moves in the opposite direction back and forth. Ligand shifting motions are rarely observed; these are like ligand rocking modes however instead of rocking around the MC, the ligand maintains its orientation, only shifting in position. These vibrational motions are illustrated using bpy by the schematics provided in Figure 6.6. Some normal modes involve separate ligands engaging in different vibrations: for example, one ligand may be rocking, while the other two exhibit flapping motions. In these cases, all present motions will be listed with the more prominent motion listed first. Next to discuss are the vibrational motions where the shape of the ligand distorts. These motions are labeled with the “Intraligand” precursor, and there are six “Intraligand” vibrational motions observed: breathing, twisting, rocking, flapping, stretching, and warping. Intraligand breathing involves breathing motions where the C-C bond between rings bends. This motion is always symmetrical for bidentate ligands, however there are two types of intraligand breathing for tridentate ligands: symmetric breathing (where both ligands move in unison) and asymmetric breathing (where two ligands breath in opposite directions, with one moving towards the MC as the other moves away). Intraligand twisting involves the rings of the ligand rotating in opposite patterns along the axis of the C-C bond between each ring. Intraligand rocking involves complete distortion of the ligand plane, where alternating atoms vibrate in opposite directions (e.g. odd numbered positions move up while even numbered positions move down). Intraligand flapping involves the ends of the ligand flapping like wings as the C-C bond between the rings moves in the opposite direction. Intraligand stretching and warping both involve stretching of the rings in the ligand, intraligand stretching sees this motion parallel to the C-C bond between rings while intraligand warping sees this motion perpendicular to the C-C bond between rings. These vibrational motions are illustrated using bpy by the schematics provided in Figure 6.7. Something important to note about intraligand vibrational motions in tridentate ligands is that the same motion may not be observed on both sides of the ligand. For example, equatorial ring 249 1 and the axial ring may be engaging in intraligand flapping motions while equatorial ring 2 engages in a twisting motion. In situations where this occurs the normal mode will be labeled with a forward slash symbol between the observed motions with the motion only one ring engages in listed second. For the example given above, the notation would be the following: Intraligand Flapping/Twisting. Figure 6.7: Schematics illustrating the “Intraligand” class of vibrational motions using bpy as an example, though these motions may be found in all ligands studied in this chapter. Intraligand breathing sees the ligand bending to accommodate Fe-N breathing vibrations. Intraligand twisting sees the rings of the ligand twisting in opposite rotational directions from one another. Intraligand rocking has neighboring atoms within the ligand vibrations in opposite directions. Intraligand flapping sees the ends of the ligand vibrating in the opposite direction of the center, giving the appearing of the rings acting as flapping wings. Intraligand stretching and warping both involve ring stretching, though in perpendicular directions. The last two vibrational motions that were observed are unique to substituted ligands: substituent bending and substituent rotating. Substituent bending involves the substituent wagging either out of or in the plane of the ligand. Substituent rotating was only observed in OMe- substituted ligands and involved the methyl group in the substituent twisting. 250 6.3 Results 6.3.1 [Fe(bpy)3](PF6)2 For each of the unsubstituted complexes, trials were performed using different pump pulse wavelengths to check if excitation energy had an impact on the vibrational coherence formed. Pump wavelengths exciting into the blue and red sides of the MLCT absorptions were chosen for each complex. In the case of [Fe(bpy)3](PF6)2, the chosen pump wavelengths were 490 nm and 550 nm. To simulate this, 1CT DFT calculations were performed using states on both the red and blue sides of the TD-DFT simulated absorption spectrum. If there is an excitation wavelength dependency in normal modes observed for a complex, it will be noted in the table listing the promoting modes. For the case of [Fe(bpy)3](PF6)2, the excitation wavelength had no impact on observed frequency values but did affect the |K| values of normal modes associated with those frequencies. Enough so that an observed normal mode may be a spectator mode with one excitation energy, but a promoting mode with another. Figure 6.8: Vibrational coherence ultrafast TA data collected for [Fe(bpy)3]2+ with a pump wavelength of 490 nm (A) and 550 nm (C). A multiexponential fit (in black) was applied to each spectrum with the residuals isolated (in blue). The FFT spectra for each respective coherent oscillation residuals are provided on the right (B and D). 251 A total of eight trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.8 provides ultrafast TA data and the FFT spectrum of the residual oscillations for trials collected with a 490 nm pump (A and B) and 550 nm pump (C and D). A 600 nm probe wavelength was used for this complex, as well as all others except for [Fe(DPA)2]2+. The observed peak frequencies across all trials are provided in Table 6.S1. The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized Table 6.3: Promoting modes identified for the excited state evolution of [Fe(bpy)3]2+. Normal Mode Frequency represents the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. †: Normal mode is only associated with higher energy excitation (490 nm pump). ‡: Normal mode is only associated with lower energy excitation (550 nm pump). Electronic State Transition Normal Mode Frequency (cm-1) Normal Mode |K| Normal Mode Description Observed Frequency (cm-1) Rate of Observation 1A1 → 1MLCT Excitation 1MLCT → 3T1 Relaxation 38→40‡/39† 42 → 43‡/41† 268 → 264‡ 128 → 117‡/122† 158‡/161† → 153 40‡/38† → 42 192‡/194† → 189 264‡/261† → 270 0.89‡ 0.44† 0.51‡ 0.31† 0.13‡ 0.33‡ 0.13† 1.51‡ 1.16† 1.17‡ 0.63† 0.61‡ 0.39† 0.27‡ 0.10† Ligand Flapping and Wagging Ligand Rocking and Wagging 42.4 42.4 50% 50% Intraligand Rocking 264.9 62.5% Intraligand Twisting 126.7 100% Ligand Intraligand Flapping/Breathing Ligand Rocking and Wagging Intraligand Flapping and Breathing 156.2 62.5% 42.4 50% 193.7 87.5% Intraligand Rocking 264.9 62.5% 153 → 148 2.27 Intraligand Flapping 156.2 62.5% 3T1 → 5T2 Relaxation 118 → 117 1.84 Intraligand Breathing 126.7 100% 189 → 193 0.14 Intraligand Flapping and Breathing 193.7 87.5% 252 for all electronic state transitions in Table 6.3. For all data tables presented in this section, normal modes are organized by electronic state transitions and ordered within categories by the strength of |K|. Table 6.3 shows that the 1A1 → 1MLCT excitation is tied to near 42 cm-1 normal modes including ligand flapping, rocking, and wagging motions; however, this frequency was only observed in 50% of trials. An intraligand twisting normal mode with frequency near 126 cm-1 was observed in all trials. While this normal mode had a lower |K| value, the 100% observance rate of this vibrational mode cannot be ignored. This high observance rate points to the intraligand twisting normal mode being the one associated with the formation of the 1MLCT state. Both excited state relaxations after excitation, 1MLCT → 3T1 and 3T1 → 5T2, exhibit a near 192 cm-1 intraligand flapping and breathing normal mode. This is similar in frequency to the 193.7 cm-1 oscillation that was observed in all but one trial (87.5%). The only trial that did not exhibit this frequency in the FFT data is shown in Figure 6.8D. This vibration involves simultaneous flapping and subtle breathing motions within each ligand. A near 156 cm-1 intraligand flapping normal mode was also observed. This normal mode has a higher |K| but was observed at a lower rate. Lastly, a near 126 cm-1 intraligand breathing mode was observed for the 3T1 → 5T2 relaxation. 6.3.2 [Fe(phen)3](PF6)2 A total of eight trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.9 provides ultrafast TA data and the FFT spectrum of the residual oscillations for trials collected with a 480 nm pump (A and B) and 510 nm pump (C and D). The observed peak frequencies across all trials are provided in Table 6.S2. The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.4. The initial 1A1 → 1MLCT excitation presents an interesting problem for this complex. While the near 170 cm-1 intraligand twisting normal mode fits the criteria for promoting mode when excited with the lower energy pump (510 nm), this same mode had a |K| less than 0.1 for the higher energy pump (480 nm), with one normal mode having a |K| if 0.004 and another having 0.014. The issue is that no normal mode with sufficient |K| value calculated for high energy excitation had an oscillation frequency that matched observed frequency values. The highest |K| values normal mode within 480 nm excitation calculations with a frequency that matches observed 253 Figure 6.9: Vibrational coherence ultrafast TA data collected for [Fe(phen)3]2+ with a pump wavelength of 480 nm (A) and 510 nm (C). A multiexponential fit (in black) was applied to each spectrum with the residuals isolated (in blue). The FFT spectra for each respective coherent oscillation residuals are provided on the right (B and D). frequency values is the near 297 cm-1 normal mode, which has a |K| = 0.062. It seems reasonable to rely on the 510 nm excitation calculations to identify the most likely promoting mode as it is the only set of data to produce a normal mode that matches the criteria. Thus, the most likely promoting mode is the near 170 cm-1 intraligand twisting normal mode. A different near 170 cm-1 normal mode appears as the highest |K| value normal mode present for the 1MLCT → 3T1 relaxation, this normal mode incorporates intraligand flapping vibrations. This mode does not appear for the 3T1 → 5T2 relaxation, instead a higher frequency (near 206 cm-1) intraligand flapping and breathing normal mode appears as the most prominent promoting mode. This normal mode still engages in intraligand flapping motions while introducing subtle breathing motions. This points to the importance of intraligand flapping motions throughout the excited state relaxation of [Fe(phen)3]2+. Lastly, the excitation wavelength does have a subtle change in the secondary promoting modes observed for the 1MLCT → 3T1 relaxation, where higher energy excitation produces an additional intraligand breathing mode and lower energy excitation 254 produces an intraligand rocking mode which does not otherwise appear in this transition but does carry through to the 3T1 → 5T2 relaxation. Furthermore, while a 246 cm-1 oscillation frequency was only observed in 510 nm pump trials, the 236 cm-1 frequency was observed at both wavelengths. This indicates that this vibration only occurs in vibrationally cool excited states which the higher energy pump pulse prevents at earlier points in excited state evolution. Table 6.4: Promoting modes identified for the excited state evolution of [Fe(phen)3]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. †: Normal mode is only associated with higher energy excitation (480 nm pump). ‡: Normal mode is only associated with lower energy excitation (510 nm pump). Electronic State Transition Normal Mode Frequency (cm-1) Normal Mode |K| Normal Mode Description Observed Frequency (cm-1) Rate of Observation 160 → 157 1A1 → 1MLCT Excitation 170 → 172‡/170† 0.20‡ 0.004† 0.19‡ 0.014† Intraligand Flapping and Twisting 172.6 Intraligand Twisting 172.6 297 → 296† 0.062† Intraligand Breathing 297.0 1MLCT → 3T1 Relaxation 167†/166‡ → 164 245 → 236‡ 296†/299‡ → 301 280 → 290† 1.44† 0.69‡ 0.39‡ 0.33† 0.17‡ 0.12† Intraligand Flapping 172.6 Intraligand Rocking 246.0‡/236.0 25%/50% Intraligand Breathing 297.0 Intraligand Breathing 284.2† 115 → 107 1.94 Ligand Breathing 109.1 3T1 → 5T2 Relaxation 236 → 237 0.23 Intraligand Rocking 236.0 204 → 209 0.13 Intraligand Flapping and Breathing 205.6 6.3.3 [Fe(4,4’-diCl-bpy)3](BF4)2 For the substituted complexes, only one excitation wavelength was used for each complex as the goal for studying these complexes was to examine differences in vibrations that arose from 255 75% 75% 50% 75% 50% 25% 25% 50% 75% the addition of a substituent. In the case of [Fe(4,4’-diCl-bpy)3](BF4)2 a pump wavelength of 550 nm was used and a total of four trials were run. Figure 6.10 provides the ultrafast TA data and the FFT spectrum of the residual oscillations for one of these trials. The observed peak frequencies across all trials are provided in Table 6.1. Figure 6.10: Vibrational coherence ultrafast TA data collected for [Fe(4,4’-diCl-bpy)3]2+ with a pump wavelength of 550 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.5. The most prominent normal modes for the 1A1 → 1MLCT excitation are near 91 cm-1 normal modes which involve intraligand twisting. The higher |K| value normal mode also involves intraligand breathing. The most prominent promoting mode for both the 1MLCT → 3T1 and 3T1 → 5T2 relaxations is a different near 91 cm-1 normal mode, which primarily involves intraligand flapping motions. Intraligand flapping motions appear in several of the promoting modes listed in Table 6.5, highlighting the importance of this vibrational motion. Low frequency ligand rocking and flapping normal modes also appear across each transition, however the frequency associated with this mode was only observed in 50% of trials while frequencies associated with intraligand flapping normal modes were observed with higher frequency (75%). A near 180 cm-1 intraligand twisting mode is also present for the 3T1 → 5T2 relaxation, the lower |K| value and 50% observation rate point to this motion likely being a spectator mode. Higher frequency normal modes, one near 216 cm-1 and one near 301 cm-1, were also observed across several trials. These higher frequency normal modes involve the ligands 256 engaging in multiple motions, with the near 216 cm-1 normal mode involving both ligand and intraligand motions. The near 301 cm-1 normal mode was only observed in 50% of trials, and both Table 6.5: Promoting modes identified for the excited state evolution of [Fe(4,4’-diCl-bpy)3]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Transition Normal Mode Frequency (cm-1) Observed Frequency (cm-1) Rate of Observation Normal Mode Description Normal Mode |K| 1A1 → 1MLCT Excitation 84 → 85 98 → 97 0.31 0.28 Intraligand Breathing and Twisting Intraligand Flapping and Twisting 23 → 26 0.12 Ligand Rocking and Wagging 303 → 299 0.10 Intraligand Flapping and Breathing 95 → 87 0.97 Intraligand Flapping 26 → 26 0.43 Ligand Rocking and Wagging 1MLCT → 3T1 Relaxation 32 → 30 0.41 Ligand Flapping and Rocking 92 → 93 0.25 Intraligand Flapping and Twisting 91.5 91.5 31.8 301.4 91.5 31.8 31.8 91.5 297 → 296 0.19 Intraligand Rocking 301.4 75% 75% 50% 50% 75% 50% 50% 75% 50% 218 → 218 0.16 Ligand Rocking, Intraligand Flapping, and Substituent Bending 216.4 75% 26 → 31 1.76 Ligand Rocking and Flapping 31.8 93 → 90 84 → 83 1.10 0.93 Intraligand Flapping Intraligand Flapping 91.5 91.5 50% 75% 75% 3T1 → 5T2 Relaxation 218 → 217 0.30 Ligand Rocking, Intraligand Flapping, and Substituent Bending 87 → 86 0.21 Intraligand Flapping 183 → 176 0.18 Intraligand Twisting 180 → 175 0.12 Intraligand Twisting 216.4 75% 91.5 180.0 180.0 75% 50% 50% 257 the higher frequency normal modes have lower |K| values than near 91 cm-1 normal modes, so these are considered less likely to be important promoting modes. 6.3.4 [Fe(4,4’-diOMe-bpy)3](PF6)2 A total of four trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.11 provides ultrafast TA data and the FFT spectrum of the residual oscillations for one of the trials collected with a 550 nm pump. The observed peak frequencies across all trials are provided in Table 6.S3. Figure 6.11: Vibrational coherence ultrafast TA data collected for [Fe(4,4’-diOMe-bpy)3]2+ with a pump wavelength of 550 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.6. All normal modes identified for the 1A1 → 1MLCT excitation have frequency near 94 cm-1 and most involve intraligand flapping motions. Two of these normal modes see one of the ligands engaging in an intraligand breathing motion, while the lowest |K| value normal mode sees two of the ligands engaging in lintraligand twisting motions. One of the near 94 cm-1 normal modes only includes intraligand twisting motions, but overall intraligand flapping motions appear to be the most prominent. The most observed promoting mode for the 1MLCT → 3T1 transition is a near 337 cm-1 intraligand breathing and stretching normal mode. This normal mode includes intraligand breathing motions with an asymmetric stretching motion where one ring shifts away from the other while the second ring does not move outside of a breathing motion. This frequency was observed in 100% of trials, but the 1MLCT → 3T1 relaxation was the only electronic state transition where 258 a normal mode with frequency near 337 cm-1 had a |K| value greater than 0.1. For the 3T1 → 5T2 relaxation, the most prominent promoting mode is a near 149 cm-1 intraligand flapping normal mode. Intraligand flapping motions appear in several normal modes at this transition. Table 6.6: Promoting modes identified for excited state evolution in [Fe(4,4’-diOMe-bpy)3]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Transition Normal Mode Frequency (cm-1) Normal Mode |K| Normal Mode Description Observed Frequency (cm-1) Rate of Observation 1A1 → 1MLCT Excitation 93 → 91 0.20 Intraligand Flapping and Breathing 94.5 100 → 101 0.14 Intraligand Flapping 94.5 88 → 89 0.13 Intraligand Twisting 94.5 91 → 94 98 → 97 0.11 0.11 94 → 100 1.03 Intraligand Flapping and Breathing Intraligand Twisting and Flapping Intraligand Flapping and Breathing 94.5 94.5 94.5 463 → 459 0.31 Intraligand Stretching 458.9 50% 50% 50% 50% 50% 50% 75% 1MLCT → 3T1 Relaxation 341 → 335 0.19 141 → 142 0.11 Intraligand Breathing and Stretching Ligand Shifting and Substituent Bending 337.0 100% 149.6 91 → 97 0.11 Intraligand Flapping 94.5 148 → 142 1.02 Intraligand Flapping 149.6 93 → 94 0.69 Intraligand Flapping and Breathing 94.5 147 → 144 0.33 Intraligand Flapping 149.6 459 → 455 0.31 Intraligand Stretching 458.9 97 → 91 0.27 Intraligand Flapping 94.5 3T1 → 5T2 Relaxation 259 75% 50% 75% 50% 75% 75% 50% 6.3.5 [Fe(5,5’-diCl-bpy)3](BF4)2 A total of three trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.12 provides ultrafast TA data and the FFT spectrum of the residual oscillations for one of the trials collected with a 510 nm pump. The observed peak frequencies across all trials are provided in Table 6.S4. Figure 6.12: Vibrational coherence ultrafast TA data collected for [Fe(5,5’-diCl-bpy)3]2+ with a pump wavelength of 510 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.7. The near 80 cm-1 intraligand twisting normal mode is prominent across all state transitions and was observed in 100% of trials conducted. This normal mode is also the only one to fit the criteria for promoting mode for the 1A1 → 1MLCT excitations and 1MLCT → 3T1 relaxation. There are several normal modes near this frequency, however they all exhibit intraligand twisting motions. The 3T1 → 5T2 relaxation has two other modes, one near 160 cm-1 and the other near 255 cm-1. The near 160 cm-1 normal mode still has intraligand twisting motions but introduces intraligand breathing motions. The near 255 cm-1 normal mode abandons the intrligand twisting motion and instead exhibits ligand rocking and substituent bending. These two normal modes are only observed in 66% of the trials, less than the observation rate of the near 80 cm-1 normal modes. This strongly suggests that the intraligand twisting vibration is of key importance for excited state evolution in [Fe(5,5’-diCl-bpy)3]2+. 260 Table 6.7: Promoting modes identified for the excited state evolution of [Fe(5,5’-diCl-bpy)3]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Normal Mode Normal Mode Observed Normal Rate of Transition Frequency (cm-1) Mode |K| Description Frequency (cm-1) Observation 1A1 → 1MLCT Excitation 1MLCT → 3T1 Relaxation 88 → 82 1.32 Intraligand Twisting 78 → 79 0.38 Intraligand Twisting 80.3 80.3 100% 100% 86 → 79 0.61 Intraligand Twisting 80.3 100% 88 → 82 0.41 Intraligand Twisting 82 → 76 0.33 Intraligand Twisting 80.3 80.3 100% 100% 82 → 75 1.97 Intraligand Twisting 80.3 100% 71 → 73 1.95 Intraligand Twisting 80.3 100% 3T1 → 5T2 Relaxation 79 → 74 1.50 Intraligand Twisting 80.3 100% 157 → 165 1.27 259 → 257 0.60 Intraligand Breathing and Twisting Ligand Rocking and Substituent Bending 159.6 66% 255.4 66% 6.3.6 [Fe(5,5’-diOMe-bpy)3](PF6)2 A total of three trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.13 provides ultrafast TA data and the FFT spectrum of the residual oscillations for one of the trials collected with a 490 nm pump. The observed peak frequencies across all trials are provided in Table 6.S5. The results of ultrafast coherent measurements were used with the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.8. 261 Figure 6.13: Vibrational coherence ultrafast TA data collected for [Fe(5,5’-diOMe-bpy)3]2+ with a pump wavelength of 490 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). Low frequency (near 15 cm-1) ligand wagging normal modes are seen across all electronic state transitions. In addition to the low frequency normal modes, a near 103 cm-1 intraligand breathing normal mode also appears across each transition. This normal mode has a high |K| value for the 1A1 → 1MLCT and 3T1 → 5T2 transitions, but it has lower |K| value at 0.1 in the 1MLCT → 3T1 transition, indicating this mode may not be important to MLCT deactivation. Furthermore, a near 221 cm-1 ligand rocking and substituent bending normal mode is prominent across the 1A1 → 1MLCT and 1MLCT → 3T1 transitions. This normal mode has a higher |K| value for the 1MLCT → 3T1 relaxation than the near 103 cm-1. This coupled with the importance of the low frequency normal modes suggests that ligand motions are more important to MLCT deactivation than intraligand motions for [Fe(5,5’-diOMe)3]2+. The same normal mode has an |K| value of 0.07 for the 3T1 → 5T2 transition, which is outside the range of consideration for 3T1 → 5T2 promoting mode. A different near 221 cm-1 normal mode which involves subtle intraligand twisting and substituent rotating motions is observed for both the 1MLCT → 3T1 and 3T1 → 5T2 transitions. There is a key distinction between the near 221 cm-1 normal modes in that one involves ligand distortions while the other includes intraligand distortions, however both do include substituent motions. However, in each transition this normal mode appears, it has a relatively low |K| value compared to other normal modes. Finally, a high frequency (near 797 cm-1) intraligand warping normal mode appears at 1A1 → 1MLCT excitation and 3T1 → 5T2 relaxation, but this normal mode has a low |K| value for the middle transition, which rules it out as a possible promoting mode. The low |K| for this normal mode at all transitions indicates it is likely a spectator mode. 262 Table 6.8: Promoting modes identified for excited state evolution in [Fe(5,5’-diOMe-bpy)3]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Transition Normal Mode Frequency (cm-1) Normal Mode |K| Normal Mode Description Observed Frequency (cm-1) Rate of Observation 24 → 19 3.91 Ligand Wagging 14.8 100% 97 →107 1.56 Intraligand Flapping and Breathing 103.8 66% 18 → 17 0.44 Ligand Wagging 14.8 100% 1A1 → 1MLCT Excitation 221 → 217 0.39 Ligand Rocking and Substituent Bending 221.4 799 → 796 0.13 Intraligand Warping 797.0 19 → 18 3.41 Ligand Wagging 15 → 12 2.54 Ligand Shifting and Wagging 17 → 15 2.54 Ligand Wagging 217 → 216 0.37 218 → 220 0.25 Ligand Rocking and Substituent Bending Intraligand Twisting and Substituent Rotating 14.8 14.8 14.8 221.4 221.4 107 → 106 0.10 Intraligand Breathing 103.8 66% 66% 100% 100% 100% 66% 66% 66% 1MLCT → 3T1 Relaxation 15 → 16 2.59 Ligand Wagging 14.8 100% 95 → 108 2.12 12 → 22 1.61 Intraligand Breathing and Substituent Bending Ligand Shifting and Flapping 103.8 66% 14.8 100% 3T1 → 5T2 Relaxation 220 → 219 0.20 Intraligand Twisting and Substituent Rotating 221.4 797 → 795 0.11 Intraligand Warping 797.0 66% 66% 263 6.3.7 [Fe(terpy)2](PF6)2 Three pump wavelengths were used for ultrafast TA vibrational coherence measurements collected on this complex. In addition to a blue-side pump (490 nm) and red-side pump (550 nm), a pump wavelength that excited into the low intensity red shoulder of the [Fe(terpy)]2+ MLCT absorption feature was used (600 nm). Between the three wavelengths, a total of eleven trials were conducted. The ultrafast TA data and the residual oscillations FFT spectrum of a trial collected with a 550 nm pump were provided in Figure 6.3. Figure 6.14 provides these spectra for trials collected with a 490 nm and 600 nm pump. The observed peak frequencies across all trials are provided in Table 6.S6. However, DFT calculations were still only performed for two 1CT states, representing a red-shifted excitation and a blue-shifted excitation. Calculations for a third state representing excitation energy near the center of the MLCT absorption feature were attempted, but never resolved. The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.9. Figure 6.14: Vibrational coherence ultrafast TA data collected for [Fe(terpy)2]2+ with a pump wavelength of 550 nm (A) and 600 nm (C). A multiexponential fit (in black) was applied to each spectrum with the residuals isolated (in blue). The FFT spectra for each respective coherent oscillation residuals are provided on the right (B and D). 264 Table 6.9: Promoting modes identified for the excited state evolution of [Fe(terpy)2]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. †: Normal mode is only associated with higher energy excitation (490 nm pump). ‡: Normal mode is only associated with lower energy excitation (600 nm pump). Observed Normal Electronic State Frequency (cm-1) Mode |K| Transition Normal Mode Frequency (cm-1) Normal Mode Description Rate of Observation 1A1 → 1MLCT Excitation 1MLCT → 3T1 Relaxation 3T1 → 5T2 Relaxation 104 → 93†/103‡ 215 → 213‡ 103‡/93† → 93 37† → 43 47 → 36 93 → 91 3.72† 0.31‡ 0.10‡ 3.92‡ 0.32† 0.12† 3.71 3.18 294 → 294 0.39 427 → 411 0.17 Intraligand Symmetric Breathing Intraligand Asymmetric Breathing Intraligand Symmetric Breathing Ligand Rocking and Intraligand Flapping Ligand Rocking and Intraligand Flapping Intraligand Symmetric Breathing Intraligand Symmetric Breathing Intraligand Asymmetric Breathing 97.4 100% 214.7 45% 97.4 42.9‡ 42.9‡ 97.4 295.0 420.5 100% 45% 45% 100% 45% 27% A near 97 cm-1 intraligand symmetric breathing normal mode was observed across each transition of excited state evolution for [Fe(terpy)2]2+. This mode was also observed in 100% of the eleven trials collected across three different pump wavelengths, highlighting the importance of this vibration. Higher frequency intraligand breathing modes (near 215 cm-1 and near 295 cm-1) were observed for the 1A1 → 1MLCT excitation and 3T1 → 5T2 relaxation, however the frequencies associated with these normal modes were only observed in 45% of trials. Lastly, a low frequency (near 43 cm-1) ligand rocking and intraligand flapping normal mode was observed for the 1MLCT → 3T1 and 3T1 → 5T2 transitions, however this oscillation was only observed in 45% of trials. 6.3.8 [Fe(DPA)2](BF4)2 As with [Fe(terpy)2]2+, ultrafast TA vibrational coherence data was collected for [Fe(DPA)2]2+ using three pump wavelengths: 480 nm, 550 nm, and 600 nm. A probe wavelength 265 of 700 nm was used for this complex. Something important to keep in mind when interpreting the data for this complex is that [Fe(DPA)2](BF4)2 is a spin-crossover complex (SCO).29 This means that the energy gap between the t2g and eg* orbitals is low enough that the 5T2 state is accessible through thermal energy at room temperature, causing the population within a sample of [Fe(DPA)2]2+ to occupy a mix of both 1A1 and 5T2 states. This causes an excitation pulse to excite a portion of the population into a 5CT state rather than only the 1MLCT. However, 5CT DFT calculation attempts failed, so analysis in this section will focus on evolution after 1MLCT Figure 6.15: Vibrational coherence ultrafast TA data collected for [Fe(DPA)2]2+ with a pump wavelength of, 480 nm (A), 550 nm (C), and 600 nm (E). A multiexponential fit (in black) was applied to each spectrum with the residuals isolated (in blue). The FFT spectra for each respective coherent oscillation residuals are provided on the right (B, D, and F). 266 excitation. The result of this is that many oscillation frequencies that are ruled out as spectator modes may be promoting modes involving the 5MLCT excited state, but for now there is no means of determining this. Figure 6.15 provides ultrafast TA data and the residual oscillations FFT spectra collected at each wavelength. The observed peak frequencies across all trials are provided in Table 6.S7. The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.10. Table 6.10: Promoting modes identified for the excited state evolution of [Fe(DPA)2]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. †: Normal mode is only associated with higher energy excitation (480 nm pump). ‡: Normal mode is only associated with lower energy excitation (600 nm pump). Observed Electronic State Frequency (cm-1) Transition Normal Mode Frequency (cm-1) Normal Mode Description Rate of Observation Normal Mode |K| 94.7 100% 99 → 101†/98‡ 332 → 331‡ 0.64† 0.15‡ 0.15‡ 152 → 158† 0.14 Intraligand Symmetric Breathing Ligand Breathing and Intraligand Stretching Intraligand Asymmetric Breathing 1A1 → 1MLCT Excitation 335.0‡ 152.7 506 → 507 0.12 Intraligand Stretching 511.7 1MLCT → 3T1 Relaxation 101†/98‡ → 94 49‡/45† → 42 3.63† 3.16‡ 0.22‡ 0.16† Intraligand Symmetric Breathing Intraligand Flapping 94.7 40.3‡ 33% 56% 33% 100% 33% 3T1 → 5T2 Relaxation 221 → 147 0.32 154 → 149 0.20 Intraligand Wagging and Asymmetric Breathing Intraligand Wagging and Breathing 224.7†/152.7 33%/56% 152.7 56% Table 6.10 shows a near 95 cm-1 intraligand symmetric breathing mode is observed for both the 1A1 → 1MLCT excitation and 1MLCT → 3T1 relaxation. This normal mode has a frequency near 77 cm-1 in the 5T2 state, a frequency that was not observed in any trial, leaving the candidates for promoting mode for the 3T1 → 5T2 transition as a near 153 cm-1 intraligand wagging and 267 breathing normal mode. This normal mode involves one ligand undergoing intraligand breathing while the other ligand engages in a vibration not replicated in any other complex called intraligand wagging. This vibration is similar to intraligand flapping, however instead of the ends of the ligand moving back and forth, it is the top and bottom of the ligand, as depicted in Figure 6.16. It is worth noting that one of the 3T1 → 5T2 promoting modes has a frequency near 221 cm-1 in the 3T1 state but relaxes to near 153 cm-1 in the 5T2 state. An oscillation near 221 cm-1 was observed in 33% of trials, but only in trials with higher energy excitation wavelengths. For this reason, the lower |K| value near 153 cm-1 normal mode is more likely to be important for 3T1 relaxation. Figure 6.16: Schematic depicting the intraligand wagging vibration of the DPA ligand. In this vibration, the top and bottom of the ligand move above the plane of the ligand while the central spine moves below the plane of the ligand, and then reverses. This wobbling of the ligand does not appear as a prominent vibrational motion in any other ligand. A normal mode of the same frequency was observed for the 1A1 → 1MLCT excitation, however this normal mode engaged in entirely different vibrations, instead being an intraligand asymmetric breathing normal mode. A final observation worth noting is the presence of the near 42 cm-1 intraligand flapping normal mode for the 1MLCT → 3T1 relaxation. This normal mode was only observed in trials using lower energy pump wavelengths and was only observed in 33% of trials. Therefore, this normal mode is likely only a spectator mode. 6.3.9 [Fe(4’-Cl-terpy)2](PF6)2 A total of four trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.17 provides ultrafast TA data and the FFT spectrum of the residual oscillations for one of the trials collected with a 550 nm pump. The observed peak frequencies across all trials are provided Table 6.S8. The results of ultrafast coherent measurements were used alongside the DFT and Duschinsky calculations results following the process outlined in section 6.2.4 to identify 268 Figure 6.17: Vibrational coherence ultrafast TA data collected for [Fe(4’-Cl-terpy)2]2+ with a pump wavelength of 550 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). Table 6.11: Promoting modes identified for the excited state evolution of [Fe(4’-Cl-terpy)2]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Transition Observed Frequency (cm-1) Normal Mode Frequency (cm-1) Rate of Observation Normal Mode Description Normal Mode |K| 1A1 → 1MLCT Excitation 1MLCT → 3T1 Relaxation 3T1 → 5T2 Relaxation 24 → 26 99 → 101 86 → 86 34 → 32 101 → 91 291 → 279 49 → 43 277 → 279 46 → 35 28 → 24 280 → 277 0.91 0.41 0.21 0.21 4.03 0.56 0.31 0.22 1.14 0.30 0.30 Ligand Rocking and Flapping Intraligand Symmetric Breathing Intraligand Flapping Ligand Flapping Intraligand Symmetric Breathing Intraligand Asymmetric Breathing Intraligand Flapping Intraligand Rocking and Asymmetric Breathing Ligand Rocking and Intraligand Flapping Ligand Wagging Intraligand Asymmetric Breathing 26.8 94.2 94.2 40.4 94.2 282.5 40.4 282.5 40.4 26.8 50% 100% 100% 50% 100% 100% 50% 100% 50% 50% 282.5 100% 269 likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.11. A near 94 cm-1 intraligand symmetric breathing normal mode is observed as a prominent promoting mode for both the 1A1 → 1MLCT excitation and the 1MLCT → 3T1 relaxation. Similar to what was observed for [Fe(DPA)2]2+, this normal mode has a lower frequency (near 78 cm-1) in the 5T2 state, which was not observed in any TA trial. Therefore, the intraligand symmetric breathing normal mode is likely not a relaxation mode for the 3T1 → 5T2 transition. Instead, a near 280 cm-1 intraligand asymmetric breathing mode appears for both the 1MLCT → 3T1 and 3T1 → 5T2 relaxations. Other low frequency normal modes appear throughout several electronic state transitions, but they were only observed in 50% of trials, whereas the near 94 cm-1 and near 280 cm-1 normal modes were both observed in 100% of trials. This indicates that intraligand breathing modes are key to excited state evolution in [Fe(4’-Cl-terpy)2]2+, with this breathing motion shifting from symmetric to asymmetric at some point along the relaxation pathway. 6.3.10 [Fe(4’-OMe-terpy)2](PF6)2 A total of three trials of ultrafast TA vibrational coherence data sets were collected for this complex. Figure 6.18 provides ultrafast TA data and the FFT spectrum of the residual oscillations for one of the trials collected with a 550 nm pump. The observed peak frequencies across all trials are provided in Table 6.S9. Figure 6.18: Vibrational coherence ultrafast TA data collected for [Fe(4’-OMe-terpy)2]2+ with a pump wavelength of 550 nm (A). A multiexponential fit (in black) was applied to this spectrum with the residuals isolated (in blue). The FFT spectra for the coherent oscillation residuals is provided on the right (B). 270 The results of ultrafast coherent measurements were used alongside the results of DFT and Duschinsky calculations following the process outlined in section 6.2.4 to identify likely promoting modes. These modes are summarized for all electronic state transitions in Table 6.12. Table 6.12: Promoting modes identified for the excited state evolution of [Fe(4’-OMe-terpy)2]2+. Normal Mode Frequency represent the DFT frequency of the normal mode in the initial state followed by the final state of the transition, Normal Mode |K| represents the Duschinsky displacement vector of that normal mode, Normal Mode Description provides the motions involved in the normal mode, Observed Frequency provides the frequency that was observed in the TA data oscillations, and Rate of Observations provides the percentage of TA trials the Observed Frequency was seen in. Electronic State Transition Normal Mode Frequency (cm-1) Normal Mode |K| Normal Mode Description Observed Frequency (cm-1) Rate of Observation 1A1 → 1MLCT Excitation 103 → 100 0.44 88 → 86 0.18 100 → 92 3.79 1MLCT → 3T1 Relaxation 86 → 90 0.23 150 → 148 0.14 148 → 124 10.47 97 → 87 7.97 92 → 87 7.90 90 → 84 1.79 3T1 → 5T2 Relaxation Intraligand Symmetric Breathing Intraligand Flapping/Twisting Intraligand Symmetric Breathing Intraligand Flapping/Twisting Intraligand Flapping/Twisting Intraligand Flapping/Twisting Intraligand Flapping/Twisting Intraligand Symmetric Breathing Intraligand Flapping/Twisting 92.4 92.4 92.4 92.4 100% 100% 100% 100% 154.0 100% 154.0/123.2 100% 92.4 92.4 92.4 100% 100% 100% A near 92 cm-1 intraligand symmetric breathing normal mode appears across all transitions of excited state evolution in [Fe(4’-OMe-terpy)2]2+, just as it has in the other bis-tridentate complexes. Additionally, an intraligand flapping/twisting vibrational motion, where one side of the ligand exhibits intraligand flapping motions while the third ring twists, was also observed in all three transitions. This vibrational motion appears in a near 92 cm-1 normal mode and a near 154 cm-1 normal mode. This normal mode appears with a near 123 cm-1 frequency value in the 5T2 state, resulting in noticeable change in frequency of the normal mode during the 3T1 → 5T2 relaxation. Furthermore, all of the normal modes listed in Table 6.12 are associated with frequency 271 values observed in 100% of TA trials. This makes it difficult to suggest that one normal mode is more relevant to excited state evolution than the other. Indicating that excited state evolution in [Fe(4’-OMe-terpy)2]2+ relies on both intraligand breathing vibrations and intraligand distortions. 6.4 Discussion 6.4.1 Tris-bidentate Normal Mode Analysis This section will examine the most prominent promoting normal modes for each tris- bidentate complex along each step of ultrafast excited state evolution, starting with the 1A1 → 1MLCT excitation. During ultrashort pulse excitation, it is possible for a wave packet generated in the excited state to be projected onto the ground state by a light field interaction some time after the formation of the initial wave packet, this process is called impulsively stimulated Raman scattering.33,34 The projected wave packet can be conserved when the curvatures of potential energy surface for the excited state and ground state match along the reaction coordinate.33 This means that normal modes projected back onto the ground state can provide information about molecular motions after excitation. After being excited into the 1MLCT state, Fe(II)-complexes rapidly undergo an intersystem crossing to the 3MLCT state. This process is incredibly fast, within 20 fs; which is faster than the duration of the excitation pulse used in experiments (40 – 60 fs). This means that the sample may be in the 3MLCT state by the time a second field interaction projects the wave packet onto the ground state, and since wave packets may be preserved even if the zero-point frequencies of the excited and ground state potential energy surfaces are displaced along the reaction coordinate,17,33 looking into these normal modes may provide insight on which vibrational modes a given complex is engaging in during this ultrafast intersystem crossing. The prominent excitation normal modes for all tris-bidentate complexes are provided in Table 6.13. Both [Fe(bpy)3]2+ and [Fe(phen)3]2+ have an intraligand twisting mode as the most prominent promoting mode for the 1A1 → 1MLCT excitation. In the case of [Fe(phen)3]2+, the frequency of this normal mode is about 40 cm-1 higher in frequency. This is likely caused by the increase in steric hindrance introduced by the addition of the C-C double bond along the backbone of the phen ligand. However, when shifting to the substituted bpy complexes, intraligand flapping and breathing motions appear as the most prominent normal modes. For [Fe(4,4’-diCl-bpy)3]2+, the intraligand twisting motion remains, however other intraligand motions become more prominent. Both the methoxy substituted complexes feature near 100 cm-1 intraligand flapping and breathing normal modes prominently, dropping the intraligand twisting motion entirely. The 272 change to intraligand flapping and breathing normal modes does not occur for the 5,5’-diCl substituted complex, however. The most prominent promoting mode for [Fe(5,5’-diCl-bpy)3]2+ remains an intraligand twisting mode, furthermore this normal mode occurs at a frequency lower than what has been observed for the unsubstituted complexes at 80 cm-1. Table 6.13: 1A1 → 1MLCT Excitation normal modes for tris-bidentate complexes. Complex 1A1 → 1MLCT Excitation Normal Mode Description Observed Frequency (cm-1) Rate of Observation [Fe(bpy)3]2+ Intraligand Twisting [Fe(phen)3]2+ Intraligand Twisting [Fe(4,4’-diCl-bpy)3]2+ Intraligand Breathing and Twisting Intraligand Flapping and Twisting Intraligand Flapping and Breathing [Fe(4,4’-diOMe-bpy)3]2+ Intraligand Flapping Intraligand Flapping and Twisting [Fe(5,5’-diCl-bpy)3]2+ Intraligand Twisting Ligand Wagging 126.7 172.6 91.5 91.5 94.5 94.5 94.5 80.3 14.8 [Fe(5,5’-diOMe-bpy)3]2+ Intraligand Flapping and Breathing 103.8 Ligand Rocking and Substituent Rotating 221.4 100% 75% 75% 75% 50% 50% 50% 100% 100% 66% 66% This change in vibrations suggests that the addition of substituents at the 4,4’-positions and a multiatomic substituent at 5,5’-positions causes the intraligand twisting motions to become less favorable, while having a heavy monoatomic substituent at the 5,5’-positions stabilizes the intraligand twisting motions. This observation is reinforced by the [Fe(5,5’-diCl)3]2+ data across all transitions pointing towards the importance of the intraligand twisting vibration. Lastly, a low frequency (near 14 cm-1) ligand wagging normal mode is present for [Fe(5,5’-diOMe-bpy)3]2+. An oscillation of the same frequency was observed in 50% of trials performed on [Fe(4,4’-diOMe- bpy)3]2+, however the normal modes associated with that frequency was ruled out as this frequency 273 was observed in the solvent blank collected with experimental settings for those trials (550/600 nm pump/probe in acetonitrile). While having a frequency that was observed in solvent scans calls the normal mode into question, it does not rule it out completely. The importance of this normal mode for the 5,5’-diOMe substituted complex should be considered and suggests that diOMe- substitution causes the low frequency ligand motion focused modes to become more favorable. These normal modes were not observed for the remainder of the complexes. The next transition to discuss is the 1MLCT → 3T1 relaxation. As stated before, Fe(II)- based chromophores relax through a 3MLCT state on the path to the ligand field excited states, however geometry and frequency DFT calculations for the 3MLCT state were not completed. For this reason, the 1MLCT and 3T1 states are used to approximate the 1MLCT → 3MLCT → 3T1 relaxation pathway. The normal modes presented in Table 6.14 are of key importance, as designing ligands to restrict these vibrations should slow the deactivation of the MLCT excited states. Table 6.14: 1MLCT → 3T1 relaxation normal modes for tris-bidentate complexes. 1MLCT → 3T1 Normal Mode Description Observed Frequency (cm-1) Complex Rate of Observation [Fe(bpy)3]2+ Intraligand Flapping and Breathing [Fe(phen)3]2+ Intraligand Flapping [Fe(4,4’-diCl-bpy)3]2+ Intraligand Flapping Intraligand Flapping and Twisting [Fe(4,4’-diOMe-bpy)3]2+ Intraligand Flapping Intraligand Breathing and Stretching [Fe(5,5’-diCl-bpy)3]2+ Intraligand Twisting Ligand Wagging Intraligand Breathing and Substituent Bending [Fe(5,5’-diOMe-bpy)3]2+ 193.7 172.6 91.5 91.5 94.5 337.0 80.3 14.8 103.8 Ligand Shifting and Flapping 14.8 Intraligand Twisting and Substituent Rotating 221.4 274 87.5% 75% 75% 75% 50% 100% 100% 100% 66% 100% 66% For [Fe(bpy)3]2+, [Fe(phen)3]2+, and [Fe(4,4’-diCl-bpy)3]2+ intraligand flapping normal modes becomes the dominant normal modes for the 1MLCT → 3T1 relaxation. This normal mode was observed at varying frequencies across the three complexes. The normal mode for the unsubstituted bpy complex incorporates an intraligand breathing motion in addition to intraligand flapping, and therefore has the highest frequency normal mode of the three complexes. Both the 4,4’-diCl substituted bpy and phen complex lack intraligand breathing motions, possibly caused by steric hindrance introduced by the substituents and more rigid backbone, respectively, restricting this motion. The fact that the intraligand flapping normal mode has significantly higher frequency (an increase over nearly 80 cm-1) for the phen-based complex indicates that the phen backbone introduces more strain on the intraligand flapping motion compared to the 4,4’-diCl substituted bpy ligand. The intraligand flapping normal mode also appears for [Fe(4,4’-diOMe-bpy)3]2+ with a similar frequency to the 4,4’-diCl substituted bpy complex, however this normal mode was only observed in 50% of trials conducted for the 4,4’-diOMe substituted bpy complex. Instead, a higher frequency intraligand breathing mode which appeared in 100% of trials becomes the dominant normal mode. Like the 1A 1 → 1MLCT transition, a near 14 cm-1 ligand wagging mode appears, similar to the 5,5’-diOMe substituted bpy complex. However, the near 337 cm-1 intraligand breathing mode appears in more trials than this low frequency mode, indicating that it may be more relevant to 1MLCT deactivation. That said, low frequency ligand wagging normal modes remain the dominant modes for [Fe(5,5’-diOMe-bpy)3]2+. As for [Fe(5,5’-diCl-bpy)3]2+, intraligand twisting remains the dominant mode, indicating that 5,5’-diCl substitution continues to stabilize this vibrational motion. The last transition examined is the 3T1 → 5T2 relaxation. The normal modes associated with this transition are summarized in Table 6.15. For [Fe(bpy)3]2+, the near 194 cm-1 intraligand flapping and bending normal mode remains the dominant promoting mode. This normal mode also becomes the dominant promoting mode for the 3T1 → 5T2 transition in [Fe(phen)3]2+, adding an intraligand breathing motion from the 1MLCT → 3T1 relaxation. However, breathing motions did not become important for this relaxation in [Fe(4,4’-diCl-bpy)3]2+, instead the near 92 cm-1 intraligand flapping normal mode remains the dominant promoting mode. A higher frequency mode was also observed in the same percentage of trials, this near 216 cm-1 normal mode involves ligand rocking and substitutent bending in addition to intraligand flapping. For [Fe(4,4’-diOMe- 275 bpy)3]2+, a near 150 cm-1 intraligand flapping normal mode becomes the most prominent normal mode, though a high frequency intraligand stretching is also present. Near 95 cm-1 intraligand flapping and breathing normal modes also appear, though this frequency was only observed in 50% of trials. For both [Fe(5,5’-diCl-bpy)3]2+ and [Fe(5,5’-diOMe-bpy)3]2+ the same dominant promoting modes carry over from the 1MLCT → 3T1 transition to the 3T1 → 5T2 transitions, namely intraligand twisting and ligand wagging (as well as intraligand breathing and substituent bending), respectively. Table 6.15: 3T1 → 5T2 relaxation normal modes for tris-bidentate complexes. Complex [Fe(bpy)3]2+ 3T1 → 5T2 Normal Mode Description Observed Frequency (cm-1) Rate of Observation Intraligand Breathing 126.7 100% Intraligand Flapping and Breathing 193.7 87.5% [Fe(phen)3]2+ Intraligand Flapping and Breathing 205.6 [Fe(4,4’-diCl-bpy)3]2+ Intraligand Flapping Ligand Rocking, Intraligand Flapping, and Substituent Bending [Fe(4,4’-diOMe-bpy)3]2+ Intraligand Flapping Intraligand Stretching [Fe(5,5’-diCl-bpy)3]2+ Intraligand Twisting Ligand Wagging 91.5 216.4 149.6 458.9 80.3 14.8 [Fe(5,5’-diOMe-bpy)3]2+ Intraligand Breathing and Substituent Bending 103.8 Intraligand Twisting and Substituent Rotating 221.4 75% 75% 75% 75% 75% 100% 100% 66% 66% Overall, normal modes that involve intraligand distortion vibrations are of key importance for excited state relaxation in tris-bidentate Fe(II)-based polypyridines. Within the 1MLCT state, intraligand twisting motions are the most prominent. However, adding substituents to the 4,4’- positions, as well as adding a multiatomic substituent to the 5,5’-positions, causes intraligand 276 flapping motions to become more prominent. However, if a heavy monoatomic substituent, such as chlorine, is added to the 5,5’-positions, the intraligand twisting motions become more favorable. As the complexes relax into to the ligand field excited states, the intraligand twisting motions become less prominent and instead intraligand flapping motions are seen across most of the complexes studied. The 5,5’-position substituted bpy complexes stand out as exceptions, both retaining intraligand twisting vibrations as prominent promoting modes. In particular, intraligand twisting vibrations appear prominently across each transition for the 5,5’-diCl bpy complex, which suggests 5,5’-substitution makes these motions more favorable than intraligand flapping or intraligand breathing motions. 6.4.2 Bis-tridentate Normal Mode Analysis This section will investigate the normal modes important to excited state evolution in bis- tridentate Fe(II)-based polypyridines. Unlike the tris-bidentate complexes, substituents appear to have little influence on the key vibrational motions important to the MLCT excited states. Tables 6.16 and 6.17 provide key promoting modes for the 1A1 → 1MLCT excitation and 1MLCT → 3T1 relaxations. All complexes feature the near 95 cm-1 intraligand symmetric breathing normal mode across both transitions. For most of the complexes, this is the only normal mode provided. [Fe(4’- OMe-terpy)2]2+ sees an additional intraligand flapping/twisting normal mode near the same frequency, but |K| values suggest the intraligand symmetric breathing normal normal mode is the more important normal mode. The importance of breathing motions is a stark difference from bis- tridentate complexes where intraligand twisting and flapping motions were important vibrations. Table 6.16: 1A1 → 1MLCT excitation normal modes for bis-tridentate complexes. Complex 1A1 → 1MLCT Normal Mode Description Observed Frequency (cm-1) Rate of Observation [Fe(terpy)2]2+ Intraligand Symmetric Breathing [Fe(DPA)2]2+ Intraligand Symmetric Breathing [Fe(4’-Cl-terpy)2]2+ Intraligand Symmetric Breathing [Fe(4’-OMe-terpy)2]2+ Intraligand Symmetric Breathing Intraligand Flapping/Twisting 97.4 94.7 94.2 92.4 92.4 100% 100% 100% 100% 100% 277 For the 3T1 → 5T2 relaxation, differences appear between each complex. The normal modes for this transition are summarized in Table 6.17. The near 95 cm-1 intraligand symmetric breathing normal mode remains the most prominent mode. This normal mode also appears for the 4’-OMe substituted terpy complex, though with a lower |K| value than the other normal modes present, it becomes less prominent. Instead, the intraligand flapping/twisting motions observed in previous transitions becomes more prominent. For the Cl- substituted terpy complex, an intraligand breathing motion remains important, however this motion shifts from symmetric to asymmetric. Table 6.17: 1MLCT → 3T1 relaxation normal modes for bis-tridentate complexes. Complex 1MLCT → 3T1 Normal Mode Description Observed Frequency (cm-1) Rate of Observation [Fe(terpy)2]2+ Intraligand Symmetric Breathing [Fe(DPA)2]2+ Intraligand Symmetric Breathing [Fe(4’-Cl-terpy)2]2+ Intraligand Symmetric Breathing 97.4 94.7 94.2 Intraligand Asymmetric Breathing 282.5 [Fe(4’-OMe-terpy)2]2+ Intraligand Symmetric Breathing Intraligand Flapping/Twisting 92.4 92.4 100% 100% 100% 100% 100% 100% Lastly, the intraligand breathing motion disappears altogether for [Fe(DPA)2]2+. The near 95 cm-1 normal mode still has a high |K| value, but this normal mode changes frequency in the 5T2 state (from 94 cm-1 in 3T1 to 77 cm-1 in 5T2) and the 5T2 frequency values was not observed in the coherent oscillations from any trial. Instead, the most prominent normal mode for [Fe(DPA)2]2+ becomes an intraligand wagging and breathing mode. This normal mode involves one ligand engaging in subtle intraligand breathing while the other engages in the intraligand wagging motions displayed in Figure 6.16. This indicates that the rigidity of the DPA ligand interferes with the intraligand breathing motions enough to disrupt relaxation from the 3T1 state, but not enough to disrupt the deactivation of the MLCT states. The results from this study show that the ligand modifications explored in this series have no impact of the vibrational modes relevant to MLCT state relaxation. 278 Table 6.18: 3T1 → 5T2 relaxation normal modes for bis-tridentate complexes. 3T1 → 5T2 Observed Frequency (cm-1) Normal Mode Description Complex Rate of Observation [Fe(terpy)2]2+ Intraligand Symmetric Breathing [Fe(DPA)2]2+ Intraligand Wagging and Breathing [Fe(4’-Cl-terpy)2]2+ Intraligand Asymmetric Breathing 97.4 152.7 282.5 Intraligand Flapping/Twisting 154.0/123.2 [Fe(4’-OMe-terpy)2]2+ Intraligand Flapping/Twisting Intraligand Symmetric Breathing 92.4 92.4 100% 56% 100% 100% 100% 100% 6.5 Concluding Thoughts Ultrafast transient absorption spectroscopy was used to impulsively generate and then study vibrationally coherent oscillations. These oscillations were used alongside DFT calculations to determine the normal modes Fe(II)-based polypyridyl complexes engage in during excited state evolution after excitation into an MLCT excited state. Two series of complexes were investigated in this study, one of tris-bidentate complexes and one of bis-tridentate complexes. This was done with the goal of identifying the vibrational motions that enable the deactivation of the MLCT excited states. This information can be used to design ligand structures that restrict these motions and slow relaxation from the charge transfer excited states as a result. It was found that MLCT deactivation of tris-bidentate complexes is dependent on vibrations that distort the ligands while MLCT deactivation of bis-tridentate complexes is dependent on vibrations where the ligands breathe along the Fe-N bonds. These vibrational motions are illustrated in the schematics shown in Figure 6.19. Ligand structure was found to have subtle influence on the vibrations engaged during relaxation in tris-bidentate complexes, where introducing rigidity and 4,4’-substituents made intraligand breathing motions less favorable compared to intraligand flapping motions, and 5,5’-substitution leads to intraligand twisting motions becoming more favorable. For bis-tridentate complexes, ligand structure and substitution had no impact on MLCT deactivation, though it did impact relaxation from ligand field excited states. 279 Figure 6.19: Schematics showing the vibrations key to relaxation from the MLCT excited states in Fe(II)-based polypyridine complexes. On the left shows the vibrations important for tris- bidentate complexes and the right shows the vibration important for bis-tridentate complexes. 6.6 Future Directions To develop a complete map of the reaction coordinate, it will be important to finish DFT calculations on the triplet charge transfer state. These calculations require a great deal of time and resources, thus completing them will be very difficult. However, the analysis performed in this work using the singlet charge transfer state can still be put to use. Past efforts to slow MLCT state deactivation focused on using “cage” ligands to restrict motions along the Fe-N bonds.19 This was primarily focused on “tris-bidentate” style ligand structures, but this study suggests that these ligands do not rely on Fe-N breathing vibrations as heavily as previously thought. Instead, bis- tridentate ligands were the structures found to rely most heavily on these breathing vibrations. Therefore, a cage strategy would be more useful if designed around tridentate ligands. 280 REFERENCES 1. Gawelda, W.; Cannizzo, A.; Pham, V.-T.; van Mourik, F.; Bressler, C. and Chergui, M. J. Am. Chem. Soc., 2007, 129, 8199–8206. 2. Lee, A.; Son, M.; Deegbey, M.; Woodhouse, M. D.; Hart, S. M.; Beissel, H. F.; Cesana, P. T.; Jakubikova, E.; James K. McCusker, J. K.; Gabriela S. Schlau-Cohen, G. S. Chem. Sci. 2023, 14, 13140-13150. 3. Zhang, W.; Alonso-Mori, R.; Bergmann, U.; Bressler, C.; Chollet, M.; Galler, A.; Gawelda, W.; Hadt, R. G.; Hartsock, R. W.; Kroll, T.; Kjær, K. S.; Kubiček, K.; Lemke, H. T.; Liang, H. W.; Meyer, D. A.; Nielsen, M. M.; Purser, C.; Robinson, J. S.; Solomon, E. I.; Sun, Z.; Sokaras, D.; van Driel, T. B.; Vankó, G.; Weng, T.; Zhu, D.; Gaffney, K. J. Nature. 2014, 509, 345–348. 4. Monat, J. E. and McCusker, J. K. J. Am. Chem. Soc. 2000, 122, 4092-4097. 5. Auböck, G. and Chergui, M. Nat. Chem. 2015, 7, 629–633. 6. McCusker, J. K. Science. 2019, 363, 484–488. 7. Wenger, O. S. Chem. - Eur. J. 2019, 25, 6043−6052. 8. Jamula, L. L.; Brown, A. M.; Guo, D. and McCusker, J. K. Inorg. Chem. 2014, 53, 15– 17. 9. Liu, Y.; Harlang, T.; Canton, S. E.; Chábera, P.; Suárez-Alcántara, K.; Fleckhaus, A.; Vithanage, D. A.; Göransson, E.; Corani, A.; Lomoth, R.; Sundström, V. and Wärnmark, K. Chem. Commun. 2013, 49, 6412-6414. 10. Liu, Y.; Persson, P.; Sundström, V. and Wärnmark, K. Acc. Chem. Res. 2016, 49, 1477−1485. 11. Khalil, M.; Marcus, M. A.; Smeigh, A. L.; McCusker, J. K.; Chong, H. H. W. and Schoenlein, R. W. J. Phys. Chem. A. 2006, 110, 38-44. 12. Gawelda, W.; Pham, V.-T.; Benfatto, M.; Zaushitsyn, Y.; Kaiser, M.; Grolimund, D.; Johnson, S. L.; Abela, R.; Hauser, A.; Bressler, C. and Chergui, M. Phys. Rev. Lett. 2007, 98, 057401. 13. Vanko, G.; Bordage, A.; Pápai, M.; Haldrup, K.; Glatzel, P.; March, A. M.; Doumy, G.; Britz; A.; Galler, A.; Assefa, T.; Cabaret, D.; Juhin, A.; van Driel, T. B.; Kjær, K. S.; Dohn, A.; Møller, K. B.; Lemke, H. T.; Gallo, E.; Rovezzi, M.; Németh, Z.; Rozsályi, E.; Rozgonyi, T.; Uhlig, J.; Sundström, V.; Nielsen, M. M.; Young, L.; Southworth, S. H.; Bressler, C. and Gawelda, W. J. Phys. Chem. C. 2015, 119, 5888–5902. 281 14. Ashley, D. C. and Jakubikova, E. Inorg. Chem. 2018, 57, 5585−5596. 15. Nance, J.; Bowman, D. N.; Mukherjee, S.; Kelley, C. T.; and Jakubikova, E. Inorg. Chem. 2015, 54, 11259–11268. 16. Zewail, A. H. J. Phys. Chem. A. 2000, 104, 5660-5694. 17. Paulus, B. C. Insights into the photophysical reaction coordinate of 1st row transition metal chromophores from vibrational coherences. Doctoral Thesis, Michigan State University. ProQuest Dissertations Publishing. 18. Heller, E. J. Acc. Chem Soc. 1981, 14, 368-375. 19. Paulus, B. C.; Adelman, S.L.; Jamula, L. and McCusker J. K. Nature. 2020, 582, 214– 218. 20. Juban, E. A. and McCusker, J. K. J. Am. Chem. Soc. 2005, 127, 6857-6865. 21. Monnia, R.; Auböck G.; Kinschela, D.; Aziz-Lange, K. M.; Gray, H. B.; Vlček, A. and Chergui, M. Chem. Phys. Lett. 2017, 683, 112-120. 22. Zhang, W.; Alonso-Mori, R.; Bergmann, U.; Bressler, C.; Chollet, M.; Galler, A.; Gawelda, W.; Hadt, R. G.; Hartsock, R. W.; Kroll, T.; Kjær, K. S.; Kubiček, K.; Lemke, H. T.; Liang, H. W.; Meyer, D. A.; Nielsen, M. M.; Purser, C.; Robinson, J. S.; Solomon, E. I.; Sun, Z.; Sokaras, D.; van Driel, T. B.; Vankó, G.; Weng, T.; Zhu, D.; Gaffney, K. J. Nature. 2014, 509, 345–348. 23. Zhang, K.; Ash, R.; Girolami, G. S. and Vura-Weis, J. J. Am. Chem. Soc. 2019, 141, 17180–17188. 24. Ando, H.; Iuchi, S. and Sato, H. Chem. Phys. Lett. 2012, 535, 177–181. 25. Sutin, N. Prog. Inorg. Chem. 1983, 30, 441– 498. 26. Gawelda, W.; Melanie, J.; de Groot, F. M. F.; Abela, R.; Bressler, C. and Chergui, M. J. Am. Chem. Soc. 2006, 128, 5001-5009. 27. Hamilton, J. M.; Whitehead, J. R.; Williams, N. J.; Ojaimi, M. E.; Thummel, R. P. and Hancock, R. D. Inorg. Chem. 2011, 50, 3785–3790. 28. Levis, M.; Lüning, U.; Müller, M.; Schmittel, M. and Wöhrle, C. Zeitschrift für Naturforschung B. 1994, 49, 675–682. 29. Lien, Y.-J., Bowers, B. and McCusker, J. K. Manuscript in process. 282 30. Brown, A. M.; McCusker, C. E. and McCusker, J. K. Dalton Trans. 2014, 43, 17635– 17646. 31. Tribollet, J.; Galle, G.; Jonusauskas, G.; Deldicque, D.; Tondusson, M.; J.F. Letard, J. F. and Freysz, E. Chem. Phys. Lett. 2011, 513, 42–47. 32. Miller, W. H.; Ruf, B. A. and Chang, Y.-T. J. Chem. Phys. 1988, 89, 6298–6304. 33. De Silvestri, S.; Cerullo, G.; Lanzani, G. Coherent Vibrational Dynamics. CRC Press: Boca Raton, FL, 2008. 34. Ruhman, S.; Joly, A. G.; Nelson, K. A. J. Chem. Phys. 1987, 86, 6563–6565. 283 APPENDIX: SUPPLEMENTAL INFORMATION 6.SI.1 Synthetic Characterization Characterization of bipyridine and terpyridine complexes is found in previous chapters. [Fe(phen)3](PF6)2 was characterized via 1H NMR using a Bruker 500 MHz spectrometer. NMR data on this complex was collected by Bekah Bowers. Characterization of [Fe(DPA)2](BF4)2 was performed by Yi-Jyun Lien. In addition to 1H NMR spectroscopy, electrospray-ionization mass spectrometry using a Waters Xevo G2-XS Quadrupole time-of-flight spectrometer (ESI-TOF) and elemental analysis were conducted. Characterization of [Fe(phen)3](PF6)2: 1H NMR (500 MHz, Acetone-d6) δ 8.84 (dd, J = 8.2, 1.2 Hz, 1H), 8.44 (s, 1H), 8.06 (dd, J = 5.2, 1.3 Hz, 1H), 7.80 (dd, J = 8.2, 5.2 Hz, 1H). Characterization of [Fe(DPA)2](BF4)2: 1H NMR (500 MHz, d3-CD3CN): δ 79.30, 31.70, 18.27, 15.94, 12.06, -4.55. HRMS (ESI-TOF) m/z: [M-2(BF4)]2+ calc’d for C38H22N6Fe: 309.0628, obs. 309.0629. Elemental Analysis: Calc’d for C38H22B2F8FeN6·2H2O: C, 55.12; H, 3.16; N, 10.15. Found: C, 54.96; H, 2.96; N, 10.20. 6.SI.2 Vibrational Coherence Oscillations This section will provide the FFT determined frequencies of the vibrational coherence oscillations found in ultrafast TA data collected for each complex, as well as those of solvent blanks collected for each pump/probe combination. This information is already provided for [Fe(4,4’- diCl-bpy)3](BF4)2 in Table 6.1, so that complex will be omitted from this section. [Fe(bpy)3](PF6)2 Vibrational Coherence Results Interestingly, a near 14 cm-1 oscillation was observed in a few trials pumped at 490 nm, but this oscillation was only seen in the 550 nm pumped solvent blank. Regardless, DFT calculations predicted no normal modes with frequency near 14 cm-1 except for within the 3T1 state, which had one normal mode at 16 cm-1, this being the lowest frequency normal mode. 284 Table 6.S1: The frequencies of vibrational coherence oscillations across eight trials of ultrafast TA data for [Fe(bpy)3](PF6)2 in acetonitrile collected with a pump wavelength of 490 and 550 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequenies observed in the solvent blanks collected at the same pump/probe combinations, while frequency values highlighted in green represent those which appeared multiple times across the eight trials and do not match frequency values observed in the solvent blanks. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. Observed Frequencies (cm-1) 13.94 [Fe(bpy)3]2+ Coherent Oscillation Frequency (cm-1) 490 nm Pump 550 nm Pump Solvent Blank Coherent Oscillation Frequency (cm-1) Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 13.671 14.015 14.134 Trial 1 Trial 2 Trial 3 490 nm Pump ––––––––––––––––––––––––––––––––––––––→ 26.473 28.031 550 nm Pump 13.671 27.341 56.061 42.046 42.402 43.133 42.046 –––––––––––––––––––––––→ 42.40675 56.061 68.353 68.353 70.076 66.183 70.076 –––––––––→ 68.353 68.6082 95.695 86.267 123.04 123.04 126.14 126.14 127.21 132.37 129.4 126.14 154.17 155.47 158.84 158.16 154.17 84.092 112.12 109.37 –––––––––––––––––––––––→ 140.15 136.71 –––––––––––––––––––––––→ 182.2 126.685 156.162 191.39 191.39 196.21 196.21 197.88 186.91 196.21 –––––––––––––––––––––––→ 193.7429 224.24 226.15 238.26 246.07 246.07 205.06 232.4 266.29 268.55 264.73 258.8 266.29 –––––––––––––––––––––––→ 264.932 252.28 287.08 300.75 280.31 341.77 336.37 273.41 280.31 304.44 301.93 ––––––––––––––––––––––––––––––––→ 302.3733 314.43 322.35 350.38 355.44 382.78 451.13 505.81 366.214 389.79 439.3267 460.0375 369.11 364.4 367.48 370.63 359.45 –––––––––→ 364.4 369.11 462.5 395.75 438.16 466.43 494.69 388.2 392.43 → 392.43 396.45 423.57 431.33 448.49 460.09 –––––––→ ––––––––→ 462.5 490.54 518.57 437.46 478.47 505.81 532.58 574.63 574.17 285 [Fe(phen)3](PF6)2 Vibrational Coherence Results Table 6.S2: The frequencies of vibrational coherence oscillations across eight trials of ultrafast TA data for [Fe(phen)3](PF6)2 in acetonitrile collected with a pump wavelength of 480 and 510 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blanks collected at the same pump/probe combinations, while frequency values highlighted in green represent those which appeared multiple times across the eight trials and do not match frequency values observed in the solvent blanks. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(phen)3]2+ Coherent Oscillation Frequency (cm-1) 480 nm Pump 510 nm Pump Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 30.602 61.204 31.174 ––––––––––––––––→ 60.102 → 74.456 Solvent Blank Coherent Oscillation Frequency (cm-1) 510 nm 480 nm Pump Pump 24.527 29.006 58.011 87.017 120.2 119.13 119.13 122.41 123.54 109.11 109.11 –––––––––––––––––––––––––––––––––––––→ 120.2 –––––––––––→ 122.63 Observed Frequencies (cm-1) 30.888 60.653 109.11 120.7683 180.31 163.8 168.31 171.46 171.46 180.31 ––––––––––––––––––––––––––––––→ 172.6083 145.03 208.48 208.48 238.26 238.26 225.38 255.43 200.76 200.76 202.63 202.63 202.63 202.63 –––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––→ 210.36 202.0067 205.5567 188.54 196.21 244.82 247.08 –––––––––––––––––––––––––––––––––––––→ 233.81 –––––––––––––––––––––––––––––––––––––––––––––––––––––→ 233.81 236.035 245.95 275.42 261.05 269.79 285.48 282.93 315.53 312.72 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––→ 297.83 300.51 293.41 296.15 ––––––––––––––––––––––––––––→ 318.85 321.32 311.74 284.205 296.975 313.33 342.5 339.74 –––––––––––––––––––––––––––––––––––––––––––––––––––––→ 341.12 348.07 330.56 390.66 387.17 402.06 367.23 370.63 ––––––––––––––––––––––––––––––––––––→ 375.64 374.09 ––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––→ 397.83 405.26 374.09 369.7 368.93 374.6067 388.915 401.7167 420.71 420.58 416.96 436.44 450.76 570.97 461.63 473.33 478.73 506.3 506.3 509.61 ––––––––––––––––––––––––→ 498.79 –––––––––––––––––––––––––––––––––––––––––––––––––––––→ 478.59 476.03 507.4033 514.38 515.06 466.01 550.98 550.84 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 550.91 571.38 576.72 –––––––––––––––––––––––––––––––––––––––––––––→ 592.31 573.0233 602.27 565.61 613.17 639.07 631.07 ––––––––––––––––––––––––––––→ 635.07 662.22 673.25 679.48 676.14 ––––––––––––––––––––––––––––→ 676.29 711.28 725.14 685.83 732.59 748.18 286 Table 6.S2 (cont’d) [Fe(phen)3]2+ Coherent Oscillation Frequency (cm-1) 480 nm Pump 510 nm Pump Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Solvent Blank Coherent Oscillation Frequency (cm-1) 510 nm 480 nm Pump Pump 760.33 Observed Frequencies (cm-1) 772.14 810.53 826.12 856.45 1096.9 1322.2 1562.6 809.38 907.49 1005.6 1103.7 1226.3 1275.4 [Fe(4,4’-diOMe-bpy)3](PF6)2 Vibrational Coherence Results Table 6.S3: The frequencies of vibrational coherence oscillations across four trials of ultrafast TA data for [Fe(4,4’-diOMe-bpy)3](PF6)2 in acetonitrile collected with a pump wavelength of 550 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the four trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(4,4’-diOMe-bpy)3]2+ Coherent Oscillation Frequency (cm-1) Trial 1 Trial 2 Trial 3 13.56 Trial 4 13.899 Solvent Blank Coherent Oscillation Frequency (cm-1) –––––––––––––––––––→ 13.671 Observed Frequencies (cm-1) 13.7295 41.012 95.695 177.72 232.4 273.41 314.43 341.77 26.685 66.713 93.398 146.77 186.8 226.82 253.51 293.54 333.56 396.45 386.93 451.13 466.99 533.16 574.17 493.67 520.36 560.39 600.42 667.13 693.81 733.84 787.21 41.696 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 41.354 68.353 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 108.48 –––––––––––––––––––→ 111.19 109.37 136.71 94.5465 109.835 149.15 176.27 203.39 230.51 271.19 311.87 338.99 366.11 393.23 433.9 474.58 501.7 583.06 152.88 180.68 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 149.6 180.3675 222.38 –––––––––––––––––––→ 205.06 232.4 –––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 264.07 291.87 –––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 333.56 361.36 –––––––––––––––––––→ –––––––––––––––––––––––––––––→ 403.06 369.11 396.45 273.41 314.43 228.0275 269.5567 292.705 313.15 336.97 363.735 392.2033 458.65 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 458.9233 437.46 –––––––––––––––––––––––––––––→ 478.47 505.81 574.17 497.685 287 80.184 128.29 160.37 320.73 352.81 400.92 449.03 529.21 577.32 [Fe(5,5’-diCl-bpy)3](BF4)2 Vibrational Coherence Results Table 6.S4: The frequencies of vibrational coherence oscillations across three trials of ultrafast TA data for [Fe(5,5’-diCl-bpy)3](BF4)2 in acetonitrile collected with a pump wavelength of 510 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the three trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(5,5’-diCl-bpy)3]2+ Coherent Oscillation Frequency (cm-1) Trial 2 Trial 1 Trial 3 16.271 31.768 79.42 111.19 40.678 56.95 81.357 97.629 122.04 146.44 Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequencies (cm-1) 24.527 –––––––––––––––––––––––––––––––––––––––––––––→ 80.32033 ––––––––––→ 122.63 125.165 158.84 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 159.605 206.49 170.85 195.26 211.53 244.07 –––––––––––––––––––––––––––––––––––––––––––––→ 209.01 196.21 256.59 254.14 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 255.365 268.48 292.89 309.16 269.79 317.68 –––––––––––––––––––––––––→ 318.85 333.56 366.11 382.38 398.65 414.92 431.19 447.46 463.73 488.14 512.55 528.82 569.5 585.77 602.04 618.31 634.59 667.13 683.4 699.67 715.94 732.21 756.52 789.16 821.71 870.52 365.33 428.87 524.17 571.83 603.59 651.24 683.01 730.67 778.32 937.16 319.205 365.72 ––––––––––→ 369.7 –––––––––––––––––––––––––––––––––––––––––––––→ 399.785 –––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––→ 430.03 448.245 416.96 466.01 515.06 –––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––→ 527.4 572.8833 –––––––––––––––––––––––––––––––––––––––––––––→ 602.815 613.17 662.22 –––––––––––––––––––––––––––––––––––––––––––––→ 683.205 –––––––––––––––––––––––––––––––––––––––––––––→ 731.44 711.28 760.33 809.38 907.49 1275.4 288 [Fe(5,5’-diOMe-bpy)3](PF6)2 Vibrational Coherence Results Table 6.S5: The frequencies of vibrational coherence oscillations across three trials of ultrafast TA data for [Fe(5,5’-diOMe-bpy)3](PF6)2 in acetonitrile collected with a pump wavelength of 490 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the three trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(5,5’-diOMe-bpy)3]2+ Coherent Oscillation Frequency (cm-1) Trial 2 14.759 Trial 3 14.759 Trial 1 14.891 Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequencies (cm-1) –––––––––––––––––––––––––––––––––––––––––––––––––→ 14.803 59.565 104.24 134.02 238.26 282.93 357.39 402.06 446.74 491.41 550.98 580.76 640.32 685 714.78 744.56 774.35 819.02 863.69 908.37 997.71 44.278 88.557 28.031 56.061 84.092 103.32 –––––––––––––––––––––––––––––––––––––––––––––––––→ 112.12 132.84 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 103.78 133.43 191.87 221.39 265.67 295.19 324.71 354.23 383.75 413.27 147.59 140.15 182.2 221.39 –––––––––––––––––––––––––––––––––––––––––––––––––→ 221.39 250.91 252.28 280.31 324.71 –––––––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 324.71 355.81 368.99 398.51 428.03 364.4 392.43 442.78 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 444.76 487.06 516.58 546.1 634.66 693.7 723.21 797.01 841.29 885.57 959.37 1003.6 1062.7 1166 1239.8 1328.4 1446.4 1535 –––––––––––––––––––––––––––––→ –––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––––––→ 574.63 462.5 490.54 518.57 –––––––––––––––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––––––––→ 457.54 516.58 546.1 575.62 634.66 664.18 693.7 723.21 489.235 516.58 547.7267 578.19 636.5467 690.8 720.4 797.01 –––––––––––––––––––––––––––––––––––––––––––––––––→ 797.01 870.81 –––––––––––––––––––––––––––––––––––––––––––––––––→ 915.09 944.61 –––––––––––––––––––––––––––––––––––––––––––––––––→ 867.25 911.73 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 1000.655 1092.2 289 [Fe(terpy)2](PF6)2 Vibrational Coherence Results Table 6.S6: The frequencies of vibrational coherence oscillations across eleven trials of ultrafast TA data for [Fe(terpy)2](PF6)2 in acetonitrile collected with a pump wavelength of 490, 550, and 600 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blanks collected at the same pump/probe combinations, while frequency values highlighted in green represent those which appeared multiple times across the eleven trials and do not match frequency values observed in the solvent blanks. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. Due to the high number of trials, values are rounded to three significant figures so they can all fit in the table; original values were used to calculate the observed frequency values. [Fe(terpy)2]2+ Coherent Oscillation Frequency (cm-1) Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequencies (cm-1) 490 nm Pump Trial 2 15.3 30.6 Trial 1 15.5 31.0 Trial 3 28.0 Trial 1 14.1 550 nm Pump Trial 3 13.9 Trial 4 15.4 Trial 2 15.1 Trial 5 15.4 53.6 62.1 46.3 46.3 37.7 52.8 67.9 69.5 600 nm Pump Trial 2 Trial 490 nm ––––––––––––––––––––––––––––→ 28.0 Trial 3 27.8 1 ––––––––––––→ 42.0 42.0 550 nm 13.7 600 nm 29.8 ––––––––––––––––––––––––→ 69.5 70.1 56.1 ––––––––––––→ 84.1 59.6 68.4 101 99.5 98.1 98.9 98.1 97.3 92.7 92.7 97.3 98.1 98.1 ––––––––––––––––––––––––→ 97.415 124 147 163 122 138 168 194 199 233 230 248 264 126 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 139 139 –––––––––––––––––––––→ 140 137 134 112 109 104 155 182 210 212 151 166 189 204 219 170 ––––––––––––––––––––––––––––→ 168 167 168 185 ––––––––––––––––––––––––→ 182 179 195 196 196 ––––––––––––––––––––––––→ 205 208 216 216 –––––––––––––––––––––––––––––––––––––––––––––––→ 236 240 241 247 222 250 –––––––––––––––––––→ 238 232 –––––––––––––––––––––––––––––––→ 252 → 252 –––––––––––––––––––––––––––––––→ 253 263 266 279 294 309 325 ––––––––––––––––––––––––––––––––––––––––––→ 280 273 293 292 ––––––––––––––––––––––––––––––––––––––→ 309 308 –––––→ 314 313 320 334 14.975 29.366 42.896 67.808 124.22 138.56 167.18 185.39 195.67 214.70 230.22 240.01 249.44 264.19 278.32 294.96 309.72 371.12 375 ––––––––––––→ 364 392 369 396 357 387 417 420 ––––––––––––––––––––––––→ 420.48 437 432 448 463 → 463 462.53 478 477 491 519 506 575 574 536 581 280 308 275 298 314 337 297 367 424 466 509 537 608 371 432 479 347 347 498 528 459 584 290 [Fe(DPA)2](BF4)2 Vibrational Coherence Results Table 6.S7: The frequencies of vibrational coherence oscillations across nine trials of ultrafast TA data for [Fe(DPA)2](BF4)2 in acetonitrile collected with a pump wavelength of 480, 550, and 600 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blanks collected at the same pump/probe combinations, while frequency values highlighted in green represent those which appeared multiple times across the nine trials and do not match frequency values observed in the solvent blanks. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. Due to the high number of trials, values are rounded to four significant figures so they can all fit in the table; original values were used to calculate the observed frequency values. [Fe(DPA)2]2+ Coherent Oscillation Frequency (cm-1) 480 nm Pump 550 nm Pump 600 nm Pump Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 9.641 Trial 1 Trial 2 Trial 3 Solvent Blank Coherent Oscillation Frequency (cm-1) 550 nm Pump 600 nm Pump 480 nm Pump Observed Frequencies (cm-1) 32.70 27.93 27.95 –––––––––––––––––––––––––––––––––––––––––––––––––––→ 29.01 29.78 16.85 19.39 18.95 ––––––––––––––––––––––––––––––––––––→ 13.67 38.56 57.84 98.11 93.17 93.17 92.66 96.97 94.41 62.1 88.71 88.71 37.91 –––––––––––––––→ 44.36 94.76 88.71 → 88.71 ––––––––––––––––––––––––––––→ 58.01 87.02 59.57 68.35 ––––––––––––––––––––––––––––→ 109.4 104.2 18.3977 29.5263 40.2747 59.9715 88.714 93.4083 126.4 135.0 145.5 136.7 134.0 145.3 152.6 149.1 158.4 151.6 163.9 168.6 151.6 ––––––––––––––––––––––––––––––––––––→ –––––––––––––––––––––––––––––––––––––––––––→ 152.666 166.225 193.7 184.2 203.6 227.4 223.0 242.3 252.7 252.1 270.2 288.8 281.2 279.6 294.8 300.6 298.9 186.3 189.5 177.4 –––––––→ 188.5 178.7 192.8 ––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 208.5 204.0 –––––––→ 205.1 208.5 212.9 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 231.4 230.7 266.1 232.4 –––––––→ –––––––––––––––→ 230.7 257.3 236.9 255.9 –––––––––––––––→ 274.8 261.1 –––––––––––––––→ 273.4 253.2 ––––––––––––––––––––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 293.8 292.8 207.1 250.7 195.7 223.6 242.3 260.9 288.8 305.2 307.5 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 304.6 –––––––––––––––→ 314.4 312.7 320.1 310.5 312.7 348.8 354.1 350.6 –––––––→ 348.1 339.4 337.4 328.2 ––––––––––––––––––––––––––––––––––––––––––––→ 381.5 425.1 501.4 378.2 400.7 404.3 395.3 390.3 417.0 388.5 369.1 ––––––––––––––––––––––→ 396.5 357.4 387.2 407.5 ––––––––––––––––––––––––––––––––––––→ 401.928 424.2 –––––––––––––––––––––––––→ 420.6 424.66 437.5 431.9 453.1 454.9 ––––––––––––––––––––––––––––––––––––→ 453.985 463.3 512.5 511.0 511.7 ––––––––––––––––––––––––––––––––––––→ 511.71 483.3 478.6 478.5 505.8 476.5 291 186.69 194.073 205.815 224.69 232.4 250.451 263.515 272.51 284.615 293.26 298.093 306.35 311.61 335.013 351.167 382.743 Table 6.S7 (cont’d) [Fe(DPA)2]2+ Coherent Oscillation Frequency (cm-1) 480 nm Pump 550 nm Pump 600 nm Pump Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3 Observed Frequencies (cm-1) Solvent Blank Coherent Oscillation Frequency (cm-1) 550 nm Pump 480 nm Pump 600 nm Pump 536.1 545.0 599.5 643.1 686.8 730.4 559.2 626.6 565.6 574.2 580.8 725.1 [Fe(4’-Cl-terpy)2](PF6)2 Vibrational Coherence Results Table 6.S8: The frequencies of vibrational coherence oscillations across four trials of ultrafast TA data for [Fe(4’-Cl-terpy)2](PF6)2 in acetonitrile collected with a pump wavelength of 550 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the four trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(4’-Cl-terpy)2]2+ Coherent Oscillation Frequency (cm-1) Trial 3 Trial 2 Trial 1 Trial 4 8.0962 Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequencies (cm-1) 15.162 31.174 45.076 32.385 48.577 72.866 ––––––––––––––––––––––––––––––––––––––––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––→ 31.7795 46.8265 13.671 68.353 93.523 90.972 90.152 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 91.549 124.7 171.46 181.94 218.22 212.27 264.98 272.92 303.24 327.33 333.56 358.5 363.89 405.26 454.86 500.35 576.72 113.35 137.64 161.92 109.37 –––––––––––––→ ––––––––––––––––––––––––––––––––––––––––––––––––––→ 136.71 136.435 163.6 178.12 ––––––––––––––––––––––––––––––––––––––––––––––––––→ 180.03 202.41 218.6 ––––––––––––––––––––––––––––––––––––––––––––––––––→ 216.3633 205.06 135.23 165.28 195.33 240.41 232.4 273.41 270.46 –––––––––––––––––––––––––→ 299.56 ––––––––––––––––––––––––––––––––––––––––––––––––––→ 315.53 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→ 314.43 340.04 364.33 388.62 –––––––––––––→ –––––––––––––→ 360.61 390.66 465.79 369.11 396.45 437.46 478.47 505.81 292 271.69 301.4 330.445 361.8325 389.64 [Fe(4’-OMe-terpy)2](PF6)2 Vibrational Coherence Results Table 6.S9: The frequencies of vibrational coherence oscillations across three trials of ultrafast TA data for [Fe(4’-OMe-terpy)2](PF6)2 in acetonitrile collected with a pump wavelength of 550 nm and probe wavelength of 600 nm. These frequency values were determined through FFT of the oscillations present in the TA data. Frequency values highlighted in red match the frequencies observed in the solvent blank collected at the same pump/probe combination, while frequency values highlighted in green represent those which appeared multiple times across the three trials and do not match frequency values observed in the solvent blank. The Observed Frequency Values column provides an averaged value of the observed frequency values from the same row. [Fe(4’-OMe-terpy)2]2+ Coherent Oscillation Frequency (cm-1) Trial 1 Trial 2 Trial 3 30.602 Solvent Blank Coherent Oscillation Frequency (cm-1) Observed Frequencies (cm-1) 13.671 68.353 92.675 91.807 92.657 –––––––––––––––––––––––––––––––––––––––––––––→ 92.37967 109.37 123.54 122.41 123.54 –––––––––––––––––––––––––––––––––––––––––––––→ 123.1633 154.43 231.64 262.53 324.3 355.18 154.43 –––––––––––––––––––––––––––––––––––––––––––––→ 153.9567 136.71 153.01 183.61 200.76 247.08 293.41 432.4 205.06 232.4 273.41 314.43 369.11 396.45 437.46 478.47 505.81 6.SI.3 Duschinsky Calculation Results This section provides the |K|, frequency, and reorganization energy calculated through Duschinsky analysis associated with each normal mode across excited state evolution for all the complexes studied in this work. These results were used to determine which normal modes qualified to be considered for promoting modes. Normal modes will be ordered by |K|, and only those with |K| greater than 0.1 will be listed unless specified otherwise. 293 [Fe(bpy)3](PF6)2 Duschinsky Analysis Results Table 6.S10: Results of Duschinsky analysis for [Fe(bpy)3]2+. For this complex, 1MLCT state calculations were performed with a low energy excitation, state 1 (s1), and a high energy excitation, state 11 (s11). Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1A1 → 1MLCT s1 (550 nm side) 1A1 → 1MLCT s11 (490 nm side) 1MLCT → 3T1 s1 (550 nm side) Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) 2.25E+01 2.48E+01 16.72192 1.03E+01 47.73719 28.06698 2.579256 7.89264 34.6810 2 38.4959 8 37.6171 8 42.3724 8 116.172 3 128.185 2 33.6232 1 84.0295 2 139.535 10.94214 268.523 3 221.778 7 421.845 5 1413.67 6 258.659 2 22.69014 14.57195 32.26771 242.2815 12.44432 2 4 3 5 11 12 1 7 13 24 20 35 141 23 0.9301 57 0.8890 71 0.7429 34 0.5098 37 0.4697 71 0.3296 92 0.3263 37 0.2483 19 0.1471 98 0.1343 7 0.11739 5 0.1085 01 0.1012 87 0.10111 3 Mod e # 3 1 4 5 10 13 11 7 8 12 2 |K| 0.4599 17 0.4471 81 0.4399 89 0.3125 66 0.3087 18 0.2483 7 0.2462 96 0.1419 11 0.1390 45 0.1269 85 0.1255 96 Frequen cy (cm-1) 37.6171 8 33.6232 1 38.4959 8 42.3724 8 113.518 8 Reorganizati on Energy (cm-1) 6.408323 4.843148 6.075328 3.878287 19.57188 139.535 31.15296 116.172 3 84.0295 2 85.1232 3 128.185 2 34.6810 2 13.12191 2.577709 2.546121 4.163726 0.409669 Mod e # 13 2 14 4 12 15 3 16 17 11 21 28 1 20 |K| 2.6758 3 2.0272 5 1.51130 9 1.17411 1 0.7445 78 0.6793 43 0.6357 43 0.6231 34 0.6103 77 0.3825 8 0.35411 6 0.3143 7 0.3088 52 0.3044 14 8 0.2816 24 18 5 22 19 31 26 6 41 49 0.2663 14 0.2630 73 0.21116 2 0.2040 72 0.1701 37 0.1680 8 0.14711 3 0.1325 22 0.1266 1 0.1059 39 Frequen cy (cm-1) 140.572 2 36.1536 8 Reorganizati on Energy (cm-1) 3548.079 119.0792 157.886 1455.903 40.0065 7 116.497 8 162.393 9 38.0427 2 47.92432 117.0399 313.6727 11.89202 167.507 274.3241 191.603 9 106.683 7 237.875 5 349.209 3 33.7336 5 217.215 6 83.1083 2 264.007 8 192.240 3 43.3951 9 243.276 9 215.439 2 321.8529 26.5044 147.7945 381.8212 2.366322 95.45201 9.894159 88.68357 60.382 1.789327 56.65478 28.71188 384.503 67.94415 342.514 3 58.5486 6 459.297 9 622.226 9 79.11866 1.192768 72.29972 126.1915 294 Table 6.S10 (cont’d) 1MLCT → 3T1 s11 (490 nm side) 3T1 → 5T2 Mode # |K| Frequency (cm-1) Reorganization Energy (cm-1) Mode # |K| Frequency (cm-1) Reorganization Energy (cm-1) 13 15 3 14 17 1 4 2 16 12 18 23 21 6 19 25 29 27 11 31 22 10 2.76149 140.5001 1.158585 161.0615 0.889726 34.84536 0.88936 159.4535 0.748097 188.439 0.684261 31.32153 0.629546 38.76227 0.574482 33.57481 0.490708 170.3016 0.404304 121.7584 0.388776 0.348433 0.342254 194.0399 247.5913 235.3424 0.316875 57.58228 0.291912 213.2084 0.259265 315.5562 0.257578 348.0877 0.234006 345.025 0.196978 99.64561 0.145954 376.9732 0.137031 245.4945 0.129974 99.25382 142 0.115424 1430.121 91 8 40 0.115386 972.8284 0.113225 84.16189 0.105768 451.699 3859.609 889.129 20.66106 491.7337 471.5343 10.01878 12.35442 7.904127 174.9168 38.51625 137.5306 174.4474 137.745 6.664409 83.03585 341.7254 250.0422 204.8953 6.227715 52.74736 21.64661 2.690373 422.441 372.0277 1.638847 49.5715 16 14 4 3 1 17 5 2 13 26 18 21 30 11 7 12 41 54 10 25 9 19 42 93 2.269605 152.871 1.838778 118.3664 1.232698 33.46924 1.185655 32.23253 1.032797 16.05992 0.937412 166.8093 0.895967 38.02106 0.817658 30.31102 0.617988 108.0075 0.456452 289.7576 0.447298 0.31087 0.297406 169.6516 218.4416 343.7637 0.241271 93.03268 0.225494 50.97643 0.217324 95.77307 0.198938 454.8643 0.18116 632.3475 0.169052 83.33058 0.148228 269.6077 0.145956 80.92387 0.140869 189.0981 0.133935 456.7062 0.117694 973.0919 3136.48 1294.576 37.72205 32.46699 6.89583 613.3318 25.55017 13.63785 73.52866 636.9603 137.4152 99.86426 323.77 8.894742 2.759752 7.397269 159.5229 375.9759 3.57633 44.47721 2.481272 17.74053 73.64775 418.1861 [Fe(phen)3](PF6)2 Duschinsky Analysis Results Table 6.S11: Results of Duschinsky analysis for [Fe(phen)3]2+. For this complex, 1MLCT state calculations were performed with a low energy excitation, state 1 (s1), and a high energy excitation, state 20 (s20). Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1A1 → 1MLCT s1 (510 nm side) 1A1 → 1MLCT s20 (480 nm side) 1MLCT → 3T1 s1 (510 nm side) Mod e # 4 |K| 1.41849 8 Frequen cy (cm-1) 42.3019 7 Reorganizati on Energy (cm- 1) Mod e # |K| 81.48799 4 0.44072 4 Reorganizati on Energy (cm- 1) Mod e # |K| 7.866269 10 3.0145 3 Frequen cy (cm-1) 129.837 8 Reorganizati on Energy (cm- 1) 4676.029 Frequen cy (cm-1) 42.3019 7 295 Table 6.S11 (cont’d): For the high energy 1A1 → 1MLCT transition, normal modes with |K| value above 0.01 are included. 1A1 → 1MLCT s1 (510 nm side) Mod e # |K| Frequenc y (cm-1) Reorganizatio n Energy (cm-1) 33.72643 7.078703 8.287503 9.423639 2.111151 19.87621 22.10546 18.43 0.493866 3.129967 0.854256 0.649095 0.419416 0.260113 0.233379 0.199903 0.196782 0.193023 0.158723 0.154618 44.7884 26.20303 45.00175 83.88393 40.65291 159.8214 134.8301 170.2433 28.31736 81.2473 0.112719 278.9032 19.25719 0.10249 413.7293 38.82161 5 1 6 8 3 14 10 16 2 7 26 34 Mode # 10 5 31 6 53 3 84 54 27 2 14 7 69 16 3 12 3 52 114 17 1 14 5 13 5 15 9 12 0 15 1 9 1A1 → 1MLCT s20 (480 nm side) Frequenc y (cm-1) |K| 0.322294 0.288109 0.143846 0.139074 0.06886 0.068139 0.067474 0.066541 0.061996 0.057482 134.8301 44.7884 409.251 45.00175 535.34 40.65291 838.2123 535.8948 297.1584 28.31736 Reorganizatio n Energy (cm-1) 59.29702 3.836266 98.92513 0.911226 41.95866 0.179966 83.728 39.23225 7.636843 0.064772 0.044162 1354.09 52.45544 0.032982 702.8323 14.85175 0.03087 1502.716 45.08727 0.030201 1151.557 16.71807 0.027197 0.02677 534.9551 1058.65 6.530951 7.460928 0.025497 1546.295 41.7236 1MLCT → 3T1 s1 (510 nm side) Frequenc y (cm-1) |K| 1.622235 1.307401 1.224441 0.839352 0.718819 0.694421 0.552369 0.535857 0.515793 0.451013 48.26077 159.4067 157.1714 45.53043 147.7888 165.5711 28.6632 43.14109 27.40497 41.62003 Reorganizatio n Energy (cm-1) 137.9751 1188.516 732.7842 34.4734 284.4083 349.9813 6.117757 12.46737 4.859765 8.265817 0.386066 245.3675 240.4235 0.199124 156.7992 19.15843 0.187085 372.7764 82.39474 0.184763 415.0838 135.2809 0.184285 0.168671 270.5605 299.0951 53.58738 57.0422 0.163886 184.6731 22.58855 Mode # 6 14 13 5 11 15 2 4 1 3 22 12 31 36 25 28 17 0.025316 1353.3 16.8586 34 0.154097 410.3909 108.6426 0.023758 1239.582 13.04901 9 0.131797 93.064 3.040281 0.02241 1439.034 16.60266 63 0.126559 690.1505 126.6977 0.01742 1105.466 1.904776 65 0.126228 691.4869 181.0122 0.016642 1382.088 6.150116 8 0.124775 85.01358 2.227754 0.016251 90.4117 0.043851 119 0.014913 1104.642 1.394562 33 16 0.014486 411.0234 0.775534 0.014069 170.2433 0.097906 Table 6.S11 (cont’d) 29 30 11 0 64 59 0.121571 341.6749 80.138 0.115973 347.1863 68.44458 0.108949 1032.719 158.5359 0.108046 690.9938 120.9442 0.104538 627.5614 145.2668 Mode # |K| 1MLCT → 3T1 s20 (480 nm side) Frequency (cm-1) 142.4887 168.5078 166.7889 169.6214 166.2851 194.8198 30.47964 49.11463 43.24198 296.0586 48.04902 155.1545 196.3056 704.8266 2.930389 1.497585 1.442153 0.626724 0.55805 0.456449 0.421984 0.417407 0.394858 0.330496 0.30114 0.281826 0.21097 0.160414 10 15 14 16 13 17 2 6 4 27 5 11 18 67 Reorganization Energy (cm-1) 5338.224 1698.521 1522.572 199.5436 221.1965 189.6651 4.031676 9.679352 6.753802 205.9687 4.696191 34.75304 40.89662 340.7306 Mode # |K| 14 11 5 3 16 13 15 1 17 18 26 23 10 36 2.250287 1.94365 1.784825 0.945776 0.780232 0.689796 0.66023 0.468615 0.439302 0.344778 0.337506 0.231849 0.229242 0.213257 3T1 → 5T2 Frequency (cm-1) 147.3493 114.6291 37.61469 31.35979 154.7078 138.9113 151.8199 22.14476 164.1789 176.2563 266.1403 235.9538 84.23317 409.2632 Reorganization Energy (cm-1) 2937.374 1599.648 103.7036 21.28593 300.7955 195.0969 201.1006 2.812545 137.7294 83.96727 176.9345 85.25684 7.663308 185.2192 296 Table 6.S11 (cont’d) 1MLCT → 3T1 s20 (480 nm side) 275.9027 410.5189 417.7868 28.19615 41.99142 429.6481 279.9164 357.9184 628.3171 82.88337 0.147332 0.142861 0.130655 0.127978 0.12788 0.127273 0.117214 0.114319 0.107949 0.107645 24 33 34 1 3 35 25 30 60 7 31.92631 97.63102 55.42912 0.318134 0.666493 82.47787 25.93723 81.69346 153.7954 1.588674 3T1 → 5T2 272.0562 78.63298 24.11159 698.9463 404.2632 204.0558 618.1864 0.195653 0.184925 0.163989 0.160705 0.129257 0.126963 0.108518 27 8 2 69 34 19 59 65.31874 4.267251 0.390935 346.8196 75.73984 22.88059 152.7157 [Fe(4,4’-diCl-bpy)3](BF4)2 Duschinsky Analysis Results Table 6.S12: Results of Duschinsky analysis for [Fe(4,4’-diCl-bpy)3]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1MLCT → 3T1 1A1 → 1MLCT 3T1 → 5T2 Mod e # 3 1 2 4 9 15 10 13 21 31 5 35 34 |K| 1.8844 4 1.1068 35 0.7197 49 0.6424 33 0.61178 1 0.5388 88 0.3079 94 0.2788 47 0.1713 72 0.1431 79 0.1245 99 0.11447 5 0.1009 67 Frequen cy (cm-1) 19.4785 2 16.2242 9 16.7616 8 20.4442 8 81.8274 2 102.271 4 83.5007 9 98.0624 2 147.595 5 291.664 6 23.2631 6 306.759 1 303.004 5 Reorganizati on Energy (cm-1) 99.73529 25.23558 11.12689 10.85317 80.02766 73.5531 36.63892 35.66772 36.2841 Mod e # 12 4 21 3 2 20 14 9 1 46.9386 15 0.552299 33.4601 29.96071 5 8 6 18 51 17 7 |K| 2.9281 6 2.4402 68 2.1516 17 2.0781 08 2.0368 18 1.7604 14 0.9658 41 0.9658 09 0.79117 7 0.7195 81 0.4290 86 0.4221 8 0.4064 7 0.38311 3 0.3663 32 0.3388 5 0.2914 82 Reorganizati on Energy (cm-1) 3051.969 201.7861 5652.652 116.9579 117.1879 2932.092 434.0066 113.7978 14.81608 234.7671 7.756072 24.25258 9.659089 79.58438 770.6382 60.78291 9.405227 Mod e # 5 4 21 3 11 6 12 2 15 13 20 16 18 51 10 8 28 |K| 3.3368 52 3.2467 2.8913 19 2.1850 94 2.0817 63 1.7608 15 1.3012 01 1.2306 2 1.1028 21 0.9333 99 0.7675 84 0.6636 87 0.5761 3 0.4564 85 0.4203 82 0.3401 05 0.3007 8 Frequen cy (cm-1) 21.2111 5 18.7439 6 136.992 9 17.2505 5 77.0820 8 26.2807 3 80.3663 6 15.9163 6 92.6256 2 83.9697 9 122.848 4 94.5772 3 Reorganizati on Energy (cm-1) 313.3627 281.7729 8040.567 104.99 930.2693 114.8644 246.2411 28.86936 485.8858 132.4948 463.2904 183.8668 105.901 156.2438 431.676 2 76.1488 8 61.9242 5 217.337 4 1286.793 20.27933 12.0365 193.5676 Frequen cy (cm-1) 84.7584 6 21.2830 1 148.485 4 20.9543 4 19.1432 4 139.183 3 95.0195 7 78.1910 1 17.5089 2 96.5126 5 25.7048 1 75.3560 8 31.8357 1 119.567 3 441.450 2 116.727 4 73.6646 5 297 Table 6.S12 (cont’d) Mode # 24 13 53 34 35 41 16 39 26 57 19 10 42 31 45 29 101 105 |K| 0.276664 0.245348 0.198468 0.188195 0.184467 0.184179 0.175854 0.171582 0.162372 0.16117 0.151404 0.150668 0.149362 0.129955 0.128124 0.124426 0.108982 0.103372 1MLCT → 3T1 Frequency (cm-1) 191.8412 91.64303 447.3027 296.8722 298.888 363.7416 104.9304 331.8482 217.5565 510.6304 138.9031 79.11323 365.0638 266.6318 405.5878 237.8201 964.7284 980.3945 Reorganization Energy (cm- 1) 63.19659 11.8051 343.8879 73.07004 110.9408 145.2403 14.26829 105.8143 54.76616 166.5826 21.58014 3.697477 89.84737 38.75613 56.75803 33.59845 332.0499 325.2784 3T1 → 5T2 Mode # 27 19 35 14 23 57 7 38 105 22 52 45 |K| 0.293535 0.219721 0.217293 0.205417 0.183655 0.182455 0.178269 0.17391 0.142129 0.11742 0.108056 0.105143 Frequency (cm-1) 214.735 111.5813 289.7782 87.04095 183.1502 499.7013 30.29177 318.0565 972.8258 179.5775 433.9856 394.8528 Reorganization Energy (cm- 1) 175.707 23.80488 125.8989 5.883143 25.63626 206.1473 1.583294 119.6132 593.9373 9.979684 31.29594 61.80523 [Fe(4,4’-diOMe-bpy)3](PF6)2 Duschinsky Analysis Results Table 6.S13: Results of Duschinsky analysis for [Fe(4,4’-diOMe-bpy)3]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1MLCT → 3T1 1A1 → 1MLCT 3T1 → 5T2 Mod e # 2 12 5 4 10 1 21 11 20 19 27 15 48 3 18 17 |K| 1.0907 97 0.7062 15 0.51311 9 0.5067 3 0.4556 99 0.3699 24 0.3527 33 0.2680 62 0.2619 85 0.2591 52 0.2159 34 0.1987 39 0.1464 66 0.1448 5 0.1420 44 0.1371 08 Frequen cy (cm-1) 17.0846 9 83.7905 8 23.1468 7 21.1792 9 80.8852 5 16.1697 5 105.700 7 83.1319 9 104.994 8 103.696 1 159.459 1 92.6227 2 Reorganizati on Energy (cm-1) 7.581827 68.88908 3.329929 2.700135 23.28545 0.8067 22.46541 8.330267 12.4893 12.79672 35.57724 5.337546 322.794 49.3401 18.5738 1 100.075 2 99.1769 3 0.162229 3.561633 3.052446 Mod e # 13 5 27 1 23 14 16 4 20 2 11 6 36 19 3 64 |K| 3.0763 18 2.8478 72 1.9451 75 1.6427 11 1.6253 43 1.1627 81 1.03117 2 0.9455 99 0.9257 49 0.8706 77 0.8598 4 0.7018 34 0.4018 29 0.3264 72 0.3191 32 0.3079 14 Mod e # 1 5 14 2 25 21 13 23 27 3 19 36 4 16 12 50 |K| 2.2224 94 2.0784 84 1.9707 16 1.9554 46 1.6874 18 1.5241 22 1.2206 99 1.0741 9 1.0210 79 0.7921 23 0.6934 88 0.5885 0.5397 62 0.4780 53 0.4339 26 0.3853 21 Frequen cy (cm-1) 10.1396 1 21.2266 6 78.9534 7 11.0780 1 142.118 7 99.7824 1 77.5983 7 124.583 2 147.998 4 15.8339 3 92.8786 9 220.884 9 17.7553 6 89.1105 7 70.8702 2 306.782 7 Reorganizati on Energy (cm-1) 11.63216 45.16431 497.2711 12.13001 1466.384 421.1347 166.1736 420.2867 495.5637 3.595815 61.0422 239.4935 2.065936 30.85837 17.91741 334.5043 Frequen cy (cm-1) 83.6199 1 22.9251 6 161.563 4 13.6438 2 135.402 8 Reorganizati on Energy (cm-1) 1252.51 96.35415 2982.275 11.04898 1125.519 89.1015 163.6537 93.8485 5 21.5143 4 102.659 1 18.9900 9 160.3667 9.364098 159.9731 6.215056 77.6568 73.53879 8.714865 117.1186 19.37275 0.842983 455.3404 28.6267 5 222.254 9 101.206 4 19.4871 8 463.262 2 298 Table 6.S13 (cont’d) 1A1 → 1MLCT 1MLCT → 3T1 3T1 → 5T2 Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) 13 14 16 0.12787 7 0.11085 6 0.10616 6 88.06414 1.920262 91.18372 1.637058 97.90292 1.763666 49 18 22 0.27202 5 0.26763 9 0.23528 3 312.2197 160.3753 31 0.34068 9 198.2334 26.45177 98.11974 12.15122 26 0.33174 146.8472 47.0013 66 41 15 17 20 22 51 78 69 8 7 18 39 57 0.30845 8 0.30102 9 0.29914 1 459.2869 456.9595 236.2144 88.30418 87.12768 11.40566 0.2691 90.6593 9.131343 0.26512 9 0.19796 5 0.19724 3 0.17994 2 0.17881 4 0.15244 6 0.11352 3 0.10415 4 0.10325 1 0.10292 7 0.10029 9 96.487 11.14483 106.1663 8.738106 308.3465 88.3337 561.5042 220.0965 516.2514 220.0719 56.26965 1.056268 33.16308 0.343508 91.7972 1.46855 230.4891 7.333748 358.7656 31.90594 977.3275 219.9527 131.783 19.96868 74 0.22953 555.4243 347.7113 10 31 66 52 0.22505 2 0.21189 8 0.20274 5 0.18852 9 75.08869 4.684132 197.8212 8.683527 471.2035 215.7466 340.8459 90.0374 125 0.18328 968.7451 966.0934 69 0.15648 527.6438 165.4476 213 51 57 44 7 9 30 24 15 8 46 65 0.15600 9 0.15353 9 0.14624 9 0.12542 9 0.12398 8 0.11948 2 0.11903 5 0.11395 5 0.10882 8 0.10697 6 0.10210 9 0.10144 9 1541.86 1635.905 320.7864 63.21723 385.4164 107.4546 262.0172 21.72137 54.11185 0.615661 63.70281 0.822739 195.794 2.743206 129 140.7856 5.843354 90.52472 1.645112 57.44008 0.529376 289.2796 22.88568 469.0367 53.78606 [Fe(5,5’-diCl-bpy)3](BF4)2 Duschinsky Analysis Results Table 6.S14: Results of Duschinsky analysis for [Fe(5,5’-diCl-bpy)3]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1MLCT → 3T1 1A1 → 1MLCT 3T1 → 5T2 Mod e # 3 12 15 |K| 3.2260 26 2.2610 18 1.6013 13 Frequen cy (cm-1) 26.3211 6 80.4319 9 92.1065 5 Reorganizati on Energy (cm-1) 421.0563 2101.921 1334.597 Mod e # 1 4 3 |K| 1.2658 48 1.0632 6 1.0406 78 Mod e # 3 15 14 |K| 2.3626 8 2.1663 52 1.9716 06 Frequen cy (cm-1) 20.4990 1 88.1681 2 82.2536 2 Reorganizati on Energy (cm-1) 140.1985 1438.736 698.3337 Frequen cy (cm-1) Reorganizati on Energy (cm-1) 19.0209 32.27741 19.47634 21.01379 23.6774 9 21.1063 2 299 27.63137 2.841421 9 0.65143 36.02717 36.65639 Table 6.S14 (cont’d) 1A1 → 1MLCT 1MLCT → 3T1 3T1 → 5T2 Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # |K| Frequen cy (cm-1) Reorganizati on Energy (cm-1) 67.81295 175.2067 11 43.56676 38.48284 2 1.94904 7 1.69733 9 70.3918 948.8487 19.22141 60.69277 19.61329 12.15025 13 1.50256 78.66945 306.1406 13 17 21 1 33 22 11 2 28 24 34 14 30 4 1.32415 2 0.98935 5 0.92487 6 0.70136 2 0.44475 1 0.39206 5 0.37819 4 0.34857 6 0.34543 6 0.34056 4 0.33587 4 0.32407 1 0.25328 1 0.24652 2 87.92662 319.1171 123.8512 752.9956 182.6391 1125.013 11 10 2 22.10535 14.55533 14 265.7931 476.2625 189.3283 129.8166 77.84992 20.42978 23.51036 3.785324 232.6842 308.3297 196.7668 116.9395 6 9 7 15 13 5 316.9914 421.9414 21 91.83815 63.2507 234.2373 138.5496 32.17464 1.524099 8 20 23 17 0.76610 3 0.72926 8 0.72715 5 0.60865 6 0.57838 3 0.53872 6 0.41063 5 0.33499 6 0.31626 6 0.30693 9 0.24114 7 0.19540 1 85.37251 89.29177 32.12101 20.87286 0.57506 42.45219 19.18488 34.26769 17.72174 87.46039 33.17337 81.62195 19.22153 16 144.7163 58.97935 36.70433 5.748182 135.9307 21.45806 0.1856 173.3318 27.52968 0.14027 3 113.4422 10.43328 42 0.23329 370.406 217.8236 7 55 51 19 69 23 46 67 81 35 0.20793 2 0.16718 7 0.16639 6 0.16405 2 0.16357 3 0.13329 9 0.12949 7 0.12208 0.10313 3 0.10263 3 36.33956 1.379123 18 0.11157 114.7421 5.970194 503.5078 158.1511 464.3181 136.3826 162.2731 19.01817 648.8342 385.4221 190.3267 16.23949 419.3418 40.97538 644.0496 214.9345 761.7472 185.6593 317.3332 38.89423 300 22 17 12 1 4 1.26844 5 1.02990 3 0.90744 7 0.89405 8 0.87125 3 0.76142 6 156.7665 1444.415 110.0584 498.2349 75.8483 99.61111 14.78215 8.043243 26.07205 16.04218 102.1974 300.0507 23 18 34 33 30 19 7 5 35 8 69 57 25 36 81 21 42 0.62682 6 0.60879 6 0.59719 6 0.46958 2 0.45921 5 0.35212 4 0.34341 8 0.33610 9 0.23351 4 0.21305 5 0.17919 6 0.14524 9 0.12814 0.12633 2 0.11453 4 0.11441 3 0.11343 4 162.5154 302.5426 117.7361 180.6611 259.0591 884.2275 247.814 587.2948 229.7029 628.3941 123.1734 66.68827 33.47346 9.02004 27.34577 2.797662 279.6915 232.1122 34.4359 2.220048 645.6348 461.0255 486.6002 101.4608 214.4062 37.55564 312.2029 76.14839 758.3992 226.194 140.8266 7.845883 352.9265 42.6974 [Fe(5,5’-diOMe-bpy)3](PF6)2 Duschinsky Analysis Results Table 6.S15: Results of Duschinsky analysis for [Fe(5,5’-diOMe-bpy)3]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1MLCT → 3T1 1A1 → 1MLCT 3T1 → 5T2 Mod e # 3 13 19 12 25 23 17 20 5 1 37 11 50 51 48 |K| 3.9066 4 1.8412 4 1.5562 65 1.4998 55 1.2301 4 1.1817 13 0.8094 76 0.6505 03 0.5452 17 0.4357 56 0.3906 05 0.3875 68 0.2988 93 0.2484 16 0.2475 23 30 0.23411 10 46 26 15 2 59 29 87 18 72 105 63 0.2253 78 0.1991 59 0.1790 2 0.1639 72 0.1575 02 0.1495 65 0.1440 01 0.1405 65 0.1401 78 0.1360 3 0.1313 26 0.1293 89 Frequen cy (cm-1) 24.3188 3 82.0706 3 97.0814 7 80.8902 9 173.065 3 131.242 4 92.7751 3 Reorganizati on Energy (cm-1) 186.501 374.1778 425.921 261.3718 929.1941 591.6444 94.23746 97.8203 62.14645 32.5534 9 18.0418 8 220.751 7 72.3930 5 6.06072 1.275532 114.8115 11.47142 358.153 363.928 358.283 4 300.135 5 200.841 1 264.107 101.1942 13.85794 70.3699 3.681196 281.749 4 179.603 7 90.1467 5 20.8767 6 429.048 9 199.829 9 658.274 5 93.8619 1 537.635 6 68.0438 13.76058 3.428117 0.225405 57.96498 4.671802 244.3167 2.751606 84.22173 798.975 292.7138 465.811 1 72.15924 Mod e # 3 1 2 23 19 10 21 9 7 5 11 20 4 30 36 17 27 44 14 26 37 41 50 16 15 51 29 13 |K| 3.4083 83 2.5434 46 2.5416 77 1.7624 13 1.7594 62 1.1602 47 1.1421 41 0.8848 89 0.8579 31 0.8549 7 0.7726 86 0.7369 22 0.5457 94 0.3984 62 0.3672 41 0.3350 87 0.3260 2 0.3259 66 0.3175 89 0.2949 03 0.2546 46 0.2528 89 0.2525 06 0.2524 04 0.21153 3 0.2080 9 0.2054 92 0.2028 33 Mod e # 2 21 14 1 26 23 4 11 12 15 27 18 28 17 25 50 37 41 8 5 10 42 44 38 56 9 36 87 |K| 2.5874 41 2.1204 88 1.8369 06 1.6083 93 1.3856 85 1.3830 88 1.3225 24 1.2472 57 0.9502 08 0.8033 39 0.7264 23 0.6731 62 0.6029 99 0.5308 54 0.4276 32 0.4193 5 0.4065 05 0.3966 52 0.3673 23 0.2576 35 0.2232 58 0.2188 76 0.2095 4 0.1987 57 0.1803 86 0.15311 4 0.1496 75 0.1459 06 Frequen cy (cm-1) 15.4155 4 94.9539 3 77.8322 9 Reorganizati on Energy (cm-1) 32.65367 798.2402 357.2303 12.396 9.849967 153.846 2 122.433 7 21.1847 7 1053.75 692.0227 16.56396 68.8323 121.2104 74.3005 2 80.2009 3 164.387 8 86.4546 1 166.800 4 85.1367 5 148.401 3 306.025 3 80.26823 65.80056 251.5865 55.57894 202.9616 31.0188 81.97828 370.6506 215.733 87.11205 230.764 6 37.8227 4 27.7634 8 45.0040 8 233.709 8 265.729 8 220.336 8 394.878 7 40.5697 3 213.124 7 163.5644 3.697485 1.105392 1.92372 56.65495 111.5513 10.97358 110.1001 0.751439 10.3697 653.964 260.1877 Frequen cy (cm-1) 18.8801 3 16.0086 7 16.8803 6 122.117 9 90.9061 1 44.8391 8 94.4484 6 42.3321 1 36.6053 9 27.4717 9 69.0783 5 92.0568 6 23.6914 1 188.568 3 216.515 6 85.9018 9 161.365 2 Reorganizati on Energy (cm-1) 86.2229 36.68226 39.16353 1089.153 458.4902 61.03264 188.0216 29.66923 19.66399 11.24999 44.81135 73.86213 3.297403 99.43707 56.04766 13.4896 57.97931 260.469 272.3919 80.7390 4 149.758 7 217.910 6 230.361 1 302.553 3 82.2725 2 81.6682 5 10.06508 39.27298 17.30968 52.49825 118.5299 6.593932 4.428109 332.103 172.1267 185.521 2 28.67579 72.8393 4.065895 301 Table 6.S15 (cont’d) 1A1 → 1MLCT 1MLCT → 3T1 3T1 → 5T2 Mode # |K| Frequen cy (cm- 1) Reorganiz ation Energy (cm-1) Mode # |K| Frequen cy (cm- 1) Reorganiz ation Energy (cm-1) Mode # |K| Frequen cy (cm- 1) Reorganiz ation Energy (cm-1) 81 14 4 8 0.1246 96 0.1240 99 0.1095 37 0.1043 63 627.42 25 87.264 44 31.182 6 39.028 48 157.2557 1.85591 0.221174 0.317982 56 28 40 8 49 65 42 83 35 22 0.1792 74 0.1759 36 0.1611 31 0.1550 16 0.1460 77 0.1332 03 0.1286 46 0.1061 68 0.1048 7 0.1011 94 394.34 39 167.02 4 228.75 08 37.843 33 289.17 94 473.66 44 233.19 11 628.05 52 213.39 57 107.40 03 104.6346 17.03046 25.07594 0.672435 33.52586 92.25132 17.76702 125.0701 6.380685 2.2867 19 65 13 49 10 5 63 20 29 0.1412 02 0.1363 85 0.1318 71 0.1176 18 0.1107 37 0.1107 29 0.1083 25 0.1071 41 89.035 57 471.72 65 75.589 32 291.41 67 797.30 46 447.08 73 91.408 17 180.54 61 2.536065 96.01039 1.606399 21.81535 140.7861 51.41951 1.491643 7.610227 [Fe(terpy)2](PF6)2 Duschinsky Analysis Results Table 6.S16: Results of Duschinsky analysis for [Fe(terpy)2]2+. For this complex, 1MLCT state calculations were performed with a low energy excitation, state 1 (s1), and a high energy excitation, state 12 (s12). Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1A1 → 1MLCT s1 (600 nm side) |K| 0.3054 26 0.2579 71 0.2099 29 0.1928 77 0.1016 Frequen cy (cm-1) 104.361 4 27.0786 7 309.246 8 31.9720 1 214.676 Reorganizati on Energy (cm-1) 24.36484 1.099308 110.2581 0.847359 13.35198 Mod e # 9 1 24 2 16 1A1 → 1MLCT s12 (490 nm side) 1MLCT → 3T1 s1 (600 nm side) Mod e # 9 8 16 24 47 1 87 55 91 |K| 3.7200 71 1.1280 24 0.9636 9 0.8028 91 0.1905 0.1758 43 0.1459 41 0.11655 3 0.1015 69 Frequen cy (cm-1) 104.361 4 102.928 1 214.676 1 309.246 8 620.259 27.0786 7 958.175 1 696.126 1 989.430 3 Reorganizati on Energy (cm-1) 3614.549 223.9432 1201.017 1612.801 418.7037 0.510772 627.2138 183.0113 295.6549 Mod e # |K| Frequen cy (cm-1) 10 17 25 18 7 2 12 5 11 6 50 47 3 3.9189 84 0.8166 56 0.6363 13 0.4882 53 0.41136 0.31211 0.2644 48 0.1872 72 0.1547 75 0.1355 95 0.1301 67 0.1280 87 0.1276 9 102.688 2 213.333 8 307.620 1 215.867 9 75.2764 27.5406 5 141.970 9 53.5508 4 120.806 6 69.1294 4 623.500 9 618.457 6 31.1742 6 Reorganizati on Energy (cm-1) 4010.954 825.1018 1009.627 263.8413 15.81656 1.652269 32.47182 2.286162 6.860537 1.698307 197.573 188.9999 0.336184 302 Table 6.S16 (cont’d) 1MLCT → 3T1 s1 (600 nm side) Mod e # 83 8 1 86 |K| 0.1233 89 0.11834 6 0.1029 96 0.1015 59 Frequen cy (cm-1) 933.075 1 81.8767 4 20.8525 7 960.445 9 Reorganizati on Energy (cm-1) 384.0903 1.616077 0.101707 279.2806 1MLCT → 3T1 s12 (490 nm side) Mod e # 11 1 18 4 2 |K| 0.3193 38 0.1667 28 0.11730 3 0.11571 3 0.1002 17 Frequen cy (cm-1) 93.1207 4 25.4017 2 180.189 5 37.0622 9 28.3512 9 Reorganizati on Energy (cm-1) 22.35689 0.388531 12.97388 0.360853 0.182437 3T1 → 5T2 |K| 3.7050 07 3.1751 95 2.8328 45 1.8696 47 1.7324 56 0.9388 7 0.9315 3 0.9166 73 0.5789 58 0.4813 25 0.4602 38 0.4088 2 0.3931 22 0.3479 9 0.3008 14 0.2994 67 0.2664 41 0.2471 67 0.1915 22 0.1847 92 0.1838 19 0.1708 65 0.1546 37 0.1355 46 0.1204 92 0.11623 4 0.11454 1 0.1055 49 Frequen cy (cm-1) 47.0895 3 93.0506 5 26.6487 3 179.250 1 23.3952 1 43.1642 4 29.8266 6 151.891 1 Reorganizati on Energy (cm-1) 667.5135 2213.254 127.2354 3136.751 36.14092 35.55972 17.36167 489.033 58.7224 28.16595 205.246 1 73.4419 3 297.783 1 294.336 9 79.0434 7 297.136 5 68.9519 9 443.960 1 77.0540 5 696.274 1 247.870 8 978.999 3 426.511 9 301.0378 20.10156 432.3746 352.2972 13.14758 230.8497 7.439173 238.1529 6.459324 439.1869 50.18644 1064.842 146.0182 639.8 282.3747 614.277 7 697.527 2 149.128 1 515.960 8 345.904 6 212.3601 53.17183 7.597304 76.15177 82.50374 Mod e # 5 11 2 16 1 4 3 15 6 18 8 26 24 10 25 7 39 9 56 22 91 35 53 47 57 14 42 29 303 [Fe(DPA)2](BF4)2 Duschinsky Analysis Results Table 6.S17: Results of Duschinsky analysis for [Fe(DPA)2]2+. For this complex, 1MLCT state calculations were performed with a low energy excitation, state 6 (s6), and a high energy excitation, state 18 (s18). Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1A1 → 1MLCT s6 (600 nm side) 1A1 → 1MLCT s18 (480 nm side) 1MLCT → 3T1 s6 (600 nm side) Mod e # 30 8 |K| 0.1526 14 0.1325 71 Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # 331.467 73.62586 98.5924 3 4.782734 8 2 16 3 12 46 1 |K| 0.6410 53 0.2713 68 0.1726 79 0.1522 16 0.1398 33 0.11810 6 0.1055 14 Frequen cy (cm-1) 98.5924 3 27.8604 9 201.040 7 30.5894 3 151.276 4 506.048 7 27.2793 2 Reorganizati on Energy (cm-1) 111.832 1.424423 35.05058 0.54241 11.17061 95.05727 0.206436 Table 6.S17 (cont’d) Frequen cy (cm-1) 98.4604 7 202.502 2 126.676 4 31.6364 4 48.6370 7 24.7729 1 330.494 2 648.783 7 205.684 1 128.835 6 208.272 5 75.0750 4 213.553 1 855.125 2 Reorganizati on Energy (cm-1) 2746.704 1273.863 33.59213 1.407549 2.672706 0.47345 90.66097 347.5053 21.50311 9.22659 17.62404 2.195165 13.86676 168.5932 Mod e # 8 16 9 3 5 2 30 63 17 10 18 7 |K| 3.1579 75 1.0613 69 0.2725 65 0.2376 45 0.2234 98 0.1764 12 0.1691 3 0.1640 55 0.1635 5 0.1401 83 0.1384 0.1382 95 20 0.11598 89 0.1007 13 3T1 → 5T2 Mode # |K| 1MLCT → 3T1 s18 (480 nm side) Frequency (cm-1) 101.3097 3.627528 1.309063 0.584236 0.306382 0.25765 0.245301 0.244059 0.203048 0.174257 0.173857 0.172333 0.15728 211.6243 157.8165 75.61326 26.74888 334.1199 28.50269 653.5131 439.3485 33.97502 208.4788 45.16974 8 19 12 7 1 29 2 63 38 3 18 4 Reorganization Energy (cm-1) 3845.779 2109.75 212.9432 10.92242 1.179936 195.1645 1.21339 529.6328 132.3743 0.877886 24.93242 1.157751 Mode # |K| 2 9 4 16 10 22 8 30 11 5 21 1 5.270697 4.774801 1.10891 0.896563 0.692705 0.546721 0.508108 0.448773 0.416138 0.405057 0.323366 0.322448 Frequency (cm-1) 28.0633 Reorganization Energy (cm-1) 531.5648 93.54852 41.45878 180.2932 118.4941 221.4144 87.96238 323.0607 121.9718 42.07811 220.5831 24.47608 5789.223 51.45831 736.6373 172.792 463.7499 54.34582 608.9898 65.35762 7.045166 164.6909 1.592453 304 Table 6.S17 (cont’d) 1MLCT → 3T1 s18 (480 nm side) Frequency (cm-1) 202.2719 0.142802 0.118212 205.0824 0.107475 652.4888 Mode # |K| 16 17 61 Reorganization Energy (cm-1) 16.97526 11.04485 125.9748 Mode # |K| 0.295728 3T1 → 5T2 Frequency (cm-1) 314.6204 Reorganization Energy (cm-1) 191.0594 28 81 6 67 14 41 13 12 39 17 0.251368 789.5344 0.218719 69.04764 0.212722 705.4086 0.203561 153.7236 0.170422 455.1624 0.13484 152.5173 0.133581 131.7393 0.125087 430.667 0.116716 188.5952 1220.133 4.736959 668.005 23.52661 114.0585 9.948773 6.971763 65.85448 10.30334 [Fe(4’-Cl-terpy)2](PF6)2 Duschinsky Analysis Results Table 6.S18: Results of Duschinsky analysis for [Fe(4’-Cl-terpy)2]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. 1A1 → 1MLCT 1MLCT → 3T1 Mod e # 2 1 10 5 8 6 9 3 4 12 16 18 |K| 0.9082 77 0.8602 66 0.41147 6 0.3496 87 0.3263 06 0.3202 84 0.2137 97 0.2101 79 0.1801 43 0.1753 68 0.1370 09 0.11008 4 20.0097 8 99.2909 3 51.0362 4 82.2856 4 76.7528 8 86.1545 6 34.0031 1 48.8142 9 134.749 9 150.761 8 184.034 7 Frequen cy (cm-1) Reorganizati on Energy (cm-1) Mod e # 24.6608 23.81091 Frequen cy (cm-1) 101.311 1 186.555 4 17.5038 2 290.901 7 26.0856 7 121.605 9 82.8235 5 49.3952 2 296.158 5 351.107 8 277.247 2 124.994 1 |K| 4.0273 81 1.0203 32 0.9975 99 0.5595 59 0.5318 87 10 18 1 27 2 12.42099 43.48676 7.493472 13.465 10.4803 11 0.3297 6.811744 1.166363 1.913096 19.75474 11.42884 14.6814 8 5 29 33 25 12 19 16 26 55 7 0.3130 58 0.3084 63 0.2542 03 0.2263 46 0.2205 53 0.2122 37 0.2098 25 0.2021 73 0.1935 63 0.1468 64 0.1388 32 Reorganizati on Energy (cm-1) 4342.108 1308.095 11.49375 829.8108 9.051878 32.9952 13.38252 5.548768 Mod e # 11 18 2 5 1 7 14 45 |K| 4.4844 94 1.9648 79 1.5215 72 1.1403 61 0.6333 34 0.5949 28 0.5148 52 0.4472 42 172.5906 32 0.41147 6 13 3 28 31 15 95 17 0.4034 6 0.3497 54 0.3025 63 0.2974 06 0.2504 12 0.2337 53 0.1783 0.1645 22 189.3264 100.1677 20.64494 195.169 44.75161 24.90109 95.86039 249.7215 2.237996 145.160 2 287.010 8 628.122 8 82.7033 5 305 3T1 → 5T2 Frequen cy (cm-1) 91.1547 1 Reorganizati on Energy (cm-1) 4512.974 162.729 3341.303 19.9220 9 45.6399 4 16.7727 7 62.6351 7 120.752 6 490.760 5 338.852 9 55.6644 7 116.397 6 28.2349 3 279.556 3 328.420 2 135.415 5 980.271 2 137.350 3 36.39645 64.96585 4.747224 28.96618 114.6668 1575.886 643.3435 12.19998 48.24541 1.659594 204.8863 245.748 38.55575 1018.415 10.74558 Table 6.S18 (cont’d) Mode # 88 |K| 0.134338 1MLCT → 3T1 Frequency (cm-1) 946.1933 Reorganization Energy (cm- 1) 515.3621 3T1 → 5T2 Mode # 26 |K| 0.155984 Frequency (cm-1) 278.5446 Reorganization Energy (cm- 1) 53.86892 32 34 28 23 44 0.132372 346.7465 0.121104 357.6261 0.10692 292.4251 0.102664 242.5238 0.100941 500.8306 66.0232 60.98066 27.89177 12.459 81.26106 71 27 23 24 0.118601 773.4899 0.11599 278.7402 0.113489 219.2304 0.103952 221.5641 332.5871 30.06725 19.48156 10.40478 [Fe(4’-OMe-terpy)2](PF6)2 Duschinsky Analysis Results Table 6.S19: Results of Duschinsky analysis for [Fe(4’-OMe-terpy)2]2+. Normal modes highlighted in green had frequencies at both states of the transition matching observed frequency, those highlighted in yellow had only the normal mode match an observed frequency only in the initial state, and those highlighted in red have frequencies that match solvent blank frequencies. Due to the drastic geometric differences between the 3T1 and 5T2 states for this complex, there were nearly 150 normal modes with |K| above 0.1. While all of the normal modes were looked at, not all of them will be shown as most did not have relevant frequency values. 1A1 → 1MLCT 1MLCT → 3T1 3T1 → 5T2 Reorganizati on Energy (cm-1) Mod e # |K| Frequen cy (cm-1) 14.18644 10 3.7908 Mod e # 1 10 4 6 5 28 11 2 14 15 9 12 32 20 44 3 35 27 |K| 1.3214 26 0.4355 81 0.4305 26 0.3843 33 0.3530 85 0.2506 64 0.2063 0.2039 0.2022 34 0.1804 95 0.1796 19 0.17011 8 0.1623 56 0.1556 95 0.1520 55 0.1319 15 0.1221 72 0.1057 Frequen cy (cm-1) 18.9014 6 103.006 5 44.4467 9 50.19257 8.385929 69.4432 10.88122 46.6880 7 6.260214 270.221 83.77781 106.711 24.0739 9 138.157 6 141.063 9 87.9162 1 112.096 9 306.307 9 193.396 6 442.690 4 30.6938 5 358.388 8 266.470 8.007012 0.556462 13.44905 11.64971 4.207128 5.780468 62.13872 20.49259 179.144 0.375126 66.24285 13.36289 1 20 5 6 32 14 11 34 29 9 36 30 3 28 15 42 37 31 1.7372 71 0.91198 9 0.4713 16 0.4387 88 0.4302 66 0.41118 0.3920 28 0.2516 31 0.2401 81 0.2303 72 0.2292 83 0.2033 62 0.1968 85 0.1842 49 0.1667 76 0.1616 94 0.1550 0.1545 29 Mod e # 30 34 17 46 83 20 38 93 28 15 24 19 51 13 12 22 32 55 |K| 34.516 77 26.328 73 25.779 74 25.377 22 21.001 57 18.645 43 16.799 16.233 51 11.1398 5 11.0245 7 11.0185 4 10.471 79 9.5870 55 8.8597 63 7.9666 88 7.9039 68 7.7676 48 7.6303 7.6096 08 Frequen cy (cm-1) 264.908 2 307.949 1 136.568 7 457.286 8 845.263 9 166.010 5 371.872 Reorganizati on Energy (cm-1) 1563318 1555042 211873.6 2983318 8594729 261170.3 1064111 943.467 7970705 240.929 1 129.009 1 213.346 8 53541.83 37593.9 162823.9 147.877 50272.35 536.117 6 1151.37 4 96.6245 6 91.9262 3 195.432 7 286.770 593.369 6 710435.8 1044373 9564.169 13079.47 19605.67 127570.3 422656.1 Reorganizati on Energy (cm-1) 3601.085 28.66854 706.2959 12.51108 12.48768 443.2177 56.58675 166.1663 68.94664 6.491731 245.7263 87.11972 100.064 7 20.4850 4 182.898 7 49.3891 7 63.7704 4 298.352 2 137.371 315.310 6 269.563 2 85.4930 9 337.074 9 280.421 1 32.9909 1 261.876 6 101.272 25.47881 0.940023 125 38.8652 138.69 9.918809 410.976 98.23599 353.170 287.786 9 105.5917 47.80224 306 Table 6.S19 (cont’d) Mode # 17 |K| 0.143164 1MLCT → 3T1 Frequency (cm-1) 150.0384 Reorganization Energy (cm- 1) 7.619506 3T1 → 5T2 Mode # 79 |K| 7.399858 Frequency (cm-1) 806.1028 Reorganization Energy (cm- 1) 1011298 61 159 18 35 102 98 0.14114 627.7834 0.135123 1451.581 0.121238 164.3025 0.11586 324.8502 0.109169 974.7895 0.100737 964.135 227.0457 980.1578 7.585093 62.50427 310.5823 239.2308 8 123 103 122 31 49 11 7.117959 66.8567 6.691106 1132.244 6.655834 986.3843 6.449163 1129.715 6.083406 269.7555 5.847504 467.7796 1.786894 89.64918 4102.398 376420.3 1543453 354195.2 58914.25 111468.2 420.5538 6.SI.4 Duschinsky Displacement Vector Calculation Script The following is the MATLAB script used to determine the Duschinsky displacement vectors for normal modes as a complex shifts between the geometries of two electronic states. As stated before, this MATLAB script was written by former group member Dr. Bryan Paulus. The script begins with instructions on how to prepare files that will be used by the script. Three files are necessary: one that contains geometric information of the initial state, one that contains Hessian information of the initial state, and one that contains geometric information of the final state. These files are all prepared from DFT calculated geometries and checkpoint files. clear all clc %This is a script that uploads the coordinates from excel files that were %created from pbd files and reorients them both so that they have no %relative linear or angular momentum. The procedure of [Miller, w. H.; Ruf, %B. A.; Chang, Y.-T. J chem phys. 89 (10), 1988] was followed for this. %Then once there is no Linear or Angular momentum between these two %geometries, it uses the hessian calculated for one geometry to calculate %the Duschinsky displacement vector that describes the displacement of each %vibrational mode necessary to distort the reactant state to the product %state geometry. The procedure of [Ando, S.; Iuchi, S.; Sato, H. Chem. %Phys. Let. 2012, 535, 177-181.] was used for this. Finally, The reduced %masses for each vibrational mode as well as the force constant was %calculated so that the innersphere reorgonization energy for each normal %mode could be obtained using lamda_i_tot=sum(0.5*f_j*(K_j)^2). %********************************************************************* %___________________________DIRECTIONS________________________________ %This script will ask for three files before doing any calculations. It %will ask for the coordinates of the reactant state followed by the %coordinates of the product state followed by the hessian for the reactant %state. These files must be in the correct format for this script to work. %To make the coordinate files, first create a pdb file that contains the %geometry of the molecule and then import the data into excell so that not %all of the data is in the first column. This will initially create 8 %columns of values. Delete the first, third and last (eigth) columns by %right clicking the top of the column in excel and selecting delete. Now, %delete the top row by selecting row 1 on the side of the excel doc and %selecting delete. Next, scroll down to the end of the coordinates and 307 %delete all of the values that describe the connectivity of atoms. We now %have a file that has columns with atom index, atom type, x coordinate, y %coordinate, and z coordinate. We also need to add the masses of each of %these atoms in amu. Add this in by hand after the z coordinate column. %Save it as an excell file and its ready for use. To get the Hessian, we %first need to create a formatted checkpoint file for the frequency %calculation of the reactant state. to do this, log onto the terminal in %HPCC and go to the directory that contains the checkpoint file (.chk) for %the relevant frequency calculation. next, type "formchk -3 filename.chk %newfilename.fchk" without the quotes. This creates a formated checkpoint %file. Now, open this document in excel and delete everything but the %cartesian force constants and save the file. Now open up a new excel doc %and import these cartesian force constants using fixed width cells to get %five columns with all of the lower triangle of the symmetric hessian %matrix. Delete everything but these hessian values (ie no text should be %present, just the five columns of data). Save this as an excel file and it %should be ready for use. %************************************************************************** %First we will import an excel file with the reactant state geometry and %save it as Reactant. [Initial_Geometry]=uigetfile; s=strcat(Initial_Geometry); Reactant=xlsread(s); %Next, we will do the same with an excel file with the product state %geometry and save it as Product. [Final_Geometry]=uigetfile; ss=strcat(Final_Geometry); Product=xlsread(ss); %Note, These excel files are arranged in columns from left to right as atom %index number, atom type (ie C, H, etc), x coord, y coord, z coord, and %mass (in amu). %The following makes column vectors of the x, y, z components of each atom %first for the reactant state and then for the product state. Rrinix=Reactant(:,3); Rriniy=Reactant(:,4); Rriniz=Reactant(:,5); mir=Reactant(:,6); %masses of each reactant atom Rpinix=Product(:,3); Rpiniy=Product(:,4); Rpiniz=Product(:,5); mip=Product(:,6);%masses of each product atom %we get rid of any linear momentum between reactant and product states by %setting the origin of each of their axis systems at the center of mass of %each. The result is colum vectors of the x, y, and z components %(separately) of all the atoms mRrinix=Rrinix.*mir; mRriniy=Rriniy.*mir; mRriniz=Rriniz.*mir; mRpinix=Rpinix.*mip; mRpiniy=Rpiniy.*mip; mRpiniz=Rpiniz.*mip; Rcomrx=sum(mRrinix)/sum(mir); Rcomry=sum(mRriniy)/sum(mir); Rcomrz=sum(mRriniz)/sum(mir); Rcompx=sum(mRpinix)/sum(mip); Rcompy=sum(mRpiniy)/sum(mip); Rcompz=sum(mRpiniz)/sum(mip); 308 Rrfinalx=Rrinix-Rcomrx; Rrfinaly=Rriniy-Rcomry; Rrfinalz=Rriniz-Rcomrz; Rpfinalx=Rpinix-Rcompx; Rpfinaly=Rpiniy-Rcompy; Rpfinalz=Rpiniz-Rcompz; %Now we must insure that there is no angular momentum between the two %structures. To do this, we must ensure that the sum of the cross products %of each of the position vectors for each element multiplied by the scalar %of their respective masses is zero. %We use the procedure of [Miller, w. H.; Ruf,B. A.; Chang, Y.-T. J. %chem. phys. 89 (10), 1988] for this. Rri_z0=[0,0,0]; Rpi_z0=[0,0,0]; %First, we find the current angular momentum of the system by summing the %mass weighted cross products of the position vectors. for i = 1: length(Rrfinalx); Rri=[Rrfinalx(i),Rrfinaly(i),Rrfinalz(i)]; Rpi=[Rpfinalx(i),Rpfinaly(i),Rpfinalz(i)]; mcross_sub=mir(i)*cross(Rri,Rpi); cross_sub=cross(Rri,Rpi); if i==1 mcrossA=mcross_sub; crossA=cross_sub; mcrossB=mcrossA; crossB=crossA; else mcrossA=mcrossB; crossA=crossB; mcrossB=mcrossA+mcross_sub; crossB=crossA+cross_sub; end end %Next, we create a unit vector, N, that points in the direction fo the sum %of the mass weighted cross product of the position vectors nx=mcrossB(1)/sqrt(mcrossB(1)^2+mcrossB(2)^2+mcrossB(3)^2); ny=mcrossB(2)/sqrt(mcrossB(1)^2+mcrossB(2)^2+mcrossB(3)^2); nz=mcrossB(3)/sqrt(mcrossB(1)^2+mcrossB(2)^2+mcrossB(3)^2); N=[nx,ny,nz]; %Now, before we determine the angle that the position vectors of one system %needs to be rotated by (along the N unit vector) we must first project all %atoms onto the plane perpendicular to the N unit vector. This will ensure %that there is no overcounting towards the sum of the mass weighted dot %product of the position vectors which is used to determine the angle of %rotation. for i=1: length(Rrfinalx); Rri=[Rrfinalx(i),Rrfinaly(i),Rrfinalz(i)]; Rpi=[Rpfinalx(i),Rpfinaly(i),Rpfinalz(i)]; Rri_projection=Rri-dot(Rri,N)*N; Rpi_projection=Rpi-dot(Rpi,N)*N; mdot(i)=mir(i)*(dot(Rri_projection,Rpi_projection)); end %phi= angle that Rr is needed to be rotated about the vector sum of the mass %weighted cross products of Rr and Rp phi=-atan(sqrt(dot(mcrossB,mcrossB))/sum(mdot)); C=cos(phi); S=sin(phi); 309 t=1-cos(phi); T=[ t*(nx^2)+C t*nx*ny-S*nz t*nx*nz+S*ny; t*nx*ny+S*nz t*(ny^2)+C t*ny*nz- S*nx; t*nx*nz-S*ny t*ny*nz+S*nx t*(nz^2)+C]; %T is the rotation matrix that rotates Rp about mcrossB by angle phi Rpi_rotx=zeros(size(Rrfinalx)); Rpi_roty=zeros(size(Rrfinaly)); Rpi_rotz=zeros(size(Rrfinalz)); for i =1: length(Rrfinalx); Rpi=[Rpfinalx(i),Rpfinaly(i),Rpfinalz(i)]; Rpi_rot=T*Rpi.'; Rpi_rotx(i)=Rpi_rot(1); Rpi_roty(i)=Rpi_rot(2); Rpi_rotz(i)=Rpi_rot(3); end %Now we will check that the rotation is done correctly. for i=1: length(Rrfinalx); Rri=[Rrfinalx(i),Rrfinaly(i),Rrfinalz(i)]; Rpi=[Rpi_rotx(i),Rpi_roty(i),Rpi_rotz(i)]; mcross_sub=mir(i)*cross(Rri,Rpi); if i==1 mcrossA=mcross_sub; mcrossB=mcrossA; else mcrossA=mcrossB; mcrossB=mcrossA+mcross_sub; end mdot(i)=mir(i)*dot(Rri,Rpi); end %Next we import the data necessary to create the hessian matrix by %importing an excel file with the lower triangular elements. This file %should have all of the data from a formated checkpoint file where each %matrix element is listed in five columns. [Initial_Hessian]=uigetfile; Hess=strcat(Initial_Hessian); H_xl=xlsread(Hess); hessLTcount=(((3*length(Rrfinalx)- 1)*(3*length(Rrfinalx)))/2)+(3*length(Rrfinalx)); %First, we create one long column array with all of the hessian elements for i=1:ceil(hessLTcount/5); if i==1 Hvect=[H_xl(i,1);H_xl(i,2);H_xl(i,3);H_xl(i,4);H_xl(i,5)]; else Htempsub=[H_xl(i,1);H_xl(i,2);H_xl(i,3);H_xl(i,4);H_xl(i,5)]; Hvect=vertcat(Hvect,Htempsub); end end %Because of how Hvect was just made, if hessLTcount is not a multiple of 5 %then there will be several NAN values in Hvect. We remove those using the %following line. Hvect(find(isnan(Hvect)))=[]; numb=0; HessLD=zeros(3*length(Rrfinalx)); %Next, we create the lower diagonal Hessian matrix for i=1:3*length(Rrfinalx) for j=1:i numb=numb+1; HessLD(i,j)=Hvect(numb); 310 end end %now, we make the full hessian. Hessfull=HessLD+tril(HessLD,-1).'; %Now we create the mass weighted cartesian hessian which we then will %diagonallize to determine the appropriate eigen vectors (vibrational %modes). %Lets first create another array that carries all the masses of our system for i=1:length(mir) if i==1 Mbig_sub=[mir(i);mir(i);mir(i)]; Mbig=Mbig_sub; else Mbig_sub=[mir(i);mir(i);mir(i)]; Mbig=vertcat(Mbig,Mbig_sub); end end %Lets now generate the mass weighted cartesian coordinates Hm=zeros(size(Hessfull)); for i=1:(3*length(Rrfinalx)); for j=1:(3*length(Rrfinalx)); Hm(i,j)=Hessfull(i,j)/sqrt(Mbig(i)*Mbig(j)); end end %Now we diagonalize the mass weighted cartesian coordinates, Hm, and %extract out the eigen vectors ,L. [V,D,L]=eig(Hm); %now, we generate the displacement vector that discribes the displacement %between the two equilibrium geometries. for i=1:length(Rrfinalx) if i==1 Rdiff_sub=[Rpi_rotx(i)-Rrfinalx(i);Rpi_roty(i)-Rrfinaly(i);Rpi_rotz(i)- Rrfinalz(i)]; Rdiff=Rdiff_sub; else Rdiff_sub=[Rpi_rotx(i)-Rrfinalx(i);Rpi_roty(i)-Rrfinaly(i);Rpi_rotz(i)- Rrfinalz(i)]; Rdiff=vertcat(Rdiff,Rdiff_sub); end end freq=real(1.890152771.*(diag(D).*(4.35974e- 18)*(1e20)*(6.022e23)*1000/(4*(pi()^2)*(299792548^2)*(1e4))).^(0.5));%array of frequencies in units of cm-1 %Now, we calculate the reduced masses, mu, of each vibrational mode lcart=zeros(length(Mbig),length(Mbig)); for i=1 :length(Mbig); for j=1:length(Mbig); lcart_small(j)=(L(j,i)^2)/Mbig(j); end lcartsum(i)=sum(lcart_small); mu(i)=(lcartsum(i)^(-1));% reduced masses end K=((diag(Mbig).^(0.5))*L)'*Rdiff; %Duschinsky displacement vector kf=diag(D).*mu.';%Force constants lamda=(219470)*0.5.*kf.*K.^2;%Reorganization energies in cm^-1 I think. %The ~220000 number is to convert from hartrees to wavenumbers. %Now, we remove the rotational and transaltional modes. 311 Kcut=K(7:length(K)); freqcut=freq(7:length(freq)); lamdacut=lamda(7:length(lamda)); figure(1) bar(abs(Kcut)) xlabel('Normal Mode Number'); ylabel('Elements of the Duschinsky K vector in absolute value (Aamu^1^/^2)') figure(2) stem(freqcut,abs(Kcut)) xlabel('Frequency (cm^-^1)'); ylabel('Elements of the Duschinsky K vector in absolute value (Aamu^1^/^2)'); figure(3) stem(freqcut,Kcut) xlabel('Frequency (cm^-^1)'); ylabel('Elements of the Duschinsky K vector (Aamu^1^/^2)'); figure(4) stem(freqcut,lamdacut) xlabel('Frequency (cm^-^1)'); ylabel('Reorgonization Energy (cm^-^1)') lamdatot=sum(lamdacut)%Total Reorganization energy in cm^-1 Kabs=abs(Kcut); %Uses the absolute displacements to compute a weighted average frequency ABSweightavg=sum(abs(Kcut).*freqcut)/sum(abs(Kcut)) rmsfreq_a=(sum((Kabs.^2).*freqcut)/sum(Kabs.^2)) rmsfreq_b=(sum((Kabs.*freqcut).^2)/sum(Kabs.^2))^(0.5) ReorgWieghtedFreq=sum(lamdacut.*freqcut)/sum(lamdacut) 6.SI.5 DFT Calculations The geometric results for the singlet and quintet states of bpy and terpy based complexes can be found in the Supplemental Information of previous chapters. This section will provide said information for the triplet and singlet charge transfer states for those complexes, as well as the geometric data across all states for [Fe(phen)3]2+ and [Fe(DPA)2]2+. In addition, orbital structures will be provided for the charge transfer states related to each excitation state used. [Fe(bpy)3]2+ DFT Results Table 6.S20: Relevant structural parameters determined via DFT calculations. State Triplet Singlet CT (s1) Singlet CT (s11) Fe-N (Å) 2.03/2.20 2.04 2.05 Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) 172 175/178 174/176 75/87/97 81/86/95 82/90/95 312 Figure 6.S1: The DFT optimized geometries of the triplet state (left), low energy singlet charge transfer state (middle) and high energy charge transfer state (right) for [Fe(bpy)3]2+. Figure 6.S2: Molecular orbital diagrams for the molecular orbitals involved in the low energy excitation (state 1) singlet charge transfer state. This excitation state involves the following transitions: HOMO to LUMO+1 (54%), HOMO to LUMO+2 (35%), and HOMO to LUMO+9 (7%). Figure 6.S3: Molecular orbital diagrams for the molecular orbitals involved in the high energy excitation (state 11) singlet charge transfer state. This excitation state involves the following transitions: HOMO-2 to LUMO+9 (31%), HOMO-2 to LUMO+10 (15%), HOMO-1 to LUMO+9 (15%), and HOMO-1 to LUMO+10 (31%). 313 [Fe(phen)3]2+ DFT Results Table 6.S21: Relevant structural parameters determined via DFT calculations. State Fe-N (Å) Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) Singlet Triplet Quintet Singlet CT (s1) Singlet CT (s20) 2.05 2.05/2.1/2.3 2.25/2.27 2.05/2.06 2.06 175 171/173/174 168/170 176/178 175 82/90/95 80/90/97 76/92/96 82/88/95 82/90/94 Figure 6.S4: The DFT optimized geometries of each state calculated for [Fe(phen)3]2+. The label for each state is provided below. Figure 6.S5: Molecular orbital diagrams for the molecular orbitals involved in the low energy excitation (state 1) singlet charge transfer state. This excitation state involves the following transitions: HOMO to LUMO (24%), HOMO to LUMO+1 (62%), and HOMO to LUMO+7 (7%). 314 Figure 6.S6: Molecular orbital diagrams for the molecular orbitals involved in the high energy excitation (state 20) singlet charge transfer state. This excitation state involves the following transitions: HOMO-2 to LUMO+4 (39%), HOMO-2 to LUMO+5 (8%), HOMO-1 to LUMO+4 (7%), to HOMO-1 to LUMO+5 (41%). [Fe(4,4’-diCl-bpy)3]2+ DFT Results Table 6.S22: Relevant structural parameters determined via DFT calculations. State Fe-N (Å) Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) Triplet Singlet CT (s1) 2.03/2.18/2.24 2.04/2.05 170/172/174 175/178 77/87/96 81/87/95 Figure 6.S7: The DFT optimized geometries of the triplet state (left) and charge transfer state (right) of [Fe(4,4’-diCl-bpy)3]2+. [Fe(4,4’-diOMe-bpy)3]2+ DFT Results Table 6.S23: Relevant structural parameters determined via DFT calculations. State Fe-N (Å) Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) Triplet Singlet CT (s1) 2.04/2.17/2.25 2.04 169/171/174 175/177 78/87/96 81/87/96 315 Figure 6.S8: The DFT optimized geometries of the triplet state (left) and charge transfer state (right) of [Fe(4,4’-diOMe-bpy)3]2+. [Fe(5,5’-diCl-bpy)3]2+ DFT Results Table 6.S24: Relevant structural parameters determined via DFT calculations. Neq-Fe-Nax (deg) Nax-Fe-Nax (deg) Fe-N (Å) State Triplet Singlet CT (s1) 2.04/2.15/2.26 2.04/2.13/2.23 170/172/174 170/175 76/87/97 77/88/96 Figure 6.S9: The DFT optimized geometries of the triplet state (left) and charge transfer state (right) of [Fe(5,5’-diCl-bpy)3]2+. [Fe(5,5’-diOMe-bpy)3]2+ DFT Results Table 6.S25: Relevant structural parameters determined via DFT calculations. Fe-N (Å) Nax-Fe-Nax (deg) Neq-Fe-Nax (deg) State Triplet 2.03/2.17/2.24 Singlet CT (s1) 2.03/2.13/2.23 78/88/97 78/88/95 172/174 170/175 316 Figure 6.S10: The DFT optimized geometries the triplet state (left) and charge transfer state (right) of [Fe(5,5’-diOMe-bpy)3]2+. [Fe(terpy)2]2+ DFT Results Table 6.S26: Relevant structural parameters determined via DFT calculations. State Triplet Singlet CT (s1) Singlet CT (s12) Fe-Nax (Å) 1.97 1.95 1.97 Fe-Neq (Å) 2.20 2.05 2.18/2.19 Nax-Fe-Nax (deg) 179.8 179.4 179.6 Neq-Fe-Neq (deg) 158.1 162.2 158.3 Figure 6.S11: The DFT optimized geometries of the triplet state (left), low energy singlet charge transfer state (middle) and high energy charge transfer state (right) for [Fe(terpy)2]2+. 317 Figure 6.S12: Molecular orbital diagrams for the molecular orbitals involved in the low energy excitation (state 1) singlet charge transfer state. This excitation state involves the following transition: HOMO to LUMO (96%). Figure 6.S13: Molecular orbital diagrams for the molecular orbitals involved in the high energy excitation (state 12) singlet charge transfer state. This excitation state involves the following transitions: HOMO-2 to LUMO+2 (67%), HOMO-2 to LUMO+3 (17%), HOMO-1 to LUMO+2 (8%), HOMO-1 to LUMO+3 (5%) . [Fe(DPA)2]2+ DFT Results Table 6.S27: Relevant structural parameters determined via DFT calculations. State Singlet Triplet Quintet Singlet CT (s6) Singlet CT (s18) Fe-Nax (Å) 1.94 1.97 2.16 1.95 1.95 Fe-Neq (Å) 2.12 2.26 2.35 2.13 2.10/2.11 Nax-Fe-Nax (deg) 179.8 179.9 168.5 179.8 179.5 Neq-Fe-Neq (deg) 159.1 156.4 147.5 159.3 159.7 318 Figure 6.S14: The DFT optimized geometries of each state calculated for [Fe(DPA)2]2+. The label for each state is provided below. Figure 6.S15: Molecular orbital diagrams for the molecular orbitals involved in the low energy excitation (state 6) singlet charge transfer state. This excitation state involves the following transitions: HOMO-2 to LUMO (38%), HOMO-2 to LUMO+1 (9%), HOMO-1 to LUMO (9%), HOMO-1 to LUMO+1 (39%), and HOMO to LUMO+3 (3%). Figure 6.S16: Molecular orbital diagrams for the molecular orbitals involved in the high energy excitation (state 12) singlet charge transfer state. This excitation state involves the following transitions: HOMO-2 to LUMO+9 (6%), HOMO-1 to LUMO+9 (5%), HOMO to LUMO+4 (49%), and HOMO to LUMO+5 (28%). 319 [Fe(4’-Cl-terpy)2]2+ DFT Results Table 6.S28: Relevant structural parameters determined via DFT calculations. State Triplet Singlet CT (s1) Fe-Nax (Å) 1.97 2.00 Fe-Neq (Å) 2.20 2.04/2.06 Nax-Fe-Nax (deg) 17.9 178.5 Neq-Fe-Neq (deg) 157.8/158.0 161.9/162.0 Figure 6.S17: The DFT optimized geometries of the triplet state (left) and charge transfer state (right) for [Fe(4’-Cl-terpy)2]2+. [Fe(4’-OMe-terpy)2]2+ DFT Results Table 6.S29: Relevant structural parameters determined via DFT calculations. State Triplet Singlet CT (s1) Fe-Nax (Å) 1.98 1.93/1.97 Fe-Neq (Å) 2.20 2.04/2.07 Nax-Fe-Nax (deg) 179.7 178.4 Neq-Fe-Neq (deg) 157.4 159.9/162.9 Figure 6.S18: The DFT optimized geometries of the triplet state (left) and charge transfer state (right) for [Fe(4’-OMe-terpy)2]2+. 320 6.SI.6 DFT Optimized Geometry Files [Fe(bpy)3]2+: Table 6.S30: The optimized geometry file for the lowest energy triplet state of [Fe(bpy)3]2+. TITLE Fe-bpy Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.688 0.829 -1.222 N HETATM 2 N 0 1.637 -0.745 0.960 N HETATM 3 N 0 -0.187 -1.912 -1.042 N HETATM 4 N 0 -1.582 -1.053 1.136 N HETATM 5 N 0 -1.446 1.013 -1.004 N HETATM 6 N 0 -0.071 1.858 1.136 N HETATM 7 C 0 1.611 1.649 -2.324 C HETATM 8 H 0 0.588 1.874 -2.699 H HETATM 9 C 0 2.762 2.188 -2.964 C HETATM 10 H 0 2.645 2.844 -3.855 H HETATM 11 C 0 4.047 1.870 -2.434 C HETATM 12 C 0 4.138 1.024 -1.293 C HETATM 13 H 0 5.134 0.772 -0.870 H HETATM 14 C 0 2.939 0.511 -0.706 C HETATM 15 C 0 2.911 -0.405 0.496 C HETATM 16 C 0 4.089 -0.912 1.127 C HETATM 17 H 0 5.094 -0.636 0.743 H HETATM 18 C 0 3.973 -1.780 2.248 C HETATM 19 C 0 2.672 -2.125 2.717 C HETATM 20 H 0 2.526 -2.801 3.589 H HETATM 21 C 0 1.539 -1.591 2.048 C HETATM 22 H 0 0.509 -1.834 2.384 H HETATM 23 C 0 0.612 -2.318 -2.088 C HETATM 24 H 0 1.329 -1.561 -2.478 H HETATM 25 C 0 0.540 -3.618 -2.658 C HETATM 26 H 0 1.207 -3.889 -3.506 H HETATM 27 C 0 -0.392 -4.549 -2.113 C HETATM 28 C 0 -1.219 -4.142 -1.030 C HETATM 29 H 0 -1.941 -4.860 -0.586 H HETATM 30 C 0 -1.103 -2.811 -0.515 C HETATM 31 C 0 -1.953 -2.291 0.632 C HETATM 32 C 0 -3.068 -3.011 1.169 C HETATM 33 H 0 -3.370 -3.993 0.744 H HETATM 34 C 0 -3.804 -2.453 2.252 C HETATM 35 C 0 -3.413 -1.185 2.773 C HETATM 36 H 0 -3.955 -0.710 3.621 H HETATM 37 C 0 -2.303 -0.522 2.180 C HETATM 38 H 0 -1.966 0.471 2.553 H HETATM 39 C 0 -2.131 0.502 -2.090 C HETATM 40 H 0 -1.834 -0.516 -2.419 H HETATM 41 C 0 -3.153 1.220 -2.766 C HETATM 42 H 0 -3.666 0.753 -3.636 H HETATM 43 C 0 -3.496 2.525 -2.308 C HETATM 44 C 0 -2.801 3.063 -1.190 C HETATM 45 H 0 -3.054 4.078 -0.816 H HETATM 46 C 0 -1.781 2.293 -0.550 C HETATM 47 C 0 -0.995 2.774 0.644 C HETATM 48 C 0 -1.163 4.059 1.249 C HETATM 49 H 0 -1.901 4.785 0.846 H HETATM 50 C 0 -0.371 4.404 2.380 C HETATM 51 C 0 0.572 3.461 2.883 C HETATM 52 H 0 1.210 3.688 3.765 H HETATM 53 C 0 0.691 2.203 2.230 C HETATM 54 H 0 1.414 1.438 2.587 H HETATM 55 Fe 0 0.035 0.040 -0.017 Fe HETATM 56 H 0 -4.673 -3.002 2.679 H HETATM 57 H 0 -0.473 -5.581 -2.522 H HETATM 58 H 0 -4.292 3.117 -2.813 H HETATM 59 H 0 -0.491 5.401 2.862 H HETATM 60 H 0 4.885 -2.180 2.745 H HETATM 61 H 0 4.971 2.277 -2.902 H END 321 Table 6.S30 (cont’d) CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 57 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 30 4 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 56 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 58 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 46 6 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 59 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 34 CONECT 57 27 CONECT 58 43 CONECT 59 50 CONECT 60 18 CONECT 61 11 Table 6.S31: The optimized geometry file for the singlet charge transfer state of [Fe(bpy)3]2+ using td=1. TITLE Fe-bpy CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.521 0.771 -1.091 N HETATM 2 N 0 1.508 -0.793 1.091 N 322 Table 6.S31 (cont’d) HETATM 3 N 0 -0.083 -1.775 -1.004 N HETATM 4 N 0 -1.516 -0.921 1.061 N HETATM 5 N 0 -1.502 0.946 -1.064 N HETATM 6 N 0 -0.058 1.778 1.002 N HETATM 7 C 0 1.438 1.656 -2.153 C HETATM 8 H 0 0.415 1.884 -2.524 H HETATM 9 C 0 2.569 2.230 -2.772 C HETATM 10 H 0 2.436 2.912 -3.640 H HETATM 11 C 0 3.882 1.904 -2.252 C HETATM 12 C 0 3.992 1.024 -1.158 C HETATM 13 H 0 4.991 0.784 -0.733 H HETATM 14 C 0 2.804 0.440 -0.565 C HETATM 15 C 0 2.796 -0.483 0.566 C HETATM 16 C 0 3.974 -1.087 1.159 C HETATM 17 H 0 4.977 -0.863 0.735 H HETATM 18 C 0 3.849 -1.965 2.253 C HETATM 19 C 0 2.530 -2.270 2.772 C HETATM 20 H 0 2.386 -2.950 3.639 H HETATM 21 C 0 1.409 -1.677 2.152 C HETATM 22 H 0 0.383 -1.890 2.522 H HETATM 23 C 0 0.735 -2.126 -2.055 C HETATM 24 H 0 1.453 -1.353 -2.400 H HETATM 25 C 0 0.664 -3.403 -2.677 C HETATM 26 H 0 1.343 -3.635 -3.526 H HETATM 27 C 0 -0.277 -4.356 -2.191 C HETATM 28 C 0 -1.122 -4.001 -1.102 C HETATM 29 H 0 -1.858 -4.731 -0.705 H HETATM 30 C 0 -1.009 -2.700 -0.525 C HETATM 31 C 0 -1.842 -2.205 0.626 C HETATM 32 C 0 -2.887 -2.953 1.248 C HETATM 33 H 0 -3.144 -3.972 0.886 H HETATM 34 C 0 -3.601 -2.381 2.338 C HETATM 35 C 0 -3.257 -1.072 2.781 C HETATM 36 H 0 -3.784 -0.586 3.631 H HETATM 37 C 0 -2.213 -0.373 2.115 C HETATM 38 H 0 -1.915 0.647 2.434 H HETATM 39 C 0 -2.205 0.408 -2.120 C HETATM 40 H 0 -1.918 -0.615 -2.441 H HETATM 41 C 0 -3.240 1.120 -2.785 C HETATM 42 H 0 -3.773 0.642 -3.636 H HETATM 43 C 0 -3.569 2.432 -2.338 C HETATM 44 C 0 -2.849 2.994 -1.247 C HETATM 45 H 0 -3.093 4.015 -0.883 H HETATM 46 C 0 -1.812 2.232 -0.627 C HETATM 47 C 0 -0.973 2.715 0.525 C HETATM 48 C 0 -1.069 4.017 1.103 C HETATM 49 H 0 -1.797 4.757 0.708 H HETATM 50 C 0 -0.218 4.361 2.191 C HETATM 51 C 0 0.713 3.396 2.674 C HETATM 52 H 0 1.397 3.620 3.522 H HETATM 53 C 0 0.767 2.119 2.051 C HETATM 54 H 0 1.476 1.337 2.394 H HETATM 55 Fe 0 -0.027 0.001 -0.001 Fe HETATM 56 H 0 -4.416 -2.953 2.834 H HETATM 57 H 0 -0.354 -5.366 -2.652 H HETATM 58 H 0 -4.378 3.015 -2.833 H HETATM 59 H 0 -0.281 5.371 2.652 H HETATM 60 H 0 4.754 -2.428 2.706 H HETATM 61 H 0 4.795 2.352 -2.705 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 323 Table 6.S31 (cont’d) CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 57 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 56 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 4 35 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 58 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 5 44 47 CONECT 47 46 6 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 59 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 34 CONECT 57 27 CONECT 58 43 CONECT 59 50 CONECT 60 18 CONECT 61 11 Table 6.S32: The optimized geometry file for the singlet charge transfer state of [Fe(bpy)3]2+ using td=11. TITLE Fe-bpy CT s11 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.549 0.775 -1.097 N HETATM 2 N 0 1.567 -0.799 1.070 N HETATM 3 N 0 -0.101 -1.775 -1.037 N HETATM 4 N 0 -1.473 -0.914 1.098 N HETATM 5 N 0 -1.419 0.946 -1.109 N HETATM 6 N 0 -0.099 1.723 1.080 N HETATM 7 C 0 1.437 1.624 -2.175 C HETATM 8 H 0 0.406 1.858 -2.511 H HETATM 9 C 0 2.573 2.172 -2.834 C HETATM 10 H 0 2.426 2.851 -3.702 H 324 Table 6.S32 (cont’d) HETATM 11 C 0 3.872 1.836 -2.359 C HETATM 12 C 0 3.996 0.962 -1.242 C HETATM 13 H 0 5.002 0.695 -0.854 H HETATM 14 C 0 2.819 0.441 -0.626 C HETATM 15 C 0 2.827 -0.489 0.562 C HETATM 16 C 0 4.011 -1.023 1.154 C HETATM 17 H 0 5.011 -0.774 0.739 H HETATM 18 C 0 3.902 -1.882 2.284 C HETATM 19 C 0 2.611 -2.188 2.801 C HETATM 20 H 0 2.477 -2.850 3.685 H HETATM 21 C 0 1.469 -1.627 2.167 C HETATM 22 H 0 0.446 -1.836 2.543 H HETATM 23 C 0 0.673 -2.137 -2.118 C HETATM 24 H 0 1.382 -1.374 -2.500 H HETATM 25 C 0 0.575 -3.416 -2.731 C HETATM 26 H 0 1.221 -3.657 -3.603 H HETATM 27 C 0 -0.356 -4.357 -2.206 C HETATM 28 C 0 -1.162 -3.989 -1.092 C HETATM 29 H 0 -1.891 -4.713 -0.669 H HETATM 30 C 0 -1.020 -2.688 -0.523 C HETATM 31 C 0 -1.832 -2.182 0.644 C HETATM 32 C 0 -2.894 -2.913 1.254 C HETATM 33 H 0 -3.178 -3.918 0.877 H HETATM 34 C 0 -3.598 -2.338 2.351 C HETATM 35 C 0 -3.224 -1.043 2.810 C HETATM 36 H 0 -3.744 -0.553 3.662 H HETATM 37 C 0 -2.161 -0.362 2.155 C HETATM 38 H 0 -1.837 0.648 2.480 H HETATM 39 C 0 -2.102 0.437 -2.200 C HETATM 40 H 0 -1.789 -0.569 -2.552 H HETATM 41 C 0 -3.134 1.137 -2.864 C HETATM 42 H 0 -3.635 0.676 -3.743 H HETATM 43 C 0 -3.508 2.448 -2.371 C HETATM 44 C 0 -2.837 2.983 -1.256 C HETATM 45 H 0 -3.124 3.981 -0.857 H HETATM 46 C 0 -1.777 2.232 -0.607 C HETATM 47 C 0 -1.007 2.684 0.544 C HETATM 48 C 0 -1.112 3.991 1.164 C HETATM 49 H 0 -1.810 4.745 0.737 H HETATM 50 C 0 -0.329 4.311 2.288 C HETATM 51 C 0 0.588 3.321 2.820 C HETATM 52 H 0 1.222 3.528 3.709 H HETATM 53 C 0 0.666 2.065 2.181 C HETATM 54 H 0 1.358 1.283 2.563 H HETATM 55 Fe 0 0.004 -0.008 0.001 Fe HETATM 56 H 0 -4.429 -2.896 2.836 H HETATM 57 H 0 -0.458 -5.369 -2.658 H HETATM 58 H 0 -4.323 3.024 -2.864 H HETATM 59 H 0 -0.410 5.316 2.761 H HETATM 60 H 0 4.818 -2.305 2.753 H HETATM 61 H 0 4.782 2.251 -2.847 H END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 325 Table 6.S32 (cont’d) CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 57 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 56 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 4 35 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 58 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 46 6 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 59 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 34 CONECT 57 27 CONECT 58 43 CONECT 59 50 CONECT 60 18 CONECT 61 11 [Fe(phen)3]2+: Table 6.S33: The optimized geometry file for the lowest energy singlet state of [Fe(phen)3]2+. TITLE Fe-phen Singlet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 Fe 0 30.205 3.538 9.094 Fe HETATM 2 N 0 29.002 3.659 7.435 N HETATM 3 N 0 28.909 2.336 10.138 N HETATM 4 N 0 31.442 3.600 10.732 N HETATM 5 N 0 31.470 4.633 7.902 N HETATM 6 C 0 31.217 4.718 11.542 C HETATM 7 N 0 30.906 1.723 8.438 N HETATM 8 N 0 29.505 5.285 9.919 N HETATM 9 C 0 31.609 5.678 5.627 C HETATM 10 C 0 32.713 5.134 8.176 C HETATM 11 H 0 33.136 4.905 9.178 H HETATM 12 C 0 32.962 4.027 13.145 C HETATM 13 H 0 33.557 4.188 14.073 H HETATM 14 C 0 30.491 -0.740 8.661 C HETATM 15 C 0 30.179 5.639 11.093 C HETATM 16 C 0 27.317 0.323 11.453 C HETATM 17 H 0 26.695 -0.450 11.961 H 326 Table 6.S33 (cont’d) HETATM 18 C 0 31.639 6.205 13.516 C HETATM 19 H 0 32.209 6.410 14.451 H HETATM 20 C 0 30.201 0.641 8.978 C HETATM 21 C 0 29.594 4.370 6.385 C HETATM 22 C 0 27.915 2.677 11.014 C HETATM 23 H 0 27.747 3.762 11.184 H HETATM 24 C 0 32.263 0.103 7.173 C HETATM 25 H 0 33.093 -0.049 6.448 H HETATM 26 C 0 33.177 2.893 12.330 C HETATM 27 H 0 33.943 2.130 12.594 H HETATM 28 C 0 28.955 4.587 5.106 C HETATM 29 C 0 29.131 0.973 9.911 C HETATM 30 C 0 31.953 4.986 12.758 C HETATM 31 C 0 29.692 -1.791 9.306 C HETATM 32 H 0 29.919 -2.854 9.060 H HETATM 33 C 0 27.103 1.700 11.688 C HETATM 34 H 0 26.310 2.050 12.385 H HETATM 35 C 0 28.359 -0.078 10.538 C HETATM 36 C 0 30.940 5.889 4.336 C HETATM 37 H 0 31.472 6.480 3.555 H HETATM 38 C 0 31.911 1.444 7.552 C HETATM 39 H 0 32.464 2.312 7.133 H HETATM 40 C 0 30.653 7.092 13.080 C HETATM 41 H 0 30.426 8.015 13.663 H HETATM 42 C 0 29.671 5.364 4.085 C HETATM 43 H 0 29.175 5.529 3.101 H HETATM 44 C 0 33.464 5.913 7.229 C HETATM 45 H 0 34.470 6.287 7.520 H HETATM 46 C 0 30.924 4.910 6.644 C HETATM 47 C 0 32.920 6.187 5.954 C HETATM 48 H 0 33.487 6.785 5.204 H HETATM 49 C 0 31.560 -0.993 7.724 C HETATM 50 H 0 31.822 -2.040 7.446 H HETATM 51 C 0 27.757 3.134 7.219 C HETATM 52 H 0 27.297 2.572 8.061 H HETATM 53 C 0 27.050 3.293 5.978 C HETATM 54 H 0 26.039 2.840 5.874 H HETATM 55 C 0 28.851 7.704 11.346 C HETATM 56 H 0 28.591 8.638 11.894 H HETATM 57 C 0 28.530 6.134 9.469 C HETATM 58 H 0 28.001 5.841 8.537 H HETATM 59 C 0 27.640 4.019 4.919 C HETATM 60 H 0 27.107 4.159 3.950 H HETATM 61 C 0 29.888 6.836 11.852 C HETATM 62 C 0 28.180 7.348 10.155 C HETATM 63 H 0 27.375 7.990 9.733 H HETATM 64 C 0 32.402 2.712 11.133 C HETATM 65 H 0 32.572 1.828 10.481 H HETATM 66 C 0 28.674 -1.475 10.207 C HETATM 67 H 0 28.078 -2.281 10.692 H END CONECT 2 21 51 CONECT 3 22 29 CONECT 4 6 64 CONECT 5 10 46 CONECT 6 4 15 30 CONECT 7 20 38 CONECT 8 15 57 CONECT 9 36 46 47 CONECT 10 5 11 44 CONECT 11 10 CONECT 12 13 26 30 CONECT 13 12 CONECT 14 20 31 49 CONECT 15 6 8 61 CONECT 16 17 33 35 CONECT 17 16 CONECT 18 19 30 40 327 Table 6.S33 (cont’d) CONECT 19 18 CONECT 20 14 7 29 CONECT 21 2 28 46 CONECT 22 3 23 33 CONECT 23 22 CONECT 24 25 38 49 CONECT 25 24 CONECT 26 12 27 64 CONECT 27 26 CONECT 28 21 42 59 CONECT 29 3 20 35 CONECT 30 6 12 18 CONECT 31 14 32 66 CONECT 32 31 CONECT 33 16 22 34 CONECT 34 33 CONECT 35 16 29 66 CONECT 36 9 37 42 CONECT 37 36 CONECT 38 24 7 39 CONECT 39 38 CONECT 40 18 41 61 CONECT 41 40 CONECT 42 28 36 43 CONECT 43 42 CONECT 44 10 45 47 CONECT 45 44 CONECT 46 9 5 21 CONECT 47 9 44 48 CONECT 48 47 CONECT 49 14 24 50 CONECT 50 49 CONECT 51 2 52 53 CONECT 52 51 CONECT 53 51 54 59 CONECT 54 53 CONECT 55 56 61 62 CONECT 56 55 CONECT 57 8 58 62 CONECT 58 57 CONECT 59 28 53 60 CONECT 60 59 CONECT 61 55 15 40 CONECT 62 57 55 63 CONECT 63 62 CONECT 64 4 26 65 CONECT 65 64 CONECT 66 35 31 67 CONECT 67 66 Table 6.S34: The optimized geometry file for the lowest energy triplet state of [Fe(phen)3]2+. TITLE Fe-phen Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 Fe 0 30.213 3.504 9.050 Fe HETATM 2 N 0 28.933 3.655 7.337 N HETATM 3 N 0 28.989 2.230 10.076 N HETATM 4 N 0 31.439 3.698 10.833 N HETATM 5 N 0 31.412 4.635 7.853 N HETATM 6 C 0 31.203 4.824 11.623 C HETATM 7 N 0 31.017 1.517 8.367 N HETATM 8 N 0 29.459 5.456 10.024 N HETATM 9 C 0 31.569 5.685 5.582 C HETATM 10 C 0 32.660 5.127 8.132 C HETATM 11 H 0 33.073 4.894 9.137 H HETATM 12 C 0 32.946 4.129 13.240 C HETATM 13 H 0 33.534 4.298 14.171 H HETATM 14 C 0 30.518 -0.917 8.638 C HETATM 15 C 0 30.163 5.770 11.183 C 328 Table 6.S34 (cont’d) HETATM 16 C 0 27.353 0.256 11.387 C HETATM 17 H 0 26.716 -0.505 11.893 H HETATM 18 C 0 31.655 6.310 13.608 C HETATM 19 H 0 32.236 6.499 14.540 H HETATM 20 C 0 30.270 0.482 8.926 C HETATM 21 C 0 29.540 4.385 6.313 C HETATM 22 C 0 27.995 2.591 10.947 C HETATM 23 H 0 27.854 3.681 11.109 H HETATM 24 C 0 32.340 -0.165 7.152 C HETATM 25 H 0 33.172 -0.363 6.441 H HETATM 26 C 0 33.164 2.986 12.440 C HETATM 27 H 0 33.925 2.222 12.714 H HETATM 28 C 0 28.910 4.627 5.032 C HETATM 29 C 0 29.195 0.861 9.848 C HETATM 30 C 0 31.944 5.090 12.842 C HETATM 31 C 0 29.686 -1.938 9.287 C HETATM 32 H 0 29.885 -3.010 9.060 H HETATM 33 C 0 27.161 1.636 11.618 C HETATM 34 H 0 26.372 2.003 12.311 H HETATM 35 C 0 28.393 -0.167 10.480 C HETATM 36 C 0 30.911 5.917 4.289 C HETATM 37 H 0 31.455 6.510 3.518 H HETATM 38 C 0 32.022 1.193 7.503 C HETATM 39 H 0 32.601 2.041 7.074 H HETATM 40 C 0 30.682 7.213 13.183 C HETATM 41 H 0 30.473 8.138 13.768 H HETATM 42 C 0 29.637 5.410 4.024 C HETATM 43 H 0 29.148 5.592 3.040 H HETATM 44 C 0 33.420 5.900 7.191 C HETATM 45 H 0 34.428 6.266 7.488 H HETATM 46 C 0 30.872 4.914 6.589 C HETATM 47 C 0 32.882 6.184 5.916 C HETATM 48 H 0 33.455 6.782 5.170 H HETATM 49 C 0 31.590 -1.221 7.718 C HETATM 50 H 0 31.817 -2.282 7.465 H HETATM 51 C 0 27.684 3.149 7.113 C HETATM 52 H 0 27.222 2.572 7.944 H HETATM 53 C 0 26.981 3.341 5.874 C HETATM 54 H 0 25.965 2.904 5.755 H HETATM 55 C 0 28.881 7.865 11.478 C HETATM 56 H 0 28.653 8.798 12.044 H HETATM 57 C 0 28.500 6.326 9.598 C HETATM 58 H 0 27.953 6.047 8.669 H HETATM 59 C 0 27.590 4.077 4.832 C HETATM 60 H 0 27.064 4.238 3.863 H HETATM 61 C 0 29.907 6.969 11.960 C HETATM 62 C 0 28.180 7.543 10.295 C HETATM 63 H 0 27.384 8.210 9.895 H HETATM 64 C 0 32.391 2.807 11.243 C HETATM 65 H 0 32.554 1.921 10.591 H HETATM 66 C 0 28.671 -1.577 10.172 C HETATM 67 H 0 28.046 -2.357 10.666 H END CONECT 2 21 51 CONECT 3 22 29 CONECT 4 6 64 CONECT 5 10 46 CONECT 6 4 15 30 CONECT 7 20 38 CONECT 8 15 57 CONECT 9 36 46 47 CONECT 10 5 11 44 CONECT 11 10 CONECT 12 13 26 30 CONECT 13 12 CONECT 14 20 31 49 CONECT 15 6 8 61 CONECT 16 17 33 35 329 Table 6.S34 (cont’d) CONECT 17 16 CONECT 18 19 30 40 CONECT 19 18 CONECT 20 14 7 29 CONECT 21 2 28 46 CONECT 22 3 23 33 CONECT 23 22 CONECT 24 25 38 49 CONECT 25 24 CONECT 26 12 27 64 CONECT 27 26 CONECT 28 21 42 59 CONECT 29 3 20 35 CONECT 30 6 12 18 CONECT 31 14 32 66 CONECT 32 31 CONECT 33 16 22 34 CONECT 34 33 CONECT 35 16 29 66 CONECT 36 9 37 42 CONECT 37 36 CONECT 38 24 7 39 CONECT 39 38 CONECT 40 18 41 61 CONECT 41 40 CONECT 42 28 36 43 CONECT 43 42 CONECT 44 10 45 47 CONECT 45 44 CONECT 46 9 5 21 CONECT 47 9 44 48 CONECT 48 47 CONECT 49 14 24 50 CONECT 50 49 CONECT 51 2 52 53 CONECT 52 51 CONECT 53 51 54 59 CONECT 54 53 CONECT 55 56 61 62 CONECT 56 55 CONECT 57 8 58 62 CONECT 58 57 CONECT 59 28 53 60 CONECT 60 59 CONECT 61 55 15 40 CONECT 62 55 57 63 CONECT 63 62 CONECT 64 4 26 65 CONECT 65 64 CONECT 66 35 31 67 CONECT 67 66 Table 6.S35: The optimized geometry file for the lowest energy quintet state of [Fe(phen)3]2+. TITLE Fe-phen Quintet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 Fe 0 30.200 3.534 9.115 Fe HETATM 2 N 0 28.945 3.765 7.260 N HETATM 3 N 0 28.783 2.102 10.130 N HETATM 4 N 0 31.554 3.762 10.903 N HETATM 5 N 0 31.526 4.691 7.712 N HETATM 6 C 0 31.285 4.880 11.692 C HETATM 7 N 0 30.918 1.480 8.461 N HETATM 8 N 0 29.482 5.443 10.121 N HETATM 9 C 0 31.631 5.697 5.429 C HETATM 10 C 0 32.790 5.154 7.944 C HETATM 11 H 0 33.225 4.924 8.943 H HETATM 12 C 0 33.079 4.258 13.279 C HETATM 13 H 0 33.677 4.454 14.199 H 330 Table 6.S35 (cont’d) HETATM 14 C 0 30.469 -0.965 8.697 C HETATM 15 C 0 30.197 5.781 11.269 C HETATM 16 C 0 27.200 0.063 11.376 C HETATM 17 H 0 26.580 -0.730 11.855 H HETATM 18 C 0 31.712 6.388 13.667 C HETATM 19 H 0 32.301 6.605 14.587 H HETATM 20 C 0 30.175 0.425 8.990 C HETATM 21 C 0 29.578 4.461 6.230 C HETATM 22 C 0 27.764 2.420 10.980 C HETATM 23 H 0 27.584 3.506 11.150 H HETATM 24 C 0 32.319 -0.163 7.275 C HETATM 25 H 0 33.178 -0.335 6.590 H HETATM 26 C 0 33.330 3.119 12.482 C HETATM 27 H 0 34.124 2.388 12.748 H HETATM 28 C 0 28.942 4.707 4.948 C HETATM 29 C 0 29.053 0.755 9.888 C HETATM 30 C 0 32.035 5.180 12.897 C HETATM 31 C 0 29.653 -2.021 9.308 C HETATM 32 H 0 29.895 -3.083 9.071 H HETATM 33 C 0 26.949 1.430 11.627 C HETATM 34 H 0 26.132 1.755 12.308 H HETATM 35 C 0 28.275 -0.313 10.486 C HETATM 36 C 0 30.957 5.931 4.145 C HETATM 37 H 0 31.503 6.498 3.357 H HETATM 38 C 0 31.955 1.183 7.626 C HETATM 39 H 0 32.527 2.048 7.220 H HETATM 40 C 0 30.693 7.248 13.258 C HETATM 41 H 0 30.455 8.165 13.845 H HETATM 42 C 0 29.668 5.454 3.914 C HETATM 43 H 0 29.163 5.633 2.937 H HETATM 44 C 0 33.544 5.897 6.973 C HETATM 45 H 0 34.570 6.244 7.228 H HETATM 46 C 0 30.942 4.959 6.472 C HETATM 47 C 0 32.966 6.169 5.713 C HETATM 48 H 0 33.525 6.740 4.937 H HETATM 49 C 0 31.577 -1.239 7.810 C HETATM 50 H 0 31.836 -2.293 7.560 H HETATM 51 C 0 27.681 3.294 7.041 C HETATM 52 H 0 27.207 2.742 7.883 H HETATM 53 C 0 26.976 3.487 5.805 C HETATM 54 H 0 25.948 3.078 5.693 H HETATM 55 C 0 28.832 7.819 11.594 C HETATM 56 H 0 28.577 8.741 12.167 H HETATM 57 C 0 28.477 6.273 9.717 C HETATM 58 H 0 27.930 5.975 8.794 H HETATM 59 C 0 27.605 4.195 4.756 C HETATM 60 H 0 27.085 4.362 3.785 H HETATM 61 C 0 29.906 6.967 12.051 C HETATM 62 C 0 28.119 7.471 10.426 C HETATM 63 H 0 27.286 8.104 10.045 H HETATM 64 C 0 32.545 2.909 11.297 C HETATM 65 H 0 32.728 2.027 10.643 H HETATM 66 C 0 28.603 -1.709 10.170 C HETATM 67 H 0 27.991 -2.515 10.637 H END CONECT 2 21 51 CONECT 3 22 29 CONECT 4 6 64 CONECT 5 10 46 CONECT 6 4 15 30 CONECT 7 20 38 CONECT 8 15 57 CONECT 9 36 46 47 CONECT 10 5 11 44 CONECT 11 10 CONECT 12 13 26 30 CONECT 13 12 CONECT 14 20 31 49 331 Table 6.S35 (cont’d) CONECT 15 6 8 61 CONECT 16 17 33 35 CONECT 17 16 CONECT 18 19 30 40 CONECT 19 18 CONECT 20 14 7 29 CONECT 21 2 28 46 CONECT 22 3 23 33 CONECT 23 22 CONECT 24 25 38 49 CONECT 25 24 CONECT 26 12 27 64 CONECT 27 26 CONECT 28 21 42 59 CONECT 29 3 20 35 CONECT 30 6 12 18 CONECT 31 14 32 66 CONECT 32 31 CONECT 33 16 22 34 CONECT 34 33 CONECT 35 16 29 66 CONECT 36 9 37 42 CONECT 37 36 CONECT 38 24 7 39 CONECT 39 38 CONECT 40 18 41 61 CONECT 41 40 CONECT 42 28 36 43 CONECT 43 42 CONECT 44 10 45 47 CONECT 45 44 CONECT 46 9 5 21 CONECT 47 9 44 48 CONECT 48 47 CONECT 49 14 24 50 CONECT 50 49 CONECT 51 2 52 53 CONECT 52 51 CONECT 53 51 54 59 CONECT 54 53 CONECT 55 56 61 62 CONECT 56 55 CONECT 57 8 58 62 CONECT 58 57 CONECT 59 28 53 60 CONECT 60 59 CONECT 61 55 15 40 CONECT 62 55 57 63 CONECT 63 62 CONECT 64 4 26 65 CONECT 65 64 CONECT 66 35 31 67 CONECT 67 66 Table 6.S36: The optimized geometry file for the singlet charge transfer state of [Fe(phen)3]2+ using td=1. TITLE Fe-phen CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 Fe 0 30.201 3.525 9.070 Fe HETATM 2 N 0 29.014 3.624 7.393 N HETATM 3 N 0 28.946 2.333 10.167 N HETATM 4 N 0 31.443 3.567 10.699 N HETATM 5 N 0 31.442 4.672 7.911 N HETATM 6 C 0 31.190 4.699 11.523 C HETATM 7 N 0 30.876 1.703 8.394 N HETATM 8 N 0 29.491 5.268 9.881 N HETATM 9 C 0 31.584 5.742 5.655 C 332 Table 6.S36 (cont’d) HETATM 10 C 0 32.659 5.206 8.228 C HETATM 11 H 0 33.059 4.976 9.238 H HETATM 12 C 0 32.873 3.951 13.178 C HETATM 13 H 0 33.447 4.092 14.122 H HETATM 14 C 0 30.473 -0.752 8.660 C HETATM 15 C 0 30.192 5.618 11.067 C HETATM 16 C 0 27.405 0.350 11.554 C HETATM 17 H 0 26.799 -0.413 12.095 H HETATM 18 C 0 31.607 6.163 13.512 C HETATM 19 H 0 32.155 6.356 14.464 H HETATM 20 C 0 30.193 0.631 8.974 C HETATM 21 C 0 29.600 4.369 6.366 C HETATM 22 C 0 27.995 2.700 11.077 C HETATM 23 H 0 27.852 3.789 11.239 H HETATM 24 C 0 32.201 0.086 7.110 C HETATM 25 H 0 33.011 -0.073 6.365 H HETATM 26 C 0 33.085 2.787 12.356 C HETATM 27 H 0 33.827 2.009 12.641 H HETATM 28 C 0 28.970 4.582 5.083 C HETATM 29 C 0 29.156 0.971 9.938 C HETATM 30 C 0 31.914 4.927 12.770 C HETATM 31 C 0 29.695 -1.796 9.339 C HETATM 32 H 0 29.911 -2.861 9.096 H HETATM 33 C 0 27.207 1.730 11.787 C HETATM 34 H 0 26.444 2.089 12.512 H HETATM 35 C 0 28.407 -0.070 10.603 C HETATM 36 C 0 30.928 5.949 4.356 C HETATM 37 H 0 31.452 6.564 3.591 H HETATM 38 C 0 31.856 1.430 7.482 C HETATM 39 H 0 32.389 2.297 7.037 H HETATM 40 C 0 30.665 7.081 13.046 C HETATM 41 H 0 30.459 8.015 13.620 H HETATM 42 C 0 29.678 5.391 4.081 C HETATM 43 H 0 29.191 5.554 3.093 H HETATM 44 C 0 33.400 6.017 7.301 C HETATM 45 H 0 34.388 6.420 7.614 H HETATM 46 C 0 30.908 4.941 6.649 C HETATM 47 C 0 32.872 6.286 6.017 C HETATM 48 H 0 33.436 6.911 5.288 H HETATM 49 C 0 31.516 -1.005 7.693 C HETATM 50 H 0 31.774 -2.053 7.417 H HETATM 51 C 0 27.787 3.066 7.170 C HETATM 52 H 0 27.339 2.481 8.001 H HETATM 53 C 0 27.088 3.228 5.925 C HETATM 54 H 0 26.089 2.754 5.806 H HETATM 55 C 0 28.924 7.735 11.272 C HETATM 56 H 0 28.680 8.682 11.805 H HETATM 57 C 0 28.562 6.166 9.391 C HETATM 58 H 0 28.032 5.875 8.458 H HETATM 59 C 0 27.672 3.982 4.882 C HETATM 60 H 0 27.142 4.119 3.911 H HETATM 61 C 0 29.911 6.849 11.801 C HETATM 62 C 0 28.252 7.388 10.046 C HETATM 63 H 0 27.474 8.051 9.609 H HETATM 64 C 0 32.359 2.635 11.145 C HETATM 65 H 0 32.533 1.754 10.490 H HETATM 66 C 0 28.708 -1.469 10.272 C HETATM 67 H 0 28.128 -2.270 10.785 H END CONECT 2 21 51 CONECT 3 22 29 CONECT 4 6 64 CONECT 5 10 46 CONECT 6 4 15 30 CONECT 7 20 38 CONECT 8 15 57 CONECT 9 36 46 47 CONECT 10 5 11 44 333 Table 6.S36 (cont’d) CONECT 11 10 CONECT 12 13 26 30 CONECT 13 12 CONECT 14 20 31 49 CONECT 15 6 8 61 CONECT 16 17 33 35 CONECT 17 16 CONECT 18 19 30 40 CONECT 19 18 CONECT 20 14 7 29 CONECT 21 2 28 46 CONECT 22 3 23 33 CONECT 23 22 CONECT 24 25 38 49 CONECT 25 24 CONECT 26 12 27 64 CONECT 27 26 CONECT 28 21 42 59 CONECT 29 3 20 35 CONECT 30 6 12 18 CONECT 31 14 32 66 CONECT 32 31 CONECT 33 16 22 34 CONECT 34 33 CONECT 35 16 29 66 CONECT 36 9 37 42 CONECT 37 36 CONECT 38 24 7 39 CONECT 39 38 CONECT 40 18 41 61 CONECT 41 40 CONECT 42 28 36 43 CONECT 43 42 CONECT 44 10 45 47 CONECT 45 44 CONECT 46 9 5 21 CONECT 47 9 44 48 CONECT 48 47 CONECT 49 14 24 50 CONECT 50 49 CONECT 51 2 52 53 CONECT 52 51 CONECT 53 51 54 59 CONECT 54 53 CONECT 55 56 61 62 CONECT 56 55 CONECT 57 8 58 62 CONECT 58 57 CONECT 59 28 53 60 CONECT 60 59 CONECT 61 55 15 40 CONECT 62 55 57 63 CONECT 63 62 CONECT 64 4 26 65 CONECT 65 64 CONECT 66 35 31 67 CONECT 67 66 Table 6.S37: The optimized geometry file for the singlet charge transfer state of [Fe(phen)3]2+ using td=20. TITLE Fe-phen CT s20 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 Fe 0 30.205 3.540 9.095 Fe HETATM 2 N 0 28.989 3.661 7.438 N HETATM 3 N 0 28.899 2.340 10.140 N HETATM 4 N 0 31.452 3.595 10.732 N HETATM 5 N 0 31.481 4.632 7.904 N 334 Table 6.S37 (cont’d) HETATM 6 C 0 31.218 4.713 11.532 C HETATM 7 N 0 30.917 1.725 8.434 N HETATM 8 N 0 29.494 5.289 9.918 N HETATM 9 C 0 31.618 5.656 5.624 C HETATM 10 C 0 32.735 5.113 8.186 C HETATM 11 H 0 33.152 4.884 9.189 H HETATM 12 C 0 32.986 4.027 13.118 C HETATM 13 H 0 33.588 4.189 14.041 H HETATM 14 C 0 30.507 -0.732 8.672 C HETATM 15 C 0 30.172 5.631 11.087 C HETATM 16 C 0 27.293 0.342 11.429 C HETATM 17 H 0 26.664 -0.429 11.932 H HETATM 18 C 0 31.646 6.197 13.503 C HETATM 19 H 0 32.221 6.407 14.434 H HETATM 20 C 0 30.209 0.654 8.978 C HETATM 21 C 0 29.590 4.368 6.398 C HETATM 22 C 0 27.893 2.698 11.002 C HETATM 23 H 0 27.729 3.783 11.168 H HETATM 24 C 0 32.300 0.119 7.195 C HETATM 25 H 0 33.138 -0.035 6.480 H HETATM 26 C 0 33.205 2.896 12.302 C HETATM 27 H 0 33.979 2.140 12.561 H HETATM 28 C 0 28.947 4.597 5.118 C HETATM 29 C 0 29.129 0.985 9.906 C HETATM 30 C 0 31.964 4.990 12.746 C HETATM 31 C 0 29.709 -1.777 9.306 C HETATM 32 H 0 29.941 -2.841 9.069 H HETATM 33 C 0 27.074 1.717 11.660 C HETATM 34 H 0 26.273 2.066 12.349 H HETATM 35 C 0 28.349 -0.071 10.522 C HETATM 36 C 0 30.951 5.870 4.344 C HETATM 37 H 0 31.485 6.451 3.557 H HETATM 38 C 0 31.941 1.461 7.559 C HETATM 39 H 0 32.488 2.333 7.142 H HETATM 40 C 0 30.638 7.081 13.074 C HETATM 41 H 0 30.407 7.998 13.663 H HETATM 42 C 0 29.661 5.360 4.100 C HETATM 43 H 0 29.163 5.532 3.119 H HETATM 44 C 0 33.488 5.876 7.229 C HETATM 45 H 0 34.499 6.240 7.513 H HETATM 46 C 0 30.928 4.900 6.652 C HETATM 47 C 0 32.942 6.150 5.956 C HETATM 48 H 0 33.515 6.740 5.204 H HETATM 49 C 0 31.594 -0.976 7.741 C HETATM 50 H 0 31.864 -2.022 7.468 H HETATM 51 C 0 27.732 3.145 7.245 C HETATM 52 H 0 27.277 2.587 8.090 H HETATM 53 C 0 27.023 3.320 6.008 C HETATM 54 H 0 26.007 2.880 5.908 H HETATM 55 C 0 28.820 7.683 11.349 C HETATM 56 H 0 28.552 8.613 11.900 H HETATM 57 C 0 28.502 6.121 9.463 C HETATM 58 H 0 27.979 5.825 8.530 H HETATM 59 C 0 27.617 4.038 4.947 C HETATM 60 H 0 27.078 4.182 3.982 H HETATM 61 C 0 29.874 6.823 11.858 C HETATM 62 C 0 28.145 7.326 10.161 C HETATM 63 H 0 27.334 7.963 9.746 H HETATM 64 C 0 32.423 2.703 11.112 C HETATM 65 H 0 32.587 1.822 10.457 H HETATM 66 C 0 28.668 -1.458 10.199 C HETATM 67 H 0 28.067 -2.266 10.676 H END CONECT 2 21 51 CONECT 3 22 29 CONECT 4 6 64 CONECT 5 10 46 CONECT 6 4 15 30 335 Table 6.S37 (cont’d) CONECT 7 20 38 CONECT 8 15 57 CONECT 9 36 46 47 CONECT 10 5 11 44 CONECT 11 10 CONECT 12 13 26 30 CONECT 13 12 CONECT 14 20 31 49 CONECT 15 6 8 61 CONECT 16 17 33 35 CONECT 17 16 CONECT 18 19 30 40 CONECT 19 18 CONECT 20 14 7 29 CONECT 21 2 28 46 CONECT 22 3 23 33 CONECT 23 22 CONECT 24 25 38 49 CONECT 25 24 CONECT 26 12 27 64 CONECT 27 26 CONECT 28 21 42 59 CONECT 29 3 20 35 CONECT 30 6 12 18 CONECT 31 14 32 66 CONECT 32 31 CONECT 33 16 22 34 CONECT 34 33 CONECT 35 16 29 66 CONECT 36 9 37 42 CONECT 37 36 CONECT 38 24 7 39 CONECT 39 38 CONECT 40 18 41 61 CONECT 41 40 CONECT 42 28 36 43 CONECT 43 42 CONECT 44 10 45 47 CONECT 45 44 CONECT 46 9 5 21 CONECT 47 9 44 48 CONECT 48 47 CONECT 49 14 24 50 CONECT 50 49 CONECT 51 2 52 53 CONECT 52 51 CONECT 53 51 54 59 CONECT 54 53 CONECT 55 56 61 62 CONECT 56 55 CONECT 57 8 58 62 CONECT 58 57 CONECT 59 28 53 60 CONECT 60 59 CONECT 61 55 15 40 CONECT 62 57 55 63 CONECT 63 62 CONECT 64 4 26 65 CONECT 65 64 CONECT 66 35 31 67 CONECT 67 66 336 [Fe(4,4’-diCl-bpy)3]2+: Table 6.S38: The optimized geometry file for the lowest energy triplet state of [Fe(4,4’-diCl- bpy)3]2+. TITLE 4,4'-Cl-bpy Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.690 0.829 -1.220 N HETATM 2 N 0 1.638 -0.746 0.957 N HETATM 3 N 0 -0.191 -1.914 -1.048 N HETATM 4 N 0 -1.581 -1.056 1.131 N HETATM 5 N 0 -1.451 1.018 -1.000 N HETATM 6 N 0 -0.072 1.861 1.135 N HETATM 7 C 0 1.617 1.653 -2.320 C HETATM 8 H 0 0.599 1.883 -2.703 H HETATM 9 C 0 2.762 2.198 -2.961 C HETATM 10 H 0 2.653 2.856 -3.850 H HETATM 11 C 0 4.037 1.869 -2.419 C HETATM 12 C 0 4.143 1.022 -1.282 C HETATM 13 H 0 5.142 0.775 -0.866 H HETATM 14 C 0 2.939 0.512 -0.706 C HETATM 15 C 0 2.912 -0.409 0.493 C HETATM 16 C 0 4.093 -0.916 1.112 C HETATM 17 H 0 5.101 -0.648 0.734 H HETATM 18 C 0 3.961 -1.785 2.229 C HETATM 19 C 0 2.671 -2.138 2.712 C HETATM 20 H 0 2.532 -2.815 3.582 H HETATM 21 C 0 1.543 -1.596 2.042 C HETATM 22 H 0 0.517 -1.841 2.386 H HETATM 23 C 0 0.606 -2.324 -2.093 C HETATM 24 H 0 1.319 -1.571 -2.497 H HETATM 25 C 0 0.547 -3.625 -2.660 C HETATM 26 H 0 1.209 -3.904 -3.507 H HETATM 27 C 0 -0.381 -4.544 -2.095 C HETATM 28 C 0 -1.215 -4.151 -1.013 C HETATM 29 H 0 -1.927 -4.877 -0.568 H HETATM 30 C 0 -1.100 -2.815 -0.515 C HETATM 31 C 0 -1.951 -2.293 0.630 C HETATM 32 C 0 -3.066 -3.018 1.157 C HETATM 33 H 0 -3.374 -3.999 0.737 H HETATM 34 C 0 -3.794 -2.445 2.234 C HETATM 35 C 0 -3.419 -1.180 2.767 C HETATM 36 H 0 -3.968 -0.711 3.611 H HETATM 37 C 0 -2.306 -0.525 2.172 C HETATM 38 H 0 -1.975 0.467 2.552 H HETATM 39 C 0 -2.142 0.509 -2.082 C HETATM 40 H 0 -1.851 -0.508 -2.420 H HETATM 41 C 0 -3.170 1.218 -2.756 C HETATM 42 H 0 -3.692 0.757 -3.622 H HETATM 43 C 0 -3.499 2.519 -2.284 C HETATM 44 C 0 -2.809 3.071 -1.171 C HETATM 45 H 0 -3.067 4.085 -0.802 H HETATM 46 C 0 -1.786 2.296 -0.546 C HETATM 47 C 0 -0.994 2.779 0.645 C HETATM 48 C 0 -1.163 4.067 1.239 C HETATM 49 H 0 -1.898 4.798 0.843 H HETATM 50 C 0 -0.360 4.399 2.364 C HETATM 51 C 0 0.587 3.468 2.878 C HETATM 52 H 0 1.225 3.704 3.756 H HETATM 53 C 0 0.694 2.212 2.224 C HETATM 54 H 0 1.418 1.451 2.588 H HETATM 55 Fe 0 0.035 0.043 -0.019 Fe HETATM 56 Cl 0 -0.543 6.004 3.135 Cl HETATM 57 Cl 0 -4.785 3.468 -3.090 Cl HETATM 58 Cl 0 -5.197 -3.324 2.920 Cl HETATM 59 Cl 0 -0.499 -6.210 -2.742 Cl HETATM 60 Cl 0 5.427 -2.435 3.024 Cl HETATM 61 Cl 0 5.523 2.528 -3.171 Cl END CONECT 1 7 14 337 Table 6.S38 (cont’d) CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 59 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 58 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 57 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 56 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 50 CONECT 57 43 CONECT 58 34 CONECT 59 27 CONECT 60 18 CONECT 61 11 Table 6.S39: The optimized geometry file for the singlet charge transfer state of [Fe(4,4’-diCl- bpy)3]2+ using td=1. TITLE 4,4'-Cl-bpy CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.518 0.774 -1.089 N HETATM 2 N 0 1.504 -0.796 1.089 N HETATM 3 N 0 -0.084 -1.772 -1.008 N 338 Table 6.S39 (cont’d) HETATM 4 N 0 -1.520 -0.925 1.055 N HETATM 5 N 0 -1.505 0.951 -1.058 N HETATM 6 N 0 -0.059 1.777 1.006 N HETATM 7 C 0 1.437 1.649 -2.154 C HETATM 8 H 0 0.418 1.884 -2.529 H HETATM 9 C 0 2.563 2.217 -2.789 C HETATM 10 H 0 2.442 2.892 -3.661 H HETATM 11 C 0 3.862 1.875 -2.255 C HETATM 12 C 0 3.995 1.008 -1.160 C HETATM 13 H 0 4.997 0.775 -0.744 H HETATM 14 C 0 2.800 0.437 -0.566 C HETATM 15 C 0 2.792 -0.482 0.566 C HETATM 16 C 0 3.976 -1.076 1.159 C HETATM 17 H 0 4.983 -0.861 0.742 H HETATM 18 C 0 3.828 -1.940 2.253 C HETATM 19 C 0 2.523 -2.258 2.788 C HETATM 20 H 0 2.390 -2.931 3.660 H HETATM 21 C 0 1.407 -1.669 2.154 C HETATM 22 H 0 0.385 -1.887 2.529 H HETATM 23 C 0 0.736 -2.124 -2.057 C HETATM 24 H 0 1.451 -1.353 -2.410 H HETATM 25 C 0 0.682 -3.400 -2.678 C HETATM 26 H 0 1.361 -3.636 -3.525 H HETATM 27 C 0 -0.256 -4.346 -2.179 C HETATM 28 C 0 -1.112 -4.006 -1.094 C HETATM 29 H 0 -1.839 -4.747 -0.701 H HETATM 30 C 0 -1.003 -2.703 -0.529 C HETATM 31 C 0 -1.843 -2.209 0.620 C HETATM 32 C 0 -2.888 -2.964 1.228 C HETATM 33 H 0 -3.147 -3.982 0.869 H HETATM 34 C 0 -3.602 -2.380 2.311 C HETATM 35 C 0 -3.273 -1.073 2.768 C HETATM 36 H 0 -3.810 -0.594 3.614 H HETATM 37 C 0 -2.226 -0.382 2.105 C HETATM 38 H 0 -1.935 0.638 2.434 H HETATM 39 C 0 -2.216 0.418 -2.111 C HETATM 40 H 0 -1.935 -0.603 -2.443 H HETATM 41 C 0 -3.255 1.123 -2.773 C HETATM 42 H 0 -3.797 0.652 -3.621 H HETATM 43 C 0 -3.571 2.432 -2.311 C HETATM 44 C 0 -2.851 3.005 -1.226 C HETATM 45 H 0 -3.100 4.025 -0.865 H HETATM 46 C 0 -1.815 2.237 -0.620 C HETATM 47 C 0 -0.969 2.719 0.531 C HETATM 48 C 0 -1.064 4.022 1.098 C HETATM 49 H 0 -1.784 4.771 0.709 H HETATM 50 C 0 -0.201 4.351 2.182 C HETATM 51 C 0 0.730 3.395 2.675 C HETATM 52 H 0 1.414 3.624 3.520 H HETATM 53 C 0 0.769 2.120 2.052 C HETATM 54 H 0 1.478 1.341 2.401 H HETATM 55 Fe 0 -0.029 0.002 -0.001 Fe HETATM 56 Cl 0 -0.290 5.973 2.917 Cl HETATM 57 Cl 0 -4.875 3.362 -3.093 Cl HETATM 58 Cl 0 -4.916 -3.294 3.095 Cl HETATM 59 Cl 0 -0.363 -5.969 -2.911 Cl HETATM 60 Cl 0 5.282 -2.694 3.004 Cl HETATM 61 Cl 0 5.329 2.601 -3.007 Cl END CONECT 1 7 14 CONECT 2 15 21 CONECT 3 23 30 CONECT 4 31 37 CONECT 5 39 46 CONECT 6 47 53 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 11 339 Table 6.S39 (cont’d) CONECT 10 9 CONECT 11 9 12 61 CONECT 12 11 13 14 CONECT 13 12 CONECT 14 1 12 15 CONECT 15 2 14 16 CONECT 16 15 17 18 CONECT 17 16 CONECT 18 16 19 60 CONECT 19 18 20 21 CONECT 20 19 CONECT 21 2 19 22 CONECT 22 21 CONECT 23 3 24 25 CONECT 24 23 CONECT 25 23 26 27 CONECT 26 25 CONECT 27 25 28 59 CONECT 28 27 29 30 CONECT 29 28 CONECT 30 28 3 31 CONECT 31 4 30 32 CONECT 32 31 33 34 CONECT 33 32 CONECT 34 32 35 58 CONECT 35 34 36 37 CONECT 36 35 CONECT 37 35 4 38 CONECT 38 37 CONECT 39 5 40 41 CONECT 40 39 CONECT 41 39 42 43 CONECT 42 41 CONECT 43 41 44 57 CONECT 44 43 45 46 CONECT 45 44 CONECT 46 44 5 47 CONECT 47 6 46 48 CONECT 48 47 49 50 CONECT 49 48 CONECT 50 48 51 56 CONECT 51 50 52 53 CONECT 52 51 CONECT 53 6 51 54 CONECT 54 53 CONECT 56 50 CONECT 57 43 CONECT 58 34 CONECT 59 27 CONECT 60 18 CONECT 61 11 [Fe(4,4’-diOMe-bpy)3]2+: Table 6.S40: The optimized geometry file for the lowest energy triplet state of [Fe(4,4’-diOMe- bpy)3]2+. TITLE 4,4'-OMe-bpy Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.717 0.688 -0.970 N HETATM 2 N 0 1.728 -0.883 1.207 N HETATM 3 N 0 -0.061 -1.878 -1.191 N HETATM 4 N 0 -1.403 -1.046 0.972 N HETATM 5 N 0 -1.488 1.069 -1.140 N HETATM 6 N 0 -0.099 1.870 1.061 N HETATM 7 C 0 1.652 1.532 -2.057 C HETATM 8 H 0 0.631 1.783 -2.415 H HETATM 9 C 0 2.788 2.070 -2.717 C 340 Table 6.S40 (cont’d) HETATM 10 C 0 4.087 1.721 -2.224 C HETATM 11 C 0 4.174 0.847 -1.091 C HETATM 12 H 0 5.180 0.580 -0.702 H HETATM 13 C 0 2.989 0.348 -0.488 C HETATM 14 C 0 2.994 -0.573 0.714 C HETATM 15 C 0 4.178 -1.089 1.305 C HETATM 16 H 0 5.188 -0.851 0.908 H HETATM 17 C 0 4.077 -1.950 2.449 C HETATM 18 C 0 2.775 -2.265 2.962 C HETATM 19 C 0 1.647 -1.710 2.302 C HETATM 20 H 0 0.621 -1.937 2.669 H HETATM 21 C 0 0.681 -2.235 -2.292 C HETATM 22 H 0 1.428 -1.489 -2.643 H HETATM 23 C 0 0.538 -3.470 -2.977 C HETATM 24 C 0 -0.436 -4.404 -2.488 C HETATM 25 C 0 -1.214 -4.047 -1.337 C HETATM 26 H 0 -1.966 -4.772 -0.958 H HETATM 27 C 0 -1.008 -2.787 -0.717 C HETATM 28 C 0 -1.781 -2.308 0.493 C HETATM 29 C 0 -2.815 -3.063 1.104 C HETATM 30 H 0 -3.111 -4.062 0.718 H HETATM 31 C 0 -3.507 -2.536 2.244 C HETATM 32 C 0 -3.124 -1.245 2.735 C HETATM 33 C 0 -2.080 -0.553 2.068 C HETATM 34 H 0 -1.760 0.448 2.426 H HETATM 35 C 0 -2.224 0.578 -2.190 C HETATM 36 H 0 -1.910 -0.414 -2.586 H HETATM 37 C 0 -3.324 1.256 -2.782 C HETATM 38 C 0 -3.699 2.528 -2.238 C HETATM 39 C 0 -2.942 3.053 -1.138 C HETATM 40 H 0 -3.242 4.033 -0.708 H HETATM 41 C 0 -1.846 2.309 -0.622 C HETATM 42 C 0 -0.989 2.802 0.536 C HETATM 43 C 0 -1.079 4.126 1.046 C HETATM 44 H 0 -1.772 4.876 0.610 H HETATM 45 C 0 -0.246 4.522 2.144 C HETATM 46 C 0 0.664 3.560 2.695 C HETATM 47 C 0 0.695 2.265 2.111 C HETATM 48 H 0 1.395 1.497 2.510 H HETATM 49 Fe 0 0.100 -0.084 -0.002 Fe HETATM 50 H 0 2.636 2.741 -3.588 H HETATM 51 H 0 2.621 -2.923 3.843 H HETATM 52 H 0 -3.611 -0.768 3.612 H HETATM 53 H 0 1.176 -3.678 -3.861 H HETATM 54 H 0 -3.861 0.786 -3.633 H HETATM 55 H 0 1.339 3.791 3.545 H HETATM 56 O 0 -4.746 3.313 -2.681 O HETATM 57 O 0 -0.396 5.825 2.579 O HETATM 58 O 0 -4.499 -3.335 2.775 O HETATM 59 O 0 -0.701 -5.642 -3.034 O HETATM 60 O 0 5.272 -2.408 2.964 O HETATM 61 O 0 5.289 2.157 -2.747 O HETATM 62 C 0 5.235 -3.309 4.147 C HETATM 63 H 0 6.308 -3.532 4.366 H HETATM 64 H 0 4.762 -2.789 5.022 H HETATM 65 H 0 4.684 -4.256 3.903 H HETATM 66 C 0 5.258 3.070 -3.919 C HETATM 67 H 0 6.333 3.274 -4.149 H HETATM 68 H 0 4.764 2.570 -4.795 H HETATM 69 H 0 4.730 4.027 -3.659 H HETATM 70 C 0 0.061 -6.065 -4.239 C HETATM 71 H 0 -0.332 -7.082 -4.481 H HETATM 72 H 0 -0.134 -5.362 -5.093 H HETATM 73 H 0 1.159 -6.119 -4.012 H HETATM 74 C 0 -5.254 -2.834 3.953 C HETATM 75 H 0 -5.796 -1.884 3.698 H HETATM 76 H 0 -5.987 -3.645 4.187 H HETATM 77 H 0 -4.564 -2.675 4.825 H 341 Table 6.S40 (cont’d) HETATM 78 C 0 -5.569 2.815 -3.814 C HETATM 79 H 0 -6.072 1.849 -3.542 H HETATM 80 H 0 -6.334 3.612 -3.982 H HETATM 81 H 0 -4.937 2.687 -4.734 H HETATM 82 C 0 0.448 6.293 3.709 C HETATM 83 H 0 0.142 7.356 3.871 H HETATM 84 H 0 0.242 5.687 4.632 H HETATM 85 H 0 1.536 6.242 3.436 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 50 CONECT 10 9 11 61 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 60 CONECT 18 17 19 51 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 53 CONECT 24 23 25 59 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 3 25 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 58 CONECT 32 31 33 52 CONECT 33 4 32 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 54 CONECT 38 37 39 56 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 5 39 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 57 CONECT 46 45 47 55 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 9 CONECT 51 18 CONECT 52 32 CONECT 53 23 CONECT 54 37 CONECT 55 46 CONECT 56 38 78 CONECT 57 45 82 CONECT 58 31 74 CONECT 59 24 70 CONECT 60 17 62 342 Table 6.S40 (cont’d) CONECT 61 10 66 CONECT 62 60 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 61 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 56 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 57 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 Table 6.S41: The optimized geometry file for the singlet charge transfer state of [Fe(4,4’-diOMe- bpy)3]2+ using td=1. TITLE 4,4'-OMe-bpy CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.614 0.766 -1.074 N HETATM 2 N 0 1.620 -0.858 1.019 N HETATM 3 N 0 -0.044 -1.805 -1.050 N HETATM 4 N 0 -1.431 -0.937 1.033 N HETATM 5 N 0 -1.393 0.898 -1.100 N HETATM 6 N 0 0.016 1.698 1.042 N HETATM 7 C 0 1.508 1.637 -2.134 C HETATM 8 H 0 0.478 1.881 -2.466 H HETATM 9 C 0 2.629 2.204 -2.794 C HETATM 10 C 0 3.941 1.860 -2.328 C HETATM 11 C 0 4.061 0.956 -1.218 C HETATM 12 H 0 5.075 0.692 -0.850 H HETATM 13 C 0 2.891 0.430 -0.618 C HETATM 14 C 0 2.895 -0.517 0.557 C HETATM 15 C 0 4.068 -1.033 1.158 C HETATM 16 H 0 5.080 -0.765 0.785 H HETATM 17 C 0 3.958 -1.926 2.278 C HETATM 18 C 0 2.649 -2.270 2.753 C HETATM 19 C 0 1.524 -1.716 2.091 C HETATM 20 H 0 0.499 -1.964 2.438 H HETATM 21 C 0 0.734 -2.174 -2.124 C HETATM 22 H 0 1.481 -1.432 -2.475 H HETATM 23 C 0 0.611 -3.421 -2.788 C HETATM 24 C 0 -0.371 -4.352 -2.313 C HETATM 25 C 0 -1.185 -3.978 -1.189 C HETATM 26 H 0 -1.947 -4.695 -0.818 H HETATM 27 C 0 -1.000 -2.711 -0.585 C HETATM 28 C 0 -1.799 -2.212 0.595 C HETATM 29 C 0 -2.840 -2.944 1.215 C HETATM 30 H 0 -3.129 -3.958 0.865 H HETATM 31 C 0 -3.550 -2.364 2.321 C HETATM 32 C 0 -3.176 -1.052 2.764 C HETATM 33 C 0 -2.121 -0.387 2.089 C HETATM 34 H 0 -1.804 0.628 2.407 H HETATM 35 C 0 -2.109 0.399 -2.159 C HETATM 36 H 0 -1.822 -0.615 -2.515 H HETATM 37 C 0 -3.152 1.095 -2.819 C 343 Table 6.S41 (cont’d) HETATM 38 C 0 -3.498 2.415 -2.320 C HETATM 39 C 0 -2.791 2.953 -1.220 C HETATM 40 H 0 -3.077 3.955 -0.832 H HETATM 41 C 0 -1.720 2.196 -0.598 C HETATM 42 C 0 -0.923 2.648 0.532 C HETATM 43 C 0 -1.019 3.958 1.150 C HETATM 44 H 0 -1.727 4.720 0.757 H HETATM 45 C 0 -0.200 4.286 2.256 C HETATM 46 C 0 0.748 3.308 2.762 C HETATM 47 C 0 0.808 2.054 2.104 C HETATM 48 H 0 1.525 1.284 2.465 H HETATM 49 Fe 0 0.064 -0.041 -0.023 Fe HETATM 50 H 0 2.457 2.895 -3.645 H HETATM 51 H 0 2.483 -2.950 3.615 H HETATM 52 H 0 -3.680 -0.540 3.609 H HETATM 53 H 0 1.274 -3.637 -3.651 H HETATM 54 H 0 -3.661 0.621 -3.682 H HETATM 55 H 0 1.412 3.500 3.628 H HETATM 56 O 0 -4.510 3.211 -2.855 O HETATM 57 O 0 -0.361 5.564 2.788 O HETATM 58 O 0 -4.547 -3.137 2.865 O HETATM 59 O 0 -0.614 -5.596 -2.842 O HETATM 60 O 0 5.142 -2.380 2.803 O HETATM 61 O 0 5.122 2.324 -2.853 O HETATM 62 C 0 5.099 -3.314 3.963 C HETATM 63 H 0 6.171 -3.526 4.193 H HETATM 64 H 0 4.603 -2.821 4.841 H HETATM 65 H 0 4.567 -4.261 3.681 H HETATM 66 C 0 5.069 3.270 -4.002 C HETATM 67 H 0 6.140 3.490 -4.235 H HETATM 68 H 0 4.572 2.787 -4.884 H HETATM 69 H 0 4.535 4.213 -3.708 H HETATM 70 C 0 0.192 -6.046 -4.011 C HETATM 71 H 0 -0.193 -7.068 -4.241 H HETATM 72 H 0 0.026 -5.362 -4.886 H HETATM 73 H 0 1.280 -6.093 -3.740 H HETATM 74 C 0 -5.331 -2.592 4.008 C HETATM 75 H 0 -5.869 -1.656 3.701 H HETATM 76 H 0 -6.065 -3.397 4.254 H HETATM 77 H 0 -4.659 -2.398 4.886 H HETATM 78 C 0 -5.295 2.683 -3.996 C HETATM 79 H 0 -5.834 1.740 -3.709 H HETATM 80 H 0 -6.035 3.488 -4.230 H HETATM 81 H 0 -4.635 2.501 -4.887 H HETATM 82 C 0 0.479 5.959 3.943 C HETATM 83 H 0 0.169 7.008 4.178 H HETATM 84 H 0 0.282 5.296 4.828 H HETATM 85 H 0 1.570 5.936 3.672 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 50 CONECT 10 9 11 61 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 60 CONECT 18 17 19 51 CONECT 19 2 18 20 344 Table 6.S41 (cont’d) CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 53 CONECT 24 23 25 59 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 3 25 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 58 CONECT 32 31 33 52 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 54 CONECT 38 37 39 56 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 5 39 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 57 CONECT 46 45 47 55 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 9 CONECT 51 18 CONECT 52 32 CONECT 53 23 CONECT 54 37 CONECT 55 46 CONECT 56 38 78 CONECT 57 45 82 CONECT 58 31 74 CONECT 59 24 70 CONECT 60 17 62 CONECT 61 10 66 CONECT 62 60 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 61 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 56 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 57 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 345 [Fe(5,5’-diCl-bpy)3]2+: Table 6.S42: The optimized geometry file for the lowest energy triplet state of [Fe(5,5’-diCl- bpy)3]2+. TITLE 5,5-Cl'-bpy Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.718 0.760 -1.046 N HETATM 2 N 0 1.730 -0.835 1.167 N HETATM 3 N 0 -0.151 -1.788 -0.971 N HETATM 4 N 0 -1.557 -0.952 1.140 N HETATM 5 N 0 -1.586 1.052 -1.260 N HETATM 6 N 0 -0.186 1.815 0.922 N HETATM 7 C 0 1.637 1.637 -2.101 C HETATM 8 H 0 0.621 1.886 -2.477 H HETATM 9 C 0 2.797 2.198 -2.703 C HETATM 10 C 0 4.087 1.863 -2.205 C HETATM 11 C 0 4.164 0.961 -1.110 C HETATM 12 H 0 5.161 0.699 -0.697 H HETATM 13 C 0 2.968 0.413 -0.545 C HETATM 14 C 0 2.972 -0.553 0.622 C HETATM 15 C 0 4.164 -1.151 1.144 C HETATM 16 H 0 5.158 -0.942 0.694 H HETATM 17 C 0 4.080 -2.041 2.249 C HETATM 18 C 0 2.795 -2.306 2.799 C HETATM 19 C 0 1.641 -1.693 2.232 C HETATM 20 H 0 0.628 -1.890 2.646 H HETATM 21 C 0 0.627 -2.142 -2.052 C HETATM 22 H 0 1.362 -1.395 -2.417 H HETATM 23 C 0 0.499 -3.409 -2.685 C HETATM 24 C 0 -0.449 -4.357 -2.207 C HETATM 25 C 0 -1.246 -3.990 -1.091 C HETATM 26 H 0 -1.994 -4.711 -0.697 H HETATM 27 C 0 -1.087 -2.706 -0.484 C HETATM 28 C 0 -1.881 -2.233 0.701 C HETATM 29 C 0 -2.892 -3.004 1.355 C HETATM 30 H 0 -3.151 -4.025 1.003 H HETATM 31 C 0 -3.578 -2.461 2.475 C HETATM 32 C 0 -3.227 -1.150 2.908 C HETATM 33 C 0 -2.217 -0.421 2.222 C HETATM 34 H 0 -1.929 0.602 2.543 H HETATM 35 C 0 -2.263 0.557 -2.346 C HETATM 36 H 0 -1.953 -0.438 -2.733 H HETATM 37 C 0 -3.318 1.287 -2.962 C HETATM 38 C 0 -3.700 2.557 -2.445 C HETATM 39 C 0 -2.996 3.062 -1.319 C HETATM 40 H 0 -3.285 4.049 -0.899 H HETATM 41 C 0 -1.935 2.292 -0.744 C HETATM 42 C 0 -1.117 2.740 0.442 C HETATM 43 C 0 -1.253 4.026 1.052 C HETATM 44 H 0 -1.988 4.761 0.660 H HETATM 45 C 0 -0.444 4.383 2.163 C HETATM 46 C 0 0.494 3.424 2.635 C HETATM 47 C 0 0.603 2.158 1.999 C HETATM 48 H 0 1.327 1.397 2.356 H HETATM 49 Fe 0 -0.055 0.006 -0.020 Fe HETATM 50 Cl 0 -4.043 -0.399 4.313 Cl HETATM 51 Cl 0 1.547 -3.787 -4.086 Cl HETATM 52 Cl 0 2.604 -3.410 4.198 Cl HETATM 53 Cl 0 2.608 3.325 -4.082 Cl HETATM 54 Cl 0 1.554 3.788 4.031 Cl HETATM 55 Cl 0 -4.154 0.587 -4.384 Cl HETATM 56 H 0 -0.541 5.381 2.645 H HETATM 57 H 0 -4.528 3.138 -2.907 H HETATM 58 H 0 5.007 2.297 -2.654 H HETATM 59 H 0 4.994 -2.517 2.670 H HETATM 60 H 0 -4.367 -3.044 2.999 H HETATM 61 H 0 -0.562 -5.351 -2.691 H END CONECT 1 7 13 346 Table 6.S42 (cont’d) CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 53 CONECT 10 9 11 58 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 59 CONECT 18 17 19 52 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 51 CONECT 24 23 25 61 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 60 CONECT 32 31 33 50 CONECT 33 4 32 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 55 CONECT 38 37 39 57 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 5 39 42 CONECT 42 41 6 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 56 CONECT 46 45 47 54 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 32 CONECT 51 23 CONECT 52 18 CONECT 53 9 CONECT 54 46 CONECT 55 37 CONECT 56 45 CONECT 57 38 CONECT 58 10 CONECT 59 17 CONECT 60 31 CONECT 61 24 Table 6.S43: The optimized geometry file for the singlet charge transfer state of [Fe(5,5’-diCl- bpy)3]2+ using td=1. TITLE 5,5-Cl'-bpy CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.696 0.763 -1.122 N HETATM 2 N 0 1.686 -0.788 1.118 N HETATM 3 N 0 -0.178 -1.799 -0.952 N 347 Table 6.S43 (cont’d) HETATM 4 N 0 -1.524 -0.982 1.214 N HETATM 5 N 0 -1.509 1.002 -1.211 N HETATM 6 N 0 -0.151 1.798 0.954 N HETATM 7 C 0 1.616 1.657 -2.161 C HETATM 8 H 0 0.610 1.854 -2.588 H HETATM 9 C 0 2.771 2.298 -2.688 C HETATM 10 C 0 4.052 2.036 -2.125 C HETATM 11 C 0 4.128 1.114 -1.047 C HETATM 12 H 0 5.113 0.907 -0.575 H HETATM 13 C 0 2.938 0.480 -0.567 C HETATM 14 C 0 2.931 -0.522 0.563 C HETATM 15 C 0 4.113 -1.172 1.043 C HETATM 16 H 0 5.100 -0.977 0.570 H HETATM 17 C 0 4.026 -2.092 2.121 C HETATM 18 C 0 2.742 -2.337 2.685 C HETATM 19 C 0 1.595 -1.681 2.158 C HETATM 20 H 0 0.586 -1.865 2.585 H HETATM 21 C 0 0.604 -2.162 -2.026 C HETATM 22 H 0 1.322 -1.408 -2.408 H HETATM 23 C 0 0.499 -3.447 -2.628 C HETATM 24 C 0 -0.427 -4.402 -2.124 C HETATM 25 C 0 -1.232 -4.023 -1.016 C HETATM 26 H 0 -1.958 -4.752 -0.598 H HETATM 27 C 0 -1.099 -2.720 -0.445 C HETATM 28 C 0 -1.907 -2.229 0.726 C HETATM 29 C 0 -2.992 -2.952 1.315 C HETATM 30 H 0 -3.307 -3.939 0.914 H HETATM 31 C 0 -3.686 -2.400 2.425 C HETATM 32 C 0 -3.271 -1.127 2.913 C HETATM 33 C 0 -2.195 -0.442 2.286 C HETATM 34 H 0 -1.856 0.551 2.652 H HETATM 35 C 0 -2.186 0.473 -2.285 C HETATM 36 H 0 -1.866 -0.526 -2.649 H HETATM 37 C 0 -3.248 1.178 -2.915 C HETATM 38 C 0 -3.640 2.459 -2.430 C HETATM 39 C 0 -2.938 2.999 -1.319 C HETATM 40 H 0 -3.235 3.993 -0.921 H HETATM 41 C 0 -1.869 2.257 -0.726 C HETATM 42 C 0 -1.054 2.734 0.445 C HETATM 43 C 0 -1.166 4.039 1.016 C HETATM 44 H 0 -1.878 4.782 0.596 H HETATM 45 C 0 -0.357 4.404 2.126 C HETATM 46 C 0 0.550 3.433 2.633 C HETATM 47 C 0 0.635 2.147 2.031 C HETATM 48 H 0 1.338 1.380 2.415 H HETATM 49 Fe 0 -0.075 -0.001 0.001 Fe HETATM 50 Cl 0 -4.089 -0.372 4.316 Cl HETATM 51 Cl 0 1.550 -3.838 -4.024 Cl HETATM 52 Cl 0 2.545 -3.469 4.062 Cl HETATM 53 Cl 0 2.590 3.433 -4.065 Cl HETATM 54 Cl 0 1.603 3.805 4.032 Cl HETATM 55 Cl 0 -4.075 0.437 -4.320 Cl HETATM 56 H 0 -0.431 5.416 2.580 H HETATM 57 H 0 -4.475 3.020 -2.905 H HETATM 58 H 0 4.964 2.542 -2.511 H HETATM 59 H 0 4.931 -2.610 2.507 H HETATM 60 H 0 -4.532 -2.946 2.897 H HETATM 61 H 0 -0.518 -5.412 -2.578 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 53 348 Table 6.S43 (cont’d) CONECT 10 9 11 58 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 59 CONECT 18 17 19 52 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 51 CONECT 24 23 25 61 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 27 4 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 60 CONECT 32 31 33 50 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 55 CONECT 38 37 39 57 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 41 6 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 56 CONECT 46 45 47 54 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 32 CONECT 51 23 CONECT 52 18 CONECT 53 9 CONECT 54 46 CONECT 55 37 CONECT 56 45 CONECT 57 38 CONECT 58 10 CONECT 59 17 CONECT 60 31 CONECT 61 24 [Fe(5,5’-diOMe-bpy)3]2+: Table 6.S44: The optimized geometry file for the lowest energy triplet state of [Fe(5,5’-diOMe- bpy)3]2+. TITLE 5,5'-OMe-bpy Triplet B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.651 0.718 -0.962 N HETATM 2 N 0 1.659 -0.865 1.223 N HETATM 3 N 0 -0.093 -1.865 -1.158 N HETATM 4 N 0 -1.464 -1.001 0.991 N HETATM 5 N 0 -1.547 1.072 -1.142 N HETATM 6 N 0 -0.153 1.904 1.058 N HETATM 7 C 0 1.567 1.556 -2.045 C HETATM 8 H 0 0.552 1.826 -2.405 H HETATM 9 C 0 2.719 2.082 -2.717 C 349 Table 6.S44 (cont’d) HETATM 10 C 0 4.016 1.718 -2.235 C HETATM 11 C 0 4.097 0.853 -1.110 C HETATM 12 H 0 5.099 0.569 -0.721 H HETATM 13 C 0 2.916 0.356 -0.479 C HETATM 14 C 0 2.919 -0.554 0.718 C HETATM 15 C 0 4.095 -1.087 1.331 C HETATM 16 H 0 5.103 -0.850 0.927 H HETATM 17 C 0 3.992 -1.932 2.470 C HETATM 18 C 0 2.691 -2.237 2.984 C HETATM 19 C 0 1.550 -1.678 2.316 C HETATM 20 H 0 0.527 -1.903 2.692 H HETATM 21 C 0 0.662 -2.206 -2.246 C HETATM 22 H 0 1.399 -1.459 -2.617 H HETATM 23 C 0 0.532 -3.469 -2.914 C HETATM 24 C 0 -0.424 -4.408 -2.409 C HETATM 25 C 0 -1.203 -4.045 -1.276 C HETATM 26 H 0 -1.948 -4.769 -0.881 H HETATM 27 C 0 -1.030 -2.767 -0.659 C HETATM 28 C 0 -1.806 -2.280 0.531 C HETATM 29 C 0 -2.835 -3.028 1.180 C HETATM 30 H 0 -3.103 -4.043 0.815 H HETATM 31 C 0 -3.533 -2.493 2.298 C HETATM 32 C 0 -3.178 -1.184 2.753 C HETATM 33 C 0 -2.137 -0.478 2.066 C HETATM 34 H 0 -1.844 0.537 2.410 H HETATM 35 C 0 -2.250 0.555 -2.193 C HETATM 36 H 0 -1.948 -0.448 -2.571 H HETATM 37 C 0 -3.339 1.248 -2.821 C HETATM 38 C 0 -3.705 2.534 -2.311 C HETATM 39 C 0 -2.969 3.063 -1.214 C HETATM 40 H 0 -3.257 4.056 -0.806 H HETATM 41 C 0 -1.887 2.323 -0.640 C HETATM 42 C 0 -1.056 2.816 0.522 C HETATM 43 C 0 -1.156 4.136 1.065 C HETATM 44 H 0 -1.863 4.875 0.630 H HETATM 45 C 0 -0.347 4.529 2.166 C HETATM 46 C 0 0.571 3.577 2.712 C HETATM 47 C 0 0.633 2.274 2.113 C HETATM 48 H 0 1.342 1.517 2.517 H HETATM 49 Fe 0 0.036 -0.047 0.009 Fe HETATM 50 H 0 -0.436 5.558 2.576 H HETATM 51 H 0 -4.544 3.119 -2.745 H HETATM 52 H 0 4.947 2.093 -2.711 H HETATM 53 H 0 4.915 -2.338 2.938 H HETATM 54 H 0 -4.330 -3.092 2.789 H HETATM 55 H 0 -0.569 -5.404 -2.880 H HETATM 56 O 0 -3.750 -0.503 3.819 O HETATM 57 O 0 1.368 -3.657 -4.005 O HETATM 58 O 0 -3.941 0.583 -3.880 O HETATM 59 O 0 1.431 3.786 3.782 O HETATM 60 O 0 2.448 2.911 -3.795 O HETATM 61 O 0 2.409 -3.036 4.082 O HETATM 62 C 0 -4.833 -1.179 4.575 C HETATM 63 H 0 -4.455 -2.126 5.047 H HETATM 64 H 0 -5.129 -0.447 5.366 H HETATM 65 H 0 -5.709 -1.395 3.904 H HETATM 66 C 0 1.276 -4.944 -4.738 C HETATM 67 H 0 0.249 -5.078 -5.174 H HETATM 68 H 0 2.032 -4.861 -5.557 H HETATM 69 H 0 1.533 -5.804 -4.063 H HETATM 70 C 0 1.401 5.112 4.447 C HETATM 71 H 0 0.385 5.314 4.883 H HETATM 72 H 0 2.161 5.039 5.264 H HETATM 73 H 0 1.689 5.924 3.726 H HETATM 74 C 0 -5.072 1.253 -4.569 C HETATM 75 H 0 -4.737 2.221 -5.029 H HETATM 76 H 0 -5.385 0.537 -5.368 H HETATM 77 H 0 -5.922 1.428 -3.856 H 350 Table 6.S44 (cont’d) HETATM 78 C 0 3.599 3.484 -4.537 C HETATM 79 H 0 4.226 2.667 -4.986 H HETATM 80 H 0 3.138 4.102 -5.347 H HETATM 81 H 0 4.220 4.134 -3.864 H HETATM 82 C 0 3.550 -3.637 4.818 C HETATM 83 H 0 4.216 -2.836 5.237 H HETATM 84 H 0 3.080 -4.218 5.649 H HETATM 85 H 0 4.134 -4.326 4.149 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 60 CONECT 10 9 11 52 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 53 CONECT 18 17 19 61 CONECT 19 2 18 20 CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 57 CONECT 24 23 25 55 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 54 CONECT 32 31 33 56 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 58 CONECT 38 37 39 51 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 50 CONECT 46 45 47 59 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 45 CONECT 51 38 CONECT 52 10 CONECT 53 17 CONECT 54 31 CONECT 55 24 CONECT 56 32 62 CONECT 57 23 66 CONECT 58 37 74 CONECT 59 46 70 CONECT 60 9 78 351 Table 6.S44 (cont’d) CONECT 61 18 82 CONECT 62 56 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 57 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 60 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 61 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 Table 6.S45: The optimized geometry file for the singlet charge transfer state of [Fe(5,5’-diOMe- bpy)3]2+ using td=1. TITLE 5,5'-OMe-bpy CT s1 B3LYP10 6-311G+dp SDD REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.696 0.758 -1.131 N HETATM 2 N 0 1.682 -0.785 1.126 N HETATM 3 N 0 -0.186 -1.802 -0.941 N HETATM 4 N 0 -1.534 -0.965 1.228 N HETATM 5 N 0 -1.515 0.988 -1.231 N HETATM 6 N 0 -0.156 1.802 0.940 N HETATM 7 C 0 1.620 1.637 -2.175 C HETATM 8 H 0 0.613 1.842 -2.604 H HETATM 9 C 0 2.771 2.285 -2.736 C HETATM 10 C 0 4.053 2.007 -2.161 C HETATM 11 C 0 4.122 1.094 -1.073 C HETATM 12 H 0 5.109 0.885 -0.606 H HETATM 13 C 0 2.937 0.469 -0.572 C HETATM 14 C 0 2.928 -0.515 0.570 C HETATM 15 C 0 4.103 -1.156 1.074 C HETATM 16 H 0 5.094 -0.961 0.609 H HETATM 17 C 0 4.018 -2.068 2.162 C HETATM 18 C 0 2.731 -2.327 2.733 C HETATM 19 C 0 1.591 -1.663 2.170 C HETATM 20 H 0 0.580 -1.853 2.596 H HETATM 21 C 0 0.579 -2.168 -2.019 C HETATM 22 H 0 1.295 -1.417 -2.416 H HETATM 23 C 0 0.485 -3.457 -2.639 C HETATM 24 C 0 -0.443 -4.407 -2.105 C HETATM 25 C 0 -1.233 -4.021 -0.987 C HETATM 26 H 0 -1.952 -4.752 -0.559 H HETATM 27 C 0 -1.103 -2.720 -0.415 C HETATM 28 C 0 -1.901 -2.226 0.757 C HETATM 29 C 0 -2.969 -2.945 1.377 C HETATM 30 H 0 -3.274 -3.945 0.998 H HETATM 31 C 0 -3.665 -2.387 2.485 C HETATM 32 C 0 -3.275 -1.093 2.960 C HETATM 33 C 0 -2.201 -0.419 2.291 C HETATM 34 H 0 -1.879 0.587 2.641 H HETATM 35 C 0 -2.189 0.454 -2.296 C HETATM 36 H 0 -1.882 -0.557 -2.646 H HETATM 37 C 0 -3.249 1.147 -2.968 C 352 Table 6.S45 (cont’d) HETATM 38 C 0 -3.617 2.448 -2.494 C HETATM 39 C 0 -2.914 2.993 -1.384 C HETATM 40 H 0 -3.202 3.998 -1.006 H HETATM 41 C 0 -1.861 2.255 -0.761 C HETATM 42 C 0 -1.057 2.735 0.413 C HETATM 43 C 0 -1.167 4.038 0.987 C HETATM 44 H 0 -1.874 4.780 0.558 H HETATM 45 C 0 -0.373 4.409 2.107 C HETATM 46 C 0 0.538 3.444 2.641 C HETATM 47 C 0 0.613 2.153 2.019 C HETATM 48 H 0 1.316 1.391 2.417 H HETATM 49 Fe 0 -0.079 -0.000 -0.001 Fe HETATM 50 H 0 -0.470 5.428 2.539 H HETATM 51 H 0 -4.435 3.033 -2.968 H HETATM 52 H 0 4.981 2.489 -2.537 H HETATM 53 H 0 4.938 -2.563 2.540 H HETATM 54 H 0 -4.494 -2.958 2.957 H HETATM 55 H 0 -0.555 -5.425 -2.536 H HETATM 56 O 0 -3.841 -0.406 4.023 O HETATM 57 O 0 1.330 -3.657 -3.721 O HETATM 58 O 0 -3.824 0.470 -4.033 O HETATM 59 O 0 1.384 3.628 3.725 O HETATM 60 O 0 2.518 3.138 -3.802 O HETATM 61 O 0 2.463 -3.176 3.799 O HETATM 62 C 0 -4.952 -1.057 4.762 C HETATM 63 H 0 -4.606 -2.019 5.229 H HETATM 64 H 0 -5.233 -0.325 5.558 H HETATM 65 H 0 -5.826 -1.240 4.081 H HETATM 66 C 0 1.279 -4.970 -4.410 C HETATM 67 H 0 0.257 -5.150 -4.843 H HETATM 68 H 0 2.034 -4.893 -5.230 H HETATM 69 H 0 1.559 -5.798 -3.705 H HETATM 70 C 0 1.349 4.940 4.418 C HETATM 71 H 0 0.330 5.133 4.848 H HETATM 72 H 0 2.101 4.849 5.240 H HETATM 73 H 0 1.643 5.766 3.716 H HETATM 74 C 0 -4.921 1.140 -4.774 C HETATM 75 H 0 -4.557 2.096 -5.239 H HETATM 76 H 0 -5.213 0.414 -5.572 H HETATM 77 H 0 -5.794 1.338 -4.096 H HETATM 78 C 0 3.672 3.836 -4.421 C HETATM 79 H 0 4.405 3.095 -4.840 H HETATM 80 H 0 3.229 4.446 -5.246 H HETATM 81 H 0 4.175 4.508 -3.674 H HETATM 82 C 0 3.605 -3.890 4.422 C HETATM 83 H 0 4.349 -3.160 4.841 H HETATM 84 H 0 3.151 -4.492 5.247 H HETATM 85 H 0 4.099 -4.571 3.677 H END CONECT 1 7 13 CONECT 2 14 19 CONECT 3 21 27 CONECT 4 28 33 CONECT 5 35 41 CONECT 6 42 47 CONECT 7 1 8 9 CONECT 8 7 CONECT 9 7 10 60 CONECT 10 9 11 52 CONECT 11 10 12 13 CONECT 12 11 CONECT 13 1 11 14 CONECT 14 2 13 15 CONECT 15 14 16 17 CONECT 16 15 CONECT 17 15 18 53 CONECT 18 17 19 61 CONECT 19 2 18 20 353 Table 6.S45 (cont’d) CONECT 20 19 CONECT 21 3 22 23 CONECT 22 21 CONECT 23 21 24 57 CONECT 24 23 25 55 CONECT 25 24 26 27 CONECT 26 25 CONECT 27 25 3 28 CONECT 28 4 27 29 CONECT 29 28 30 31 CONECT 30 29 CONECT 31 29 32 54 CONECT 32 31 33 56 CONECT 33 32 4 34 CONECT 34 33 CONECT 35 5 36 37 CONECT 36 35 CONECT 37 35 38 58 CONECT 38 37 39 51 CONECT 39 38 40 41 CONECT 40 39 CONECT 41 39 5 42 CONECT 42 6 41 43 CONECT 43 42 44 45 CONECT 44 43 CONECT 45 43 46 50 CONECT 46 45 47 59 CONECT 47 6 46 48 CONECT 48 47 CONECT 50 45 CONECT 51 38 CONECT 52 10 CONECT 53 17 CONECT 54 31 CONECT 55 24 CONECT 56 32 62 CONECT 57 23 66 CONECT 58 37 74 CONECT 59 46 70 CONECT 60 9 78 CONECT 61 18 82 CONECT 62 56 63 64 65 CONECT 63 62 CONECT 64 62 CONECT 65 62 CONECT 66 57 67 68 69 CONECT 67 66 CONECT 68 66 CONECT 69 66 CONECT 70 59 71 72 73 CONECT 71 70 CONECT 72 70 CONECT 73 70 CONECT 74 58 75 76 77 CONECT 75 74 CONECT 76 74 CONECT 77 74 CONECT 78 60 79 80 81 CONECT 79 78 CONECT 80 78 CONECT 81 78 CONECT 82 61 83 84 85 CONECT 83 82 CONECT 84 82 CONECT 85 82 354 [Fe(terpy)2]2+: Table 6.S46: The optimized geometry file for the lowest energy triplet state of [Fe(terpy)2]2+. TITLE Fe-terpy Triplet B3LYP10 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.006 3.979 6.158 N HETATM 2 N 0 0.038 1.596 5.483 N HETATM 3 N 0 1.794 5.225 3.945 N HETATM 4 C 0 0.552 3.219 7.233 C HETATM 5 C 0 0.620 3.723 8.567 C HETATM 6 C 0 1.160 5.022 8.793 C HETATM 7 C 0 0.006 1.864 6.848 C HETATM 8 C 0 -0.438 0.389 5.031 C HETATM 9 C 0 -0.507 0.912 7.783 C HETATM 10 C 0 -0.963 -0.606 5.903 C HETATM 11 C 0 -0.996 -0.336 7.303 C HETATM 12 C 0 1.534 5.250 6.372 C HETATM 13 C 0 1.622 5.794 7.689 C HETATM 14 C 0 1.982 5.956 5.114 C HETATM 15 C 0 2.169 5.780 2.745 C HETATM 16 C 0 2.554 7.266 5.090 C HETATM 17 C 0 2.744 7.078 2.646 C HETATM 18 C 0 2.939 7.831 3.841 C HETATM 19 N 0 0.772 2.541 2.492 N HETATM 20 N 0 -1.100 3.974 3.726 N HETATM 21 N 0 2.840 2.250 4.143 N HETATM 22 C 0 -0.372 2.768 1.732 C HETATM 23 C 0 -0.479 2.263 0.401 C HETATM 24 C 0 0.604 1.520 -0.150 C HETATM 25 C 0 -1.437 3.576 2.437 C HETATM 26 C 0 -1.999 4.721 4.449 C HETATM 27 C 0 -2.697 3.922 1.855 C HETATM 28 C 0 -3.269 5.104 3.934 C HETATM 29 C 0 -3.622 4.694 2.615 C HETATM 30 C 0 1.834 1.814 1.961 C HETATM 31 C 0 1.771 1.292 0.633 C HETATM 32 C 0 3.005 1.646 2.900 C HETATM 33 C 0 3.851 2.155 5.068 C HETATM 34 C 0 4.199 0.932 2.571 C HETATM 35 C 0 5.067 1.462 4.809 C HETATM 36 C 0 5.241 0.840 3.537 C HETATM 37 H 0 0.256 3.114 9.421 H HETATM 38 H 0 1.220 5.430 9.826 H HETATM 39 H 0 -0.392 0.223 3.931 H HETATM 40 H 0 -0.526 1.136 8.871 H HETATM 41 H 0 -1.335 -1.568 5.487 H HETATM 42 H 0 -1.399 -1.090 8.017 H HETATM 43 H 0 2.044 6.808 7.855 H HETATM 44 H 0 1.999 5.155 1.839 H HETATM 45 H 0 2.700 7.841 6.030 H HETATM 46 H 0 3.031 7.483 1.650 H HETATM 47 H 0 3.386 8.850 3.804 H HETATM 48 H 0 -1.393 2.448 -0.203 H HETATM 49 H 0 0.538 1.121 -1.186 H HETATM 50 H 0 -1.680 5.018 5.473 H HETATM 51 H 0 -2.957 3.596 0.825 H HETATM 52 H 0 -3.961 5.709 4.560 H HETATM 53 H 0 -4.607 4.972 2.178 H HETATM 54 H 0 2.622 0.715 0.211 H HETATM 55 H 0 3.673 2.654 6.046 H HETATM 56 H 0 4.319 0.451 1.576 H HETATM 57 H 0 5.856 1.414 5.592 H HETATM 58 H 0 6.177 0.287 3.298 H HETATM 59 Fe 0 0.891 3.258 4.327 Fe END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 355 Table 6.S46 (cont’d) CONECT 5 4 6 37 CONECT 6 5 13 38 CONECT 7 2 4 9 CONECT 8 2 10 39 CONECT 9 7 11 40 CONECT 10 8 11 41 CONECT 11 9 10 42 CONECT 12 1 13 14 CONECT 13 12 6 43 CONECT 14 3 12 16 CONECT 15 3 17 44 CONECT 16 14 18 45 CONECT 17 15 18 46 CONECT 18 16 17 47 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 48 CONECT 24 23 31 49 CONECT 25 22 20 27 CONECT 26 20 28 50 CONECT 27 25 29 51 CONECT 28 26 29 52 CONECT 29 27 28 53 CONECT 30 19 31 32 CONECT 31 24 30 54 CONECT 32 21 30 34 CONECT 33 21 35 55 CONECT 34 32 36 56 CONECT 35 33 36 57 CONECT 36 34 35 58 CONECT 37 5 CONECT 38 6 CONECT 39 8 CONECT 40 9 CONECT 41 10 CONECT 42 11 CONECT 43 13 CONECT 44 15 CONECT 45 16 CONECT 46 17 CONECT 47 18 CONECT 48 23 CONECT 49 24 CONECT 50 26 CONECT 51 27 CONECT 52 28 CONECT 53 29 CONECT 54 31 CONECT 55 33 CONECT 56 34 CONECT 57 35 CONECT 58 36 Table 6.S47: The optimized geometry file for the singlet charge transfer state of [Fe(terpy)2]2+ using td=1. TITLE Fe-terpy CT s1 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.992 3.981 6.143 N HETATM 2 N 0 0.104 1.665 5.344 N HETATM 3 N 0 1.744 5.071 3.900 N HETATM 4 C 0 0.536 3.207 7.190 C HETATM 5 C 0 0.597 3.707 8.526 C HETATM 6 C 0 1.137 5.011 8.740 C HETATM 7 C 0 0.019 1.867 6.728 C HETATM 8 C 0 -0.327 0.473 4.812 C 356 Table 6.S47 (cont’d) HETATM 9 C 0 -0.510 0.863 7.589 C HETATM 10 C 0 -0.863 -0.568 5.621 C HETATM 11 C 0 -0.957 -0.369 7.028 C HETATM 12 C 0 1.522 5.241 6.329 C HETATM 13 C 0 1.607 5.793 7.643 C HETATM 14 C 0 1.957 5.875 5.030 C HETATM 15 C 0 2.110 5.543 2.662 C HETATM 16 C 0 2.542 7.168 4.914 C HETATM 17 C 0 2.698 6.826 2.484 C HETATM 18 C 0 2.916 7.649 3.626 C HETATM 19 N 0 0.762 2.559 2.517 N HETATM 20 N 0 -0.986 3.951 3.838 N HETATM 21 N 0 2.705 2.340 4.224 N HETATM 22 C 0 -0.396 2.797 1.770 C HETATM 23 C 0 -0.508 2.290 0.450 C HETATM 24 C 0 0.590 1.533 -0.100 C HETATM 25 C 0 -1.396 3.600 2.530 C HETATM 26 C 0 -1.839 4.697 4.629 C HETATM 27 C 0 -2.675 4.012 2.035 C HETATM 28 C 0 -3.114 5.125 4.186 C HETATM 29 C 0 -3.539 4.775 2.861 C HETATM 30 C 0 1.848 1.823 2.004 C HETATM 31 C 0 1.766 1.300 0.674 C HETATM 32 C 0 2.951 1.704 2.972 C HETATM 33 C 0 3.678 2.287 5.205 C HETATM 34 C 0 4.199 1.021 2.745 C HETATM 35 C 0 4.915 1.626 5.025 C HETATM 36 C 0 5.178 0.980 3.762 C HETATM 37 H 0 0.235 3.100 9.384 H HETATM 38 H 0 1.193 5.421 9.773 H HETATM 39 H 0 -0.235 0.365 3.710 H HETATM 40 H 0 -0.572 1.039 8.685 H HETATM 41 H 0 -1.199 -1.514 5.143 H HETATM 42 H 0 -1.373 -1.164 7.687 H HETATM 43 H 0 2.031 6.806 7.815 H HETATM 44 H 0 1.920 4.865 1.803 H HETATM 45 H 0 2.706 7.793 5.818 H HETATM 46 H 0 2.976 7.164 1.461 H HETATM 47 H 0 3.375 8.658 3.519 H HETATM 48 H 0 -1.423 2.469 -0.156 H HETATM 49 H 0 0.518 1.129 -1.134 H HETATM 50 H 0 -1.475 4.950 5.648 H HETATM 51 H 0 -2.982 3.731 1.003 H HETATM 52 H 0 -3.760 5.722 4.866 H HETATM 53 H 0 -4.535 5.098 2.482 H HETATM 54 H 0 2.609 0.716 0.244 H HETATM 55 H 0 3.440 2.797 6.164 H HETATM 56 H 0 4.383 0.530 1.764 H HETATM 57 H 0 5.658 1.616 5.852 H HETATM 58 H 0 6.144 0.452 3.589 H HETATM 59 Fe 0 0.886 3.265 4.327 Fe END CONECT 1 4 12 59 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 38 CONECT 7 2 4 9 CONECT 8 2 10 39 CONECT 9 7 11 40 CONECT 10 8 11 41 CONECT 11 9 10 42 CONECT 12 1 13 14 CONECT 13 12 6 43 CONECT 14 3 12 16 CONECT 15 3 17 44 CONECT 16 14 18 45 357 Table 6.S47 (cont’d) CONECT 17 15 18 46 CONECT 18 16 17 47 CONECT 19 22 30 59 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 48 CONECT 24 23 31 49 CONECT 25 22 20 27 CONECT 26 20 28 50 CONECT 27 25 29 51 CONECT 28 26 29 52 CONECT 29 27 28 53 CONECT 30 19 31 32 CONECT 31 24 30 54 CONECT 32 21 30 34 CONECT 33 21 35 55 CONECT 34 32 36 56 CONECT 35 33 36 57 CONECT 36 34 35 58 CONECT 37 5 CONECT 38 6 CONECT 39 8 CONECT 40 9 CONECT 41 10 CONECT 42 11 CONECT 43 13 CONECT 44 15 CONECT 45 16 CONECT 46 17 CONECT 47 18 CONECT 48 23 CONECT 49 24 CONECT 50 26 CONECT 51 27 CONECT 52 28 CONECT 53 29 CONECT 54 31 CONECT 55 33 CONECT 56 34 CONECT 57 35 CONECT 58 36 CONECT 59 1 19 Table 6.S48: The optimized geometry file for the singlet charge transfer state of [Fe(terpy)2]2+ using td=12. TITLE Fe-terpy CT s12 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.007 3.969 6.161 N HETATM 2 N 0 0.020 1.611 5.458 N HETATM 3 N 0 1.803 5.211 3.954 N HETATM 4 C 0 0.540 3.209 7.228 C HETATM 5 C 0 0.599 3.708 8.564 C HETATM 6 C 0 1.147 5.002 8.798 C HETATM 7 C 0 -0.011 1.861 6.827 C HETATM 8 C 0 -0.461 0.411 4.990 C HETATM 9 C 0 -0.528 0.899 7.749 C HETATM 10 C 0 -0.991 -0.591 5.850 C HETATM 11 C 0 -1.023 -0.341 7.253 C HETATM 12 C 0 1.545 5.233 6.381 C HETATM 13 C 0 1.626 5.774 7.700 C HETATM 14 C 0 2.008 5.934 5.125 C HETATM 15 C 0 2.188 5.762 2.756 C HETATM 16 C 0 2.610 7.230 5.106 C HETATM 17 C 0 2.790 7.048 2.661 C HETATM 18 C 0 3.004 7.792 3.858 C HETATM 19 N 0 0.766 2.549 2.488 N 358 Table 6.S48 (cont’d) HETATM 20 N 0 -1.078 3.996 3.730 N HETATM 21 N 0 2.818 2.252 4.142 N HETATM 22 C 0 -0.374 2.789 1.729 C HETATM 23 C 0 -0.484 2.291 0.396 C HETATM 24 C 0 0.594 1.541 -0.156 C HETATM 25 C 0 -1.428 3.603 2.442 C HETATM 26 C 0 -1.966 4.748 4.461 C HETATM 27 C 0 -2.689 3.961 1.870 C HETATM 28 C 0 -3.237 5.142 3.956 C HETATM 29 C 0 -3.603 4.738 2.638 C HETATM 30 C 0 1.822 1.817 1.956 C HETATM 31 C 0 1.759 1.300 0.627 C HETATM 32 C 0 2.987 1.641 2.903 C HETATM 33 C 0 3.824 2.154 5.074 C HETATM 34 C 0 4.176 0.915 2.585 C HETATM 35 C 0 5.036 1.449 4.825 C HETATM 36 C 0 5.211 0.819 3.558 C HETATM 37 H 0 0.223 3.099 9.414 H HETATM 38 H 0 1.201 5.408 9.832 H HETATM 39 H 0 -0.417 0.259 3.888 H HETATM 40 H 0 -0.545 1.110 8.841 H HETATM 41 H 0 -1.368 -1.546 5.421 H HETATM 42 H 0 -1.429 -1.101 7.958 H HETATM 43 H 0 2.053 6.785 7.872 H HETATM 44 H 0 2.003 5.145 1.848 H HETATM 45 H 0 2.773 7.796 6.049 H HETATM 46 H 0 3.083 7.451 1.665 H HETATM 47 H 0 3.474 8.801 3.824 H HETATM 48 H 0 -1.396 2.486 -0.208 H HETATM 49 H 0 0.527 1.146 -1.194 H HETATM 50 H 0 -1.638 5.039 5.484 H HETATM 51 H 0 -2.959 3.638 0.841 H HETATM 52 H 0 -3.920 5.751 4.588 H HETATM 53 H 0 -4.589 5.025 2.209 H HETATM 54 H 0 2.606 0.718 0.204 H HETATM 55 H 0 3.644 2.660 6.049 H HETATM 56 H 0 4.296 0.428 1.593 H HETATM 57 H 0 5.820 1.400 5.613 H HETATM 58 H 0 6.145 0.258 3.327 H HETATM 59 Fe 0 0.891 3.254 4.326 Fe END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 38 CONECT 7 2 4 9 CONECT 8 2 10 39 CONECT 9 7 11 40 CONECT 10 8 11 41 CONECT 11 9 10 42 CONECT 12 1 13 14 CONECT 13 12 6 43 CONECT 14 3 12 16 CONECT 15 3 17 44 CONECT 16 14 18 45 CONECT 17 15 18 46 CONECT 18 16 17 47 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 48 CONECT 24 23 31 49 CONECT 25 22 20 27 CONECT 26 20 28 50 CONECT 27 25 29 51 359 Table 6.S48 (cont’d) CONECT 28 26 29 52 CONECT 29 27 28 53 CONECT 30 19 31 32 CONECT 31 24 30 54 CONECT 32 21 30 34 CONECT 33 21 35 55 CONECT 34 32 36 56 CONECT 35 33 36 57 CONECT 36 34 35 58 CONECT 37 5 CONECT 38 6 CONECT 39 8 CONECT 40 9 CONECT 41 10 CONECT 42 11 CONECT 43 13 CONECT 44 15 CONECT 45 16 CONECT 46 17 CONECT 47 18 CONECT 48 23 CONECT 49 24 CONECT 50 26 CONECT 51 27 CONECT 52 28 CONECT 53 29 CONECT 54 31 CONECT 55 33 CONECT 56 34 CONECT 57 35 CONECT 58 36 [Fe(DPA)2]2+: Table 6.S49: The optimized geometry file for the lowest energy singlet state of [Fe(DPA)2]2+. TITLE Fe-DPA Singlet B3LYP10 SDD 6-311+G(d,p) REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.399 3.651 4.703 N HETATM 2 N 0 3.355 2.171 6.637 N HETATM 3 N 0 2.544 0.833 4.562 N HETATM 4 N 0 4.128 4.497 5.770 N HETATM 5 N 0 3.312 3.392 2.961 N HETATM 6 C 0 3.811 2.990 7.632 C HETATM 7 C 0 4.245 4.292 7.155 C HETATM 8 C 0 2.210 4.026 2.459 C HETATM 9 C 0 4.395 3.186 2.155 C HETATM 10 C 0 2.915 0.915 6.947 C HETATM 11 C 0 2.128 4.501 1.091 C HETATM 12 C 0 4.441 3.612 0.769 C HETATM 13 C 0 2.119 0.145 3.460 C HETATM 14 H 0 2.181 0.675 2.483 H HETATM 15 C 0 1.137 4.175 3.427 C HETATM 16 C 0 1.969 -1.179 5.974 C HETATM 17 C 0 3.854 2.590 9.025 C HETATM 18 C 0 6.273 1.508 4.810 C HETATM 19 H 0 6.098 1.237 5.874 H HETATM 20 C 0 5.488 2.498 2.820 C HETATM 21 C 0 6.693 2.214 2.086 C HETATM 22 C 0 -1.107 4.942 4.044 C HETATM 23 H 0 -2.078 5.436 3.809 H HETATM 24 C 0 2.906 0.394 8.301 C HETATM 25 C 0 4.528 5.717 5.299 C HETATM 26 H 0 4.437 5.887 4.203 H HETATM 27 C 0 3.275 4.281 0.245 C HETATM 28 H 0 3.261 4.629 -0.814 H HETATM 29 C 0 -0.083 4.827 3.033 C HETATM 30 C 0 7.738 1.527 2.807 C 360 Table 6.S49 (cont’d) HETATM 31 H 0 8.699 1.275 2.303 H HETATM 32 C 0 5.690 3.308 0.044 C HETATM 33 H 0 5.779 3.615 -1.023 H HETATM 34 C 0 6.753 2.647 0.672 C HETATM 35 H 0 7.684 2.430 0.100 H HETATM 36 C 0 0.864 5.166 0.717 C HETATM 37 H 0 0.749 5.551 -0.322 H HETATM 38 C 0 3.389 1.263 9.346 C HETATM 39 H 0 3.403 0.908 10.403 H HETATM 40 C 0 4.376 3.600 9.967 C HETATM 41 H 0 4.431 3.345 11.050 H HETATM 42 C 0 0.404 3.779 5.632 C HETATM 43 H 0 0.606 3.366 6.645 H HETATM 44 C 0 -0.178 5.318 1.640 C HETATM 45 H 0 -1.121 5.825 1.330 H HETATM 46 C 0 2.460 0.161 5.792 C HETATM 47 C 0 1.955 -1.723 7.350 C HETATM 48 H 0 1.577 -2.761 7.497 H HETATM 49 C 0 7.517 1.180 4.164 C HETATM 50 H 0 8.299 0.649 4.751 H HETATM 51 C 0 1.532 -1.870 4.784 C HETATM 52 H 0 1.139 -2.911 4.847 H HETATM 53 C 0 2.397 -0.984 8.454 C HETATM 54 H 0 2.369 -1.435 9.473 H HETATM 55 C 0 -0.853 4.416 5.335 C HETATM 56 H 0 -1.618 4.486 6.140 H HETATM 57 C 0 4.797 4.859 9.522 C HETATM 58 H 0 5.186 5.602 10.256 H HETATM 59 C 0 1.611 -1.200 3.537 C HETATM 60 H 0 1.282 -1.700 2.599 H HETATM 61 C 0 5.044 6.748 6.162 C HETATM 62 H 0 5.351 7.717 5.709 H HETATM 63 C 0 4.748 5.255 8.097 C HETATM 64 C 0 5.159 6.530 7.557 C HETATM 65 H 0 5.558 7.325 8.228 H HETATM 66 N 0 5.259 2.160 4.164 N HETATM 67 Fe 0 3.333 2.785 4.800 Fe END CONECT 1 15 42 CONECT 2 6 10 67 CONECT 3 13 46 CONECT 4 7 25 CONECT 5 8 9 67 CONECT 6 2 7 17 CONECT 7 4 6 63 CONECT 8 5 11 15 CONECT 9 5 12 20 CONECT 10 2 24 46 CONECT 11 8 27 36 CONECT 12 9 27 32 CONECT 13 3 14 59 CONECT 14 13 CONECT 15 1 8 29 CONECT 16 46 47 51 CONECT 17 6 38 40 CONECT 18 19 49 66 CONECT 19 18 CONECT 20 9 21 66 CONECT 21 20 30 34 CONECT 22 23 29 55 CONECT 23 22 CONECT 24 10 38 53 CONECT 25 4 26 61 CONECT 26 25 CONECT 27 11 12 28 CONECT 28 27 CONECT 29 22 15 44 CONECT 30 21 31 49 361 Table 6.S49 (cont’d) CONECT 31 30 CONECT 32 12 33 34 CONECT 33 32 CONECT 34 32 21 35 CONECT 35 34 CONECT 36 11 37 44 CONECT 37 36 CONECT 38 17 24 39 CONECT 39 38 CONECT 40 17 41 57 CONECT 41 40 CONECT 42 1 43 55 CONECT 43 42 CONECT 44 36 29 45 CONECT 45 44 CONECT 46 16 3 10 CONECT 47 16 48 53 CONECT 48 47 CONECT 49 18 30 50 CONECT 50 49 CONECT 51 16 52 59 CONECT 52 51 CONECT 53 47 24 54 CONECT 54 53 CONECT 55 22 42 56 CONECT 56 55 CONECT 57 40 58 63 CONECT 58 57 CONECT 59 51 13 60 CONECT 60 59 CONECT 61 25 62 64 CONECT 62 61 CONECT 63 7 57 64 CONECT 64 63 61 65 CONECT 65 64 CONECT 66 18 20 CONECT 67 2 5 Table 6.S50: The optimized geometry file for the lowest energy triplet state of [Fe(DPA)2]2+. TITLE Fe-DPA Triplet B3LYP10 SDD 6-311+G(d,p) REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.276 3.718 4.646 N HETATM 2 N 0 3.351 2.162 6.663 N HETATM 3 N 0 2.494 0.694 4.597 N HETATM 4 N 0 4.180 4.582 5.877 N HETATM 5 N 0 3.307 3.401 2.933 N HETATM 6 C 0 3.812 2.983 7.667 C HETATM 7 C 0 4.266 4.306 7.243 C HETATM 8 C 0 2.200 4.041 2.424 C HETATM 9 C 0 4.398 3.198 2.119 C HETATM 10 C 0 2.906 0.898 6.982 C HETATM 11 C 0 2.129 4.516 1.051 C HETATM 12 C 0 4.434 3.631 0.731 C HETATM 13 C 0 2.063 -0.037 3.529 C HETATM 14 H 0 2.115 0.458 2.532 H HETATM 15 C 0 1.087 4.217 3.356 C HETATM 16 C 0 1.953 -1.248 6.092 C HETATM 17 C 0 3.851 2.573 9.062 C HETATM 18 C 0 6.417 1.505 4.681 C HETATM 19 H 0 6.277 1.223 5.749 H HETATM 20 C 0 5.530 2.511 2.739 C HETATM 21 C 0 6.719 2.253 1.961 C HETATM 22 C 0 -1.178 5.015 3.880 C HETATM 23 H 0 -2.132 5.516 3.595 H HETATM 24 C 0 2.901 0.386 8.343 C HETATM 25 C 0 4.601 5.809 5.454 C HETATM 26 H 0 4.527 6.014 4.362 H HETATM 27 C 0 3.273 4.298 0.209 C 362 Table 6.S50 (cont’d) HETATM 28 H 0 3.260 4.646 -0.850 H HETATM 29 C 0 -0.117 4.877 2.910 C HETATM 30 C 0 7.799 1.572 2.634 C HETATM 31 H 0 8.742 1.344 2.086 H HETATM 32 C 0 5.664 3.351 -0.033 C HETATM 33 H 0 5.712 3.676 -1.098 H HETATM 34 C 0 6.748 2.696 0.553 C HETATM 35 H 0 7.668 2.494 -0.042 H HETATM 36 C 0 0.885 5.187 0.631 C HETATM 37 H 0 0.811 5.560 -0.416 H HETATM 38 C 0 3.385 1.252 9.383 C HETATM 39 H 0 3.399 0.899 10.440 H HETATM 40 C 0 4.369 3.547 10.039 C HETATM 41 H 0 4.406 3.253 11.113 H HETATM 42 C 0 0.252 3.866 5.535 C HETATM 43 H 0 0.418 3.462 6.560 H HETATM 44 C 0 -0.181 5.357 1.516 C HETATM 45 H 0 -1.113 5.869 1.179 H HETATM 46 C 0 2.437 0.092 5.856 C HETATM 47 C 0 1.949 -1.760 7.477 C HETATM 48 H 0 1.575 -2.794 7.653 H HETATM 49 C 0 7.643 1.202 3.992 C HETATM 50 H 0 8.457 0.676 4.541 H HETATM 51 C 0 1.505 -1.989 4.937 C HETATM 52 H 0 1.120 -3.028 5.056 H HETATM 53 C 0 2.399 -0.985 8.547 C HETATM 54 H 0 2.387 -1.396 9.583 H HETATM 55 C 0 -0.988 4.509 5.189 C HETATM 56 H 0 -1.785 4.599 5.960 H HETATM 57 C 0 4.805 4.812 9.640 C HETATM 58 H 0 5.193 5.534 10.395 H HETATM 59 C 0 1.562 -1.379 3.660 C HETATM 60 H 0 1.225 -1.923 2.750 H HETATM 61 C 0 5.120 6.803 6.355 C HETATM 62 H 0 5.448 7.785 5.947 H HETATM 63 C 0 4.770 5.235 8.227 C HETATM 64 C 0 5.207 6.522 7.741 C HETATM 65 H 0 5.606 7.281 8.452 H HETATM 66 N 0 5.374 2.146 4.077 N HETATM 67 Fe 0 3.329 2.783 4.799 Fe END CONECT 1 15 42 CONECT 2 6 10 67 CONECT 3 13 46 CONECT 4 7 25 CONECT 5 8 9 67 CONECT 6 2 7 17 CONECT 7 4 6 63 CONECT 8 5 11 15 CONECT 9 5 12 20 CONECT 10 2 24 46 CONECT 11 8 27 36 CONECT 12 9 27 32 CONECT 13 3 14 59 CONECT 14 13 CONECT 15 1 8 29 CONECT 16 46 47 51 CONECT 17 6 38 40 CONECT 18 19 49 66 CONECT 19 18 CONECT 20 9 21 66 CONECT 21 20 30 34 CONECT 22 23 29 55 CONECT 23 22 CONECT 24 10 38 53 CONECT 25 4 26 61 CONECT 26 25 CONECT 27 11 12 28 363 Table 6.S50 (cont’d) CONECT 28 27 CONECT 29 22 15 44 CONECT 30 21 31 49 CONECT 31 30 CONECT 32 12 33 34 CONECT 33 32 CONECT 34 32 21 35 CONECT 35 34 CONECT 36 11 37 44 CONECT 37 36 CONECT 38 17 24 39 CONECT 39 38 CONECT 40 17 41 57 CONECT 41 40 CONECT 42 1 43 55 CONECT 43 42 CONECT 44 29 36 45 CONECT 45 44 CONECT 46 16 3 10 CONECT 47 16 48 53 CONECT 48 47 CONECT 49 18 30 50 CONECT 50 49 CONECT 51 16 52 59 CONECT 52 51 CONECT 53 47 24 54 CONECT 54 53 CONECT 55 22 42 56 CONECT 56 55 CONECT 57 40 58 63 CONECT 58 57 CONECT 59 13 51 60 CONECT 60 59 CONECT 61 25 62 64 CONECT 62 61 CONECT 63 7 57 64 CONECT 64 63 61 65 CONECT 65 64 CONECT 66 18 20 CONECT 67 2 5 Table 6.S51: The optimized geometry file for the lowest energy quintet state of [Fe(DPA)2]2+. TITLE Fe-DPA Quintet B3LYP10 SDD 6-311+G(d,p) REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.047 3.860 4.362 N HETATM 2 N 0 3.396 2.054 6.830 N HETATM 3 N 0 2.354 0.565 4.811 N HETATM 4 N 0 4.037 4.575 6.034 N HETATM 5 N 0 3.213 3.497 2.773 N HETATM 6 C 0 3.919 2.878 7.790 C HETATM 7 C 0 4.272 4.245 7.367 C HETATM 8 C 0 2.159 4.172 2.214 C HETATM 9 C 0 4.345 3.277 2.033 C HETATM 10 C 0 3.033 0.773 7.149 C HETATM 11 C 0 2.181 4.676 0.848 C HETATM 12 C 0 4.479 3.736 0.654 C HETATM 13 C 0 1.833 -0.178 3.793 C HETATM 14 H 0 1.751 0.325 2.802 H HETATM 15 C 0 0.979 4.374 3.072 C HETATM 16 C 0 2.072 -1.404 6.332 C HETATM 17 C 0 4.123 2.441 9.166 C HETATM 18 C 0 6.223 1.469 4.639 C HETATM 19 H 0 6.032 1.154 5.690 H HETATM 20 C 0 5.435 2.547 2.697 C HETATM 21 C 0 6.656 2.285 1.967 C HETATM 22 C 0 -1.293 5.247 3.441 C HETATM 23 H 0 -2.204 5.785 3.091 H HETATM 24 C 0 3.191 0.228 8.491 C 364 Table 6.S51 (cont’d) HETATM 25 C 0 4.358 5.837 5.630 C HETATM 26 H 0 4.165 6.082 4.561 H HETATM 27 C 0 3.371 4.440 0.078 C HETATM 28 H 0 3.433 4.808 -0.973 H HETATM 29 C 0 -0.168 5.080 2.550 C HETATM 30 C 0 7.695 1.564 2.663 C HETATM 31 H 0 8.655 1.335 2.146 H HETATM 32 C 0 5.739 3.449 -0.053 C HETATM 33 H 0 5.848 3.798 -1.105 H HETATM 34 C 0 6.774 2.758 0.575 C HETATM 35 H 0 7.722 2.548 0.027 H HETATM 36 C 0 0.992 5.389 0.351 C HETATM 37 H 0 1.003 5.776 -0.693 H HETATM 38 C 0 3.747 1.094 9.492 C HETATM 39 H 0 3.887 0.718 10.533 H HETATM 40 C 0 4.697 3.403 10.121 C HETATM 41 H 0 4.857 3.076 11.174 H HETATM 42 C 0 -0.032 4.033 5.178 C HETATM 43 H 0 0.048 3.611 6.205 H HETATM 44 C 0 -0.126 5.581 1.163 C HETATM 45 H 0 -1.017 6.124 0.772 H HETATM 46 C 0 2.470 -0.042 6.059 C HETATM 47 C 0 2.241 -1.939 7.696 C HETATM 48 H 0 1.930 -2.992 7.890 H HETATM 49 C 0 7.473 1.157 4.000 C HETATM 50 H 0 8.251 0.599 4.569 H HETATM 51 C 0 1.522 -2.158 5.229 C HETATM 52 H 0 1.198 -3.213 5.383 H HETATM 53 C 0 2.774 -1.164 8.727 C HETATM 54 H 0 2.891 -1.589 9.750 H HETATM 55 C 0 -1.222 4.723 4.753 C HETATM 56 H 0 -2.070 4.833 5.464 H HETATM 57 C 0 5.034 4.697 9.723 C HETATM 58 H 0 5.467 5.414 10.458 H HETATM 59 C 0 1.405 -1.541 3.962 C HETATM 60 H 0 0.986 -2.092 3.090 H HETATM 61 C 0 4.919 6.822 6.518 C HETATM 62 H 0 5.158 7.835 6.125 H HETATM 63 C 0 4.832 5.159 8.336 C HETATM 64 C 0 5.158 6.487 7.871 C HETATM 65 H 0 5.593 7.233 8.576 H HETATM 66 N 0 5.217 2.148 4.014 N HETATM 67 Fe 0 3.089 2.796 4.816 Fe END CONECT 1 15 42 CONECT 2 6 10 CONECT 3 13 46 CONECT 4 7 25 CONECT 5 8 9 CONECT 6 2 7 17 CONECT 7 4 6 63 CONECT 8 5 11 15 CONECT 9 5 12 20 CONECT 10 2 24 46 CONECT 11 8 27 36 CONECT 12 9 27 32 CONECT 13 3 14 59 CONECT 14 13 CONECT 15 1 8 29 CONECT 16 46 47 51 CONECT 17 6 38 40 CONECT 18 19 49 66 CONECT 19 18 CONECT 20 9 21 66 CONECT 21 20 30 34 CONECT 22 23 29 55 CONECT 23 22 CONECT 24 10 38 53 365 Table 6.S51 (cont’d) CONECT 25 4 26 61 CONECT 26 25 CONECT 27 11 12 28 CONECT 28 27 CONECT 29 22 15 44 CONECT 30 21 31 49 CONECT 31 30 CONECT 32 12 33 34 CONECT 33 32 CONECT 34 32 21 35 CONECT 35 34 CONECT 36 11 37 44 CONECT 37 36 CONECT 38 17 24 39 CONECT 39 38 CONECT 40 17 41 57 CONECT 41 40 CONECT 42 1 43 55 CONECT 43 42 CONECT 44 29 36 45 CONECT 45 44 CONECT 46 16 3 10 CONECT 47 16 48 53 CONECT 48 47 CONECT 49 18 30 50 CONECT 50 49 CONECT 51 16 52 59 CONECT 52 51 CONECT 53 47 24 54 CONECT 54 53 CONECT 55 22 42 56 CONECT 56 55 CONECT 57 40 58 63 CONECT 58 57 CONECT 59 51 13 60 CONECT 60 59 CONECT 61 25 62 64 CONECT 62 61 CONECT 63 7 57 64 CONECT 64 63 61 65 CONECT 65 64 CONECT 66 18 20 Table 6.S52: The optimized geometry file for the singlet charge transfer state of [Fe(DPA)2]2+ using td=6. TITLE Fe-DPA CT s6 B3LYP10 SDD 6-311+G(d,p) REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.395 3.655 4.709 N HETATM 2 N 0 3.356 2.171 6.647 N HETATM 3 N 0 2.544 0.835 4.556 N HETATM 4 N 0 4.132 4.506 5.771 N HETATM 5 N 0 3.313 3.400 2.953 N HETATM 6 C 0 3.816 3.001 7.635 C HETATM 7 C 0 4.247 4.295 7.162 C HETATM 8 C 0 2.199 4.030 2.462 C HETATM 9 C 0 4.406 3.182 2.155 C HETATM 10 C 0 2.912 0.908 6.942 C HETATM 11 C 0 2.118 4.500 1.094 C HETATM 12 C 0 4.448 3.605 0.770 C HETATM 13 C 0 2.123 0.157 3.452 C HETATM 14 H 0 2.187 0.690 2.477 H HETATM 15 C 0 1.135 4.178 3.425 C HETATM 16 C 0 1.967 -1.185 5.965 C HETATM 17 C 0 3.858 2.598 9.026 C HETATM 18 C 0 6.269 1.509 4.819 C HETATM 19 H 0 6.091 1.240 5.883 H HETATM 20 C 0 5.491 2.499 2.817 C 366 Table 6.S52 (cont’d) HETATM 21 C 0 6.702 2.207 2.086 C HETATM 22 C 0 -1.112 4.939 4.053 C HETATM 23 H 0 -2.085 5.430 3.823 H HETATM 24 C 0 2.903 0.388 8.294 C HETATM 25 C 0 4.528 5.720 5.296 C HETATM 26 H 0 4.436 5.888 4.199 H HETATM 27 C 0 3.273 4.275 0.249 C HETATM 28 H 0 3.257 4.619 -0.811 H HETATM 29 C 0 -0.094 4.829 3.035 C HETATM 30 C 0 7.740 1.518 2.817 C HETATM 31 H 0 8.702 1.262 2.317 H HETATM 32 C 0 5.694 3.297 0.048 C HETATM 33 H 0 5.784 3.600 -1.021 H HETATM 34 C 0 6.762 2.634 0.677 C HETATM 35 H 0 7.691 2.415 0.102 H HETATM 36 C 0 0.855 5.160 0.722 C HETATM 37 H 0 0.736 5.542 -0.318 H HETATM 38 C 0 3.389 1.264 9.341 C HETATM 39 H 0 3.403 0.908 10.397 H HETATM 40 C 0 4.379 3.607 9.965 C HETATM 41 H 0 4.434 3.353 11.048 H HETATM 42 C 0 0.408 3.779 5.640 C HETATM 43 H 0 0.614 3.367 6.653 H HETATM 44 C 0 -0.193 5.314 1.647 C HETATM 45 H 0 -1.136 5.819 1.335 H HETATM 46 C 0 2.461 0.161 5.792 C HETATM 47 C 0 1.952 -1.730 7.334 C HETATM 48 H 0 1.574 -2.768 7.481 H HETATM 49 C 0 7.516 1.175 4.173 C HETATM 50 H 0 8.294 0.644 4.765 H HETATM 51 C 0 1.533 -1.866 4.768 C HETATM 52 H 0 1.139 -2.907 4.825 H HETATM 53 C 0 2.396 -0.987 8.442 C HETATM 54 H 0 2.366 -1.441 9.460 H HETATM 55 C 0 -0.854 4.415 5.345 C HETATM 56 H 0 -1.615 4.483 6.153 H HETATM 57 C 0 4.803 4.872 9.522 C HETATM 58 H 0 5.191 5.613 10.258 H HETATM 59 C 0 1.613 -1.192 3.523 C HETATM 60 H 0 1.287 -1.686 2.582 H HETATM 61 C 0 5.045 6.755 6.159 C HETATM 62 H 0 5.350 7.722 5.702 H HETATM 63 C 0 4.753 5.266 8.103 C HETATM 64 C 0 5.161 6.538 7.554 C HETATM 65 H 0 5.560 7.336 8.222 H HETATM 66 N 0 5.264 2.159 4.168 N HETATM 67 Fe 0 3.335 2.789 4.801 Fe END CONECT 1 15 42 CONECT 2 6 10 67 CONECT 3 13 46 CONECT 4 7 25 CONECT 5 8 9 67 CONECT 6 2 7 17 CONECT 7 4 6 63 CONECT 8 5 11 15 CONECT 9 5 12 20 CONECT 10 2 24 46 CONECT 11 8 27 36 CONECT 12 9 27 32 CONECT 13 3 14 59 CONECT 14 13 CONECT 15 1 8 29 CONECT 16 46 47 51 CONECT 17 6 38 40 CONECT 18 19 49 66 CONECT 19 18 CONECT 20 9 21 66 367 Table 6.S52 (cont’d) CONECT 21 20 30 34 CONECT 22 23 29 55 CONECT 23 22 CONECT 24 10 38 53 CONECT 25 4 26 61 CONECT 26 25 CONECT 27 11 12 28 CONECT 28 27 CONECT 29 22 15 44 CONECT 30 21 31 49 CONECT 31 30 CONECT 32 12 33 34 CONECT 33 32 CONECT 34 32 21 35 CONECT 35 34 CONECT 36 11 37 44 CONECT 37 36 CONECT 38 17 24 39 CONECT 39 38 CONECT 40 17 41 57 CONECT 41 40 CONECT 42 1 43 55 CONECT 43 42 CONECT 44 36 29 45 CONECT 45 44 CONECT 46 16 3 10 CONECT 47 16 48 53 CONECT 48 47 CONECT 49 18 30 50 CONECT 50 49 CONECT 51 16 52 59 CONECT 52 51 CONECT 53 47 24 54 CONECT 54 53 CONECT 55 22 42 56 CONECT 56 55 CONECT 57 40 58 63 CONECT 58 57 CONECT 59 51 13 60 CONECT 60 59 CONECT 61 25 62 64 CONECT 62 61 CONECT 63 7 57 64 CONECT 64 63 61 65 CONECT 65 64 CONECT 66 18 20 CONECT 67 2 5 Table 6.S53: The optimized geometry file for the singlet charge transfer state of [Fe(DPA)2]2+ using td=18. TITLE Fe-DPA CT s18 B3LYP10 SDD 6-311+G(d,p) REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.416 3.644 4.706 N HETATM 2 N 0 3.352 2.172 6.641 N HETATM 3 N 0 2.539 0.853 4.545 N HETATM 4 N 0 4.113 4.501 5.751 N HETATM 5 N 0 3.308 3.405 2.936 N HETATM 6 C 0 3.814 2.992 7.621 C HETATM 7 C 0 4.240 4.293 7.135 C HETATM 8 C 0 2.205 4.033 2.451 C HETATM 9 C 0 4.398 3.187 2.152 C HETATM 10 C 0 2.919 0.919 6.935 C HETATM 11 C 0 2.121 4.510 1.079 C HETATM 12 C 0 4.453 3.610 0.761 C HETATM 13 C 0 2.115 0.178 3.435 C HETATM 14 H 0 2.174 0.716 2.463 H HETATM 15 C 0 1.141 4.173 3.434 C 368 Table 6.S53 (cont’d) HETATM 16 C 0 1.979 -1.166 5.947 C HETATM 17 C 0 3.870 2.590 9.014 C HETATM 18 C 0 6.240 1.500 4.868 C HETATM 19 H 0 6.045 1.236 5.930 H HETATM 20 C 0 5.478 2.495 2.842 C HETATM 21 C 0 6.685 2.204 2.120 C HETATM 22 C 0 -1.110 4.932 4.080 C HETATM 23 H 0 -2.084 5.423 3.858 H HETATM 24 C 0 2.920 0.388 8.286 C HETATM 25 C 0 4.502 5.720 5.269 C HETATM 26 H 0 4.402 5.886 4.174 H HETATM 27 C 0 3.280 4.281 0.244 C HETATM 28 H 0 3.268 4.628 -0.816 H HETATM 29 C 0 -0.080 4.824 3.045 C HETATM 30 C 0 7.730 1.505 2.871 C HETATM 31 H 0 8.697 1.243 2.384 H HETATM 32 C 0 5.701 3.300 0.052 C HETATM 33 H 0 5.810 3.597 -1.015 H HETATM 34 C 0 6.755 2.629 0.715 C HETATM 35 H 0 7.692 2.405 0.153 H HETATM 36 C 0 0.860 5.171 0.713 C HETATM 37 H 0 0.730 5.560 -0.322 H HETATM 38 C 0 3.409 1.259 9.329 C HETATM 39 H 0 3.431 0.900 10.384 H HETATM 40 C 0 4.397 3.601 9.950 C HETATM 41 H 0 4.463 3.350 11.032 H HETATM 42 C 0 0.417 3.763 5.666 C HETATM 43 H 0 0.630 3.346 6.673 H HETATM 44 C 0 -0.178 5.312 1.661 C HETATM 45 H 0 -1.125 5.817 1.358 H HETATM 46 C 0 2.463 0.175 5.773 C HETATM 47 C 0 1.973 -1.721 7.320 C HETATM 48 H 0 1.598 -2.761 7.457 H HETATM 49 C 0 7.481 1.169 4.228 C HETATM 50 H 0 8.254 0.636 4.828 H HETATM 51 C 0 1.541 -1.846 4.751 C HETATM 52 H 0 1.152 -2.888 4.806 H HETATM 53 C 0 2.417 -0.991 8.430 C HETATM 54 H 0 2.393 -1.452 9.443 H HETATM 55 C 0 -0.836 4.399 5.367 C HETATM 56 H 0 -1.596 4.464 6.178 H HETATM 57 C 0 4.810 4.861 9.497 C HETATM 58 H 0 5.203 5.607 10.225 H HETATM 59 C 0 1.613 -1.168 3.507 C HETATM 60 H 0 1.283 -1.664 2.568 H HETATM 61 C 0 5.020 6.751 6.128 C HETATM 62 H 0 5.319 7.721 5.675 H HETATM 63 C 0 4.746 5.257 8.072 C HETATM 64 C 0 5.146 6.532 7.524 C HETATM 65 H 0 5.547 7.330 8.189 H HETATM 66 N 0 5.228 2.166 4.184 N HETATM 67 Fe 0 3.324 2.793 4.793 Fe END CONECT 1 15 42 CONECT 2 6 10 67 CONECT 3 13 46 CONECT 4 7 25 CONECT 5 8 9 67 CONECT 6 2 7 17 CONECT 7 4 6 63 CONECT 8 5 11 15 CONECT 9 5 12 20 CONECT 10 2 24 46 CONECT 11 8 27 36 CONECT 12 9 27 32 CONECT 13 3 14 59 CONECT 14 13 CONECT 15 1 8 29 369 Table 6.S53 (cont’d) CONECT 16 46 47 51 CONECT 17 6 38 40 CONECT 18 19 49 66 CONECT 19 18 CONECT 20 9 21 66 CONECT 21 20 30 34 CONECT 22 23 29 55 CONECT 23 22 CONECT 24 10 38 53 CONECT 25 4 26 61 CONECT 26 25 CONECT 27 11 12 28 CONECT 28 27 CONECT 29 22 15 44 CONECT 30 21 31 49 CONECT 31 30 CONECT 32 12 33 34 CONECT 33 32 CONECT 34 32 21 35 CONECT 35 34 CONECT 36 11 37 44 CONECT 37 36 CONECT 38 17 24 39 CONECT 39 38 CONECT 40 17 41 57 CONECT 41 40 CONECT 42 1 43 55 CONECT 43 42 CONECT 44 36 29 45 CONECT 45 44 CONECT 46 16 3 10 CONECT 47 16 48 53 CONECT 48 47 CONECT 49 18 30 50 CONECT 50 49 CONECT 51 16 52 59 CONECT 52 51 CONECT 53 47 24 54 CONECT 54 53 CONECT 55 22 42 56 CONECT 56 55 CONECT 57 40 58 63 CONECT 58 57 CONECT 59 51 13 60 CONECT 60 59 CONECT 61 25 62 64 CONECT 62 61 CONECT 63 7 57 64 CONECT 64 63 61 65 CONECT 65 64 CONECT 66 18 20 CONECT 67 2 5 [Fe(4’-Cl-terpy)2]2+: Table 6.S54: The optimized geometry file for the lowest energy triplet state of [Fe(4’-Cl-terpy)2]2+. TITLE 4’-Cl-terpy Triplet B3LYP10 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.007 3.979 6.159 N HETATM 2 N 0 0.035 1.599 5.486 N HETATM 3 N 0 1.794 5.227 3.949 N HETATM 4 C 0 0.555 3.219 7.234 C HETATM 5 C 0 0.621 3.714 8.570 C HETATM 6 C 0 1.165 5.012 8.779 C HETATM 7 C 0 0.006 1.864 6.852 C HETATM 8 C 0 -0.443 0.393 5.034 C HETATM 9 C 0 -0.506 0.913 7.788 C 370 Table 6.S54 (cont’d) HETATM 10 C 0 -0.968 -0.602 5.908 C HETATM 11 C 0 -0.998 -0.334 7.307 C HETATM 12 C 0 1.536 5.248 6.374 C HETATM 13 C 0 1.630 5.797 7.688 C HETATM 14 C 0 1.983 5.958 5.117 C HETATM 15 C 0 2.167 5.783 2.750 C HETATM 16 C 0 2.552 7.269 5.096 C HETATM 17 C 0 2.740 7.083 2.652 C HETATM 18 C 0 2.934 7.836 3.846 C HETATM 19 N 0 0.772 2.541 2.494 N HETATM 20 N 0 -1.101 3.976 3.721 N HETATM 21 N 0 2.842 2.248 4.141 N HETATM 22 C 0 -0.371 2.767 1.732 C HETATM 23 C 0 -0.484 2.266 0.402 C HETATM 24 C 0 0.606 1.525 -0.133 C HETATM 25 C 0 -1.439 3.574 2.433 C HETATM 26 C 0 -2.001 4.724 4.441 C HETATM 27 C 0 -2.698 3.916 1.849 C HETATM 28 C 0 -3.272 5.104 3.924 C HETATM 29 C 0 -3.624 4.689 2.606 C HETATM 30 C 0 1.833 1.815 1.961 C HETATM 31 C 0 1.779 1.289 0.636 C HETATM 32 C 0 3.007 1.645 2.899 C HETATM 33 C 0 3.854 2.152 5.064 C HETATM 34 C 0 4.199 0.930 2.567 C HETATM 35 C 0 5.070 1.459 4.803 C HETATM 36 C 0 5.242 0.837 3.532 C HETATM 37 H 0 0.260 3.111 9.429 H HETATM 38 H 0 -0.400 0.227 3.935 H HETATM 39 H 0 -0.524 1.135 8.876 H HETATM 40 H 0 -1.342 -1.563 5.491 H HETATM 41 H 0 -1.400 -1.087 8.022 H HETATM 42 H 0 2.054 6.808 7.862 H HETATM 43 H 0 1.998 5.160 1.843 H HETATM 44 H 0 2.698 7.844 6.035 H HETATM 45 H 0 3.025 7.489 1.656 H HETATM 46 H 0 3.379 8.856 3.810 H HETATM 47 H 0 -1.394 2.446 -0.208 H HETATM 48 H 0 -1.683 5.025 5.464 H HETATM 49 H 0 -2.960 3.587 0.820 H HETATM 50 H 0 -3.964 5.710 4.548 H HETATM 51 H 0 -4.610 4.965 2.168 H HETATM 52 H 0 2.625 0.713 0.208 H HETATM 53 H 0 3.678 2.651 6.044 H HETATM 54 H 0 4.320 0.449 1.572 H HETATM 55 H 0 5.859 1.411 5.586 H HETATM 56 H 0 6.178 0.285 3.291 H HETATM 57 Fe 0 0.891 3.259 4.328 Fe HETATM 58 Cl 0 1.266 5.668 10.443 Cl HETATM 59 Cl 0 0.501 0.879 -1.801 Cl END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 371 Table 6.S54 (cont’d) CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 CONECT 58 6 CONECT 59 24 Table 6.S55: The optimized geometry file for the singlet charge transfer state of [Fe(4’-Cl- terpy)2]2+ using td=1. TITLE 4’-Cl-terpy CT s1 B3LYP10 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 0.993 3.996 6.122 N HETATM 2 N 0 0.118 1.675 5.343 N HETATM 3 N 0 1.762 5.078 3.891 N HETATM 4 C 0 0.522 3.202 7.192 C HETATM 5 C 0 0.570 3.741 8.527 C HETATM 6 C 0 1.095 5.040 8.710 C HETATM 7 C 0 0.037 1.893 6.755 C HETATM 8 C 0 -0.300 0.463 4.829 C HETATM 9 C 0 -0.482 0.853 7.613 C HETATM 10 C 0 -0.817 -0.579 5.632 C HETATM 11 C 0 -0.907 -0.373 7.060 C HETATM 12 C 0 1.513 5.268 6.314 C HETATM 13 C 0 1.586 5.841 7.605 C HETATM 14 C 0 1.961 5.893 5.021 C HETATM 15 C 0 2.150 5.547 2.653 C HETATM 16 C 0 2.547 7.187 4.897 C HETATM 17 C 0 2.738 6.826 2.472 C HETATM 18 C 0 2.939 7.661 3.614 C HETATM 19 N 0 0.769 2.576 2.493 N HETATM 20 N 0 -0.999 3.931 3.852 N 372 Table 6.S55 (cont’d) HETATM 21 N 0 2.689 2.305 4.219 N HETATM 22 C 0 -0.376 2.822 1.765 C HETATM 23 C 0 -0.500 2.333 0.432 C HETATM 24 C 0 0.596 1.592 -0.105 C HETATM 25 C 0 -1.397 3.616 2.545 C HETATM 26 C 0 -1.856 4.639 4.659 C HETATM 27 C 0 -2.663 4.023 2.039 C HETATM 28 C 0 -3.137 5.070 4.212 C HETATM 29 C 0 -3.545 4.759 2.884 C HETATM 30 C 0 1.828 1.855 1.982 C HETATM 31 C 0 1.775 1.337 0.654 C HETATM 32 C 0 2.950 1.705 2.980 C HETATM 33 C 0 3.631 2.220 5.216 C HETATM 34 C 0 4.172 1.018 2.733 C HETATM 35 C 0 4.871 1.546 5.030 C HETATM 36 C 0 5.145 0.938 3.772 C HETATM 37 H 0 0.205 3.153 9.396 H HETATM 38 H 0 -0.215 0.339 3.727 H HETATM 39 H 0 -0.538 1.025 8.711 H HETATM 40 H 0 -1.142 -1.530 5.157 H HETATM 41 H 0 -1.306 -1.176 7.720 H HETATM 42 H 0 2.000 6.855 7.780 H HETATM 43 H 0 1.980 4.870 1.789 H HETATM 44 H 0 2.698 7.817 5.800 H HETATM 45 H 0 3.030 7.155 1.451 H HETATM 46 H 0 3.398 8.669 3.506 H HETATM 47 H 0 -1.413 2.513 -0.174 H HETATM 48 H 0 -1.495 4.863 5.685 H HETATM 49 H 0 -2.961 3.771 0.998 H HETATM 50 H 0 -3.795 5.640 4.904 H HETATM 51 H 0 -4.539 5.084 2.504 H HETATM 52 H 0 2.614 0.753 0.220 H HETATM 53 H 0 3.375 2.710 6.180 H HETATM 54 H 0 4.364 0.549 1.744 H HETATM 55 H 0 5.602 1.506 5.867 H HETATM 56 H 0 6.106 0.405 3.596 H HETATM 57 Fe 0 0.889 3.264 4.318 Fe HETATM 58 Cl 0 1.167 5.743 10.368 Cl HETATM 59 Cl 0 0.481 0.960 -1.772 Cl END CONECT 1 4 12 57 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 12 6 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 373 Table 6.S55 (cont’d) CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 1 19 CONECT 58 6 CONECT 59 24 [Fe(4’-OMe-terpy)2]2+: Table 6.S56: The optimized geometry file for the lowest energy triplet state of [Fe(4’-OMe- terpy)2]2+. TITLE 4’-OMe-terpy Triplet B3LYP10 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.081 3.800 6.229 N HETATM 2 N 0 -0.069 1.493 5.568 N HETATM 3 N 0 1.835 5.047 4.011 N HETATM 4 C 0 0.641 3.046 7.317 C HETATM 5 C 0 0.813 3.500 8.648 C HETATM 6 C 0 1.457 4.762 8.884 C HETATM 7 C 0 -0.017 1.737 6.936 C HETATM 8 C 0 -0.648 0.330 5.121 C HETATM 9 C 0 -0.550 0.804 7.880 C HETATM 10 C 0 -1.199 -0.643 6.001 C HETATM 11 C 0 -1.147 -0.398 7.405 C HETATM 12 C 0 1.703 5.022 6.450 C HETATM 13 C 0 1.909 5.534 7.766 C HETATM 14 C 0 2.126 5.736 5.185 C HETATM 15 C 0 2.177 5.611 2.806 C HETATM 16 C 0 2.771 7.012 5.159 C HETATM 17 C 0 2.821 6.877 2.704 C HETATM 18 C 0 3.121 7.587 3.904 C HETATM 19 N 0 0.582 2.434 2.553 N HETATM 20 N 0 -1.138 3.972 3.877 N HETATM 21 N 0 2.705 2.008 4.085 N HETATM 22 C 0 -0.581 2.737 1.843 C HETATM 23 C 0 -0.789 2.258 0.526 C HETATM 24 C 0 0.217 1.441 -0.093 C HETATM 25 C 0 -1.560 3.605 2.604 C HETATM 26 C 0 -1.955 4.764 4.647 C HETATM 27 C 0 -2.822 4.032 2.086 C HETATM 28 C 0 -3.223 5.228 4.197 C 374 Table 6.S56 (cont’d) HETATM 29 C 0 -3.661 4.852 2.893 C HETATM 30 C 0 1.559 1.648 1.959 C HETATM 31 C 0 1.411 1.133 0.637 C HETATM 32 C 0 2.768 1.401 2.835 C HETATM 33 C 0 3.754 1.848 4.957 C HETATM 34 C 0 3.899 0.616 2.446 C HETATM 35 C 0 4.911 1.082 4.637 C HETATM 36 C 0 4.980 0.457 3.358 C HETATM 37 H 0 0.465 2.905 9.519 H HETATM 38 H 0 -0.666 0.185 4.017 H HETATM 39 H 0 -0.503 1.008 8.972 H HETATM 40 H 0 -1.657 -1.569 5.588 H HETATM 41 H 0 -1.567 -1.135 8.126 H HETATM 42 H 0 2.411 6.512 7.909 H HETATM 43 H 0 1.922 5.020 1.898 H HETATM 44 H 0 2.999 7.556 6.101 H HETATM 45 H 0 3.077 7.291 1.704 H HETATM 46 H 0 3.624 8.580 3.866 H HETATM 47 H 0 -1.712 2.497 -0.045 H HETATM 48 H 0 -1.570 5.031 5.657 H HETATM 49 H 0 -3.151 3.731 1.068 H HETATM 50 H 0 -3.847 5.868 4.859 H HETATM 51 H 0 -4.648 5.194 2.505 H HETATM 52 H 0 2.209 0.508 0.189 H HETATM 53 H 0 3.654 2.355 5.944 H HETATM 54 H 0 3.941 0.133 1.446 H HETATM 55 H 0 5.734 0.983 5.380 H HETATM 56 H 0 5.868 -0.151 3.070 H HETATM 57 Fe 0 0.831 3.122 4.389 Fe HETATM 58 O 0 1.587 5.129 10.208 O HETATM 59 O 0 -0.063 1.014 -1.375 O HETATM 60 C 0 2.244 6.422 10.530 C HETATM 61 H 0 2.220 6.482 11.645 H HETATM 62 H 0 3.307 6.425 10.167 H HETATM 63 H 0 1.669 7.278 10.085 H HETATM 64 C 0 0.933 0.162 -2.077 C HETATM 65 H 0 0.473 -0.046 -3.074 H HETATM 66 H 0 1.901 0.715 -2.207 H HETATM 67 H 0 1.094 -0.798 -1.518 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 9 10 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 375 Table 6.S56 (cont’d) CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 58 6 60 CONECT 59 24 64 CONECT 60 58 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 59 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 Table 6.S57: The optimized geometry file for the singlet charge transfer state of [Fe(4’-OMe- terpy)2]2+ using td=1. TITLE 4’-OMe-terpy CT s1 SDD 6311G+dp REMARK 1 File created by GaussView 6.1.1 HETATM 1 N 0 1.062 3.776 6.207 N HETATM 2 N 0 0.031 1.532 5.446 N HETATM 3 N 0 1.813 4.850 3.966 N HETATM 4 C 0 0.597 3.034 7.279 C HETATM 5 C 0 0.719 3.492 8.607 C HETATM 6 C 0 1.357 4.786 8.826 C HETATM 7 C 0 -0.011 1.728 6.831 C HETATM 8 C 0 -0.478 0.363 4.922 C HETATM 9 C 0 -0.576 0.748 7.698 C HETATM 10 C 0 -1.052 -0.650 5.737 C HETATM 11 C 0 -1.104 -0.454 7.148 C HETATM 12 C 0 1.692 5.022 6.387 C HETATM 13 C 0 1.844 5.546 7.732 C HETATM 14 C 0 2.117 5.629 5.133 C HETATM 15 C 0 2.183 5.331 2.726 C HETATM 16 C 0 2.800 6.898 4.985 C HETATM 17 C 0 2.846 6.566 2.539 C HETATM 18 C 0 3.162 7.367 3.707 C HETATM 19 N 0 0.578 2.439 2.577 N HETATM 20 N 0 -1.010 3.945 4.009 N HETATM 21 N 0 2.614 2.089 4.141 N HETATM 22 C 0 -0.599 2.764 1.910 C HETATM 23 C 0 -0.830 2.290 0.601 C HETATM 24 C 0 0.178 1.465 -0.028 C 376 Table 6.S57 (cont’d) HETATM 25 C 0 -1.513 3.640 2.740 C HETATM 26 C 0 -1.753 4.742 4.844 C HETATM 27 C 0 -2.773 4.132 2.300 C HETATM 28 C 0 -3.021 5.269 4.465 C HETATM 29 C 0 -3.538 4.959 3.176 C HETATM 30 C 0 1.559 1.659 1.989 C HETATM 31 C 0 1.394 1.146 0.676 C HETATM 32 C 0 2.748 1.459 2.900 C HETATM 33 C 0 3.638 1.994 5.049 C HETATM 34 C 0 3.916 0.718 2.562 C HETATM 35 C 0 4.834 1.268 4.773 C HETATM 36 C 0 4.974 0.621 3.513 C HETATM 37 H 0 0.357 2.907 9.478 H HETATM 38 H 0 -0.425 0.244 3.819 H HETATM 39 H 0 -0.600 0.927 8.795 H HETATM 40 H 0 -1.450 -1.573 5.261 H HETATM 41 H 0 -1.547 -1.228 7.814 H HETATM 42 H 0 2.337 6.528 7.884 H HETATM 43 H 0 1.925 4.690 1.853 H HETATM 44 H 0 3.038 7.498 5.892 H HETATM 45 H 0 3.112 6.898 1.512 H HETATM 46 H 0 3.687 8.342 3.595 H HETATM 47 H 0 -1.756 2.531 0.037 H HETATM 48 H 0 -1.310 4.958 5.840 H HETATM 49 H 0 -3.155 3.878 1.288 H HETATM 50 H 0 -3.583 5.911 5.178 H HETATM 51 H 0 -4.526 5.355 2.848 H HETATM 52 H 0 2.183 0.521 0.211 H HETATM 53 H 0 3.488 2.516 6.019 H HETATM 54 H 0 4.003 0.223 1.571 H HETATM 55 H 0 5.635 1.220 5.543 H HETATM 56 H 0 5.894 0.046 3.268 H HETATM 57 Fe 0 0.844 3.099 4.367 Fe HETATM 58 O 0 1.434 5.163 10.162 O HETATM 59 O 0 -0.123 1.045 -1.292 O HETATM 60 C 0 2.074 6.458 10.488 C HETATM 61 H 0 2.014 6.536 11.602 H HETATM 62 H 0 3.149 6.464 10.160 H HETATM 63 H 0 1.515 7.309 10.012 H HETATM 64 C 0 0.852 0.190 -2.033 C HETATM 65 H 0 0.359 -0.006 -3.015 H HETATM 66 H 0 1.815 0.745 -2.185 H HETATM 67 H 0 1.020 -0.772 -1.482 H END CONECT 1 4 12 CONECT 2 7 8 CONECT 3 14 15 CONECT 4 1 5 7 CONECT 5 4 6 37 CONECT 6 5 13 58 CONECT 7 2 4 9 CONECT 8 2 10 38 CONECT 9 7 11 39 CONECT 10 8 11 40 CONECT 11 10 9 41 CONECT 12 1 13 14 CONECT 13 6 12 42 CONECT 14 3 12 16 CONECT 15 3 17 43 CONECT 16 14 18 44 CONECT 17 15 18 45 CONECT 18 16 17 46 CONECT 19 22 30 57 CONECT 20 25 26 CONECT 21 32 33 CONECT 22 19 23 25 CONECT 23 22 24 47 CONECT 24 23 31 59 377 Table 6.S57 (cont’d) CONECT 25 20 22 27 CONECT 26 20 28 48 CONECT 27 25 29 49 CONECT 28 26 29 50 CONECT 29 27 28 51 CONECT 30 19 31 32 CONECT 31 24 30 52 CONECT 32 21 30 34 CONECT 33 21 35 53 CONECT 34 32 36 54 CONECT 35 33 36 55 CONECT 36 34 35 56 CONECT 37 5 CONECT 38 8 CONECT 39 9 CONECT 40 10 CONECT 41 11 CONECT 42 13 CONECT 43 15 CONECT 44 16 CONECT 45 17 CONECT 46 18 CONECT 47 23 CONECT 48 26 CONECT 49 27 CONECT 50 28 CONECT 51 29 CONECT 52 31 CONECT 53 33 CONECT 54 34 CONECT 55 35 CONECT 56 36 CONECT 57 19 CONECT 58 6 60 CONECT 59 24 64 CONECT 60 58 61 62 63 CONECT 61 60 CONECT 62 60 CONECT 63 60 CONECT 64 59 65 66 67 CONECT 65 64 CONECT 66 64 CONECT 67 64 378