PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE I n I i - lrfirtti MSU Is An Affirmative Action/Equal Opportunity Institution _ ___.—-_ .»_.—-—_..__ EXCITE. EXCITED-STATE PROPERTIES OF TRANSITION-METAL COMPLEXES IN SOLUTION AND THE SOLID STATE BY Mark David Nevehem A DISSERTATION submitted to Michigan state University in partial fulfillment of the requirements for the degree or DOCTOR OP PHILOSOPHY Department of Chemistry 1988 EXCITED- ABSTRACT EXCITED-STATE PROPERTIES OF TRANSITION-METAL COMPLEXES IN SOLUTION AND THE SOLID STATE BY Mark David Newsham The intra- and intermolecular excited-state nonradiative decay processes of two classes of luminescent transition-metal complexes, the M6 cluster systems of molybdenum (II) and tungsten (II) , and the t_r_a_n§- dioxorhenium(V) complexes Re02L4z (L = pyridine, CN'; 2 = +1, -3, respectively), in homogeneous solution and the solid state are investigated. Intramolecular nonradiative decay of the M6 clusters is primarily controlled by a metal- localized deactivating mode. An emission bandshape analysis of [M06C18]C162' assigns the dominant accepting mode to be a 14-14 core vibration with a minor contribution from a [Méxs]4+ (M = M0, W: X = halide) breathing mode. These low-energy vibrational modes are manifested in extremely long emission lifetimes and high quantum yields of these systems. The long-1 ived emissive excited states of the M6 systems are efficientlY quen are investigate‘ pyridinium-relab behavior are 0 :cntrolled elem :‘e'.'iation of m icn-pair format mashing betvee pcsitive deviati range electrost. 'ze effect of :‘ecay processes fiscussed. p ::n:aining the itinescence of :2th b" ' “Meal 5‘; E'Er Ed OXides 103 h ‘ l 112.! Mark David Newsham efficiently quenched by electron-transfer reactions, which are investigated by the Stern-Volmer (SV) method using pyridinium-related quenchers. Deviations from classical SV behavior are observed in the activated and diffusion- controlled electron-transfer rate regimes. The negative deviation of the former regime is ascribed to luminescent ion-pair formation and, by using a recent model for quenching between charged reactants, it has been shown that positive deviations in the latter regime arise from long- range electrostatic interactions in homogeneous solution. The effect of heterogeneous environments on nonradiative decay processes of electronically excited M6 systems is discussed. Polymer films and ceramic-oxide matrices containing the hexanuclear cluster retain the intense luminescence of the M6 core in these reactive environments. photophysical studies on these two cluster materials show that guest-host interactions in luminescent solid-state materials cannot be evaluated owing to the amorphous nature of the host. In an effort to better define the perturbations of solid-state environments on luminescent guest trans it ion-metal complexes , the photophysical properties of the m-ReOZT core in crystalline complex- layered oxides (CLOs) are investigated. The layered silicate CLOs hectorite and fluorohectorite adsorb LIED—S- + o o 0 R302 (py) 4 to their negatively charged interlayers and a :‘icxorhenium C in“ (CV) 3" b« .eez A 4 like Hg/Al do at these thr perturbed by :‘isparate pro rteractions 3er core an Mark David Newsham dioxorhenium CLO of complementary charge incorporates trans- Reoz (CN)43' between the positive layers of a hydrotalcite- like Mg/Al double hydroxide. The photophysical properties of these three Re02+-intercalated CLOs are significantly perturbed by the host oxide to different extents. These disparate properties are attributed to specific guest-host interactions which mediate the reaction between the t_ran_s- Re02+ core and water in CLO interlayers. To My Mom and Dad :rcv both been 3 gr great fr: finp‘v‘. "“.I‘E 'VU-ou‘ th‘b life in c‘: :leege 1 fzsthall g t° 3r. C. lv~.?.lng Or 2v- va. Bhe re "Widing I dcr 1.:‘1‘ ‘M‘enCe ‘ 3:! , ACKNOWLEDGEMENTS I owe a great deal to so many people during the last 5 years at MSU. Above all, I would like to thank Dan Nocera for his unending guidance and support which have helped me grow both as a person, and as a scientist. Not only has Dan been a great adviser, but more importantly, he has been a great friend (and an adequate left fielder) and has contributed to numerous unique experiences outside of my life in chemistry. Indeed, there are very few advisers, or college freshman for that matter, that could enjoy a football game the way Dan does. I would also like to thank John Allison for his support, especially during my first two years at MSU. Special thanks to Emmanuel Giannelis and Dr. Pinnavaia for their collaboration on the rhenium-oxo paper, to Kris Berglund and Mike Cerreta for their sol-gel technology, and to Dr. Cukier for the countless number of hours spent working on the electron-transfer quenching project. I-Jy and the rest of the Nocera group deserve a lot of credit for providing a scientifically stimulating working atmosphere. I don't know how I can even begin to describe the influence that "seven marguerites" Bob, "what are you gettin at" Joe, "sprained thumbs" Dan, and "green shorts" Randy have had on my life in East Lansing (and Windsor, Chicago, New Orleans, Toronto, etc.). We’ve had so many good times, I can’t recall them all (probably for good reason). Who could forget the football, basketball, and hockey games, vi parties, Jason’s, Omar's, and downtown East Lansing. Perhaps I could have graduated one year earlier if it were not for these four great friends. Finally, my family deserves the biggest acknowledgement of all; their support of my career has been tremendous. Especially my Mom and Dad, who I can’t even begin to thank. It sure is great having parents like them that have helped me so much for so long. vii TABLE OF CONTENTS Page LIST OF TABLES...................... ...... .... ..... xi LIST OF FIGURES........................ ............ xv CHAPTER I - INTRODUCTION........................... 1 CHAPTER II - EXPERIMENTAL...... ..... ...... ......... 22 A. Materials................................ 22 l. Hexanuclear Clusters.. .............. 22 2. Cluster Modified Polymers... ........ 28 3. Trans-Dioxorhenium(V) Compounds ..... 29 4. Complex-Layered Oxide Dioxorhenium(V) Intercalates........ 29 5. Quenchers................... ........ 30 6. Supporting Electrolyte..... ......... 31 7. Solvents... ...... . .................. 31 B. Methods........................ .......... 32 1. Spectroscopic Measurements.. ........ 32 2. Quantum Yield Measurements.......... 32 3. Molar Absorptivity Measurements ..... 34 4. Quenching Experiments............... 34 i. Hexanuclear Clusters. ...... .... 34 ii. Complex-Layered Oxide Dioxorhenium(V) Intercalates... 35 5. Electrochemical Measurements........ 35 C. Instrumentation.......................... 36 1. High-Resolution Emission Spectrometer........................ 36 viii SEWER III - S? A. MC CL Introd“ Intram: 1. B- 2 Re Intern: 1 Be 2 Re Cluster 1 Ba 2 Re i. ii 5:) y,“ “QR Iv - TR IL 1” IL 0? IntrOdLI 2. Time-Resolved Luminescence Spectrometer................ ........ 3. Cryogenic Refrigeration System...... 4. Other Instrumentation........ ....... CHAPTER III - SPECTROSCOPY OF HEXANUCLEAR CHAPTER IV MOLYBDENUM(II) AND TUNGSTEN(II) CLUSTERSOO0.0000000000000000.0 ...... Introduction............................. Intramolecular Photophysical Properties.. 1. Background ......... ... .............. 2. Results and Discussion... ........... Intermolecular Quenching Processes....... 1. Background.................. ........ 2. Results and Discussion.. ............ Cluster Modified Polymers.. ...... ........ 1. Background........ .................. 2. Results and Discussion. ............. 1. ‘Organic Polyvinylpyridine Polymers................. ...... ii. Inorganic Silicon-Oxide Polymers................. ...... - THE INFLUENCE OF GUEST-HOST INTERACTIONS ON THE EXCITED-STATE PROPERTIES OF DIOXORHENIUM(V) IONS IN INTRACRYSTALLINE ENVIRONMENTS OF COMPLEX-LAYERED OXIDES..... ...... IntrOduction...........0.0.0000... ....... ix Page 46 50 51 52 52 66 66 79 117 117 126 161 161 165 165 170 176 176 B. Resul C. Discx mmv - F1! “'§F\" A: T L.‘ MIX ...... I--!‘ 'fl'h 1;:;:.....\,Es _ . _ . Page B0 ResultSOO..........OOOOOOOOOOOOOOOO....O. 179 1. Synthesis and Characterization...... 179 2. Electronic Absorption and Emission Spectroscopy............... 187 C. Discussion..................... ..... ..... 201 CHAPTERV-FINALREMARKS.................... ...... 215 APPENDIXCOOOOOOOOOOO0............OOOOOOOOOOOOOOOOO. 218 REFERENCES.......OOOOOOOOOOOOOOOOOO......OOOOOO.... 221 Ca.) ...». S"- M‘bStie. LIST OF TABLES Table Page 1 Emission Spectroscopic Data of (NBu4)2M5X14. .................................... 60 2 Electrochemical Properties of (NBu4) 2M6x14 0 o o e e eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 63 3 Emission Spectroscopic Data of All-Halide M06 Clusters... ................................... 80 4 Emission Spectroscopic Data of Organic Substituted Mo6 Clusters.... ..................... 81 5 Excited-State Decay Rates of M06 Clusters in Acetonitrile... ................... . .............. 82 6 Excited-State Decay Rates of M06 Clusters in Dichloromethane.......................... ........ 83 7 Emission Spectroscopic Data of Alkoxide Substituted Na2M06C18(OR)6 Clusters....... ....... 90 xi U) (H 4 Exiss in A: Excit in 8: £215: in t1 EXCi‘ in t] 10 11 12 13 14 15 Temperature-Dependent Second-Moment Data of (NBU4)2M06C114..............o..........o.oo...... Emission Spectroscopic Data of (NBu4)2W6X8Y6 in Acetonitrile............................. ..... Emission Spectroscopic Data of (NBu4)2M6X8Y6 in Butyronitrile at 300 K and 77 K. .............. Excited-State Decay Rates of (NBu4)2M6X8Y6 in Butyronitrile at 77 K. ....... ........ ......... Emission Spectroscopic Data of (NBu4)2M6X8Y6 in the Solid State at 300 K and 77 K... .......... Excited-State Decay Rates of (NBu4)2M6X8Y6 in the Solid State at 300 K ..... ...... ........... Electrochemical Properties of Ground-State and Electronically Excited (NBu4)2M6X14 in Acetone........... ...... ............. ............ Reduction Potentials of Pyridinium and Bipyridinium Quenchers in Acetone... ............. xii Page 96 103 107 109 115 116 127 128 \t" 1‘ 4.; Drivir Transi Bicvrj .‘ 16 17 18 19 20 21 22 Page Driving Forces and Quenching Rate Constants of M6x142-/Pyridinium Systems in Acetone... ...... 130 Driving Forces and Calculated Electron- Transfer Rate Constants of M6X142'/ Bipyridinium Systems in Acetone... ............... 143 Diffusion Coefficients and Effective Distances Determined by Stern-Volmer Fits, and Diffusion Coefficients Measured by Potential Step Chronoamperometry.............. .................. 150 Effective Distances Determined from Stern-Volmer Fits and from Debye-Hfickel Calculations as a Function of Ionic Strength.. ................... 156 Emission Lifetimes of Cluster-Incorporated Polyvinylpyridine Polymers........... ...... ...... 166 Emission Lifetimes and Quantum Yields of M06 Clusters in Methanolic and Sol-Gel EnVironments.......OOOOOOOOOOOCOOOOO......OOOOOOO 173 Idealized Structural Formulas of Complex- Layered Oxides....................... ............ 180 xiii 23 K.) n"- Solid, Lifeti: Emissic ontair 23 24 Page Solid, Solution, and Intercalate Luminescence Lifetimes of the Re02(py)4+ Ion.................. 195 Emission Lifetime of Re02(py)4-Hectorite + Containing Co-Intercalated Re02(en)2 Ions ....... 203 xiv Figure A . lL‘ .o- Schema funda: Solid Single triple abbrew inters LIST OF FIGURES Figure 1 Schematic energy level diagram showing the fundamental intramolecular decay pathways. Solid and dashed lines represent radiative and nonradiative processes, respectively. Singlet states are denoted by 8’5 and triplet states by T's. IC and ISC are abbreviations for internal conversion and intersystem crossing, respectively ............... Schematic representation of the simplified molecular orbital description of the three types of transition-metal excited states: (a) ligand field (LF); (b) intraligand (IL); and (c) charge transfer (CT) ..................... Diagram of the three most common types of quenching pathways: (a) electron transfer, (b) photochemical reaction, and (c) energy transfer........................... ...... . ....... Schematic diagram of the high-resolution emission spectrometer..................... ....... XV Page 13 37 Block diaqr emission s; The chip ar. appropriate Spectral ir standard tu Schematic d luminescenc IdealiZEd 1 0 = “C(11) m£CtrOnic H6Xl42- ic CCEIOnitr: Mo 2 - 6C114 (-... . )w6 “DISCular Modified in aesto] ( QIQCtrO: EMCitEd~ Page Block diagram of the hardware interface for the emission spectrometer to the Zenith computer. The chip and pin numbers are indicated by the appropriate components. ............ .............. 40 Spectral irradiance profile of the Optronics standard tungsten halogen lamp ................... 44 Schematic diagram of the time-resolved luminescence spectrometer ........................ 47 Idealized structure of M6X8Y62- ions: 0 = Mo(II), W(II) ................................ 53 Electronic absorption and emission spectra of M6X142' ions, as tetrabutylammonium salts in acetonitrile at room temperature: (a) (----) 2-. 00.0 2-. ____ 2-. M06C114 p ( ) M068r14 ' (b) ( ) W6C114 , ('°°°) W68r142'; (———) W61142’..... ............... 55 Molecular orbital diagram of M6X142' ions... ..... 58 Modified Latimer diagram of (NBu4)2M06Cl14 in acetonitrile at room temperature (electrode potentials in V vs SCE; excited-state energy in eV)...................... 64 xvi H '4 A (3" Potential e excited-sta and definit Potential e coupling; ' Plot of In clusters in All-halide PlOt of In Clusters ir Table 6. 1 substituteg Ir reSpeCt: Plot 0f I212 data are are disola on.) The (\) he o...) Le 12 13 14 15 16 Potential energy diagram for nonradiative excited-state decay. and definitions...... ............. ...... ....... .. See text for abbreviations Potential energy diagram of the: (a) weak- coupling: (b) strong-coupling limit .............. Plot of ln knr vs A em,max of the M06 clusters in acetonitrile shown in Table 5. All-halide and organic-substituted clusters are represented by u and I, respectively ......... Plot of In knr vs A em,max of the M06 clusters in dichloromethane shown in Table 6. All-halide and organic- substituted clusters are represented by u and -, respectively ..... . .............. . ............. Plot of m2 vs ZkBT of (NBu4)2M06C114 with single mode theoretical fits. The experimental data are represented by a and the fits to eq 28 are displayed for the parameters as follows: (0 e e e) hweff <-—> twee: ("") 1.)“err 320 cm'l, seff 184 cm'l, seff 120 CID-1 , Seff xvii 157 38; 66. O 0000000000000 Page 68 71 85 87 97 l7 Plot of 1:12 21 with the t'. experimenta the solid C eq 29 with ilwz= 320 c Plot of In (H8114)2W6XE1 shown in T5 MOI: of In (N8u4) ZMOG) at 77 K sh: “(it of In EXempl a ry g ( (III) invei 17 18 19 20 21 Page Plot of m2 vs 2kBT of (NBu4)2M060l14 with the two mode theoretical fit. The experimental data are represented by a and the solid curve corresponds to the fit to eq 29 with hwl = 120 cm'l, 51 = 39 and hmz = 320 cm’l, $2 = 6 ........................... 100 Plot of ln knr vs A of the em,max (NBu4)2W6X8Y6 clusters in acetonitrile shown in Table 9. .......... ........... ........... 105 Plot of 1n knr vs A of the em,max (NBu4)2M06X8Y6 clusters in butyronitrile at 77 K shown in Table 11 ........................ 110 Plot of 1n knr vs Aem,max of the (NBu4)2W6X8Y6 clusters in butyronitrile at 77 K shown in Table 11............ ............ 112 Exemplary plot of RT ln ke-t vs AG° showing three rate regimes: (I) normal: (II) diffusion-controlled; and (III) inverted regime............................ 119 xviii 22 ‘0 £3 Stern-Volt: 2-: "6C114 at low conc (I: 0.001 Stern-Vole: 2-. "6C114 at high Co Stern-Volt: bEt'u’een 2 ' aCEtone at 0'1 ( 0 ) i pyridiniu‘ solid Cur kqio) == 8 The dotte com-011.E 22 23 24 25 Stern-Volmer plots of M06C1142'* (o) and 2-* (e) quenching by 2,2’-bipyridinium at low concentrations in acetone (I = 0.001 M) .................................... Stern-Volmer plots of M06C1142"* (o) and W6C1142'* (I) quenching by 2,2'-bipyridinium at high concentrations in acetone (I = 0.001 M).. Stern-Volmer plots of quenching reaction between 2,2'-bipyridinium and M06C1142'* in acetone at ionic strengths of 0.001 (a) and 0.1 (<0) M ....................................... Plot of RT ln kq vs -AG° for M6x142'*/ pyridinium systems shown in Table 16. The solid curve is the plot of eq 30 with kq(0) = 8.8 x 108 n'ls‘l and A = 1.04 eV. The dotted line represents the diffusion- controlled limit (~ 2 x 1010 M'ls'l) ............. xix Page 131 134 136 139 26 Stern-Volme between W6] (b) l-deute (c) l-nethf experiment; best fits l Stern-Vol: between we bipyridini 0.011 (+) . experiment Stern-vol: between we bipyridim‘ o), and c. strength c el'fperiment are “St 1 Emission c m methanc SiliCon‘o) Ho 6C112(05 26 27 28 29 Stern-Volmer plots of quenching reaction between W61142-* and (a) 2,2'—bipyridinium: (b) 1-deuterium-2,2’-bipyridinium: and (c) 1-methyl-2,2’-bipyridinium. a represent experimental data and the solid curves are best fits by using eq 40 (see text).............. Stern-Volmer plots of quenching reaction between W6I142'* and 1-methyl-2,2'- bipyridinium at ionic strengths of 0.001 (a), 0.011 (+), and 0.102 M (e). Symbols represent experimental data and solid curves are best fits by using eq 40 (see text) ................... Stern-Volmer plots of quenching reaction between W6Il42-* and 1-methy1-2,2’- bipyridinium in acetone (a), acetonitrile (+), and dichloromethane (e) at an ionic strength of 0.001 M. Symbols represent experimental data and solid curves are best fits by using eq 40 (see text) .......... Emission spectra of: (-———) M06C112 in methanol; (----) M06C112 in silicon-oxide gel: ("”) M06C112(OSi(CH3)3)22 in methanol ....... ......... XX Page 144 153 157 171 30 O...) ...—A Infrared sfi‘ (a) (""i (—-) Rec: fluoronect: X'ray patte (3) R602 (p) hectorite; Electronic 0f nonaqde (a) (\) in") Rec fluorOhect UHF; (-_‘I Low tempe Solid: (a hectorite Lifetime beetOrite the best equation 30 31 32 33 34 Infrared spectra on KBr pellets of: (a) (----) [R802(PY)4]I3 0————) Re02(py)4-hectorite: (....) Re02(py)4- fluorohectorite: (b) (----) K3[Re02(CN)4]: (————) Re02(CN)4-hydrotalcite......... ........... X-ray patterns of the three CLO intercalates: (a) ReOz(py)4-hectorite: (b) Re02(py)4-f1uoro- hectorite; (c) ReOZ(CN)4-hydrotalcite ............ Electronic absorption and emission spectra of nonaqueous solutions of the following: (a) (————) [Re02(py)4]I in pyridine; (----) Re02(py)4-hectorite; (....) Re02(py)4- fluorohectorite: (b) (————) K3[Re02(CN)4] in DMF; (----) Re02(CN)4-hydrotalcite ............... Low temperature (9 K) emission spectra of solid: (a) [Re02(py)4]I; (b) Re02(py)4- hectorite: (c) ReOZ(py)4-fluorohectorite ......... Lifetime decay curve of solid Re02(py)4- hectorite. The smooth curve represents the best fit to the biexponential rate equation.‘ ..... 0.... OOOOOOOOOOOOOOO 00...... ...... xxi Page 182 185 188 192 196 Stern-Vol componen1 (I) with in DMF (< Model (19} axes for Proposed rhenium(' (b) fluo the idea Work of CoiPrisi and C91] PGCQsE, rhenim The hi’d rEpreSe iOnS by EPR SP6 35 36 37 38 39 Stern-Volmer plot of the long lifetime component of Re02(py)4-hectorite in DMF (u) with H20 as the quencher and Re02(py)4+ in DMF (o) with H20 as the quencher.... .......... Model depicting the C4 and C2" molecular axes for Re02(py)4+ .............................. Proposed orientations of the trans-dioxo- rhenium(V) core in (a) hectorite and (b) fluorohectorite. The circles represent the idealized geometry for the oxygen frame- work of the LSC galleries. The oxygens comprising a hexagonal cavity of the CLO floor and ceiling are indicated by black circles ....... Proposed orientation of the trans-dioxo- rhenium(V) core in an idealized LDH gallery. The hydroxyl oxygens composing the LDH are represented by large circles and the metal ions by the smaller circles ...................... EPR spectra of electrochemically generated (a) M06C114' and (b) Mo6Cll43' in frozen dichloromethane solution at 4 K.. ................ xxii Page 199 205 208 212 219 CHAPTER I INTRODUCTION Electroni been used ex electron-trans reduction chen This latter i: of schemes photocatalysts transition-ram aerqed more areas of Optical tran: :pti:s.15 In .3359 and . PCLarity, rig lie enviromne :learli’. the P'Ctential a; electronicall “ierstandinc i" “pertiES Electronically excited transition-metal complexes have been used extensively to study fundamental aspects of electron-transfer reactions and to effect oxidation- reduction chemistry unique to the excited-state molecule.1'5 This latter issue has formed the foundation for the design of schemes employing transition-metal complexes as photocatalysts.6'11 The possibility of using excited transition-metal complexes for practical applications has emerged more recently. Potential applications lie in the areas of solid-state lasers,12 Optical and electro- 13 14 optical transducers, and nonlinear sensing devices, optics.15 In addition, luminescent molecules may be used to probe and characterize molecular properties such as polarity, rigidity, acidity, and basicity of surfaces and of the environments of amorphous and crystalline solids.16'17 Clearly, the development of new chemical reactivity and the potential applications of transition-metal complexes in electronically excited states demands a thorough understanding of the factors that govern the excited-state properties in a variety of solution and solid-state environments. Several fundamental processes determine the fate of a molecule after absorbing a photon. A general model displaying excited-state processes is depicted by the Jablonski diagram in Figure 1. Excitation occurs via a Spin-allowed Franck-Condon transition: because most tIz‘ansition-metal ground states are spin singlets, the l Figure 1. Schematic energy level diagram showing the fundamental intramolecular’ decay’ pathways. Solid and dashed lines represent radiative' and nonradiative processes, respectively. Singlet states are denoted by S's and triplet states by T's. IC and ISC are abbreviations for internal conversion and intersystem crossing, respectively. n E” ITATION XC E FLUORESCENC 55 F- mozwommeDDm ZO_.P<._._0xw Figure 1 initially pre] Internal com produce then: and differen population 0 directly on 1 two conpetit excited state PhCSphorescen t0 ground st energy. Be Substantial Processes ar Reduction 01 excited.“at e PhCSphoresCen transition-me The eXCi are related 1 iecay Channel lifetime, k r . initially prepared state is primarily of singlet character. Internal conversion (IC) and intersystem crossing (ISC) produce thermally-equilibrated excited states of the same and different spin multiplicities, respectively. The population of these different excited states depends directly on the relative magnitudes of the rates of these two competitive kinetic channels. Once produced, the excited state may radiatively decay by fluorescence and/or phosphorescence, or it may simply relax nonradiatively back to ground state with the concurrent release of thermal energy. Because most transition-metal complexes have substantial spin-orbit coupling, intersystem crossing processes are typically very fast, resulting in the production of triplet excited states. Consequently, the excited-state lifetime will usually be characterized by the phosphorescent lifetime if emission is observed from a transition-metal complex. The excited-state lifetime and emission quantum yield are related intrinsically to the radiative and nonradiative decay channels by eq 1 and 2 where r is the excited-state o lifetime, kr and knr are the radiative and nonradiative rate constants, respectively, and ¢e, the quantum yield, is a measure of the efficiency of emission (the number of photons To = 1/(kr + knr) (1) ¢e = kr'o (2) emit $2.23: emitted per number of photons absorbed). Because the successful utilization of electronically excited molecules requires long-lived excited states and high emission quantum yields , trans ition-metal complexes possessing large radiative rates (with respect to nonradiative rates) are generally sought for most excited-state chemistry. The radiative rate constant is related to the oscillator strength for the transition from electronic state 1 to 2 (£12) as follows,9 £12 = 4.32 x 10’9F f edu (3) where F is a correction factor for the refractive index of the medium and the measurable molar extinction coefficient (6) is integrated over the energy limits defined by the absorption band. Nonradiative processes occur by isoenergetic electronic transitions from the upper electronic excited state to a vibrationally excited ground state, followed by relaxation to ground state with the medium absorbing the dissipated vibrational energy. For a series of similar transition- metal complexes possessing similar electronic structures, the oscillator strength, and hence, the radiative decay rates, remain relatively constant. However, nonradiative decay channels are perturbed by intramolecular (electronic structure related) and intermolecular (environmental) processes. The perturbations resulting from intramolecular truc hvnn revr these explo “3"“ V“ Alla elect processes can be manipulated by chemically tailoring the first coordination sphere of the transition-metal complex, while the intermolecular perturbations are affected by the structure of the second coordination sphere and bulk properties of the environment. A thorough understanding of these factors is essential for the chemical and practical exploitation of light-driven systems. The contributions of different intramolecular decay channels can only adequately be evaluated within an electronic structural framework. The electronic structures of transition-metal complexes are perhaps most simply understood by using molecular orbital theory in which excitation results in the promotion of a single electron between two orbitals of different parentages. This approach gives rise to three distinct excited states. First, if the parentages of the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) are primarily of metal character, transitions between metal— based orbitals will result. These are designated as ligand field (LF) or drd transitions. Conversely, an intraligand (IL) transition results from promotion of an electron between two ligand-based orbitals. Finally, an electron can be promoted between ligand- and metal-based orbitals which results in a transfer of charge from the metal to the ligand, or vice versa. Although the classic types of charge transfer excited states are metal-to-ligand (MLCT) and ligand-to-metal (LMCT), transition-metal complexes possessing metal-to-metal and second-sphere charge transfer excited states have been more recently identified. These three types of excited states are shown in Figure 2. The ligand-based orbitals are either 0, or, or 1* in character; metal-based d-orbital manifolds are typically composed of nonbonding (nb), 1r, or «* orbitals, and at higher energy are 0* orbitals. Figure 2a shows the configuration for a LF excited-state complex. The HOMO is contained in the metal-d1, d1r*, nb manifold. Depending on the parentage of the LUMO, two deactivation pathways from this excited state can result. Excitation into a metal 0* orbital provides a very efficient nonradiative decay channel via ligand dissociation or isomerization and hence these transitions do not usually lead to luminescence. Alternatively, population of a LUMO residing in the metal- dx, d1r*, nb manifold will produce only slight changes in molecular bond lengths and angles. For this case, the radiative processes can compete effectively with the radiationless processes and luminescence can be observed. The prototypes for such transitions are those of chromium (III) complexes . 18 Metal-to-metal charge transfer (MMCT), or intervalence transfer (IT), are also types of LF transitions, and arise in complexes containing two or more metals in which electronic transitions occur between different metal-centered orbitals. These excited states are usually derived from metal orbitals which are not strongly coupled, such as the 6 and 6* orbitals in multiply-bonded Figure 2. Schematic representation of the simplified molecular orbital description of the three types of transition-metal excited states: (a) ligand field (LF); (b) intraligand (IL); and (c) charge transfer (CT). MdAstMEmHh. § :ch§ CO_~G_OOmw_U : ..o..-_2_ _ .5— n m... h... u 1 L e a ha 1.. .3 we. so» .91 = M. a ensure r9511" 2: 1 ‘81 a d .‘h N4 1 a 3 €§ .: 512$ s§ cEmflEm e _ _ .3; _ 5...). K (HO—2.. JN mwmmzék m0m5 23 acid salt was heated to 100 °C for ~ 1.5 h in an oil bath under reduced pressure. The final product, Mo6C112, was obtained by heating the resulting material at 210 'C in the Lindberg furnace, again under reduced pressure. Mo6Br12 was obtained by a similar procedure to M06C112 except NaCl and AlCl3 were replaced by 14.6 g of NaBr and 26.7 g of AlBr3, respectively. HBr was used instead of HCl in the purification of the product. The tetrabutylammonium salts of Mo6X8Y62' (X = Cl, Br: Y = Cl, Br, I) were prepared by a modification of Sheldon's procedure.47 (NBu4)2Mo6Cll4 (Bu = n-C4H9) was obtained by dissolving Mo6C112 in 6 M HCl, filtering this solution, and adding tetrabutylammonium chloride, which precipitated the solid product. This crude compound was dissolved in dichloromethane and filtered to remove insoluble impurities. Large orange crystals then formed as the solvent slowly evaporated, which were collected by filtration. The tetrabutylammonium salts of Mo6C18Y62' (Y = Br, I) were prepared similarly except the appropriate hydrohalogeno acids and tetrabutylammonium salts were used. (NBu4)2MosBr8Y6 (Y = Cl, Br, I) complexes were prepared as described for Mo6C18Y62' except Mo6Br12 was used as the starting cluster. The tetrabutylammonium salts of M°6XnY14-n2- (n = 9-13) were obtained from Robert D. Mussell and their detailed preparation has been described elsewhere.48 the rea< was} 53711 C) I“ A) UV .— 24 M06C112(PMe3)2, M06C112(PPhMe2)2, Mo6C112(PPh2Me)2, and M05C112(PPh2Et)2 (Me = CH3, Et = CHZCH3, Ph = phenyl) were prepared by a similar procedure as that described by Walton e1; 11.49 In a typical reaction, 0.1 g of Mo6C112 was dissolved in 20 mL of degassed absolute ethanol and ~ 0.2 mL of the appropriate phosphine (all phosphine ligands were purchased from Strem Chemicals) was syringed into the solution under an argon atmosphere. Upon addition of the phosphine, a yellow precipitate formed immediately, however the solution was refluxed for ~ 0.5 h to insure complete reaction. The resulting product was filtered under argon, washed with petroleum ether, dried in vacuo, and stored in a glove box. 1335- and _c_i_§-[Mo6C112(PPr3)2] (Pr = CHZCHZCHB) were synthesized according to literature procedures.50 M06C112 (0.5 g) and ~ 0.8 mL of tgi-n-propylphosphine were added to 25 mL of THF under an argon atmosphere. This solution was stirred for ~ 2 days during which time the solution had turned clear yellow-orange. The solvent was removed under reduced pressure and the residue was washed with hexane, and dried in vacuo. This crude product was dissolved in chloroform ( 150 mL) and chromatographed with a chloroform:absolute ethanol (20:1) solvent mixture on a Florisil (Aldrich) column. The first pale yellow band was t1§n§r[Mo6C112(PPr3)2] and the second pale yellow band was identified as gis-[Mo6C112(PPr3)2]. Exams-[M06C112(P8u3)2] was synthesized in a similar' manner to that of (trans- fl!!! UV. M. h.“ ‘9. (l7 (I! (I ll) 25 [Mo6C112(PPr3)2], except m-n-butylphosphine was used in place of tri-n-propylphosphine. The cis isomer of this compound did not elute from the Florisil column. Mo6Clll(PPr3)3 was prepared by a modified procedure described previously.49 M06C112 (0.1 g) was dissolved in lOmL. of degassed absolute ethanol by refluxing a MoGCllz/ethanol slurry. 131-n-propylphosphine (0.5 mL) was added to this solution and a pale yellow solid immediately formed. However, after ~ 5 min of refluxing, the yellow solid had dissolved and the solution turned orange. This orange solution was refluxed for 24 h during which an orange precipitate formed. The precipitate was collected under an argon atmosphere, washed with absolute ethanol and petroleum ether, and transferred to a glove box. Na2[M06C18(OMe)6] and Na2[Mo6C18(OEt)5] were synthesized according to previously described methods.51 Na2[Mo5C18(OCD3)6] was synthesized similarly except M06C112 was dissolved in CD3OD (Aldrich) and a sodium d3-methoxide solution, made from sodium and CD3OD, was added. Several attempts were made to synthesize Na2[M06(OMe)14] , however, this compound could not be obtained by published methods.51 In a typical reaction, to a slurry of 0.5 g of M06C112 in 2mL of methanol (distilled from sodium), was added 4 mL of a sodium methoxide solution prepared by dissolving ~ 0.5 g of sodium in 10 mL of methanol. This solution was heated in an oil bath to 150 °C under reduced pressure to distill off the solvent, leaving a dark brown residue. Upon cooling, 4 mL f1 26 of methanol were added, stirred for ~ 0.5 h, and filtered to remove sodium chloride. At this point, the filtrate appeared murky dark brown and although a solid was obtained upon addition of ether, a pure crystalline product was not obtained. Removal of solvent under reduced pressure yielded a black-brown solid ‘with the texture of ‘mud. Several attempts were made to prepare the fully substituted methoxide cluster employing a variety of conditions including large scale reactions starting with 5 g of Mo6C112, varying the temperature the original slurry was heated to, and varying the time heated. However the compound was never obtained. In addition, the synthesis of Na2[W6C18(OMe)6] was attempted unsuccessfully using the same procedure as that for Na2 [M06Cl8 (OMe) 6] . The methanol/W6C112 slurry turned dark brown immediately after adding the sodium methoxide solution. K2 [M06C112(OSiMe3)2] was prepared by adding twice the number of moles of potassium trimethylsilanolate (Aldrich) to an acetonitrile solution of M06C112. In a typical reaction, to a slurry of Mo5Cl12 (0.1 g, 0.1 mmol) in 5 mL of acetonitrile was added 25.6 mg (0.2 mmol) of K[OSiMe3]. The solution immediately turned yellow-orange and was stirred for ~ 0.5 h. The solvent volume was decreased under reduced pressure until a solid, crystalline product precipitated. Precipitation by the addition of ether to reacted solutions yielded a compound which would not dissolve in acetonitrile. 27 (NEt4)2[Mo6C18(SEt)6] was prepared with the following procedure. The product was synthesized by reacting Na(SEt) with MoGCllz in methanol. Mo6Cl12 (0.2 g) was added to Na(SEt) (0.17 g) in 10 mL of degassed methanol under an argon atmosphere. This solution was stirred for ~ 3 h resulting in an. orange solution, 'which. was filtered to remove sodium chloride. (NEt4)Cl (0.17 g) was added to this solution and sodium chloride immediately precipitated, which was filtered. The filtrate was reduced in volume, and the precipitated product was washed with methanol and air dried. Sodium ethanethiolate was prepared by syringing ~ 3 mL of ethanethiol (Aldrich) into 5 mL of THF containing 0.6 g of sodium under an argon atmosphere. This solution was stirred for ~ 3 h, filtered to remove sodium chloride, and the resulting filtrate was pumped to dryness to yield the desired product. Mo6C112py2 (py = pyridine) was prepared by dissolving 0.06 g of Mo6C112 in 15 mL of methanol. Addition of 0.1 mL of pyridine yielded a yellow solid immediately, which was collected by filtration, washed with methanol, and air dried. Tungsten(II) chloride and tungsten(II) bromide were 46 synthesized by literature methods which were analogous to those used for the synthesis of molybdenum(II) chloride and bromide. The tetrabutylammonium salts of W6X8Y62 (X = Cl, Br, I: Y = Cl, Br, I) were prepared by a modified procedure 28 of Hogue and McCarley.52 A detailed description of these tungsten cluster syntheses is described by Mussell.48 2. Cluster Modified Polymers Polyvinylpyridine was modified with M06C112 by mixing a 1:1 ratio of polyvinylpyridinezMo6C112 in 20 mL of ethanol or methanol for approximately 1.1L. The resulting product, which contained Mo6C112 covalently linked to pyridine of the PVP backbone, was collected and washed with the appropriate alcohol. The percent of bound cluster was determined by measuring the absorbance of the cluster solution at 340 nm before and after the reaction. Typically, a 90% bound product was obtained from reactions carried out in methanol, while a 10% bound polymer was obtained from ethanol solutions. Silicon-oxide gels containing Mo6C112 were prepared by using formamide as a drying control additive as described by Hench.53 Tetramethylortho-silicate (Aldrich) (1.5 mL), formamide (1.25 mL), concentrated nitric acid (0.2 mL), and a methanolic solution of Mo6C112 (4.89 mg in 1.25 mL) were mixed by magnetic stirring in a disposable beaker. The hydrolysis reaction was initiated by adding 1.8 mL of H20 to this solution and stirring for 10 min. The final liquid phase was poured into disposable test tubes or 1 cm spectroscopic cells. As the hydrolysis reaction proceeded, the liquid sample became more viscous until a gel formed. Typical reaction time of the gelation process was 8-9 h. H“; Ah pr: or (J f} (I 29 3. Trans-Dioxorhenium(V) Compounds Rhenium oxo compounds [Re02(py)4]I,54 K3[Re02(CN)4],55 and [ReOz(en)2]Cl56 (en = ethylenediamine) were prepared as previously described. 4. Complex-Layered Oxide Dioxorhenium(V) Intercalates Natural sodium hectorite (San Bernardino County, CA) was obtained in spray-dried form from the Source Clay Mineral Repository, University of Missouri. The mineral was suspended in water (1 wt%) and allowed to sediment to remove carbonate impurities. The clay fraction containing particles less than 2 pm was collected, saturated with Na+ ions. by the addition of sodium chloride, dialyzed, and freeze dried. The cation-exchange capacity of the hydrated mineral was 70 meq/100 g. Fluorohectorite was obtained from Dow Corning and used as received. The particle size of this layered silicate clay (LSC) is >> 2 pm, and its cation- exchange capacity is 190 meg/100 g. The ReOz (py)4-LSC intercalates were prepared by using simple ion-exchange methods. In a typical experiment Na- hectorite (0.1 g, 0.070 meq) in 10 mL of water was added with stirring to a solution of the rhenium oxocation complex (0.070 meq) dissolved in the minimum amount of acetone. The clay immediately flocculated and turned yellow upon contacting the metal complex solution. After a reaction time of 10 min, the product was collected by centrifugation, D1 ‘4. I m). u . c “‘11. 30 washed several times with acetone to remove physically adsorbed cluster, and air dried. The Re02(CN)4-LDH (LDH = layered double hydroxide) intercalate was most successfully prepared by coprecipitating the hydrotalcite in the presence of the complex anion. K3[Re02(CN)4] (0.44 g, 1.0 mmol), MgC12-6H20 (1.83 g, 9.0 mmol) and AlCl3o6H20 (0.73 g, 3.0 mmol) were dissolved in 100 mL of degassed water. The solution pH was adjusted to 10 with l N NaOH and the suspension was refluxed for 24 h and aged at room temperature for 3 days. The resulting precipitate was collected by centrifugation, washed and dried in air. 5. e rs The chloro and iodo pyridinium salts and 1,1’-dimethyl- 2,2'-bipyridinium were synthesized by addition of either excess methyl iodide or benzyl chloride to a 1:1 acetone:ethanol solution of the appropriately substituted pyridine or bipyridine, respectively. The monosubstituted bipyridines, l-methyl-4,4'-bipyridine and 1-methyl-2,2'- bipyridine, were synthesized by the addition of one equivalent of methyl iodide to 4,4’-bipyridine and 2,2’- bipyridine, respectively. The pyridines and bipyridines were purchased from Aldrich and used without further purification. Pyridinium and bipyridinium hexafluorophosphate salts, obtained by the addition of ammonium hexafluorophosphate to In ‘il 54. 31 aqueous solutions of the chloro and iodo salts, were twice recrystallized from acetone-water solutions. The hexafluorophosphate salts of 2,2’-bipyridinium and 1- deuterium-2,2’-bipyridinium were synthesized by addition of ammonium hexafluorophosphate to solutions of 2,2’- bipyridinium in 6 M HCl and 6 M DCl, respectively. These crude products were recrystallized twice from acetone-water mixtures. 6. Supporting Electrolyte Tetrabutylammonium hexafluorophosphate (Southwestern Analytical Chemicals) was dissolved in ethyl acetate, dried over M9804 and recrystallized from pentane-ethyl acetate solution and dried in vacuo for 12 h at 60 °C. 7. §QlEQDE§ All solvents used for spectroscopic measurements were subject to seven freeze-pump-thaw (fpt) cycles and vacuum distilled into flasks containing high-vacuum Teflon valves. Acetonitrile, dichloromethane, acetone, methanol, and tetrahydrofuran were obtained from Burdick and Jackson, and butyronitrile and ethanol were purchased from Aldrich (Gold Label). Acetonitrile was distilled onto 3-A molecular sieve: dichloromethane, methanol, butyronitrile, and ethanol were distilled onto 4-A sieve, and tetrahydrofuran was distilled onto benzophenone ketyl radical (generated from sodium and benzophenone) . Acetone was removed from 4-11 sieve sieve 1 periods Al contain. consist; terninat were $11 {iiStille Sol SamPies, CUEParis. aPPIOpz-ii The enis Han Spectra a $011 :ethiid de standard SPectra We sMelly positioninc 32 sieve by distillation into a high vacuum flask to prevent sieve catalyzed decomposition of acetone over long storage periods. B. Methods 1. as o ' e s e e ts All spectroscopic measurements were made on samples contained in a specially constructed high-vacuum cell, consisting of a 1-cm quartz cuvette attached to a sidearm terminating with a 10-mL round-bottom flask. The samples were subject to 4 fpt cycles after the pure solvent was distilled from the solvent pot into the sample cell. 2. Qpaptum Yield Measupements Solution quantum yields were measured on dilute samples, with absorbances less than 0.1 at 436 nm, by comparison to 140601142' (458 = 0.19 in CH3CN) with appropriate corrections described by Demas and Crosby.57 The emission spectra were corrected for monochromator and PMT (Hamamatsu R316) response, and converted to energy spectra as described in Chapter II.C.1. Solid state quantum yields were determined by the method described by Wrighton pp pl.58 using Ru(bpy)32+ as a standard “e = 0.003 in the solid state). The emission spectra were collected on powdered samples contained in a specially designed sample holder to allow for reproducible positioning of the sample, corrected appropriately, and used to det equati vhere respec th i) 1P is NaZSO4 The re Slowl) the e; detery Vere 1 Scans “as us 33 to determine the quantum yields, ox, by using the following equation: emx LP-RsS )(----) LP-Rsx ¢x=¢s( ems where s and x refer to the standard and unknown, respectively, o5 is the quantum yield of the standard, em is the integrated area under the corrected emission spectrum, LP is the reflectance spectrum intensity (vide infza) of NaZSO4 and RS is the sample reflectance spectrum intensity. The reflectance spectra were recorded on powdered samples by slowly scanning (l A/sec) the emission monochromator over the exciting wavelength (436 nm) and the intensities were determined by integrating under the band profile. The slits were narrowed to 100 pm and the average of three successive scans was used in the quantum yield determination. Na2S04 was used because this salt does not absorb 436-nm light. Emission spectra for quantum yields at 77 K were measured on dilute samples with absorbances less than 0.1 at 436 nm in EPR tubes containing high-vacuum stopcocks. The absorbance of these samples was measured at 436 nm on room temperature solutions and corrected appropriately for the absorbance changes between room temperature and 77 K samples. .Absorption. spectra at 77 K: were ‘measured in specially constructed high-vacuum cells containing pyrex ~1-cm cuvettes that could be frozen, in a specially constructed vacuum dewar, without cracking. standa were a exchan soluti vere c disper: of the SPectra interca Vere pe of him. e"Perm range C 5e11, idemChe “with were 5d Haailtor the flag 34 3- WW Molar absorptivities were experimentally determined from suspensions of clay intercalates. Volumes of a standardized aqueous solution of the dioxorhenium(V) ion were added to aqueous suspensions of the clay to yield 15%- exchanged CLO intercalates. The absorbance of these solutions was measured and molar extinction coefficients were calculated assuming the suspensions were homogeneously dispersed. After all absorbance measurements, suspensions of the clay intercalates were centrifuged and the absorption spectra of the supernatants were recorded to ensure complete intercalation of the rhenium complex. 4. Qpepghipg Expepiments i. flexanuclear Clusters Quenching experiments of the hexanuclear clusters were performed by using the Stern-Volmer quenching method59 of luminescence intensities and lifetimes. Stern-Volmer experiments were performed over a quencher concentration range of 10'6 - 10"2 torr by using a specially constructed cell, described in Chapter II.B.1 which permitted all quencher additions to be performed under high-vacuum conditions. Aliquots of a standard quencher solution (~2mM) were added to the round-bottom flask by using a 100 uL Hamilton syringe and the solvent was distilled off. After the flask pressure was ~ 5 x 10'6 torr, the cell was removed dioxo: the S vas m< Sarple soluti suspen CU) 5L Syrin Scluti hexanu using 175 Prl the dj IRStrm '4st i electr: 30‘ . tile 35 from the vacuum line and the cluster solution isolated in the 1-cm quartz cuvette was mixed with the quencher. ii. Complex-Layered Oxide Qioxpphepium(V) ca e The reaction of electronically excited dioxorhenium(V) ions with proton donors was studied by using the Stern-Volmer quenching method. Luminescence quenching was monitored with time-resolved luminescence spectroscopy. Samples were prepared by adding dimethylformamide (DMF) solutions of the ppppp-dioxorhenium(V) ion to DMF, suspensions of the CLO. Aliquots of H20 were delivered to CLO suspensions of the intercalate with a 100 pL Hamilton syringe. All lifetime measurements were performed on solutions thoroughly degassed with argon. 5. Elepppoghemical Measurements Formal reduction potentials of the quenchers and hexanuclear clusters were determined by cyclic voltammetry using a Princeton Applied Research (PAR) 173 potentiostat, 175 programmer, and a 179 digital coulometer. The output of the digital coulometer was fed directly into a Houston Instrument 2000 x-y recorder. A three-electrode system was used with a standard H-cell configuration. The working electrode was a Pt button, the auxiliary electrode was Pt gauze and a Ag wire served as an adequate reference potential by using ferrocene as an internal standard.60 Potenti a ferrc Di by usix of the purchas experim experin 36 Potentials were related to the SCE reference scale by using a ferrocenium—ferrocene couple of 0.31 V vs SCE. Diffusion coefficients were experimentally determined by using the Cottrell equation61 which relies on knowledge of the electrode surface area. A calibrated electrode purchased from Bioanalytical Systems (BAS) was used in the experiments, but the area of the electrode was checked exPerimentally by using ferrocyanide as a standard (D0 = 0.662 2 0.004 x 10'5 cmz/s at an ionic strength of 0. 05 M) .62 Tetrabutylammonium hexafluorophosphate was used as a supporting electrolyte to give the desired ionic strength, and the solution concentrations were ~ 3 mM. c ' Instrumentation l. - es tion Em'ss'on S ectr meter Luminescence spectra were obtained by using a high- res<>lution emission spectrometer designed and constructed at Michigan State; the block diagram is shown in Figure 4. The excitation beam originates from a Hanovia 200 W Hg-Xe arc lamp mounted in a Spex 1909 lamp housing (f/4), and is f':><-‘—\lssed onto the entrance slit of a Spex 16808 double I11<>r1<>chromator (0.22 m, f/4). The wavelength-selected ethi‘lmtion light, collimated by a f/4 fused silica lens, is focussed onto the sample cell with a f/l fused silica lens. The emitted light is collected at 90° to the excitation beam with a f/1.5 collimating lens and then is focussed by a Se Q011d lens (f/7) through a Corion color glass cut-off Pig‘are 4. £31351”: 3 37 Figure 4. Schematic diagram of the high-resolution em; 8 s i on spectrometer . 38 .5an5 2925 v Gunman llw 359:8 ..mEoomm :ofiébm EEEE 22 :32: -8“. .55 = 9.5.. 2925 .8896 £93 9.4 E o o HUGH.” < 0 fl ,0 8689.285: 25on .853me 3658598: 228.25. filter nonoch monoch: The d photomm to -70 signal single- constrx to the to the Signal between The out into a digital A Conpute numbers “Wm a(“11115 1 digital amdeve, 39 filter onto the entrance slit of a Spex 0.5 m 187GB scanning monochromator (f/7) . The wavelength scan of the monochromator is controlled by a Spex 1673C Mini-Drive 2. The dispersed emission is detected by a Hamamatsu photomultiplier tube (R1104, R316, or R374) which is cooled to -70 °C in a Products for Research TE241RF housing. The signal from the PMT is passed through a LeCroy VVIOOB 8 ing file-channel fast pulse ampl i f ier , des igned and constructed by Martin Rabb (Sketch #A399) at Michigan State, to the input of an EG&G lock-in amplifier. The input signal to the lock-in amplifier is phase matched to the reference Signal generated by a PAR 125A light chopper situated bat-Ween the excitation monochromator and the sample chamber. The output from the lock-in amplifier may be fed directly ‘1JTtZCD a Soltec 124A strip chart recorder and/or collected in digital format by a Zenith ZQ-151-52 minicomputer. A schematic of the hardware interface to the Zenith computer is shown in Figure 5; the chip and connector pin numbers are also indicated . Data acquisition is a"'3'53C31lnplished using a Metrabyte Corporation DSH-16 data alcz‘illisition board which houses the required analog-to- dig ital converter (ADC) and clocks. Access to this board is ac=llieved externally using a Metrabyte Corporation STA-16 £3c=l=fiflv terminal accessory board. Data acquisition is initiated by the gated output from the Spex Mini-Drive 2. The gate of the ADC requires a TTL 'hi' signal for data a czqmlisition. Because the gated output of the Mini-Drive 2 40 Figure 5. Block diagram of the hardware interface for the emission spectrometer to the Zenith computer. The chip and pin numbers are indicated by the appropriate components. 41 m madman 6 aocoawoé opzu o... R :52 N588 o : 5&8 _ 325 2353 .1 on ~¥r N N 2. 3:. _ Em: “2812.: .. LI IF ocEcnou 955.8“ _ isT CORDEC ‘. Nu tr“ 42 is TTL ‘lo’ when the monochromator is scanning, this signal is inverted using a 7404 hex-inverting integrated circuit chip and input to the ADC gate via the 1P2 terminal of the STA-16. With a ’hi' TTL signal at the ADC gate, data acquisition occurs at the clock rate of the ADC trigger input (IPO terminal of Figure 5) . This rate is determined by the scan rate of the Mini-Drive 2 to give data collection of one point per angstrom scanned; the 100 KHz internal clock, accessed at CNTO, is frequency divided through software to provide the appropriate rate at the trigger input. The lock-in output, which is a negative signal, is connected to the low input of the double—ended ADC Channel 1 input and ground (TTL ’lo') is connected to the high input to give a positive signal so that the spectra may be displayed in standard orientation. The software that drives the data acquisition, program "getdat”, is a Fortran 77 program that uses a function library provided by Metrabyte Corporation to access the DSH‘IG board via Fortran subroutine calls. Data Ina"i-‘Lpulation and analysis are performed from the Fortran 77 s<’ftWare program "em", which is an acronym for emission. A prilhary function of "em" is to correct emission spectra for the response of the detection system, namely the In<>ric><3hromator gratings and PMT.63 The spectral responses of the monochromator and PMT were calibrated with a standard of spectral irradiance Model 245C 45 W tungsten halogen lamp 15 rth Optronics Laboratories (Serial Number L-374) . The lamp spec Rabb 450 ‘ prof: Corre respc cclle tenor pect relat area 1 energj aCCOmE Square becaug Va" 91¢ enerq 43 spectrum was recorded using a constant current source of 6.500 z 0.002 A at 6.7 V designed and constructed by Martin Rabb (Sketch #A409) , and corrections in 1 A intervals from 450 nm to 1050 nm were made using the spectral irradiance profile of the lamp provided by Optronics (Figure 6). Correction files were calculated by dividing the standard response curve of the Optronics lamp by the lamp spectrum collected, and a point by point multiplication, followed by spectra. Once a the renormalization yields the corrected Spectrum has been corrected for the detector response, relative quantum yield may be determined by integrating the area under the emission band which has an abscissa linear in energy rather than wavelength. The axis conversion is accomplished by multiplying the data set by A2 (wavelength 8qll«'-.'Ired)64 and renormalizing. The A2 factor is required because the spectrum is collected in units of quanta per wavelength interval, (dQ/di) , which is related to quanta per energy interval, (dQ/du) , by dQ/du = (dQ/dx) x (dA/dv) = - (dQ/dA) x Az/C since y = c/A; (dz/d») = -C/u2 = -A"’/c. The integration is Performed by adding the intensities of each data point, which approximates the area by a sum of rectangles 1 A wide. The data analysis software allows for several data manipulation and displaying options, however, a detailed discussion of the software specifics is beyond the scope of t his thesis. Therefore, for further details describing the 44 Figure 6. Spectral irradiance jprofile: of the Optronics standard tungsten halogen lamp. 45 2250 1750 1250 750 00 250 1.0 l l l l m. w. V. N. O O O O (wusz/M9-0L) eouegpeul lenoads klnm Figure 6 softvaz Acquisi spectrc 2. Th the bl: tate 1 excitat sa-YAG sac-u 532‘ o; kitted Excitat lens (1 Silica entranc. U4). CCTIESp. dEteQa 3923) . tezpem respect]. 46 software packages, the reader is referred to the Data Acquisition/Analysis Manual (DAM) for the emission spectrometer . 2. Time-Resolved Luminescence Spectrometer The time-resolved luminescence spectrometer depicted by the block diagram of Figure 7 was constructed at Michigan State to obtain excited-state lifetime measurements. The excitation source is either a Quanta Ray DCR-l or DCR-Z Nd-YAG laser frequency doubled or tripled with a Quanta Ray SHG-II second harmonic generator to produce a respective 532- or 355-nm pulse of 8—ns (fwhm) duration at 2 Hz. Emitted light from the sample is collected at 90° to the excitation beam by a borosilicate crown glass collimating lens (f/1.5) and then focussed by a second lens (fused Silica f/4) through a Schott 06-570 color filter onto the entrance slit of a Spex 1680A double monochromator (0.22 m, f/“) . Luminescence is monitored at wavelengths corresponding to the emission maximum of the sample and detecited by a Hamamatsu photomultiplier tube (R1104 or R928) . The R1104 and R928 PMTs are operated at room temperature in Pacific Instruments 3378 or 3150RF housings, resPQCtively. The signal from the PMT is passed through a Cu:- r 8:11: sensitive preamplifier designed and constructed by Martin Rabb (Sketch #A414), employing a LeCroy VVIOOB fast Pulse linear amplifier, to the 50-0 impedance input of a Le Cray 6102 dual amplifier/trigger. To prevent signal 47 Figure 7. Schematic diagram of the time-resolved £ luminescence spectrometer. ET} [mD 48 Nd:VAG laser SHG Emission Double Monochromator Trigger circuit -—‘——>_..: --—-Al :3: \L Sample El 0 0 E w J: PMT Pre- Transient Digitizer System AV— . . Transient Amplifiers Digitizer eh memory ailnterfoce Computer Figure 7 ringf firsi are 1 The c trans two confi gener at th 49 ringing through cable connections, a 50-0 2 W terminator is first attached to the PMT output and 50-0 impedance cables are used extensively for signals from the PMT to digitizer. The output of the amplifier is passed into a LeCroy TR88288 transient recorder, and the digitized signal is stored in two LeCroy 1048104 memory modules arranged in a series configuration. The amplifier, digitizer, memory modules, and a LeCroy 8901 GPIB interface are housed in a LeCroy 8013A mini-crate. The data acquisition is initiated by a trigger generator (Martin Rabb; Sketch #A421) that utilizes a quartz plate to deflect ~2% of the laser light to a fast photodiode circuit that outputs trigger pulses compatible with the input of the trigger generator of the LeCroy 6102. A 93-0 cable is used to connect the home-built trigger generator to the 6102, therefore, a 93-0 terminator is used at the cable/6102 interface to prevent signal ringing. If this terminator is not used, spikes in signal averaged data may occur due to multiple trigger pulses occurring for a single laser pulse. ‘ Data, acquired and processed by a Zenith 151-52 minicomputer equipped with a 10 megabyte hard disk. are typically averaged over 1000 pulses. Data analysis may then be performed using program "1t" which is similar to program "em" described previously. UnieDiponential decay curves are fit by using standard least- squaill‘es linear regression statistics. Plots of log (intensity) vs time are linear over at least three 1 ifetimes for a satisfactory analysis. Multiexponential decay by u progr. compr. cave: of th detail found Spectx ‘L’ (I) 2023 < VaCUUm IROZA CCPper filing the 3a the c< Conduci Mating indiCat Mich eXpande three 5 e eXC light t 50 decay curves may be fit to the equation y = ae't/’1 + bet/'2 by using the Kinfit65 general nonlinear curve-fitting program where r, and 12 are the excited-state lifetimes comprising the decay of fraction a and b, respectively. Convergence of the fits monitored by the sum of the squares of the residuals, yields values for a, b, 11, and 12. A detailed description of this data analysis package may be found in DAM for the time-resolved luminescence spectrometer . 3. Cryogenic Refrigeration System Samples may be cooled to 9 K using an Air Products CSA- 202E cryogenic refrigeration system equipped with a DMX-l vacuum shroud interface, DE-202 expander module, and an IR02A air-cooled compressor. Solid samples are adhered to a Copper block with copper grease (a mixture of fine copper filings and Apiezon H grease), and the block is mounted to the sample head of the DE-202. An iridium gasket between the copper block, and the DE-202 ensures good thermal °°nductivity. The temperature is varied with the resistive heating element of an Air Products APD-E digital indicator/controller and is monitored with an iron-doped goldrchromel thermocouple mounted at the end of the DE-202 expander module. The DMX-l vacuum shroud interface has three Suprasil windows surrounding the sample which allow the exciting light to impinge on the sample and emitted l e lght to reach the detector. The entire system must be connect operati 4. Ir records meters, using a perfor: associa ion 1a: 50‘60 I 51 operated at ~ 1 x 10’4 torr, therefore, the unit is connected to a vacuum line for several hours prior to operation. 4- MW Infrared and electronic absorption spectra were recorded on Perkin Elmer 599 and Cary 17D spectrophoto- meters, respectively. Proton NMR spectra were obtained using a Bruker WM-250 spectrometer. Raman experiments were performed with a Spex 1401 double monochromator and associated Ramalog electronics. A Spectra Physics 165 argon ion laser was the excitation source and incident powers were 50-60 mW. CHAPTER III SPECTROSCOPY OP HEXANUCLEAR MOLYBDBNUM(II) AND TUNGSTBN(II) CLUBTBRB A. ta EV" idem A. Introduction The hexanuclear molybdenum(II) and tungsten(II) clusters possess unique excited-state properties and chemistry.66 The idealized structure of these clusters, shown in Figure 8, consists of six metal atoms arranged in an octahedral core ligated by eight face-bridging and six axial ligands. The absorption and emission spectra of the molybdenum(II), M06x142’ (X = Cl, Br), and tungsten(II), W6X142' (X = Cl, Br, I), halide clusters are shown in Figure 9. All of the cluster compounds are yellow to yellow-orange owing to a broad, featureless absorption in the blue to near UV region of the spectrum. The absorption energy increases for the clusters along the series I < Br < Cl, consistent with the general trend expected for LMCT transitions. Excitation into these broad absorption bands yields intense luminescence in the red and near-infrared as depicted by the emission spectra in Figure 9. The trend in emission energies for both the M06 and W6 clusters indicates that the origin of the emissive state is not of LMCT parentage. For the case of the M06 clusters, the emission is relatively °°nstant and does not vary with halide substitution. The evidence against an LMCT parentage for the W6 halide clusters is even more compelling; the emission energies follow a halide ordering inverse to that expected for an LMCT emissive excited state. That the emission does not arise from an LMCT excited St . . . . . ate is consistent with theoretical calculations. Extended 52 Figure 8. Idealized Mo(II), W(II). structure 53 . 2_ of M6X8Y6 ions: um. ...»i... .353 55 Figure 9. Electronic absorption and emission spectra of M6X142' ions, as tetrabutylammonium salts in acetonitrile at room temperature: (a) (----) Mo6C1142'; ("'°) Mo6Brl42'fi’ (b) (--—-) ‘w6c1142'3 (°°°°) w6Br142‘; (———) w6114 ° i7/ 11m" 56 o 9.50.: 85x 005 000 00¢ d 000 _ 000 ....J . \li . . \\ ...... d..fl.¥..h .i DIV .7 00m . MN 0.0 0N oo. 0N. 0.0. 0.. 0.N 0.m 0.v on 0.0 ‘ oooooo ‘l’, d d a I I ll 1 I I I I. I, I, ‘I '5‘ ’I l o .‘o to t ' ...... I I‘. ‘ Iv. ca... ... s r .. - .u . m E M k p P p h n b n b p b p n b L p p n P 0N L-WOi-Weo I./3 are c112: HUM: Spe: tet: \ Jan: that file: dee c I E,f 57 Hficke167 and SCF-SW-Xa68 calculations predict that the LUMO and the HOMO possess largely metal character. The results of these molecular orbital calculations are summarized in Figure 10. The highest energy filled molecular orbitals are metal-localized and the twelve bonds in the octahedral metal core result from the filled alg' t2u' tlu' tzg, and eg levels: the LUMO is of 32g symmetry. EPR spectra of the monoanion and trianion cluster compounds support these electronic structure calculations.66d'69 The M6X142' ions are diamagnetic. Moreover, the one-electron oxidized clusters are predicted to occur with depopulation of the eg HOMO to produce a 2B9 parent state (Oh symmetry); EPR Spectra show a axial doublet signal consistent with a tetragonally distorted cluster core as predicted by the Jahn-‘I‘eller theorem (see Appendix for further details). These spectroscopic and theoretical results suggest that, although excitation involves the promotion of an e:l-etztron from a low-lying ligand-based orbital to the metal- based LUMO, ensuing IC and ISC processes are extremely efficient resulting in an excited-state configuration which is Primarily metal localized. Thus, luminescence occurs from a spin and Laporte forbidden metal-based electronic exczited state. The luminescence intensity and excited-state lifetime of the M06 and W6 clusters in solution and solid state are paralleled by few other transition-metal complexes. Table 1 di sI’lays the emission maxima, quantum yields, and lifetimes 58 Figure 10. Molecular orbital diagram of M6X142- ions. 59 V/ unoccupied om:- . DOUGIHQ metal- bosed OfbliOlS —029 ‘i‘r‘ ii- 99 «$44—44;- {fl— ‘H‘%} T2'.J,Tlu,t2l (12) where 'l’gs and ‘l’es are the ground- and excited-state electronic wavefunctions, respectively, and Q is the transition dipole moment operator. The transition dipole moment and emission energy for a series of compounds containing a common chromophore are expected to be similar, and hence, the radiative rate constants should be similar. For this reason, it is the nonradiative rate constant which is intrinsically informative about intramolecular decay processes. In recent years, experimental studies on osmium, 71 rhodium, 72 ruthenium, 54 ' 73 and rhenium?4 luminescent transition-metal complexes have indeed verified these predictions. The problem in nonradiative decay is energy disposal, or redistribution of the excited-state energy, and may be discussed in terms of the potential energy curves of the ground and excited states (Figure 12). Nonradiative decay Occurs by an electronic transition from the thermally- ecIllilibrated excited state (V850) to an isoenergetic Vibrationally-excited level of the ground state (”gs“) followed by ensuing vibrational relaxation (v.r.) to yield the molecule in its original ground-state configuration (”gs°). As depicted in Figure 12, the efficiency of this lscenergetic transition is dependent on the vibrational 68 Figure 12. Potential energy diagram for nonradiative excited-state decay. See text 'for abbreviations and definitions. ENERGY-—+ 69 Earn n 0 A l ,1 ‘. ”gs'Ves v ' : \ei'vr $7 I I I I I I V.I’.l I I A5 I I I I I 0 _. - z/gs ‘ kAQA NUCLEAR COORDINATE —>* Figure 12 7O wavefunction overlap of these two states, ”es. and ”gsn° The extent of overlap is dominated by the fact that vibrational wavefunction magnitudes increase near the potential curve as the vibrational quantum number (n) increases. As a result, the vibrational overlap, and hence, the nonradiative decay rate, are increased: i) for large distortions between the ground and excited states (large AQ): ii) for high-frequency ground-state acceptor vibrations; and iii) for small energy gaps, AB.4 The coupling strength (G) is a measure of the wavefunction overlap of states ”gen and ”es. and can be determined from eq 13,75 (13) with 3 = l/kBT, and the mean vibrational frequency of the accepting mode, com, which may be determined from vibrational fine structure of the emission spectrum. For many systems, including the M6X142' cluster Systems, Em, which is related to the Stokes shift, is difficult to determine because the transition leading to direct population of the excited state is forbidden. Hence, an estimate of the coupling strength must be made. Two liniting coupling regimes are depicted in Figure 13. The weal{-coupling limit occurs when AQ is small, and the strong couPling limit when AQ is large. Because Em is related to 49! from eq 13, weak and strong coupling will occur for 71 Figure 13. Potential energy diagram of the: (a) weak- coupling; (b) strong-coupling limit. 72 mq \\ / nu shaman A3 ... To. m4 73 compounds exhibiting small and large Stokes shifts, respectively. Although absolute values for Stokes shifts may not be determined for several systems, relative magnitudes may be considered to determine the coupling limit. It has been estimated that the Stokes shift would have to exceed 10,000 cm'"1 for the strong coupling limit to be valid76 and therefore, this limit is expected to be important for activated processes such as photochemical rearrangements and reactions. For this case, the nonradiative decay constant represents the rate of reaction. For nonradiative decay of electronically excited molecules which have small configurational changes such as internal conversion and intersystem crossing, the weak-coupling limit will apply since these processes typically have much smaller Stokes shifts (< 4000 cm":").76 Consequently, the ensuing discussion will be limited to the weak-coupling limit. Englman and Jortner7S have treated two limiting cases of the weak-coupling limit for nonradiative decay. The most commonly used, in which hum >> kBT, is referred to as the low-temperature limit and applies to systems with accepting modes of high energy relative to kBT. Conversely, deactivating modes such as low-frequency intramolecular and solvent vibrations where ham 3 kBT are described by a temperature-dependent limit. The nonradiative decay rate constant in the low-temperature, weak vibrational coupling limit is 74 C2 21! 1/ 2 -7AE k = exp <-s ) exp < "r h mm“ m ) (14) m In eq 14, C is the electron tunnelling matrix element, AE is the energy gap between excited and ground states, and 8m and 7 are defined by eq 15 and 16, respectively, sm = 1/2 X AQjZ (15) 1 = 1n (ZAE/Zhijsz) — 1 (16a) 1 ~ 1n (AE/Sfihwm) - 1 (16b) in which AQj is the dimensionless fractional displacement of the ground- from excited-state potential energy curve along the normal coordinate (Figure 12). Equation 16a reduces to eq 16b if only a single molecular vibration is treated as the critical deactivating mode. Furthermore, eq 14 may be rewritten as In knr = (In a - s - 7AE/hwm (17) m) where a is the pre-exponential factor defined by eq 18 a = 02/h (Zr/hwmAE)1/2 (18) 75 Far a series of compounds containing a common chromophoric constituent, variations in C and Sm will be small, and In a and 1 will be weakly varying functions of AE. Consequently, In a, Sn, and 1 may be treated as constants, and ln knr will exhibit a linear dependence on AE. This relationship is known as the energy gap law. Although the low-temperature, weak-coupling limit of excited-state nonradiative decay has been applied to organic fluorophores for several years77 owing to the fact that excited-state deactivation processes of organic molecules are typically dominated by high-frequency C—H vibrations (hum ~ 3000 cm'1 and thus hem >> kBT), the application of the energy gap law to transition-metal excited-state nonradiative decay has only recently been developed by Meyer and co-workers. One of the first observations of the energy gap law for luminescent transition-metal complexes was for a series of osmium(II) polypyridyl complexes OsAL42+, where A is the bidentate ligand 2,2'-bipyridine or 1,10- phenanthroline.71c Experimentally, the emission maximum, “em,max' is a measure of AE and a linear plot of ln knr vs Aem,max for the two series of osmium(II) polypyridyl complexes indicated a common luminescent core for each series of compounds. Vibrational fine structure in the emission. spectra of these compounds revealed. that. high- frequency skeletal vibrations (~ 1350 cm'l) of the 78 polypyridyl ligands were the critical deactivating modes of the MLCT excited states, thereby allowing for the 76 application of the low-temperature limit of the energy gap law. The substitution of "em,max for AB is not strictly correct because the emission energy is related to AE by a solvent dependent term, x,4r74 Aem’max = AE - X (193) x = xi + x0 (19b) where Xi and x0 include contributions from low-frequency intramolecular and solvent vibrations (10-30 cm'l), respectively. Interestingly, assuming that the low- frequency intramolecular vibrations remain relatively constant for a series of similar chromophores, a linear relationship is still predicted between 1n knr and the emission energy using the high-temperature limit for the solvent vibrations (I‘mm << kBT) .79 Indeed, solvent perturbations of the nonradiative decay pathways of [Os(phen)32"']80 and Ru(II) polypyidyl complexesm"82 have been observed. Subsequent to the initial studies by Meyer and coworkers, energy gap law behavior has been verified for several other systems.54'71b'73'74'83'85 A common characteristic of these systems has been the presence of high-frequency accepting modes. Detailed investigations applying the energy gap law to nonradiative decay processes 77 of systems which are not characterized by the low- temperature limit have heretofore been unexplored. If the critical accepting vibrational energy is not much greater than kBT, then the more general relationship represented by eq 2075 k = —"( I exp ["S (2 + 1)] exp ( I (20) nr h hum/1E m I"m mm 7' = 1n [ZAE / {hijsz (nm + 1)] - 1 (21) nm + 1 = 1 / 1-exp(-fihwm) (22) must be used to take into account the temperature dependence Of the coupling parameter. It is noteworthy that eq 20 reduces to the low-temperature result (eq 14), because “m -> 0 as hum -> on. As in the low-temperature limit, eq 20 Predicts a linear relationship between 1n knr and AE, as shown by eq 23, ln knr = [ln a - sm (2nm + 1)] - 1’AE/hwm (23) if In a, 1', and Sm are weakly varying functions of AE, and C: is relatively constant for a series of lumophores. The utilization of this more general expression for the weak- vihrational coupling limit of nonradiative decay is needed for investigations of the factors controlling the intramolecular excited-state deactivation of the Moe and W6 78 halide clusters because the highest-energy vibrations in these cluster systems are ya1g (M - X) ~ 300 cm'l. The first- and second-coordination spheres of the hexanuclear clusters may be systematically varied to identify the intramolecular processes important in the excited-state deactivation of M06 and W6 molecules. For the former, there are two types of ligands, axial and face- bridging, which may contribute to the excited-state properties; for the latter, solvent may perturb these properties. In addition to the halide substituted clusters, derivatives such as PR3, SR", and OR", which provide facile introduction of high energy metal-ligand and intraligand vibrations into the cluster core, can potentially provide valuable information regarding the critical processes that contribute to the deactivation of the excited states of these compounds. The high energy vibrations of these ligands will necessarily mean that such vibrations, if they effectively couple the excited- and ground-electronic states, will dominate the deactivation process. Hence, ligands containing high energy vibrations can act as triggers for nonradiative decay. To this end, phosphines, thiolates, and alkoxides may be used to probe the effect of ligands in the axial and face-bridging positions on the cluster excited-state properties. Furthermore, the deuteration of alkyl side-chains in these organic ligands will allow the influence of high energy C-H vibrations on the photophysical properties to be evaluated. 79 2- WW Table 3 displays the excited-state lifetimes and emission maxima for substituted molybdenum(II) halide clusters in acetonitrile, acetone, and dichloromethane. These properties do not vary substantially in the three solvents, although slightly longer lifetimes and higher emission energies are observed for the all-halide clusters in the higher dielectric solvents (acetonitrile (38.8) > acetone (20.7) > dichloromethane (9.08)). The luminescence of M06 clusters is maintained upon substitution of halide with donor ligands, but the photophysical properties are perturbed to different extents. Although the emission energies of thiocyanate (SCN') and ethanethiolate (SEt') substituted clusters remain relatively constant, the lifetimes follow an opposite solvent dependence than those of the all-halide substituted clusters (Table 4). Specifically, the excited-state lifetimes are significantly shorter in acetonitrile than acetone or dichloromethane. In contrast, the lifetimes and emission energies of the disubstituted phosphine clusters in dichloromethane are comparable to those of the halide clusters. A more quantitative understanding of the effect of ligand substitution can be achieved by using eq 1 and 2 to calculate the radiative and nonradiative rate constants. These rates are shown in Tables 5 and 6 along with the emission quantum yields for the hexanuclear molybdenum(II) 80 mmm mm awn mHH mus NNH one mma mmw as how am can mHH mus new can wmfi awe va "as HVH ems mva mus mmfi one was own mmfi E:\xsa.ao« m:\oP oescuoaouoH50w: .E: m H Hue vmh mun mvb mam How mm» men can ku bwh awn Nun 05> mmb E:\xsE.EO« n me.Em4. ”fin « as New mm mfifi mwb mmfi me mus svH swfi ave mmfi mm mam um Hos can moH mNH can mma mmfi one wee mmfi awn svH HvH own mmfi one «we ova mad men msH Goa mus mus 05H won whfl aha Hes owH m:\o » ...—{nae ...—0‘ we}: occaoo< washawoouoo< u op “mm coussfiumo muwefla possm d INQHwhmmoz -mvflrmmoz INNHUNHhmooz Immaowemmoz -monHomoz ImmHHHHUQOE INNHNHHUGOE -msmesomos Immhmwfiomoz Immhmmaomos INthOHaumoz INMhmHHHU®02 Immhmmaaoooz umsmmfifiomoz INVHHOQOE ccsoeaoo whoamsao mo: mowasalna< Ho «as: OweoowOAaovem newmmwam d m odndh 81 .m menus to muocuoom as umaeemu masses horse '0 ..nmeOww WCNHH .HO wa U .Hscmca I an .mmvoI: :2 .mzomzom:o I rd .mzomzo I s: .m:o I 0: ”macsonfl>0nnn< .u:o>flom were cw mansmomcfi mew msobmsso cousufiamosm mcfisomoze n d mus mm I I I I mflmtadvfisfioooz ems cos I I I I mfimzmmvmesomos use mes I I I I omfimramcmfisomoz mws mes I I I I mxmomcamvmesomos mom and I I _ I I onmmcaavaHomos ems ass I I I I mfimmzzaavmesooos cow HNH I I I I mxmzmcadvmsfiomos cow omfi com mm cow Hm Immflommvmfiooos was Hes own mm was v Immxzomcwsomos Hos was was so" was m Imvxzomvofisomos see ass was was was em Immxzomvmssomos Ec\xca.ao« m:\op Ec\xsa.soe w:\op E:\xda.ao« m:\o~ accsoeaoo meanumaOHOHcowD Deccamo< nmafihuw:0aoo< cwuoamsao mo: causewamnsm Owesmuo mo dawn owcoom0haooam eonmem v menus 82 hm.0H mv.NH HN.NH 50.0H 00.0 05.0 00.0 hm.m 00.0 00.0 Nw.0 «0.0 00.0 00.0 No.0 00.0 0v.w 0v.w av.w an: :H .smHIoH « 0.Hm 0cm 00w H.0m 0.0m N.0 0.0 0.0 n.0H 0.0 0.0 ~.m v.0 N.n 0.0 0.v b.v >.v n.v nlw NOH\h:x 0:: .rx neon « NOMNNHQDOOO HOOHHHHHHNHHH HCDG‘. 0303 CO 0.0 0.0 0.H ~.H Him ”0n\hx we ”we coawsfiumo mufiEHH hosum a H00.0 v00.0 000.0 N00.0 000.0 0H.0 w~.0 mm.0 v00.0 m~.0 N~.0 VH.0 0H.0 mn.0 n~.0 0—.0 0~.0 >~.0 0H.0 0. Imoxsmmcmsomos Immxzomvmsooos Imvxzomvoesomos Immazomomseooo: Immuwsmmos ImvHrmmos INNHONHLQQOS Immfiuwhmmos Imosmsomoz INmHHHHUQOE Immmmnaomo: Imsmesooos Immrmwsooo: Immrmmsomo: INthOHHUQOS Inner—Havoc: Immrmmssomos ImhmmH~U®OE Imvamuwoz cesoeaoo .aoafluawc0aooc cw muoumsuo 00: mo mead: >s000 0ada0160auoxm .0H .0H .hfi .0H .0H .VH .mH .NH .HH .0H .0 HNWVLOQDNCD henna Entry (OH ILL») b Error limits defined in Footnote of Table 5. Excited-State Decay Compounda M06C1142- M05C1138r2' M06C1128r22‘ M06C1118r32- M05C1108r42’ M06C198r52‘ M05C188r62- M06C11312- M05C112122‘ M06C111132‘ M05C18162‘ M068r8C152- MOSBrlzClgz- M068r142’ M06Br8152- Mo6c112(SCN)22‘ M06C110(SCN)42‘ M06C18(SCN)52- Mo6018(srt)62‘ M05C112(PPh2Me)2 M06C112(PPhMeg)2 M06C112(PPh23t)2 M05C112(PPhEt2)2 M06C112(PPr3)2 M06C112(PBu3)2 M06C111(PPr3)3 Rates of I06 Clusters in Dichloronethsne OOOOOOOOOOOOOOOOOOOOOOOOOO ¢e .18 .17 .16 .14 .14 .14 .145 .14 .12 .12 .08 .26 .20 .17 .12 .21 .28 .30 .12 .06 .05 .05 .05 .06 .20 .004 233 Table 6 b kr/103 s'l Rut/103 s-1 In kn, IJWU‘CDUTI-‘LJOOHOOOHHH O r4 0 C) O (D O r4 N r* H r‘rd H r4 H +4 H r4 H r4 H F‘ H r4 H IbIAIbUIOHOO H0301 Abbreviations defined in Footnote of Table 4. 8.52 8.59 8.61 8.67 8.72 8.76 8.76 8.68 8.94 9.20 9.39 8.39 8.79 8.88 9.47 8.41 8.41 8.52 8.90 8.96 9.02 8.97 9.04 8.99 8.52 10.46 0| 0] 0‘3 O -Q 0 1.»4 0 «JCfi 0 03¢» 0| m C I O N O I I O O I I O N 0) Ian 0 03 0 I0- It.- H 00 U! r we o w mmflmflflmk OOIbQmebJOU‘UI (A) U‘ U] o 84 clusters in acetonitrile and dichloromethane, respectively. The radiative rates (~ 103 s'l) are relatively constant for all these cluster compounds, but the nonradiative rate constants vary as a function of the substituted ligand. Closer inspection of these data reveals that the nonradiative rates increase as the emission energies decrease. These data are a signature of energy gap law behavior. Plots of ln knr vs emission maximum are displayed in Figures 14 and 15 for the molybdenum(II) clusters in acetonitrile and dichloromethane, respectively. A linear relationship, obeying the energy gap law, is observed in both solvents (the anomalous behavior of points 16-19 in Figure 14 will be discussed below), indicating that emission arises from an emitting state of common parentage for the series of lumophores. Because the metal core is the common substituent for the entire series, these data imply that the excited state is primarily metal-localized. Consistent with this ‘hypothesis is the observation ‘that deviations from linear behavior do not occur for axially-substituted phosphine, thiocyanate, or ethanethiolate complexes in dichloromethane, thereby indicating that ligands occupying axial positions do not greatly influence the nonradiative decay processes of the luminescent excited state. Rather, the ligands in axial positions have only a secondary effect by slightly perturbing the energy gap and thus, by eq 20, knr' It is noteworthy that the SCN' and SEt' complexes in acetonitrile (data points 16-19 in Figure 14) deviate from 85 Figure 14. Plot of In knr vs Aem,max of the Mo6 clusters in acetonitrile shown in Table 5 . All-halide and organic-substituted clusters are represented by a and s, respectively. In km 13.0 86 I18 I19 I17 I16 7.0 ‘ l I 1 12.2 12.6 13.0 13.4 )‘em,max/kK Figure 14 87 Figure 15. Plot of In knr vs Aem,max of the M06 clusters in dichloromethane shown in Table 6. All-halide and organic-substituted clusters are represented by u and I, respectively. In km 88 9.8 12.0 12.4 12.8 13.2 13.6 Kermrnax/kK Figure 15 89 the linear relationship of the energy gap law, indicating that these complexes possess additional nonradiative decay pathways in this solvent which are not operative in dichloromethane. This anomalous behavior parallels the results for other luminescent transition-metal complexes that contain ligands such as cyanide,86'87 and aza-88'90 and carboxyl-substitutedgl'93 pyridyl ligands. Ligands such as these, which are strong Lewis bases, may alter the luminescent properties of the parent complex via second- sphere donor-acceptor (SSDA) interactions.24 Thus, it is not unreasonable to expect a donor solvent, such as acetonitrile, to interact with the SCN" and SEt' acceptor ligands to perturb the nonradiative decay of these clusters. That a linear relationship is obeyed in dichloromethane, a very poor Lewis acid, further supports this analysis. The ligation of alkoxide substituents in axial positions provides an additional means of incorporating high energy vibrations in these positions of the cluster molecules. Table 7 shows the excited-state lifetimes, emission maxima, and quantum yields of three alkoxide substituted clusters. The emission energies are characteristic of the all-halide cluster compounds, but the excited-state lifetimes and quantum yields are significantly shorter than those of the halide systems in aprotic solvents. The high energy C-H vibrational modes (~ 3000 cm'l) of the axial methoxide ligands are not coupled to the excited-state nonradiative decay because the deuterated 90 Table 7 Emission Spectroscopic Data of Alkoxide Substituted Na2I06C13(OR)6 Clusters N32I06C18(0R)6 R CH3 CD3 CH2CH3 Solvent Methanol *em,max/nm 767 776 785 T O/llS 1 . 9 1 o 9 1 o 8 ¢e 0.0012 0.0007 0.0009 dl-Methanol “em,max/Hm 761 764 765 ro/us 4.3 4.0 4.5 09 0.0022 0.0015 0.0018 H20 Aem,max/nm 768 771 767 ro/us 0.745 0.74 0.70 ¢e 0.0004 0.0003 0.0003 D20 Aem,max/nm 766 767 766 ro/us 3.32 3.55 3.76 ¢e 0.0015 0.0014 0.0015 Ethanol Aem’max/nm 790 790 795 rO/us 1.81 1.80 1.80 0e 0.0011 0.0007 0.0010 91 compound, which has much lower energy C-D vibrations (~ 2200 cm'l) , has similar excited-state properties. In addition, the fact that the ethoxide and methoxide derivatives also have similar' photophysical properties suggests that, if these types of ligands contribute to the excited-state nonradiative deactivation, then the Mo-O bonding interaction must be the dominant contributor. This axial ligand influence of the emissive behavior may arise from the higher energy' of the Mo-O vibration as compared to the Mo-Cl vibrational energy. Infrared spectra of Naz [M06C18 (0Me)6] indicate a Mo-O stretching vibration of ~ 520 cm'1 while a peak at 240 cm"1 in the Raman spectrum of Mo6C1142' has been assigned to valg(Mo-Cl).94 0f further interest is that the alkoxide clusters exhibit longer lifetimes in deuterated solvents (Table 7). This result is surprising in view of the metal-localized nature of the excited state, but parallels the observation of maverick that the lifetime of MoGC1142' in aqueous 6M HCl is 20 ps compared to 180 us in acetonitrile.668 The origin of the contribution of protic solvents to inducing shorter lifetimes and smaller quantum yields of these complexes is unknown, but it is not unreasonable to expect that proton quenching, which is a well-established decay pathway of transition-metal excited states, may be an important nonradiative deactivation process in these systems. Face-bridging ligands also perturb the excited-state properties of the M06 clusters, but the effect is not as 92 easily discerned with the data presented in Tables 5 and 6. Infrared spectra show the face-bridging stretching modes of Mo-Br and Mo-Cl to be 270 cm'1 and 320 cm’l, respectively.95 On the basis of straightforward energy gap law considerations, if face-bridging vibrational modes were strongly coupled to the metal—localized excited state, then smaller nonradiative rate constants would be expected for the compounds with chloride in the face-bridging position. This is not reflected in the data shown in Tables 5 and 6. Although a clear-cut trend cannot be discerned for the two types of substituted compounds, i.e., Mo6C18Y62' and MoéBrBYGZ', a difference of 50 cm"1 will not necessarily be reflected in experimental energy gap data. Better model compounds incorporating ligands containing high energy vibrational modes in the face-bridging positions would allow for a more detailed understanding of face-bridging contributions to knr' Unfortunately, clusters coordinating face-bridging ligands with vibrational energies significantly greater than Mo-X stretches, such as M05(0Me)142', could not be synthesized as described in Chapter II.A.1. Thus, on the basis of the halide data, it is difficult to ascertain the effect of face-bridging ligands on the decay properties of the Mos clusters. These ligand substitution studies clearly show that the emitting excited state is not coupled to ligands in axial coordination sites, and therefore identify vibration(s) localized in the M06 or [M°6x8]4+ cores as the 93 predominant accepting vibration(s) in the excited-state decay pathways. The dynamics can further be defined by quantifying the results of energy gap law measurements. According to eq 23, the intercept and the slope are equal to eq 24 and 25, respectively, Intercept = ln 0 - Sm (an + 1) (24a) o = cz/h.(2«/homnr)1/2 (24b) Slope = -1’/hwm (25a) 1' = 1n [AE / smhommm + 1)] - 1 (25b) where Sm, the dimensionless Huang-Rhys parameter, is related to the equilibrium bond length changes between the excited- and ground-state species (re* - re) 1/2 pwmsz sm = (26a) hwm n R = [.21 (re* - re)j211/2 (26b) 3: In the above equations, which assume a single accepting mode, mm is the excited-state vibrational frequency and p is the reduced mass of the oscillator. Equation 26 assumes that the excited-state mode is the same as the ground-state accepting mode. If the critical accepting mode is known, then from eq 22 and 25, the Huang-Rhys parameter can be 94 directly calculated from the slope and emission energy; the electronic matrix coupling element, C, follows directly from eq 24. Although this mathematical treatment is straightforward and allows qualitative estimates of Sm and C, experimental studies on polypyridyl complexes have shown that these parameters cannot accurately be determined by this method.71a This analysis, therefore, is inappropriate for the purpose of addressing the most fundamental question regarding the nonradiative decay of these clusters: what is the critical accepting vibrational mode giving rise to this energy gap law behavior of these hexanuclear molybdenum cluster systems? This can most directly be addressed by determining the Huang-Rhys parameters and excited-state vibrational modes from a Franck-Condon analysis of emission spectra displaying vibrational fine structure.71a'73“"96 Results from this type of analysis compare favorably with those of other methods including calculation of bond length displacements from ground- and excited-state Raman spectra . 97 Unfortunately, the molybdenum and tungsten cluster systems lack.'vibrational fine structure in ‘their’ absorption and emission spectra, and excited-state Raman data of the ”6 systems has not been obtained. Therefore, S and hem must m obviously be determined by an alternative approach. One experimental handle that the M6 clusters do offer is a temperature-dependent emission band shape. The parameters of interest, namely 8m and hum, may be determined from Assul Morec eq 2E “here and a gene: mOdes 95 from an analysis of the emission band’s second moment.98'99 Assuming a Gaussian bandshape, the normalized second moment (m2) is related to the full-width at half-maximum (fwhm) by98 m2(Gaussian) = (fwhm)2/81n2 (27) Moreover, m2 varies as a function of temperature as shown by eq 28,98 where Seff and hweff are an effective Huang-Rhys parameter and an excited-state vibrational energy, respectively. In general, a band will be composed of several vibrational modes having different values of 8m and hwm, but if the vibrational frequencies are similar, this single mode analysis will provide effective values for these parameters. Therefore, Seff and hweff may be determined from a series of temperature-dependent full-width at half-maximum (fwhm) data calculated from the emission spectra. These data for M06C1142' are shown in Table 8. By using the Kinfit65 general nonlinear curve fitting program, Seff and hweff may be determined from the best fit of the data in Table 8 to eq 28: the best-fit plot is shown in Figure 16 which yields Seff = 38 2 l and hweff = 184 z 3 cm'l. This result directly conflicts with Raman94 and infrared95 studies which exhibit Fwy o o s I 0 O 80 O 9 O 9. no .50 C0 A5 :0 :0 1.. 96 Table 8 Temperature-Dependent Second-Moment Data of (NBu4)2uo60114a T/K fwhm/cm‘l III2/106oIII-2 ZkBT/cm‘l 9.1 2779 1.39 12.6 20.7 2778 1.39 28.8 30.0 2752 1.37 41.7 40.6 2685 1.30 56.4 50.0 2670 1.29 69.5 60.4 2661 1.28 84.0 69.9 2708 1.32 97.2 80.4 2708 1.32 111.8 89.9 2763 1.38 125.0 100.5 2769 1.38 139.7 120.5 2960 1.58 167.5 140.4 3032 1.66 195.2 160.4 3206 1.85 223.0 180.3 3313 1.98 250.6 200.3 3516 2.23 278.4 229.3 3712 2.49 318.7 260.3 3957 2.82 361.8 291.8 4142 3.09 405.6 a Error limits estimated as: T = t 0.5 K; fwhm = i 2%; m2 = t 5%. 97 Figure 16. Plot of m2 vs 2kBT of (NBu4)2M06C114 with single mode theoretical fits. The experimental data are represented by n and the fits to eq 28 are displayed for the parameters as follows: ('°") hweff = 320 cm-l : 15 z 0.7: (————) hoeff = 184 e 3 cm’l, seff = 38 2 1; ---- _ -1 _ + ( ) hoeff — 120 cm , seff — 66 2. m2/106cm'2 {712/1 050m-2 98 3.5 / / / . _ / / I o / 2.5 " l - / / I I I _ / /. , 1' 1.5 b I i} l’ / / I/ _ / _ —"’/ 0.5 n l l n I n I 0 100 200 300 400 ZkBT/CI‘II'1 Figure 16 500 n0 vibra' 11060114} to highe metal-me spectra lower er very p00 these re Cm-l mus Vibratic and the decay p] of a me meta1-ha The general: 99 no vibrational modes of this energy in the spectra of the Mo6C1142’ irun However, molecular metal-halide vibrations to higher energy and lower energy modes originating from metal-metal vibrations are observed in the vibrational spectra of Mo6C1142". Substitution of these higher and lower energy vibrational modes for tweff in eq 28 yields very poor fits as depicted in Figure 16. In this context, these results suggest that the effective vibration of 184 cm'1 must be a combination of two or more types of molecular vibrations. Because the excited state is metal localized, and the axial ligands do not contribute to the nonradiative decay processes, a logical starting point is the inclusion of a metal-metal core vibrational mode and a face-bridging metal-halide mode in the second moment analysis. The simple single mode model of eq 28 can be more generally expressed as99b m2 = ESm(hwm)2 coth (hum/ZRBT) (29) m=1 where the sum is over the n vibrational modes. Substitution of the vibrational energy for the symmetric metal-metal core and face-bridging metal-halide breathing modes (hwl (Mo-Mo) = 120 cm-1; hwz (Mo-C1) = 320 cm'l)loo in eq 29 allows the Huang-Rhys parameter corresponding to each of these modes (SMo-Mo and SMo-Cl) to be determined with a two parameter Kinfit65 analysis of m2 vs ZkBT. The fit, shown in Figure 17, yields SMo-Mo and SMo-Cl values of 39 2 3 and 6 z 0.7, 100 2.5, 9 Figure 17. Plot of m2 vs 2kBT of (NBu4)2M06C114 with E the two mode theoretical fit. The experimental data are i) represented by a and the solid curve corresponds to the E11 fit to eq 29 with hol = 120 cm‘l, s1 = 39 e 3 and 1562 = 320 cm’l, 82 = 6 z 0.7. 1.5_ 0.5L~ [112/1 060m-2 101 3.5 2.5 1.5 0.5 l A l 100 200 300 2kgT/CITI'1 Figure 17 400 500 respecti‘ related state vi the weig excited-: law, the acceptinI processe: therefor-I 140601142‘ Vibratiox metabha: VibratioI Studies luminesC, HowevEr' halides 1 PreviOuS: 5and 6) not be ““9380,“ ‘Iribraticl symmetric these re; include a The hexanuc 1 e 102 respectively. These Huang-Rhys parameters, which are related to the bond length displacements in the excited- state with respect to the ground-state, effectively reflect the weighted contributions of the different modes to the excited-state distortion.101 In terms of the energy gap law, these excited-state distortions compose the critical accepting modes which control the nonradiative decay processes of the Mo6C1142' cluster system. These results, therefore, imply that excited-state deactivation of Mo6C1142' is dominated by the metal-metal core breathing vibration with a smaller contribution of the face-bridging metal-halide vibration to the overall decay. This detailed vibrational analysis is concordant with the photophysical studies of the substituted halide clusters which show the luminescent excited state to be primarily metal-localized. However, a significant contribution of the face-bridging halides to the decay, which could not be discerned with the previously described data of the M06 halide systems (Tables 5 and 6), is important to radiationless decay. This should not be surprising in view of the fact that it is unreasonable to expect the totally symmetric Mos breathing vibration to be uncoupled from the [Mo6C18]4+ totally symmetric vibration. All photophysical data suggest that these results for the Mo6C1142' system can be generalized to include all Mos cluster compounds. The photophysical properties of the tungsten( II) hexanuclear clusters, displayed in Table 9, exhibit distinct 103 0.0 v.0 0.0 0.0 5.0 0.0 H0 00 00 ¢OH\an Him ¢0H\hx 00.0 00.0 HH.0 00.0 0H.0 0H.0 50.0 00.0 00.0 0+ 00 00 0H 0H 0H v.0 0.0 0.0 0.H m:\0 0 @Hnuh .800 oocmpwwmm anm cmxmp wanes d 000 000 How 00b 00h 00v 000 0H0 000 Imefisms ImmhmeQB Imofioweos 100H0h003 INvHHmma Imofiowemos Imosmfioos Imoemwsooz INVHHOOB cssoeaoo stwhuHooHoo< ow wwmxwkmavsmzv Ho dean oHeoomouaoeam eofimmwam v-INCOV'LDQDL‘CDQ differe radiati order 1 hexanuc attribu tungste. molybde: the tun; are due 0f thesq emissior room ten follows Cluster emission face-bri deactiVa photephy manifold enel‘geti ChanQES nus end the hon] States In‘. State, C 104 differences to those of the molybdenum(II) clusters. The radiative and nonradiative decay rates are approximately an order of magnitude greater than those for the molybdenum hexanuclear clusters. The larger radiative rates have been attributed to the greater spin-orbit coupling constant of tungsten as compared to molybdenum.66b Similar to the molybdenum complexes, k1. is relatively constant throughout the tungsten series and differences in the observed lifetime are due to the variation of knr across the series. Analysis of these data is shown in Figure 18 which plots 1n knr vs emission maximum for the tungsten(II) halide clusters in room temperature acetonitrile. Unlike the Moe system, which follows energy gap law behavior across the entire halide cluster series, a linear relationship between ln knr and the emission energy exists for only W6 clusters with common face-bridging ligands. These results indicate that a common deactivating pathway is specific to a [W6x8]4+ core. Photophysical studies have shown that the excited-state manifold of the W6 series is congested by a series of energetically similar excited-state levels whose spacing changes as a function of the face-bridging halide.66a To this end, the data in Figure 18 reflect the perturbations of the nonradiative decay pathways by differing W6 excited states. In an effort to "freeze out" the lowest energy excited state, the photophysical properties of the W6 halides were investigated in butyronitrile glasses at 77 K. Table 10 105 Plot of ln knr vs Aem,max of the Figure 18. (NBu4)2W6X8Y6 clusters in acetonitrile shown in Table 9. In knr 106 15.0 14.0 ' 1 13.0 - 2 I. \ 12.0 - \ ‘ h \ O t: \ ‘ x L c I: 4 a 7 _ I I \ 11.0 - 5 q . n 6 ‘ x 1. . \ \ O 8 ( I 10.0 ~ ° 19 9.0 I A l ‘ l ‘ 11.8 12.6 13.4 ”'2 lemmaxle Emis Compoun: Emission Spectroscopic Data of (NBu4)2l6X8Y6 107 Table 10 in Butyronitrile at 300 K and 77 Ka Compound M06C1142‘ M06C188r62_ M05C18162‘ M06Br8C162_ M06Br142- M06Br8162_ W6C1142- W6C188r62‘ w5018162' WeBrgClgz- W58r142‘ W6Br8162- W618C162- W6IgBr62_ W61142' a Error limits defined in Footnote of Table 3. 140 118 98 157 98 55 1.8 3.1 4.4 11. 15. 15. 11. 21. 24. 1400004300 300K 77X 247 200 188 185 156 102 23. 25. 24. 36. 33. 26. 38. 34. 32. (fiwI-‘OCXJQOCDNQ Aem,max/nm 300K 762 789 809 744 782 822 816 810 792 756 744 739 708 695 693 77X 814 838 859 828 848 870 820 795 787 749 738 731 682 680 677 shows th the W6 1'. collecte1 77 K arc excited-1 red-shif1 vibratiol these 77 radiative temperatl differenc the nonr ObSerVed reProduce Slope anI PlOts at localized rate meas temperatu 0f 1n km smaller [Wexelu striking SEIfec°ns 108 shows the excited-state lifetimes and emission maxima for the W6 halide clusters. For the sake of comparison, data collected fOr the Mos halide clusters in butyronitrile at 77 K are also included. For each cluster compound, the excited-state lifetimes are longer and emission maxima are red-shifted at low temperatures owing to depopulation of hot vibrational levels. The calculated decay rate constants for these 77 K glass systems are shown in Table 11. The radiative rates are approximately the same as the room temperature data for both the M05 and W6 clusters. However differences between the Mo6 and W6 series are observed in the nonradiative decay rates. These are most clearly observed for the plots of ln knr vs emission maximum reproduced in Figures 19 and 20. Only slight changes in the slope and intercept of the molybdenum(II) halide cluster plots at room temperature and 77 K again support the metal- localized nature of the excited state. Parallel to previous rate measurements of the tungsten(II) halide clusters at low temperatures,66a the nonradiative rates yield a linear plot of In knr vs emission maximum at 77 K with a significantly smaller intercept and slope as compared with the distinct [W6X8]4+ series at room temperature. This data is in striking contrast to the room temperature measurements. The self-consistent behavior of the W6 series indicates that states energetically proximate to the emitting state are not being populated efficiently at low temperature and the compounds now share the common "6 emitting lumophore. as. E. 3: 0:329:32. 5 osxxozmrfizc to 23:: .300: 3.30 too: 7 I n H xu‘ d H QHQSL. 109 00.0 00.0 00.0 05.0 $0.0H 00.0 00.0H 50.0H >0.0H 00.0 00.0 v0.0 0v.0 v0.0 00.5 sex s 0H 0H 0H NH 00 0H N0 00 00 0.h H.0 0.0 v.0 0.0 0.H H H10 moH\oox .em_es we oeeooeooosoom or ommxoamavomzo to mouse swoon ooeemIooaeoxm .m oflooe Ho oooeooom ea ooofieoo mosses noeom a NH 0H HH 00 0.0 H.0 0.0 0.5 0H 0.0 0.H ®© r4 0 NCOY‘I Hlm 00H\Hx MH THDdB 00.0 00.0 av.0 00.0 00.0 00.0 00.0 0H.0 00.0 00.0 00.0 00.0 HH.0 ha.0 00.0 09 INvHng Immomwsos ImoaomHoa Imoswomos INvHHmQB Immsomomos ImonHoos Imoommflooa Imeefioos 100H0s0002 lmvfihmmos Imosowomooz Imosmsooos Imoommsoooz ImeHUQOS csooaaoo .0H .VH .0H .0H .HH .0H .0 r-INC'JSI‘LOQDFQ hyanm 110 Figure 19. Plot of ln knr vs Aem,max of the (NBu4)2Mo6X8Y6 clusters in butyronitrile at 77 K shown in Table 11. In kn, 111 9.5 70 A 1 1 l 1 l 1 L n 11.4 11.6 11.8 12.0 12.2 12.4 Xem,max/"‘"< Figure 19 112 Figure 20. Plot of In knr vs "em,max of the (NBu4)2W6X8Y6 clusters in butyronitrile at 77 K shown in Table 11. In km SE In km 113 10.6 10.2 9.8 12.0 13.0 Xommax/kK Figure 20 14.0 15.0 Lif clusters observed lifetime hexanuc] The roor SOlUtiO] samples emissio- decay 1 Cluster Solids nonradi radiati laCk of to 114 Lifetime and quantum yield data of the Moe and W6 clusters in the solid state qualitatively parallel those observed in solution. Table 12 shows the excited-state lifetimes and emission maxima at 300 K and 77 K for the hexanuclear molybdenum(II) and tungsten(II) halide clusters. The room temperature properties follow the same trend as the solution data, and as with the 77 K glass data, cooling samples to 77 K lengthens the lifetime and decreases the emission energy. Table 13 shows the quantum yields and decay rate constants for the room temperature all-halide cluster solids. The quantum yields are much smaller for the solids than in solution owing to the fact that the nonradiative rate constants are much larger than the radiative rate constants. This result may be due to the lack of solvent molecules in the second coordination sphere to accept the dissipated vibrational energy upon radiationless deactivation. In summary, the intramolecular excited-state properties of the molybdenum(II) hexanuclear clusters are dependent on the nonradiative decay rates, which obey energy gap law behavior. The deactivating accepting mode is primarily a low frequency metal-metal core breathing vibration and it is the low energy nature of this accepting mode that is the overriding reason for the long-lived excited states and high emission quantum yields for these cluster systems. The tungsten(II) halide clusters, on the other hand, exhibit more complicated behavior due to the presence of (NBU4)2 Compound :O'3C1142- 110915513: Lh6C181 2- 310681" (I) 6 0163 1053r142- 105Br8162- 15C1142— WGCISBI‘BB‘ w50.8162— 163F8C162~ IeEr142— 165r816e- ‘ leI8C162~ 16188r62~ 1.61142- a r Compound M06C1142- MO6C188r62- Mo6018162- Mo68r80162‘ M06Br142- MOeBrgIGZ- W6C1142' w60188r62- W6C18162- W6Br8C162- W68r142' W68r8162- W6I8C162— WGISBrGZ' W61142' Table 12 Emission Spectroscopic Data of (NBu4)2l6X8Y6 in the Solid State at 300 K and 77 K3 to/us 300K 118 100 I-‘(DOJNID-OONIIZ- 77X 213 183 114 147 135 81 21. 22. 19. 31. 29. 23. 28. 29. 30. I'D-KIN m +4 q -0 e no Aem,max/nm 300K 77X 765 822 785 835 850 878 749 837 784 848 851 893 800 816 800 804 878 778 742 767. 732 751 728 748 702 678 698 679 695 681 aError limits defined in Footnote of Table 3. duv— CCmu dew 3442—7. How—CV. CZ.— Cm Q>~uxue=0A§222V LC 7.»... n... H Qua-eh. 116 0N.0H 0v.0H 00.0H 50.0H 50.0H HN.HH 05.HH mm.HH 0H.NH 00.0H ¢H.m b¢.0 00.0 0N.0 No.0 Max ad .au can as oaaum cwfiom may aw wwwxmanA¢=mz0 no moaam mm mm vv mm mm vb ONH 00H 00H mm 0.0 ma 0H 0.0 m.w Hum moH\u=x .m mfinme Ho mpoapoom a“ umcfiwmo mpfiefia poupm a Hm mm mm vm um um on mm ma m.H m.m m.N 0.H 0.H w.H le NGH\H£ ma mafia? ma.0 0H.0 N00.0 000.0 mv0.0 000.0 vN0.0 Nm0.0 500.0 000.0 0N0.0 0H0.0 0H0.0 0H0.0 HN0.0 o. INvHHmz -mmgmmHo; -mmfiomHms ummngmms INVHHmmk -mmfiomumms ummeHoms -mmgmwfioos -mvflflomz INGHmeOOS -mvfiumoos Immdowhmmos -mmHmHumoz -mopmwfioooz INvHHUQOS unsonaoo muom: madamlumawoxm energet nonradj tempera depopui excitec clustel W6 me‘ vibrat: these nonradj behavic 117 energetically close-lying excited states that mediate nonradiative decay to the ground state. At low temperatures, these nearby excited states can be depopulated, and the nonradiative decay pathways of the W6 excited complexes, therefore, are similar to the M06 clusters in that the excited-state decay is dominated by a W6 metal-metal core breathing mode as the accepting vibration. At higher temperatures, a Boltzman population of these proximate excited states provides additional nonradiative decay channels and linear energy gap law behavior is not observed. c. Intermolecular Quenching Processes 1. M9129!!! Contributions to the nonradiative decay channels of electronically excited molybdenum(II) and tungsten(II) Clusters arise not only from intramolecular pathways but from intermolecular energy- and electron-transfer quenching processes as well. For example, a number of organic triplet species are efficient quenchers of the M06 and W6 electronic exerted states via energy transfer.6613 More prominent (Wenching processes of these cluster systems occur by ele(flit-on transfer. As shown in Figure 11, the electronic excited state of M06C1142' will reduce and oxidize species With reduction potentials greater than -0.54 V and less than +0- 20 V vs SCE, respectively, and similar schemes may be derived for other hexanuclear clusters. Thus, the eXCitEG states of transfer q cluster ex utilized t range of <11 The r described ‘ 7. For a rewritten RT Ink “here kq‘ (A). Thr this rat >> ‘X contriby the dECreaS electro actiVat maximun decreaE regime: Expert limitii: diffus 118 states of these clusters may be used to probe electron- transfer quenching processes.66d'102 Specifically, the cluster excited states are powerful reducers and may be utilized to investigate quenching reactions over a wide range of driving forces. The rate of electron-transfer quenching is adequately described by the semi-classical rate expression given by eq 7. For a series of similar acceptors and donors eq 7 can be rewritten as103 AG'2 AG° RT 1n ke-t = RT 1n kq(0) '- 4A - —;— (30) where kq(0) includes work terms and reorganizational energy (A). Three regions of electron transfer are described by this rate expression. For mildly exergonic reactions, AG' >> -x, and the quadratic term of eq 30 insignificantly contributes to the electron-transfer rate. In this regime, the electron-transfer rate increases linearly with decreasing AG°. This is called the normal regime of Electron transfer. When AG° = -A, the reaction is activationless and the electron-transfer rate is at a maximum; for more exergonic reactions, AG' < -A and ke-t decreases with a further decrease in AG° (i.e., the inverted regime). Figure 21 shows a plot of RT ln ke-t vs AG'. Experimentally, the diffusion-controlled limit will be rate- limiting if electron transfer is faster than reactant diffusion in solution. Accordingly, the three rate regimes, Figure 21. showing three rate regimes: 119 Exemplary plot (1) controlled; and (III) ke-t vs AG° of RT In normal; (II) inverted diffusion- regime. kact 120 K-Vomu< K-A00< 9. ~05v— "931 uLLB the r signi: 21. T elect: the 1 system: behavi< require regime activat establi equilib: mechani: be desc electro] mechani1 BecaUSe c°"trol: dynamic: electroI in“ dii Po] will So] 121 the normal, diffusion-controlled, and inverted regime are signified by regions I, II, and III, respectively, in Figure 21. The normal and inverted regimes comprise activated electron-transfer reactions. Although electron transfer in the normal regime has been identified for several systems'2,lO4-108 few systems displaying inverted-regime behavior have been observed owing to the high driving forces required for such reactions. Only recently has the inverted regime been identified experimentally.109 In these activation regimes, an equilibrium. between reactants is established, and electron transfer proceeds within an equilibrium precursor complex. This has allowed the mechanistic details of electron transfer in these regimes to be described. In contrast, although many systems react by electron transfer in the diffusion-controlled regime, the mechanistic details remain for the most part undefined. Because electron transfer is not activated in the diffusion- controlled regime, the overall rate is governed by a dynamical process. Experimental and theoretical activated electron-transfer studies, therefore, do not provide insight into diffusion-controlled reactivity. For the diffusion-controlled regime, the reaction rate will solely depend on the rate of diffusion (kD) given by for un accept: center- reactic also th Although can ech Electron regimes reaC-‘l’iOh: reclimes c states. The transfer Velmer (: steadY-st 122 for uncharged reactants where D is the sum of donor and acceptor diffusion coefficients (D = DD + DA) and r is the center-to-center distance. Therefore, the overall rate of reaction not only depends on the electron-transfer rate, but also the diffusion rate as shown by eq 32 k k D -t kobs = e (32) kD + ke_t Although electron-transfer reactions in the normal regime can experimentally be interrogated by thermal reactions, electron transfer in the diffusion-controlled and inverted regimes are best studied by photoinduced electron-transfer reactions because the driving forces characteristic of these regimes can be achieved with molecules in electronic excited states. The observed rate constant of photoinduced electron- transfer reactions may be determined by using the Stern- Volmer (SV) quenching techniques9 Time-resolved and/or steady-state quantum yield emission data are recorded as a function of quencher concentration and applied to eq 33 Io/I = l + KSV [Q] (33a) ro/r = 1 + KSV [Q] (331)) st = kq'o (33c) where 70 and emi: quencher, constant transfer of a plo‘ The: photoind: plots. ' linear b Controll. by grow CIllt’éncher Static distingu: Because 4 unc°mple1 linear. be 9r9at pairs d dQViatiOr behaViOr 123 where re (1) and I0 (I) refer to the excited-state lifetime and emission intensity in the absence (presence) of quencher, respectively, and KSV is the Stern-Volmer constant. From eq 33, kg, which is the observed electron- transfer rate constant, is simply determined from the slope of a plot of ro/r or Io/I vs [Q]. There are, however, experimental conditions in which photoinduced electron transfer does not lead to linear SV plots. The majority of experiments that show deviation from linear behavior are those in which static quenching is the controlling mechanism.11°"113 This often involves quenching by ground-state ion-pair formation between lumophore and quencher to yield a complex that does not luminesce. A static quenching mechanism can be experimentally distinguished by time-resolved and steady-state SV plots. Because the lifetime decay only reflects contributions from uncomplexed lumophores, the time-resolved SV plots will be linear. Conversely, relative luminescence intensities will be greatly diminished by the presence of nonemissive ion pairs and steady-state SV plots will show positive deviations from linearity. This deviation from linear behavior is described by eq 34 Io/I = (1 + KSV[Q])(1 + KquQl) (34) where Keq the ion p ground-st.- will occu: transfer ion pair . course, rates, t efficient Devi.- for sever. Pair form.- from reac are in Processes derived diffusio quenchin incl-gage therefor other no excited With qu those luIIIOphOr Nor diffuSio 124 where Keq is the equilibrium constant for the formation of the ion pair. More complicated behavior is observed if the ground-state ion pair emits.11‘1'115 This type of quenching will occur forion pairs that have relatively slow electron- transfer rates such that the radiative decay of the excited ion pair can compete with the electron-transfer process. Of course, for ion pairs with very fast electron-transfer rates, the nonradiative decay pathways will be very efficient and static quenching will result. Deviations from linear SV behavior have been observed for several years that can not solely be attributed to ion- pair formation.116 Commonly such behavior has been observed from reactions in the diffusion-controlled regime and hence are interesting because they focus on dynamical 117 Baird and Escott,118 and Keizer119 have processes. derived models, based on the Smoluchowski theory of diffusion-controlled reactions, which show that the quenching rate constant increases as quencher concentration increases. Positive deviations from linear SV behavior, therefore, are predicted at high quencher concentrations. Other models used to explain SV deviations propose that excited lumophores possess higher probabilities for reaction with quenchers adjacent to the lumophore as compared to those that must diffuse through solution to the lumophore.116e'120 More recently, Cukier has addressed the problem of diffusion-controlled quenching reactions between charged 125 reactants.121 By using the Smoluchowski-Debye diffusion equation, deviations from linear SV behavior are predicted for oppositely charged reactants. The magnitude of this deviation is dependent on the solvent dielectric constant and solution ionic strength. This theory is important because most biological and small molecule activation electron-transfer reactions occur between charged reactants. Accordingly, the experimental design of systems probing this theory could lead to greater insight of dynamical electron- transfer in a wide variety of systems. The hexanuclear molybdenum(II) and tungsten(II) clusters are well-suited to study the fundamentals of electron-transfer reactions. Firstly, the quenching reaction driving forces may be systematically varied over a wide energy range. Therefore, in principle activated and dynamical electron transfer can be investigated by simply tuning the cluster redox potential. Secondly, the clusters possess long-lived excited states which is a necessary prerequisite ‘when studying' diffusion-controlled. quenching reactions because low concentrations of quencher will significantly attenuate the luminescence lifetimes and intensities. Compounds possessing short-lived excited states pose a problem to electron-transfer studies because the lifetimes and emission intensities will be difficult to measure at high quencher concentrations. Finally, the clusters are charged and hence provide the opportunity to study dynamical electron transfer theories and in 126 particular, that developed by Cukier. In addition to exploring dynamical electron transfer, intermolecular electron-transfer quenching processes that may contribute to nonradiative decay channels of these electronically excited clusters in heterogeneous environments may be identified. Herein is a report of the oxidative quenching of the electronically excited hexanuclear’ molybdenum(II) and tungsten(II) clusters by pyridinium- and bipyridinium- related electron-accepting quenchers. 2. gesglts and Qiscussion Table 14 displays the excited-state M6X14'/2'* and ground-state M6X14’/2' reduction potentials for the molybdenum(II) and tungsten(II) clusters used in the oxidative quenching studies. Excited-state energies (AE(2/2-*)) are estimated from the 77 K emission spectra by determining the energy at 5% of the peak maximum on the high energy side of the emission band.122 From these data, reduction_ potentials of the 2M6X14-/2'* couples are determined from the appropriate modified Latimer diagrams. The series of electronically- and structurally-related pyridinium ions used as electron-accepting quenchers (A+) of the M06 and W6 electronically excited clusters are listed in Table 15. Quenchers were electroanalytically characterized by their E1/2(A+/o) reduction potentials and in some cases, NMR spectra were recorded to verify the structure. The hexafluorophosphate salt of 2,2'-bipyridinium was also .sppooew conmHEo x be Eosm oocflsuouoc mowmeoco oumumnooufioxm o .mamsmafio posfipmq ooflmfloos ousfieooeeem one Eoem moaosoo *Im\|vax®2 one no“ counsfipmo mfieflucouoe coHuosoom a modesoo Im\|vax02 on» no“ mamflucouoo cOaposcom o m.H ve.o- mv.H+ -mvaflomoz m.H me.o- mo.H+ -mvfifiooa m.H oo.H- om.o+ -mvfipmoz m H.m H¢.H- mm.o+ -mvHHos > \ mom w> o>o\A*IN\INVm< nA*IN\IvN\HM oAIN\I0N\Hm openness; onoaoo< as vfixoamavsmzc tenuous saueoeaopaoofim one madamI6:50hu Ho moduthOHQ HdofiaoQOOHHOOHm vd wand? 128 Table 15 Reduction Potentials of Pyridinium and Bipyridinium Quenchers in Acetone quencher E1/2(A+/0)/Va l-methy1-2,2'-bipyridinium -O.38 1-deuterium-2,2'-bipyridinium -0.46 2,2'-bipyridinium -O.47 4-cyano-N-methylpyridinium -O.78 4-carboethoxy-N-benzylpyridinium -O.84 4-amido-N-benzylpyridinium -1.00 4-amido-N-methylpyridinium -1.05 l-methy1-4,4'-bipyridinium -1.08 a Reduction potentials for the A+/O quencher couples. 129 identified by elemental analysis (Galbraith Laboratories #X- 2203). The driving forces (AG°) listed in Table 16 of the electron-transfer reactions 2-* + .- are related to the formal redox potentials by the simple thermodynamic relation103 (36) AG° = El/2(M6x14‘/2‘*) - E1/2(A+/0) + w - wr P where wp and wr respectively. The minor entropic contributions to AG° have are the product and reactant work terms, been ignored. Solving eq 10 by using a center-to-center distance of 9.5 A for the cluster/quencher precursor complex yields wr = -0.14 eV in acetone (I = 0.001 M); and wp = 0 eV for all quenchers. From_ these data, and quencher and cluster reduction potentials, the electron-transfer driving forces shown in Table 16 directly follow from eq 36. Quenching rate constants, displayed in Table 16, were calculated from Stern-Volmer (SV) analysis of lifetime quenching data. SV plots for systems 1-4 and 7-11 in Table 16 are linear' to a quencher' concentration of .. 1. mM. Standard linear SV behavior is observed for systems 5 and 6 at quencher concentrations less than 3 x 10-5 M. Figure 22 shows the SV plots at low quencher concentrations for these 130 OHOH x H.m OHOH x H.v 00H x 0.0 OHOH x m.m OHOH x m.N 00H x v.H wOH x N.H 50H x m.m 00H x v.H 00H x m.H mOH x ®.H Humensxux 0v.0I mv.0I 0m.0I NN.0I 0H.0I ¢H.01 5H.0+ wv.0+ vm.0+ 0>.0+ mh.0+ >®\oU< EsficfioweseaanuoEIZIOcs>0I¢\*INvHH03 EsflsfiofiemeaansooIZImxonpooohaOIv\*INVHH03 Eoficflofiexewou.N.N\*Imvapm®3 as“cficfip>oachoEIZIocflEslv\*ImwaH03 esflaflospsaen-.e.vuasnpme-fi\*umeaems ssficflcfitsaflnn.m.m\*umefleoms esfiastfinsafiou.m.m\*umeaeooo= EsficficfipzeamnpoEIZIOcmzolv\*IN¢HH0002 Esflcwcfiemea>Nson|21>xocuoooesolv\*Imvfiaomoz EoflcfioweseamNconIZIocHEwlv\*lmvHH0002 EsflcflcfipmoamapoEIZIooHEslv\*nmNHH0002 Honooosa\ohoneoadq HH 0H «3 co 5 V N H hyanm occaooe cu maouwhm aswowofinhm\IN¢HN®I Ho maneuwoou ovum mafiaooood one weapon weabuhn 0H @Hndfi 131 Figure 22. Stern-Volmer plots of M06C1142'* (o) and W6C1142'* (n) quenching by 2,2'-bipyridinium at low concentrations in acetone (I = 0.001 M). «a museum 132 .2 wow \ .9 3 3 ... o . . . . . . . . o3 ' e _ I . 1 m0; . i H\oP . 1 OF.— m0— and 1W 133 two systems; quenching rate constants determined from the slopes of these plots are 1.2 x lo8 and 1.4 x 109 M'ls'1 for systems 5 and 6, respectively. At higher quencher concentrations, however, systems 5 and 6 exhibit negative deviations from standard SV behavior as depicted in Figure 23. The deviation from linearity is not the result of static quenching due to the fact that ro/r and Io/I plots yield identical results. This observation establishes that the quenching reaction is a dynamical process. The general shapes of these plots are consistent with the formation of emissive ion pairs.11°brll4 Further support of ion-pair formation is shown in Figure 24, which displays the SV plots for the Mo6C1142'/2,2’-bipyridinium system in acetone at ionic strengths of 0.001 and 0.1 M. Qualitatively, the expected dependence of ion-pair formation on ionic strength is shown in that the deviation from linearity is suppressed with increasing ionic strength. Contributions of ion-pair formation to the observed quenching rate constant may be quantitated with recent mechanistic modelslll'114 that allow for several excited- state decay pathways for the ion pair as in the following scheme, 134 Figure 23. Stern-Volmer plots of M06C1142'* (a) and W6C1142'* (I) quenching by 2,2i—bipyridinium at high concentrations in acetone (I = 0.001 M). ..mu 135 an unseen s. e: \ .o. m F n.. d .u.p II 1— F.F A P\oH .m.F and high 136 Figure 24. Stern-Volmer plots of quenching reaction between 2,2'—bipyridinium and M06C1142'* in acetone at ionic strengths of 0.001 (n) and 0.1 (o) M. 137 up or vu Gunmah .2 m2 \ 5. d F; N; 0.. v.9 m4 H\oP 138 (A*-Q) 1“ > (No) + hu (A*°o) km, > (No) (A*'Q) kq' > A + Q (A*°Q) + Q kq" > A + 20 Scheme I where (A'Q) and (A*'Q) represent the ground-state and emissive ion-pairs, respectively. By using this model, the equilibrium constant for ion-pair formation and the quenching rate constants for the "free" and complexed lumophores may be analytically determined from the SV plots. The calculated equilibrium constants from this analysis are > 103 than those calculated with a standard Debye-Hfickel formalism.123 This discrepancy may be due to the fact that, unlike Scheme I, several equilibria exist between the dianionic cluster and monocationic acceptor. Under these conditions, SV expressions comprised of several equilibrium constants are derived that cannot be deconvoluted without independent equilibrium constant measurements. For the purposes of this discussion, the quenching rate constants determined from the low concentration measurements of systems 5 and 6 will suffice. A plot of RT 1n k vs -AG° for the data shown in Table q 16 is reproduced in Figure 25. A slope of -o.5 is predicted 139 Figure 25. Plot of RT 1n k vs -AG° for M6X142'* q / pyridinium systems shown in Table 16. The solid curve is the plot of eq 30 with kq(0) = 8.8 x 108 N’ls‘l and A = 1.04 eV. The dotted line represents the diffusion- controlled limit (~ 2 x 1010 M-ls'l). 140 m.— 0... mu change >0 \60 <- m6 0.0 0.7 0.0 ‘ 1‘ a o.— Ae / bx ULLH 141 by eq 30 for low driving force activated electron-transfer reactions. The observed slope of -0.44 in the linear regime agrees well with this prediction and thus establishes an electron-transfer mechanism for the quenching of M6 clusters by pyridinium-related quenchers. This analysis relies on the assumption that the electronic and molecular structures within a cluster series and within a quencher series are similar; the results of Chapter III.B. support this assumption for the cluster reactants. For both the M06 and W6 halide clusters an electron is removed from the metal- based aZg orbital in the reaction given by eq 35. As is generally observed for bimolecular electron- transfer reactions, activated electron transfer is circumvented as the driving force is increased by the diffusion-controlled limit, which is responsible for the leveling of the rate at AG° = ~ -0.25 eV in Figure 25. That the diffusion limit is achieved at relatively low driving forces suggests that the M6X142'/pyridinium systems offer the ideal opportunity to study dynamical electron-transfer processes. The theoretical dependence of RT 1n kq on AG° can be obtained from eq 30 with a knowledge of kq(0) and A (reorganizational energy). The electron-transfer rate at AG° = 0 eV, defined by kq(0), is estimated by the average rate constant of entries 5 and 6 in Table 16 to be 8.8 x 108 M'ls'l; A for M6X142'7pyridinium electron transfer is 1.04 eV.124 From these values, the solid curve in Figure 25 is obtained. Comparison of the theoretical rates with the 142 observed rate constants show that the electron-transfer rates for systems 9-11 (~ 7 x lo11 M"1 s'l) are an order of magnitude greater than the diffusion-controlled rate (~ 2 x 1010 M'ls'l) indicated by the dashed line in Figure 25. This result suggests that reactions with driving forces more negative than systems 9-11 are necessary to satisfy the necessary criterion, ke-t >> kD, for exploration of the dynamical electron-transfer regime. Three systems satisfying this criterion are shown in Table 17. Calculated activated electron-transfer rates for these systems are ~ 1 x 1013 M'ls'1 which is three orders of magnitude greater than the diffusion rate and thus, may be used to study dynamical quenching processes. Stern-Volmer plots for the. quenching reactions of ‘W6I142'* by 2,2'- bipyridinium(I), 1-deuterium-2,2'-bipyridinium(II), and l-methyl-2,2'-bipyridinium(III) are displayed in Figure 26. The fact that. these SV’ lifetime plots exhibit positive deviations establish a dynamical electron-transfer process (emission intensity and time-resolved plots are identical). These experimental observations for quenching between the oppositely charged clusters and pyridinium acceptors are consistent with the recent predictions of Cukier.121 By using an effective medium theory, positive deviations from linear SV behavior are predicted with introduction of an effective volume fraction (3) which is dependent on the quencher concentration [Q], 143 .>o vo.H u e new Humfiuz woe x m.w H onex nee; cm at ween: so nopeesoeeo a mace x e.H mm.o- eeeefltfiesefle-.m.mueeeeme-H\*umeHHoe mHoH x m.H Hw.o- eefiefieeeeese-.«.mueeeeoeeoele\*umeHHos mHoH x m.H ow.o- seflefleeeseflel.m.m\*-mvHHma ufilmHII \Aoasovalox >o\o0< Honooosa\ohoneoasq occaooe ow maoumhm Esuowcwhhnwm\lmvdxwl Ho mueeamooo ovum hommodhhlsouuooam coweasoado one woohom wcfi>whn NH mfindh vH ma NH macaw 144 Figure 26. Stern-Volmer plots of quenching reaction between W61142-* and (a) 2,2'Fbipyridinium; (b) 1- deuterium-Z,2’-bipyridinium; and (c) 1-methyl-2,2’- bipyridinium. :1 represent experimental data and the solid curves are best fits by using eq 40 (see text). 145 eon ousmum s. @210. ov on cm 0— co on See 146 he" ousuam .2 o2 \ 5. 00 cm 0v cm 00 v\ov 147 new enemah .2 we \ 5. on om ov om co :3. 148 3 = 4xNAf’[Q]/3ooo (37) where i': is an effective distance. This effective volume fraction arises from electrostatic interactions of charged reactants which results in the nonlinear SV relation, at small 3, described by eq 38 3/2 ro/r = l + kq r0 [Q] l + < > (33)1/2 (38) 1+u where r is the usual lifetime in the absence of quencher, O and u = ke-t/kD with kD defined as k0 = 4nNADf/1000 (39) Two limiting quenching regimes are defined by this master equation (eq 38). In the activated electron-transfer regime (u << 1), the standard SV equation is obtained and linear quenching behavior will be observed. Alternatively, in the diffusion-controlled regime (u 4 co) , eq 38 becomes ro/r = l + kDr°[Q][l + (33)1/2] (40) and positive deviations will result. For larger 3, ro/r and [Q] are related by the quadratic expression 149 ro/r = l + kDro[Q](3$) (41) Experimentally, the quencher concentrations required to obtain large 3 are difficult to achieve, and eq 40 describes most quenching reactions and certainly the M6 cluster/pyridinium systems. Substitution of eq 37 and 39 into eq 40 yields 1000 1000 where eq 37 and 39 have been substituted for 3 and k0, respectively (SV represents parameters of Stern-Volmer analysis). Thus, a fit of 10/1' vs [Q] yields DSV and fsv. The optimum value of DSV' iteratively determined by fitting the data shown in Figure 26 to eq 42, is 1 x 10"6 cmz/s. Using this value in a Kinfit65 analysis yields the fSV values listed in Table 18 and the solid curves in Figure 26. These *malues from SV measurements may be compared to independent determinations of D and f for the W6I142'/bipyridinium systems. The diffusion coefficient for W6I142- and quenchers I, II, and III can directly be measured with potential step chronoamperometry. A plot of i vs l/tl/2 (i = current; t = 150 .Amv gov cofipssoo HHoeuuoo one ween: so «use mepoeonoeeeocoeco eopm Hewpcouoe some cocHEeoqu a .me an op om mesmHe eH meet meHeeHe en emeHseoemo a muoH x m.e emH @10H x H seHthHeeeHe-.m.m-Heeeme:H\*umvHHee vH mIOH x w.m mmH mnoH x H eeHeHeHeeeHeu.«.museHemeeoeuH\*nmeHHo3 mH muoH x e.m mmH ouoH x H eeHeHeHeeeHeu_m.m\*nmeHHoe NH nHImNEo \ouaon om\>mm «HIwNEO\>m0 Honocosd\ouoneoasu heanm havoaonoeado:0hao Qoam Howaoouom ha consume! maaofiowuuooo nofimsuwwn one .mauh hoauo>lahoam ha ooewahouon moondumwn 0>Haoouum one maooaowuuooo nowwsmmwn 0H GHDGE 151 time) yields D for a species of concentration Co* as described by the Cottrell equation61 i(t) = nFAD1/2C°*/xl/2t1/2 (43) where n is the number of equivalents, F is Faraday’s constant (9.648 x 104 C/eq), and A is, the area of the 3 electrode. If i(t) is expressed in amps, Co* in mol/cm , t in seconds, and A in cmz, then the diffusion coefficient has the standard units of cmz/s. As predicted by eq 43, plots of i(t) vs l/tl/Z for "5114 ', I, II, and III are linear with slopes of 2.81, 20.3, 25.4 and 6.58 Mel/2, respectively, and intercepts of 0.00 e 0.7 uA. The calculated diffusion coefficients125 of 1.5 x 10'5, 1.9 x 10'5, 2.3 x 10's, and 2.8 x 10"5 cmz/s for W61142-, I, II, and III, respectively, in acetone (I = 0.01 M) are consistent with calculated values using the Stokes-Einstein equation.126'127 The overall diffusion coefficient, D, of the three W6I142'/bipyridinium systems (D = D(W61142') + D(A+)) are given in Table 18. The effective distance (i‘) for electron-transfer quenching between the oppositely charged W61142' and bipyridinium reactants may be calculated from a Debye-Hfickel potential fDH = -zAercexp(-nrc) (44a) 152 where 2A and zQ are the reactant charges, rc is the Onsager length128 (560.6 A/er with c the dielectric constant), and r n2 is defined by n2 = (ZezNApfi/IOOOeocr)I (44b) In eq 44b, p is the solution density, p = 1/kBT, e is the o permittivity' of a vacuum, and I is the solution ionic strength (in room temperature water (er = 78.5) , for I = 0.01 M, :c = 3.3 x lo6 cm‘l). For the W61142'/bipyridinium systems in acetone (I = 0.001 M), fDH = 33 A. The experimental values of DSV and ESV determined from SV analysis with eq 42 (see Table 18), qualitatively agree well with those calculated from eq 43 and 44, respectively. Moreover, Cukier’s theory of diffusion-controlled quenching makes two other striking predictions about the dependence of ro/r on [Q]. Namely, the magnitude of positive SV deviation is predicted to depend markedly on solution ionic strength and solvent dielectric. Figure 27 displays the effect of solution ionic strength on the magnitude of deviation for ’* and III in acetone. the quenching reaction between W6Il42 Because EDH decreases with increasing ionic strength (eq 44), the positive deviation of ro/r is predicted by eq 42 to monotonically decrease with increasing I. The data in Figure 27 confirm this prediction. The solid curves in Figure 27 represent the Kinfit65 best fits using eq 42: the results from this analysis, along with the calculated values 153 Figure 27. Stern-Volmer plots of quenching reaction 2-* between W6114 and 1-methyl-2,2'-bipyridinium at ionic strengths of 0.001 (a), 0.011 (+), and 0.102 M (u). Symbols represent experimental data and solid curves are best fits by using eq 40 (see text). 154 co ow nu unseen s. e: e 5. ov 1 cm 1 om Sow :ion mic ov 155 of ion using eq 44, are shown in Table 19. As previously observed, the experimentally determined values of rsv are larger than the calculated values, but the expected ionic strength trends are followed. Unlike the ionic strength effect, the contributions of the solvent dielectric constant on SV deviations is not straightforward. With increasing er, eq 44 predicts the increase in EDH derived from the exponential to be countered by a concurrent decrease in the pre-exponential factor (the Onsager length decreases). The value of er yielding the maximum EDH’ and hence the largest deviation from linear SV behavior, can be determined by differentiating eq 44 with respect to er. Solving eq 44 for di‘DH/der = 0 yields er = 18, and thus predicts that quenching reactions in solvents with er's nearest 18 will exhibit the largest deviations from linear behavior. Figure 28 displays the SV plots for the quenching reaction between W6I142- and III in acetone (er = 20.7), acetonitrile (err = 38.8), and dichloromethane (er = 9.08) at I = 0.001 M. As predicted, the quenching reaction in acetone shows the greatest deviation from linear behavior. Thus, the first theory to predict positive deviations from SV behavior resulting from electrostatic interactions in dynamical quenching regimes, has been successfully verified by the W6I142'/bipyridinium electron-transfer chemistry. The important parameter, the effective capture distance r is qualitatively predicted. Quantitative 156 mm 0.0 N.0 sw\mom boa 50H 0v s¢\>mm 0fl0900¢ UH fl Scum one .vv co wsHm: he ooumaooaso n .mv so 05 sm oesmHm ea wees one msHeeHe es soeHssmeoo a Hoo.o ssHeHsHeseHe-.m.muHsseosuH\*umeHHos HHo.o seHeHsHessHsu_«.muHseeosuH\*umeHHos moH.o seHeHeHeseHs-.m.m-HseeesuH\*umeHHee :\H no:ooo=0\ohoaeoasq awooham owsoH Ho ooHuoosh a we mooHucasoan Hoxoanlohaon mawh hoEHo>lohoam EOHH coofiahoaon mooseumwn obuuoowmm 0H @Hndh 157 Figure 28. Stern-Volmer plots of quenching reaction between W6I142'* and 1-methyl-2,2’-bipyridinium in acetone (o), acetonitrile (+), and dichloromethane (e) at an ionic strength of 0.001 M. Symbols represent experimental data and solid curves are best fits by using eq 40 (see text). 158 Dow on auscuh .2 we \ .o. S 8 3 em 6 . . . 3 . . . o I“ I .+ a + A ... D I 1 S D + . See 1 8 D L 8 in m at ed am 159 discrepancies between 2’ determined from SV analysis and Debye-Huckel theory is difficult to rationalize. The calculation of {DH involves integration over all distances (x) of a electrostatic potential interaction (V(x)) between the charged reactants, co l/i‘on = f expwvmn/x? dx (45) r In the present analysis, V(x) was a Debye-Hfickel potential. More generally, V(x) may include a Debye-Huckel term as well as other potential interactions, V(x) = V1(x) + V2(x) + ..... (46) The importance of ion-dipole interactions on trans ition- metal excited-state nonradiative decay processes have been considered for polypyridyl complexes of ruthenium and osmium.71b The inclusion of an ion-dipole interaction,129 V2(x) = rID4/x4 (47) in V(x) resulted in less than 10% increase of 1'? than that calculated considering only Debye-Hfickel formalisms. Additionally, dielectric saturation, for which the dielectric constant appears smaller near the ion than the bulk dielectric of the solvent, may contribute to V(x) in 160 130 This effect was found to be minimal ionic environments. at best. The fact that short-range interactions, such as ion-dipole interactions and dielectric saturation, provide negligible contributions to the magnitude of E is not surprising based on the fact that the Debye-Hi'lckel interactions are relatively large and dominate at long distances. Although some intimate details relating theory and experiment presently remain unresolved, the theory developed by Cukier can be implemented, for the first time, to describe the observed results of electron-transfer reactions in dynamical quenching regimes. The intermolecular electron-transfer reactivity of the hexanuclear’ molybdenum(II) and tungsten(II) clusters are important considerations for the nonradiative decay of the electronic excited state. Not only do emissive ion pairs form between the clusters and charged species in solution, but enhanced quenching occurs in the diffusion-controlled regime when the quenchers are cations. In contrast to the insensitivity of the intramolecular nonradiative decay channels of the metal-localized M6 excited state to environment described in Chapter III.B., intermolecular decay via quenching is extremely efficient. Therefore, the excited-state properties of the M5 clusters in heterogeneous environments containing electron sources or traps will be greatly perturbed. To this end, environment may play an important role in the design of novel solid-state cluster- based luminescent materials. 161 D. Cluster Modified Polymers 1- we The properties of polymer-bound luminescent compounds afford these materials for application in photocatalysis , 13 1'135 electrocatalysis , 136-138 molecular electronics ' 13 6 and sensor technologies . Sol id-state materials have found particular importance in the design of catalytic139 and technological140 schemes. A popular approach for the synthesis of luminescent solid-state materials is to electrostatically or covalently bind the lumophore to an inorganic or organic host structure. In this regard, polymers have been particularly useful host matrices. Polymer supported luminescent materials provide several advantages over conventional solution-based systems. Site isolation of photoreactive centers applied to polymer backbones has been successful in inhibiting back-electron transfer reactions in photocatalytic cycles.141'142 For instance, polypyridyl complexes of ruthenium( II) have been studied extensively for use in water-splitting schemes in which the metal complex is covalently attached to pyridyl- "_- - 143 type polymers as shown below (N N — blpyrldyl). ----(CHmzh----(mmz» ----(CW---- 29 8 N N r / R: N N —N k. \) '2Cl' \l/Z 162 A main theme that has emerged for systems such as these is that the excited-state properties of the polymer materials are similar to those of the monomers. Photoproducts generated from excited-state electron-transfer reactions are available for useful chemistry and. not depleted. by the energy-wasting back reactions that are common to the homogeneous environment. Polymers have also found success not only in electron- transfer schemes but energy-transfer schemes as well. Photo-oxidation systems based on energy-transfer reactions of sensitizers to produce singlet 02 have particularly benefited from the use of polymer-based materials in which the sensitizer (e.g., Rose Bengal) is covalently or electrostatically attached to the polymeric backbone.132 These materials outperform the analogous homogeneous systems because photobleaching of the sensitizer over long periods of time, reactions between the sensitizer and substrates and/or products, and sensitizer depletion is minimized for polymer-supported systems. Perhaps the greatest contribution of polymers to chemistry in recent years has been in the area of 4 Polymer films containing electroactive electrochemistry.14 transition-metal complexes deposited on electrode surfaces 'engender unique electrochemical properties. Electrodes Inodified with electroactive polymers provide a concentrated and motionally constrained environment in which reactivity they be chemically tailored for specific electrocatalytic 163 applications. Additionally, the ability to manipulate current and voltage characteristics with the microstructural design and chemical composition of the polymer on the electrode surface has led to the development of molecularly— based devices for application in microelectronics technologies.l36 In the systems described above, the polymer is typically constructed from organic monomers. Recently, there has been renewed interest in all-inorganic matrices of ceramic oxides, which are simply polymeric silicate structures. A relatively new area of ceramic development is the use of sol-gel processing145 for the synthesis of ceramic materials incorporating luminescent complexes for applications in optical data storage, dye lasers, fluorescent probes, and photoconductive devices.146 The sol-gel process involves the hydrolysis and polymerization of’ metal and metalloid. alkoxides in the presence of a catalyst at or near room temperature. The hydrolysis and polymerization reactions proceed until a "wet solid", or gel, is formed. This gel is composed of an interconnected oxide phase and pores containing a reaction solvent mixture. Drying the gels results in the expulsion of this solvent phase with substantial volume shrinkage of the gel, leaving a dry porous solid. In principal, the incorporation of essentially an llnlimited number of luminescent compounds may be achieved by addition of the desired compound to the initial reaction 164 mixture. However, many of these lumophores are extremely unstable in the harsh sol-gel environment. For instance, organic dyes decompose readily in the sol-gel matrix.147 Even organometallic and coordination compounds possess limited stability despite recent optimistic reports.148 The hexanuclear molybdenum(II) and tungsten(II) clusters are well-suited for inorganic and organic polymeric environments because the ease of ligand substitution should permit the clusters to be incorporated in a wide variety of matrices. Furthermore, the clusters are extremely robust and therefore should be stable in the polymer environment. The photophysical studies described in Chapters III.B. imply that the excited-state properties of the M6 core will be maintained in the reactive environments of organic polymers and ceramic oxides. Accordingly, the practical application of polymer-bound clusters for light-emitting devices, photo- oxidation chemical schemes, and sensing devices warrants an investigation of the effect of polymer and ceramic environments on the Mo6 and W6 cluster photophysical properties. Herein is a report of my efforts to examine the photophysical properties of polyvinylpyridine polymers and silicon-oxide gels containing the luminescent core of the model M6 system, molybdenum(II) chloride. i. 'c P ' ' ' e r Table 20 displays the luminescent lifetime properties of three cluster-incorporated polyvinylpyridine polymers. These properties are similar to the model monomer complex MoSCllzpyz (py = pyridine) which exhibits broad, featureless emission maximizing at 766 nm with an emission decay profile that can adequately be fit by a biexponential rate law (r1 = 19 us (29%) and 72 = 43 us (71%) in the solid state). Solids and solutions saturated with N2 and 02 of polyvinylpyridine (PVP) containing covalently-linked cluster (PVP-M06C112 at loadings of 10% and 90%) also exhibit multiexponential luminescence decays which may be fit by a biexponential rate law. That the solid cluster modified polymer emission lifetimes are not attenuated by pure 02, indicates that the triplet excited state is not being quenched by molecular oxygen. In striking contrast, the long lifetime component of the emission decay of polymers dispersed in solution is noticeably quenched by molecular oxygen. Although the relative percentages of these solution data are somewhat inconsistent across the series of solvents, the results suggest that the polymer is swelled by solvent and M06C112 sites become accessible to freely diffusing molecular oxygen. The attenuation of the long- lifetime component decreases along the series water > toluene ~ DMF. These data suggest that water swells the polymer to the greatest extent, and DMF and toluene to 1136 as as Hfimnv no wanv mm Hfimvv NH Hfimvv NH Hfiouv mm Hfiosv mm Hesse mH Heesc mH «no owe-ea .-mvHHooosuHweHsHeseHseH>sHoeucoeeHseemsc mH +e>e-ms 00H Afiwmv mm Afimvv NH Aflmbv mm Hfivmv mH vl==oe> wuuum tumom .NO so Nz :uHB couscousm can aco>mow :H coccoemsm we em Hesse eh Hesse em Hesse e.e Asses m.e Asses we Hesse eoH Hesse H.n Heomc s.v No a: oouHaeaeoeHseeoaH: .oeonemosue mo when s OH oomooxo ofiosdm co concedes oEHuoufiq .osocomosue Eooe one Cu comooxm museum :0 consumes oEHuOHHq .huOH 0|0H x v 2 an cohamdms mEHummuq gossHoo cHHom one uocHoHssoH2cH>2HoeuNHHomox we e>e .m: eH omsaHseee mosHeoqu a £3 0 1! fl *4 - - Hesse em Hesev we H+e>e-ezcm. me he Hesse s.h Hesse H.o H+e>e-o=cH. Hesse He lemme es Hesse hm Hesse vs Hess .esevm. Ashes m.e Hesse w.n Hesse m.e Hammc e.v Heom .e>evH. Hemmc me Hesse coH HsHec mm Heomc me HeoH .e>e0m. lanes n.m Agony v.4 Hemmc m.v seems m.m HaoH .e>ecH. no «2 No a: eHwoasHoec emceeHoe oeoeu- uoIHuouHH whoahmoe oououuhQH>eH>h~oe oceanoeuoooulhoamsuo no moamuouwd seamed-fl 0N wand? 167 lesser extents, which is consistent with the expected behavior based on solvent dielectric.149 Lifetime data also correlate with the morphology of the cluster modified polymer. The M06C112 cluster contains two coordinatively unsaturated sites available to bind pendent pyridine of the PVP polymer backbone. As the incorporation of MoGCllz into the polymer proceeds, greater cross-linking of the PVP chains will occur, thereby allowing the polymer to be less susceptible to swelling by solvent. To this end, clusters incorporated in polymers at high loadings should be the least accessible to quenchers contained in solution. This appears to be the case; the 10%-bound polymer material as compared to the 90%-bound polymer possesses a significantly greater percentage of the long-lifetime component which is quenched by oxygen. The data is not explained by a simple surface loading model wherein the quenchable clusters are covalently attached to the polymer surface because similar quenching effects would be observed for polymers in solution and the solid state. Examination of the data in Table 20 shows that this is not the case. The ionically-bound methylated-PVP Mo6C1142' polymer ((PVP-Me+)2M06C1142-) exhibits different behavior than that observed for the covalently-bound materials. While the polymer suspended in H20 displays biexponential lifetime behavior parallel to that of the covalently-bound materials, the lifetimes of the compound suspended in toluene and DMF show predominantly uniexponential decays. The luminescence 168 from the polymers in toluene and DMF is quenched by oxygen. Further differences between the covalently- and electrostatically-bound polymers are observed in the solid- state luminescence lifetime properties. In contrast to PVP- M06C112, the emission decay of (PVP-Me+)2M06Cll42' is described by a uniexponential lifetime profile and the luminescence is efficiently quenched by oxygen. The lifetime decreases from 168 ps in vacuo to 93 ps when measured in the room atmosphere, and decreases further to 75 us when subject to a pure atmosphere of oxygen. That the solid polymer exhibits oxygen quenching indicates distinct separation of luminescent centers which allows oxygen to freely diffuse to the cluster core for a quenching reaction to occur. The reasons for the substantially different oxygen quenching behavior of the ionically-bound polymer as compared to the covalently-bound material may be that these polymers possess different microstructural domains. Electrochemical studies on charged Nafion150 and polystyrene-relatedlSI'152 polymers have suggested the existence of hydrophobic and hydrophilic polymeric phases between which incorporated ions may partition. Using these studies as models, the charged methylated-PVP polymer is expected to form hydrophilic and hydrophobic regions when suspended in an aqueous environment. Consequently, the observed biexponential lifetime behavior of (PVP-Me+)2- Mo6C1142' in water may arise from clusters situated in each 169 of these two distinct types of regions. The microstructure of each of these regions presently remains unresolved. That the excited-state lifetime of the solid polymer, and the polymer suspended in DMF and toluene exhibits predominantly uniexponential decay behavior, further substantiates this model. As a solid, and in. nonaqueous solvents, these hydrophobic and hydrophilic phases are not expected to form, thereby providing a unique cluster site, and hence uniexponential lifetime decay. Electrochemical experiments on these methylated-PVP cluster- containing polymers underway in these laboratories also support this proposed model. The cluster modified PVP and PVP-Me+ polymers may have applications in photo-oxidation schemes in which the cluster is quenched by 02 via energy transfer to produce the powerful oxidant 102. An attractive property of these materials as photosensitizers is the stability of the luminescent cluster cores to singlet oxygen and the polymer environment. Although quantitative studies have not been undertaken, these clusterbmodified polymers are stable for months, far-outlasting the organic substituted materials which have been studied extensively. Moreover, singlet oxygen production from solid-state materials such as the ionically-bound cluster polymer is uncommon, consequently these materials lend themselves to a variety of industrial processes which require the production of powerful oxidizers in the solid state. 170 ii. a ' ' ' -O ' s Molybdenum(II) chloride can be readily incorporated in sil icon-oxide matrices by using standard 53 The characteristic luminescence of sol-gel techniques. the M06 core is preserved in these reactive environments as depicted by the steady-state luminescence spectra of Figure 29. Comparison of the emission spectra of M06C112 in methanol and the silica gel show that the band shapes, maxima, and relative intensities are similar. This is not the case for the model cluster complex in which two silanolate (-OSiR3) groups are occupying axial positions. Although the emission energies and band shapes of this model Complex are similar to those of the unmodified cluster in methanol and silica gel, the emission intensity is S ignif icantly attenuated . These observations are illustrated by the relative intensities of the emission spectra shown in Figure 29, and quantitated by the absolute gnantum yields shown in Table 21. Interestingly, the quaI‘ltzum yield of the hexanuclear molybdenum cluster prior to gelation is comparable to the silanolate substituted cluster and not similar to M06C112 in methanol or the silica gel. Paralleling the results of steady-state luminescence e)‘1>e:riments, are the emission lifetime data displayed in Table 21. The luminescence decays of MoGC112(OSi(CH3)3)22' and MoGCllz in methanol, in solution prior to gelation, and 1“ dried gel exhibit multiexponential behavior, and 171 Figure 29. Emission spectra of: (——) M06C112 in methanol; (----) M06C112 in siliCon—oxide gel: ("’°) Mo6c112(osi(cn3)3)22' in methanol. RELATIVE INTENSITY 172 1”». I ’5 I 1‘ .. I, K t I \ I l J x \ I \ I d \ I " ‘\ °". \ " \ "' \ ,.: x f: ’ \N ..s ‘ " .... .34. \v ""~~-:‘- ‘ ’ '. aka-'3» ,fi. .... l l I 1 l a 1 M ,9 550 650 750 850 950 X/nm Figure 29 I050 Emission Lifetimes and Quantum Yields of 173 Table 21 I06 Clusters in Methanolic and Sol-Gel Environments Cluster/Environment M06C112/Methanol M06C112/precursor solution to gelation MOgCllZ/gel Mo60112(081(CH3)3)22"/ Methanol T1/118 5.3 (67%) 12 (48%) 3.6 (67%) 5.5 (38%) Tzlns 23 (33%) 18 (52%) 14 (33%) 37 (62%) 0.03 0.002 0.02 0.001 174 excellent fits can be achieved with a biexponential rate law. The luminescence decay profiles of hexanuclear molybdenum halide cluster in methanol and silica gel are similar. Namely, the decay curves primarily reflect a short lifetime component. Conversely, the decay of M06C112 in solution prior to gelation comprises a much longer component and is comparable to that for clusters substituted with trimethylsilanolate. The steady-state and time-resolved experiments show a consistent trend, that is, the quantum yields and lifetimes of methanolic solutions of M05C112 are similar to those of silanolate-substituted clusters upon addition of the tetramethylorthosilicate (TMOS) required for the gelation process, but return to values characteristic of Mb6C112 in methanol after gelation. A.reasonable explanation is that reactive intermediates, formed upon addition of the gelation reagents, substitute chlorides in the axial coordination sites to lead to the formation of clusters resembling the silanolate model compounds. During the gelation, the intermediates are converted to yield a native halide cluster in a methanolic environment. The results of the luminescent oxides in conjunction with the organic polymer data, suggest the applicability of the hexanuclear clusters for potential uses as unique luminescent materials. The main theme of these studies is that the luminescent properties of the cluster core are maintained in these various heterogeneous environments. 175 Although the preliminary studies have focussed on modifying inorganic and organic polymer host structures with the hexanuclear molybdenum(II) chloride cluster, the approach is completely general, and a variety of M06 and W6 clusters may be utilized to tune the emission energy and intensity of these polymer materials. Furthermore, since the cluster core is stable even in the silicon-oxide environment, other ceramic materials may be incorporated with M6 luminescent cores, opening the door for the design of new ceramic oxides exhibiting novel luminescent properties. CHAPTER IV THE INFLUENCE OF GUEST-HOST INTERACTIONS ON THE EXCITED- STLTE PROPERTIES OP DIOXORHENIUN(V) IONS IN INTRACRYSTALLINE ENVIRONMENTS OP CONPLEX-LAYERED OXIDES A. Introduction Host structures can significantly perturb the excited- state properties of guest transition-metal complexes by altering the nonradiative decay processes that deactivate the excited state. A thorough understanding of the interaction of guest lumophores with host structures is essential to the ultimate design and construction of practical devices utilizing luminescent transition-metal complexes. Ideally, photophysical studies will be facilitated by well-characterized host structures, and lumophores which are sensitive to the host environment. In this context, although molybdenum(II) chloride incorporated in ceramic and polyvinylpyridine environments yield interesting new materials, they are a poor choice for the study of guest-host interactions owing to the fact that (i) the molybdenum(II) and tungsten(II) clusters are insensitive to environment: and (ii) the structure of the ceramic and polyvinylpyridine matrices are poorly defined. To this end, layered silicate clays (LSCs) and layered double hydroxides (LDHs) are complex-layered oxides (CLOs) which are highly ordered and allow for the incorporation of charged species in well-defined orientations in the environment. The structure of LSCs consists of negatively charged, two-dimensional silicate layers which are separated by sheets of hydrated cations in the galleries, while LDHs are complementary structures to the LSCs in that the charge of the layers and gallery ions is reversed. Simple ion- 176 177 exchange procedures permit a variety of cations and anions of virtually any size to be accommodated in the galleries of LSCs and LDHs, respectively; To date, inorganic photochemical studies have primarily centered on LSC and LDH intercalates of metal polypyridyl complexes.‘u'43'449'45 Principal themes that have emerged from consideration of the luminescence properties of these intercalates is that excited-state properties are generally preserved upon intercalation and the photoactive ions are accessible for excited-state electron- and energy-transfer reactions. Specific effects of LSC and LDH host structures on the dynamics of excited-state processes, however, have been experimentally difficult to assess owing to the relative insensitivity of the excited-state properties of metal polypyridyl complexes to environmental effects. Consequently, the extent to which excited-state properties such as lifetime, energy, and geometry are perturbed by guest-host interactions in LSC and LDH intercalates has, for the most part, remained undefined. In an effort to better understand guest-host interactions of photoactive CLOs, the spectroscopy and photophysics of high-valent transition metal-oxo compounds in LSC _and LDH galleries were explored. The results of electronic absorption and emission studies have provided a detailed account of the electronic structure of metal-oxo 84,153-161 compounds and. suggest. that this class of photoreagents, and particularly, d2 trans-dioxo species,161 178 are well-suited to probe the effect of LSC and LDH intracrystalline environments on excited-state properties. In particular, recent investigations of trans- dioxorhenium(V) complexes have demonstrated that the lowest energy ligand field transitions, which involve the promotion of an electron from. the bzg(xy) orbital to the doubly degenerate Re-O x-antibonding eg(xz,yz) level, produce 1Eg[(bzg)1(eg)1] and 3Eg[(bzg)1(eg)l] states, the latter of 84,161 which is highly emissive. The excited-state lifetime and energy of the 3Eg state is quite sensitive to environment, and protic solvents efficiently quench the molecular luminescence. Along these lines, it is reasonable to expect that the extent of complex formation between the photoactive gallery ion and the interlayer water molecules of LSCs and LDHs, and hence the luminescence properties of the intercalates, will be extremely sensitive to the orientation and specific interaction of the LIQDS-R802+ cores within the intracrystalline environment. Moreover, the ability to vary the charge of the dioxorhenium(V) complexes with the ancillary ligands in the equatorial coordination sites while maintaining the luminescent properties of the metal-oxo core allows for a comparative study of the effect of the anionic LSC and cationic LDH environments on the excited-state dynamics of structurally and electronically similar gallery ions to be undertaken. Reported in this Chapter are results of spectroscopic and photophysical investigations of the LSC and LDH intercalates 179 containing the trans-Re02(py)4+ (py = pyridine) and trans- Re02(CN)43' complexes, respectively, as well as the discovery that guest-host interactions can engender luminescence from dioxorhenium(V) species in highly protic environments. B. Results 1. Synthesis and Characterizatign The structural formulas of the CLO host structures used in our studies are presented in Table 22. The LSCs are 2:1 phyllosilicates whose structures consist of elementary Vlayers composed of an octahedral sheet of oxygen and hydroxyl ions coordinating Mg2+ or Li+ ions sandwiched 4+ 162 A between two tetrahedral sheets coordinating Si ions. net negative charge on these elementary layers, resulting from the isomorphic substitution of Li+ for Mg2+ ions in the octahedral sites is balanced by a sheet of hydrated sodium or lithium ions. As indicated by the formulas listed in Table 22, fluorohectorite is distinguished from hectorite simply by increased Li+ substitution in the octahedral layer. The increased layer charge of fluorohectorite is manifested in a concomitant increase in the interlayer cation concentration. In contrast to the properties of LSCs, the reversed layer and gallery charges of the LDHs result from the replacement of divalent cations by trivalent cations in brucite-like (i.e., Mg(OH)2-like) octahedral sheets. Table 22 lists the chemical composition of a 180 Table 22 Idealized Structural Formulas of Complex-Layered Oxides Complex- Layered Oxide Formula hectorite Nao 67[Mg5 33110.67](818.00)020(OH,F)4a fluorohectorite Lil.6[M94.4Lil.6](Si8.oo)°20F4 hydrotalcite [Mgo.75Alo.25(OH)2]Clo.25-H20b a In this CLO some framework hydroxyls are replaced by fluoride. b Synthetic hydrotalcites are characterized by the generalized formula [MIII_XMIIIX(OH)2]An'x{nfyH20, where x = 0.20-0.33, An" is the gallery anion and MI MIII and tripositive ions. In the mineral hydrotalcite M11 = Mg2+, MIII= Al3+, and An' = C032“. and are di- 181 synthetic hydrotalcite, which is characterized by substitution of Al3+ for Mg2+ in the octahedral sites.163' 165 The LDH host structure is typically prepared by precipitating' the double Ihydroxide from. an aqueous NaOH solution containing the aluminum and magnesium chloride salts. The ion-exchange properties of LSCs and LDHs suggested that the appropriately charged dioxorhenium(V) ions would be incorporated readily into the galleries of the host structures. Addition of aqueous suspensions of Na-exchanged hectorite and fluorohectorite LSCs and Cl-exchanged hydrotalcite LDH to aqueous solutions of Re02(py)4+ and ReOz(CN)43', respectively, yielded yellow solids. All spectroscopic and photophysical studies employed samples with loadings of the metal-oxo complex corresponding to 15% of the ion-exchange capacity of the CLO; chemical compositions of the exchanged CLOs were confirmed by spectrophotometric analysis. The IR spectra of these solids, reproduced in Figure 30, indicate that the dioxorhenium(V) complexes are immobilized on the CLO supports. The ReOz(py)4-LSC solids exhibit an IR spectrum consisting of the superposition of the metal-oxo cation spectrum on that of the LSC support. Bands attributable to ring stretching vibrations of the pyridine at 1490 and 1502 cm"1 are clearly apparent in the KBr pellet spectra of the Re02(py)4-LSC solids. Similarly, the C-N stretching vibrations of ReOz(CN)43' adsorbed to LDH are not shifted V 182 Figure 30. Infrared spectra on KBr pellets of: (a) (----) [Reoz(py)41I: (——) ReOZ(py)4-fluorohectorite; ( (—) ReO2 (CN) 4-hydrotalcite . ReOZ (py) 4-hectorite; (. . - .) b) (----) K3[ReOZ(CN)4]; 183 2000 50C) T H300 ISOC) (.l u u ) \\\ . I . Q 0 b .s e! ( '-" ( ||||| I. s . . II [I D ’ «It! I lllllll IUIII III "‘ |||||||| w " I'l- ”' """""""""""""" "' ‘O"l""'l|"fld ......... \n. ’I ....... """"""" . n ...... Chl|l \\\ I ...... \ a \ f e I .II 1 e l ‘lllf . ‘| \I\ . . II _ \\I- ‘ ell . "-Ill" - I. — z . .I _ I _ I llllllllllll . JPIPFINIHI lllll 'lll. ......... g 2 ........... . p l . l I! h. 2 \IIIII ’ I II s\ \\l a \ I s a . 2 ~ 2 . 2 . .. _ a . a . e _ 2 H I T l I; ’l'il. I IIII IIHII.IIlh.IJI \ _ ~ . a ’l b l l500 IOOO 500 L 2000 MUZHHHEEmZguH fi/cm'l Figure 30 184 significantly relative to that observed for KBr pellet spectra of the native anion. Although the infrared active O-Re-O vibrations are obscured by absorptions of the metal- oxo lattice vibrations of the aluminosilicate and double hydroxide host structures, Raman spectra obtained by using excitation frequencies coincident with the 1Eg « 1A19 absorption leads to a significant enhancement of the Raman peaks associated with metal-oxygen vibrations. A prominent band corresponding to the totally symmetric O-Re-O stretching vibration is observed at 916 and 919 cm”1 for the ReOz(py)4-hectorite and Re02(py)4-fluorohectorite adsorbates, respectively. The energy of the metal-oxygen vibrations is only marginally shifted upon CLO immobilization (”alg (Re-O) = 916 cm'1 for Re02(py)4+ in H20 at 25°C) and this result in conjunction with those from infrared spectroscopy reveal minor structural distortions, if any, of the bound complexes. That the dioxorhenium(V) ions indeed occupy the LSC and LDH galleries is confirmed by the x-ray patterns of the reacted solids (see Figure 31). Basal spacings of the three layered products clearly reveal an expansion of the gallery upon. adsorption of the dioxorhenium(V) complex. If ‘we account for the van der Waals thickness of the LSC and LDH layers (dlayer(hectorite and fluorohectorite) = 9.5 A: dlayer(hydrotalcite) = 4.8 A), the 002 reflections for Re02(py)4-hectorite, ReOZ(py)4-fluorohectorite and ReOz(CN)4-hydrotalcite correspond to gallery heights of 6.7, 185 Figure 31. X-ray patterns of the three CLO intercalates: (a) Re02(py)4-hectorite; (b) Re02(py)4- fluoro—hectorite; (c) ReOZ(CN)4-hydrotalcite. INTENSITY 186 4.0 8.0 I2.0 I6.0 20.0 I I 1 T t l t I (0) l l 1 l 1 l l L (b) I l L l l I 4 l (c) 7 1 l 1 l 1 l 1 E 4.0 8.0 I2.0 |6.0 20.0 29/ ° Figure 31 187 9.0, and 4.3 A, respectively. It is clearly apparent from these results that the dioxorhenium(V) ions do not reside in the CLO galleries with similar molecular orientations (yige ipfipg). As typically observed in CLO intercalation chemistry, the relatively low order of 001 reflections for the three solids is indicative of interstratified intercalation of the trans-dioxorhenium(V) complexes. 2. Electronic Absopption and Emission Spectroscopy The room temperature electronic absorption spectra of ReOz(py)4-hectorite and ReOZ(py)4-fluorohectorite are reproduced in Figure 32a. The absorption profiles of these ReOZ (py)4+ intercalates are similar at wavelengths longer than 300 nm being characterized by a single intense absorption band between 350-400 nm and a less intense shoulder on the low energy side. The shoulder is energetically coincident with the 1Eg « 1Alg transition of Re02(py)4+ at 420 nm. Conversely, the 320—nm band shifts to lower energy upon intercalation of the complex. Disparate absorption cross-sections and slightly different energies of this band for the hectorite (A == 377 nm; e = 29,700 max M'lcm‘l) and fluorohectorite intercalates (Amax = 347 nm; e = 13,100 M'lcm-l) suggest some type of perturbation of the electronic structure of the mpg-Reoz+ core by the negatively charged silicate layers. Spectroscopic evidence implies that the 320-nm band of the native complex arises from either an alg(zz) « bzg(xy) transition or oxygen LMCT. 188 Figure 32. Electronic absorption and emission spectra of nonaqueous solutions of the‘ following: (a) (————) [Re02(py)4]I in pyridine; (----) Re02(py)4-hectorite; (....) ReOZ(py)4-fluorohectorite; (b) (————) K3[ReOZ(CN)4] in DMF: (----) ReOz(CN)4-hydrota1cite. 189 v/pnf m_<__mm_OZ HZHmZm3.< ‘_---3----- 500 600 700 800 X/nm 400 300 7.5.2 no: . _IEU-I2 MO- \W Figure 32 190 The data are consistent with the latter assignment because the LMCT should energetically be stabilized by the proximity of the oxygens of the trans-ReOz+ core to the negative silicate layers. The physical significance of the intensity variation of this transition presently remains unresolved.166 The LSC intercalates luminesce at room temperature with blue and near UV-excitation. The red emission observed from these solids is characteristic of ReOz(py)4+. The uncorrected emission spectra for Re02(py) 4+ in deaerated pyridine and aqueous suspensions of the hectorite and fluorohectorite intercalates are reproduced in Figure 32a. Emission spectra of the intercalated solids are independent of the excitation wavelength over the range of 313-436 nm and are insensitive to the loading of the oxocation. The emission maximum of the fluorohectorite intercalate is 685 nm and the emission band is broad and featureless. In contrast, luminescence from ReOZ(py)4-hectorite is shifted to the blue (*em,max = 630 nm) and vibrational fine structure can be resolved. As observed from Figure 32a, the luminescence properties of the hectorite intercalate as opposed to the fluorohectorite intercalate, more closely resemble that of the native ReOz(py)4+ ion. Relative quantum yield measurements of aqueous suspensions of the LSC intercalates reveal that the luminescence intensity of Re02(py)4+ in fluorohectorite is fifty times less than that in hectorite. 191 Comparisons between the emission of Re02(py)4+ and Re02(py)4-LSC intercalates are more readily accomplished at low temperature. Electronic emission spectra of ReOz(py)4I, ReOz (py)4-hectorite, and ReOZ(py)4-fluorohectorite at 9 K are shown in Figure 33. While the emission band of the fluorohectorite intercalate remains featureless, the emission band of the hectorite intercalate consists of a distinct progression of 900 cm'1 subdivided by a less pronounced progression of 200 cm'l. Similar progressions in 900- and 210-cm"1 modes are observed in the luminescence spectrum of crystalline ReOz(py)4I at 9 K and isotopic labelling studies have shown that these modes correspond to the symmetric rhenium-oxygen and rhenium-pyridine stretching vibrations, respectively.84 Electronic absorption and emission spectra of the ReOz(CN)4-LDH (Figure 32b) are much less informative than those of the ReOz(py)4-LSC intercalate compounds. The pale yellow color of the LDH intercalate arises from the absorption ‘tail between 300 and 350 nm, IHowever, the maximum of the ultraviolet peak responsible for this absorbance cannot be recorded due to significant scattering of light at wavelengths shorter than 300 nm. Additionally, the low loadings of the LDH galleries 'with ReOz(CN)43' precluded observation of the weak 1Eg « lAlg transition. In contrast to the ReOz(py) 4-LSC compounds, no luminescence (*exc = 365, 405 and 436 nm) is detected from Re02(CN)4-LDH over the temperature range of 9 to 300 K. 192 Figure 33. Low temperature (9 K) emission spectra of solid: (a) [Re02(py)4]I; (b) Re02(py)4-hectorite; (c) ReOz(py)4-fluorohectorite. EMISSION INTENSITY 193 17/ um'l |.7 l.6 I.5 L4 L3 (0) (b) (c) 550 600 650 700 750 800 A/nm Figure 33 194 Further insight into the excited-state properties of dioxorhenium(V) ions in CLO intracrystalline environments is provided by time-resolved luminescence measurements. Emission lifetimes of solids and suspensions of hectorite and fluorohectorite are listed in Table 23: for purposes of comparison, the lifetimes of ReOz(py)4+ as the iodide salt and in DMF solution are also included. For the latter two systems, the measured decay rates are in excellent agreement with previously reported lifetimes of solids and solutions of the ReOZ(py)4+ ion.84 The luminescence decay of Re02(py)4+ in hectorite exhibits multiexponential behavior and can be fit reasonably well with a biexponential rate law with a = 0.37, r1 3.9 ps, b = 0.63, and 72 13.0 us where a and b are the fraction of molecules with luminescence decays r1 nd 12, respectively. A typical lifetime decay curve and the theoretical biexponential fit are shown in Figure 34. The lifetime of the major component of the luminescence decay is comparable to that for the ion in homogeneous solution and remains relatively constant when ReOz (py)4-hectorite is suspended in H20 and D20 solutions. The short lifetime component of the decay curve, however, is quite sensitive to the nature of the solvent as evidenced by reduction of the lifetime by a factor of 2 when the intercalate is suspended in H20. Even more significant is the fact that this short lifetime component. exhibits a distinct increase when Re02(py)4-hectorite is suspended in D20. Interestingly, ReOz(py)4+ residing in fluorohectorite nnnnnnnn we: hsu ter 33 ;ree:e: *1". u :eb“’ ,I0 0 7' 24» J A'; 5 5.1:" 1.. A‘ i 195 Table 23 Solid, Solution, and Intercalate Luminescence Lifetimes of the ReOz(py)4+ Iona Medium. a rllpS b Tz/pS iodide saltb 1.00 33.0 - - dimethylformamideb 1.00 9.6 - - hectorite 0.37 3.9 0.63 13.0 hectorite/H20c 0.41 1.9 0.59 11.0 hectorite/020d 0.43 3.1 0.57 12.0 fluorohectoriteb > 0.97 0.63 - - fluorohectorite/Hzob'c > 0.99 0.35 - - fluorohectorite/D20d 0.86 1.2 0.14 2.8 aLuminescence decay curves were fit to the multiexponential equation y = ae't/'1 + bet/'2 where a and b represent the fractions of total emission decay described by the short lifetime component r1 and long-lifetime component r2, respectively. bSystem exhibits uniexponential emission decay kinetics. cOne wt% H20 suspension of a 15%-exchanged LSC. dOne wt% DZO suspension of a 15%-exchanged LSC. 196 Figure 34. Lifetime decay curve of solid Re02(py)4- hectorite. The smooth curve repreSents the best fit to the biexponential rate equation. 197 OfiN vn eufluwh m1\k OQO OdN u 00. Q0. q _ IOQm IOOO. 1 COO. IOOdN 1OQmN 000m 1 ALISNEINI NOISSIWH 198 galleries exhibits a short, predominantly single-exponential decay, and the trend of the excited-state lifetime on medium (e.g., solid, H20, D20) is parallel to that observed for the short lifetime component of the hectorite intercalate. Although. the jperturbation. of 'the luminescence from Re02(py)4+ ion by H20 and 020 in CLO environments is expected in view of the extreme sensitivity of dioxorhenium(V) excited states towards proton donors,84 the relative insensitivity of a significant fraction of ReOz(py)4+ ions residing in the hectorite galleries to interstitial water is a surprising and unforseen result. In order to quantitate this result, the H20 quenching of the luminescence from DMF suspensions of Re02(py)4- hectorite by the Stern-Volmer method was investigated. A plot of the ratio of the long lifetime component of the Re02(py)4-hectorite luminescence decay in the absence of H20 to that in the presence of H20 is shown in Figure 35; a Stern-Volmer plot obtained for ReOz(py)4+ in DMF with H20 as the. quencher' is also illustrated" Consistent. with the Stern-Volmer equation, rb/r varies linearly with the concentration of H20 for the two systems. The Stern-Volmer constant for the luminescence quenching of Re02(py)4+ in homogeneous solution is two orders of magnitude greater than that of the ion incorporated in hectorite galleries.167 199 Figure 35. Stern-Volmer plot of the long lifetime component of Re02(py)4-hectorite in DMF (I) with H20 as the quencher and ReOz(py)4+ in DMF (o) with H20 as 'the quencher. 200 v._ N._ 0.. an uneven _2\ 8.; 0.0 0.0 A q 0.0 0.0 201 C. Discussion The excited-state properties of dioxorhenium(V) ions exhibit a pronounced dependence on the CLO interlayer environments. Time-resolved emission and steady-state luminescence experiments show that the behavior of intercalated dioxorhenium(V) ions can be classified into three distinct categories: (i) the luminescent hectorite intercalate is largely unperturbed and excited—state decay channels are similar to those of the dioxorhenium(V) ion in aprotic solution; (ii) emission from the ion in the fluorohectorite intercalate is significantly attenuated in intensity and a corresponding decrease in the emission lifetime is observed; and (iii) no luminescence is detected from the dioxorhenium(V) core intercalated in LDH. These discrepancies in the luminescence properties of the structurally and electronically related rhenium-oxo complexes in CLO environments are significant and suggest unique Iand specific lifetime limiting* processes for ‘the three CLO systems. Luminescence quenching of photoactive ions in CLO interlayer environments has previously been attributed to either the presence of impurity ions isomorphically substituted into the octahedral sites in the layers,168'169 or efficient excited—state self-quenching processes promoted by the high local concentration of ions within the 43 interlayer galleries. However, several pieces of evidence suggest that neither of these two quenching mechanisms are 202 operative in the dioxorhenium(V) CLO intercalate systems. First, quenching by impurity ions is precluded by the virtual absence of transition metal ions (e.g., Fe3+, Cr3+) in the octahedral sites of the LSC and LDH host structures. Indeed, the synthetic hydrotalcite intercalate from which no emission is detected, has only Al3+ and Mg2+ ions composing the sheet structure: and for the hectorite and fluorohectorite intercalate compounds, the concentration of cations other than Mg2+ or Li+ is less than 0.1%. Second, luminescence decay kinetics of all the intercalates investigated do not exhibit a dependence on the presence of the non-emissive trans-Re02(en)2+ ions in the gallery.170 The data presented in Table 24 for the hectorite intercalate are exemplary. If the intercalate was exhibiting self- quenching, then the emission decay time should become longer as the concentration of the non-emissive co-intercalate is increased owing to an increase in the center-to-center distance between lumophores in the gallery. Of course, this conclusion is predicated on the tacit assumption that the Re02(py)4+ and ReOz(en)2+ ions do not segregate in the CIO galleries. Considering the similar charges, sizes, and structures of these two oxocations, the assumption is a reasonable one. Finally, the short lifetime component of the hectorite intercalate and the fluorohectorite intercalate exhibits a pronounced deuterium isotope effect which parallels that observed for the ion in nonaqueous solution. This result clearly demonstrates that the 203 Table 24 Emission Lifetime of Re02(py)‘-Hectorite Containing Co-Intercalated Re02(en)2+ Ions~ Re02(py)4+: 100 : 0 0.41 1.9 0.59 11.0 50 : 50 0.44 2.6 0.56 11.0 10 : 90 0.45 2.4 0.55 10.0 a Hectorite intercalate with 15% of the exchange capacity replaced by Re02(py)4+ and Re02(en)2+ ions in the given molar ratios. b Defined in the footnote of Table 23. 204 quenching decay channels of the intercalated ions are intimately related to the presence of proton donors in the interlayer region and suggests that the unique emission behavior of the three CLO systems is engendered by different guest-host interactions. The d-spacing of the three intercalate compounds offers additional evidence for the dioxorhenium(V) complexes occupying unique sites in the CLO interlayer environments. We first consider the ReOZ(py)4-hectorite compound. A d- spacing of 6.7 A is consistent with the guest complex assuming an orientation with the O-Re-O axis perpendicular to the layers of the host structure.171 Figure 36 depicts this molecular axis (C4) along with the C2" axis. Interestingly, an average density of 80 A2 per unit negative charge on the aluminosilicate layer is calculated from the unit cell dimensions of hectorite. When one considers that the cross-sectional area of the univalent oxocation situated along its C4 axis is 89 A”, the ReOZ(py)4+ ion resides in the hectorite gallery with the orientation which most effectively counterbalances the layer charge. A parallel charge effect governs the intercalative reaction of fluorohectorite but a higher charge density (27 Az/unit charge) for this CLO precludes intercalation of the oxocation along its C4 axis. Alternatively, positioned on the C2" axis (bisecting the pyridine-rhenium-pyridine equatorial axes), the oxocation’s cross-sectional area of 30 A2 is compatible with the layer charge density of 205 Figure 36. Model depicting the C4 and C2" molecular axes for Re02(py)4+. 206 o py,“ | | py ., Re . / ll \ W W 0 Figure 36 207 fluorohectorite. The molecular dimension of 9.43 A for the Re02(py)4+ ion along its C2" axis accounts nicely for the observed d-spacing of the fluorohectorite intercalate. In contrast to the LSC intercalates, the observed d-spacing of the LDH intercalate is in accordance with the effective C3 axis of the pseudo-octahedral ReOz(CN)43' complex aligned normal to the host layers. Previous studies of hydrotalcites have demonstrated that the preferred orientation of intercalated anions either: (i) maximizes the hydrogen bonding interactions of the protons of the hydroxide layers with the guest species, and/or (ii) minimizes the charge separation distance between the positive layers and gallery anions.172 Both of these criteria are fulfilled by the observed orientation of the ReOz(CN)43' anion. In view’ of these X-ray diffraction results in conjunction with the different luminescence properties. of the three CLO intercalates, an intriguing issue is whether the unique orientations of the dioxorhenium(V) ions in the galleries can give rise to the disparate photophysical properties. This question can be addressed by considering the structures of the CLOs. The ‘ceiling' and ‘floor’ of the LSC galleries are composed of basal planes of SiO4 tetrahedra. An idealized geometry for the oxygen framework of LSC galleries is illustrated in Figure 37. The basal oxygens comprising SiO4 tetrahedra are linked. with neighboring tetrahedra to form hexagonal cavities of 208 Figure 37 . Proposed orientations of the gag-dioxo- rhenium(V) core in (a) hectorite and (b) fluorohectorite. The circles represent the idealized geometry for the oxygen framework of the LSC galleries. The oxygens comprising a hexagonal cavity of the CLO floor and ceiling are indicated by black circles. 210 appreciable dimension (diametrically opposed vertices of the hexagon are separated by 5.28 A). The model proposes that Re02(py)4+ in hectorite, lying on its C4 axis, traverses the gallery with oxygens ‘keyed’ into the hexagonal cavities of the CLO layers as shown in Figure 37a. This model accounts for the photophysical properties of the hectorite intercalate. In a keyed configuration, access of the hydroxyl protons of interlayer water molecules to the oxygens of the Eggps-ReOZT core is inhibited and quenching of Re02(py)4+ luminescence will be precluded. In accordance with the results for the long lifetime component of the emission decay of the hectorite intercalate, guest ions keyed into CLO layers are not expected to exhibit a deuterium isotope effect and will retain the long-lived, highly-emissive excited-state properties characteristic of ReOZ(py)4+ ions in the solid state and nonaqueous solution. Additionally, if the proposed model is correct, then ReOz(py)4+ ions which are not keyed into CLO layers should efficiently be quenched by interlayer H20. This is observed for the fluorohectorite intercalate. Indeed, the original. motivation for investigating fluorohectorite was that this CLO provided an easy route to rotating the oxygens of the trans-Re02+ core away from the layers without altering the structural integrity of the LSC gallery. By assuming an orientation 'with the O-Re-O axis parallel to the fluorohectorite layers, as depicted in Figure 37b, the oxo core of the Re02(py)4+ ion is clearly accessible to water in 211 the interlayer environment and hence, the emission lifetime and. intensity’ are significantly attenuated” 'The single exponential decay kinetics imply a predominantly uniform orientation of the ReOz(py)4+ ions in the fluorohectorite gallery. Along these lines the observation that the luminescence of a fraction of Re02(py)4+ ions incorporated in hectorite is quenched, giving rise to the short lifetime component of the emission decay profile, is consistent with incomplete keying of all the Re02(py)4+ ions in the interlayer environment. The quenchable ions may be aligned along their C4 axis, but with only one oxygen keyed into the CLO interlayer or they may be in an orientation similar to that observed in the fluorohectorite intercalate. Although these two limiting orientations presently cannot unequivocally be distinguished, one is inclined to favor the former because ions with similar orientations in hectorite and fluorohectorite galleries would be expected to exhibit similar decay kinetics. Comparison of the data for the short lifetime component of hectorite and that of fluorohectorite shows that this is not the case. The inability to detect luminescence from ReOz(CN)4-LDH can now easily be understood within the context of the LSC results. As illustrated in Figure 38, the LDH gallery is bound by a cubic closed packed hydroxide sheet, and there are no cavities in which the rhenium oxo core can key. Conversely, the oxygens of the trans-Re02+ core are probably 212 Figure 38. Proposed orientation of the trans-dioxo- rhenium(V) core in an idealized LDH gallery. The hydroxyl oxygens composing the LDH are represented by large circles and the metal ions by the smaller circles. 214 hydrogen bonded directly to the hydroxide layer. Because proton donating solvents efficiently quench dioxorhenium(V) excited states by presumed hydrogen bonding interactions,84 the nonradiative decay rates of electronically excited ReOz(CN)43‘ ions in LDH are expected to be exceedingly fast. Thus, specific guest-host interactions in CLO intercalates can significantly' mediate the excited-state properties of transition-metal complexes. By controlling subtle charge balancing effects between host and guest structures, highly-emissive long-lived excited states can be preserved in reactive environments. The observations are completely general and can be extended to include many other metal-oxo compounds in CLO environments. The ability to use the CLO as a template to specifically align immobilized photoactive reagents in a gallery which can be accessed by a variety of reactants provides the opportunity to explore, and perhaps exploit, the effects of orientation on excited- state chemical reaction pathways. CHAPTER V FINAL REMARKS Photophysical and photochemical studies provide the framework in which to develop excited-state chemistry. The hexanuclear cluster systems of M06 and W6 possess properties that are promising for their use in photoactive systems. The ability to introduce a variety of pendant groups onto the metal core while maintaining the excited-state properties permits the design of interesting surface and intralayer cluster-modified materials for diverse applications. For instance, the. polymer' materials incorporated with the cluster core may have applications in photo-oxidation chemistry. These materials were shown to produce singlet oxygen via the electronic excited state, but further studies quantifying the efficiency of singlet oxygen production remain to be performed before solid-state photocatalysts will be realized. The stability of the luminescent cluster core in reactive environments has more far-reaching consequences. Unpublished results have shown that the clusters’ characteristic red and near-infrared luminescence is generated not only upon excitation with blue and UV light, but upon excitation with X-rays as well. Potentially, materials such as thin films and coatings derived from cluster-incorporated silicon-oxides may be synthesized which could be utilized as X-ray sensors. Moreover, the intense emissive properties of the M06 coordinatively unsaturated polynuclear cores offer exciting possibilities for the development of new multielectron catalysts.173'175 The design of useful multielectron 215 216 schemes will ultimately require that a photoreagent not only exhibit multielectron reactivity, but that the complex be returned to its original parentage. The design of solid- state photochemical systems consisting of macromolecular assemblies containing' multiple redox. centers of complementary function is one approach to the realization of multielectron photocatalytic systems. In these systems, the oxidation-reduction multielectron photochemistry of an intercalated metal complex functions in concert with redox sites intrinsic to the intracrystalline environment. The first step in the design of these supermolecular assemblies is an understanding of the excited-state properties of the intercalated catalyst provided by studies such as the type described in Chapter IV. However, the use of complex- layered oxides as intracrystalline environments is hindered by two important factors: (1) it is difficult to insert redox active metal ions into CLO layers (which is ultimately required for a functional multielectron photocatalytic system), and (ii) the basic properties of the CLO are not compatible with the acidic properties of many multielectron photoreagents. 'Therefore, alternative inorganic layered hosts such as the Zr(HPO4)2 and Ti(HPO4)2 layered 34b'176 which are stable in acidic solutions and materials, provide apical oxygens from individual phosphate groups for coordinating to photoactive metal cores, offer interesting opportunities to develop solid-state photocatalysts. Newly 217 synthesized materials such as these will open new avenues in the area of solid-state photochemistry. The goal of the research described herein was to provide background information for the ultimate design of improved photoact ivated systems . Obviously , the dioxorhenium(V) and M6 clusters described in this dissertation represent only two classes of luminescent compounds of several hundred systems which could be utilized for the design of novel photochemical materials. For example, the preliminary sol-gel research described in Chapter III.D. has led to the incorporation of lanthanide complexes in oxide matrices possessing photophysical properties more suitable for the design of light-emitting systems with energy transduction applications. Research at frontiers of the chemistry presented in this thesis will continue to provide the necessary foundation for the successful utilization of luminescent compounds in solving problems on new horizons of excited-state chemistry. APPENDIX The EPR spectra of MoGClM' and M06C1143" are reproduced in Figure 39. These spectra were recorded at ~4K with a Bruker ER200 EPR spectrometer equipped with an Oxford Instruments ESR9 flow cryostat and the samples were prepared in dichloromethane from Mo6C1142 by using bulk electrolysis. The peaks at 3354 G and 3344 G in Figures 39a and 39b, respectively, indicated by the arrows are due to an impurity in the EPR sample tube. As described in Chapter III.A., the EPR spectrum of M06C114' (Figure 39a) is expected to exhibit an axial doublet; this has previously been observed.66d'69 Conversely , the Mo6Cl 14 3 '- ion is generated by electrochemically placing an electron in the azg orbital. Figure 39b shows the EPR spectrum which is consistent with an isotropic zAzg ground state configuration. Further experimental verification of the proposed electronic structure was attempted by measuring the EPR spectrum of electronically excited Mo6C1142-. Samples at 4 K were irradiated with a home-built flash lamp system equipped with a 1000 W Xe lamp and EPR spectra were recorded during irradiation. With the proposed model, the excited state is expected to arise from a 3E9 parent state with unpaired electrons in the 329 LUMO and e HOMO. Several attempts g were made to observe a triplet EPR signal on 4 K samples. Unfortunately, no signals were observed. It is assumed that the concentration of excited molecules was too low for the detection of an EPR signal from these samples. 218 219 Figure 39. EPR spectra of electrochemically generated (a) M06C114- and (b) Mo6Cll43' in frozen dichloromethane solution at 4 K. 220 h (0) 4.. ‘— (b) 2000 REFERENCES 100 (a) Newton, M.D.; Sutin, N. Anne, Rev. Phys. Chem, 1984, 3;, 437-480. (b) Sutin, N. P o . o . 1983, §_Q, 441-498. (c) Sutin, N. In "Tunneling in Biological Systems"; Chance, 3., DeVault, D.C., Frauenfelder, H., Marcus, R.A., Schrieffer, J.B., Sutin, N., Eds.; Academic Press: New York, 1979; pp 201-227. Kavarnos, G.J.: Turro, N.J. Chem, Bey, 1986, fig, 401- 449. Guarr, T.; McLendon, G. Qoopg, Chem, Rev, 1985, fig, 1- 52. Meyer, T.J. Prog. Inorg, Chem, 1983, eg, 389-440. Whitten, D.G. Ace, Chem. Bee, 1980, 1;, 83-90. (a) Kalyanasundaram, K.; Gratzel, M.; Pelizzetti, E. 99913, Qpem. Rev. 1986, _6_2, 57-125. (b) Gr'atzel, M. (Ed.) "Energy Resources Through Photochemistry and Catalysis"; Academic Press: New York, 1983. Fendler, J.H. J. Phys, Chem, 1985 fig, 2730-2740. (a) Gray, H.B.; Maverick, A.W. Seiepce (Weehipgpen, DLQLL 1981, 211, 1201-1205. (b) Maverick, A.W.; Gray, H.B. W 1980, 53,, 2339-2348. Balzani, V.: Carassiti, V. "Photochemistry of Coordination Compounds"; Academic Press: New York, 1970. Adamson, A.W.: Fleischauer, P.D. (Eds.) "Concepts of Inorganic Photochemistry": J. Wiley: New York, 1975. 221 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 222 The state of the art of inorganic photochemistry has recently been summarized in J. Chem. Ed. 1983, 59 (10). (a) Schaper, H.; Schnedler, E. J. Phys, Chem, 1982, £57 4380-4385. (b) Schaper, H.; Koestlin, H.; Schnedler, E. Philips Tech. Rev. 1982, 59, 69-80. (c) Laser, 0.: Bard, A.J. J. Electrochem. Soe, 1975, 12.2.. 632-640. (d) Brilmyer, G.H.; Bard, A.J. W 1980, 1gl, 104-110. (a) Heller, C.A.; Jernigan, J.L. Appl. Opp. 1977, 1e, 61-66. (b) Measures, R.M. Appl. Opt. 1975, 1.5, 909- 916. (c) Measures, R.M. Appl. Opt. 1974, Q, 1121- 1133. Sheppard, L.M. Adv. Map, Free. Metal Prog, 1987, g, 73. Jain, R.K.: Lind, R.C. J. Opp, §oc. Am, 1983, 1;, 647- 653. Thomas, J.K. J. Phys, Chem, 1987, e1, 267-276. Oelkrug, D.; Flemming, W.; Fullemann, R.; Gunther, R.: Honnen, W.; Krabichler, G.; Schafer, M.; Uhl, S. £_I._1_r_e Appl, Chem. 1986, §§, 1207-1218. Endicott, J.F.: Ramasami, T.; Tamilarasan, R.; Lessard, R.B.: Ryu, C.K. Coord, Chem. hev, 1987, 11, 1-87. (a) Hopkins, M.D.; Miskowski, V.M.; Gray, H.B. J. . Chem. Soc. 1988, 1 0, 1787-1793. (b) Manning, M.c.: Trogler, W.C. J, Am, Chem, Sec, 1983, 105, 5311-5320. (c) Togler, W.C.; Gray, H.B. Acc. Chem Res 1978, ll, 232-239. Creutz, C. Prog. Ihorg. Chem, 1983, 29. 1-73. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 223 (a) Vogler, A.; Osman, A.H.: Kunkely, H. geepgy_§hemy Beym 1985, 54. 159-173. (b) Schanze, K.S.: Meyer, T.J. W 1985, 2A, 2121-2123. (C) Curtis, J.C.; Bernstein, J.S.; Meyer, T.J. w 1985, A. 385-397. Carstens, D.H.W.; Crosby, G.A. J. Mel, Spee, 1970, 34. 113-135. Paris, J.P.: Brandt, W.W. J, Am. Qhem. goe, 1959, g1, 5001-5002. Balzani, V.; Sabbatini, N.: Scandola, F. 9hem,_3ey,_ 1986, ee, 319-337. Endicott, J.F. J, Chem, Ed, 1983, he, 824-829. Sutin, N.; Creutz, C. J, Chem, Be. 1983, fig, 809-814. Scandola, F.; Balzani, V. g, ghem, Ed: 1983, fig, 814- 823. Turro, N.J. Pure Appl. Chem. 1977, 52, 405-429. F6rster, T. Qiec, Eepeday Soe, 1959, 21, 7-17. Zemel, H.; Hoffman, B.M. J. Am, Chem, See, 1981, ;g;, 1192-1201. Dexter, D.L. Q, ghem. Phys. 1953, 21, 836-850. Turro, N.J. "Modern Molecular Photochemistry": Benjamin/Cummings: Menlo Park, CA, 1978; Chapter 9. (a) Marcus, R.A. Anne, Rev, Ehye. Qhem. 1964, 1;, 155- 196. (b) Marcus, R.A. J, Qhem, Ehye, 1956, L4,, 966- 978. (a) Pinnavaia, T.J. Seiehce (Weshihgpoh, Q,C,) 1983, gee, 365-371. (b) Thomas, J.M. In "Intercalation 35. 36. 37. 38. 224 Chemistry"; Whittingham, M.S., Jacobson, A.J., Eds.; Academic Press: New York, 1982; pp 55-99. (c) Laszlo, D. W 1987, 215, 1473-1477. (d) Ballantine, J.A. In "Chemical Reactions in Organic and Inorganic Constrained Systems"; Setton, R. , Ed.; Reidel: Boston, 1986: pp 197-212. (e) Fripiat, J.J. gleye Qley Miher. 1986, 15, 501-506. Suib, S.L.: Tanguay, J.F.; Occelli, M.L. M 890. 1986, 108, 6972-6977. (a) Turro, N.J. Pupe Appl, ghem, 1986, ee, 1219-1228. (b) Turro, N.J.: Lei, x.-G.; Cheng, C.-C.: Corbin, D.R.: Abrams, L. J, M, Chem. Soc. 1985, 191, 5824- 5826. (c) Turro, N.J.: Cheng, C.-C.; Lei, X.-G.: Flanigen, B.M. q, Am, Qhem, Soe, 1985, 197, 3739-3741. (d) Turro, N.J.: Wan, P. g, Am, Chem, 599- 1985, 191, 678-682. (e) Turro, N.J.: Wan, P. W 1984, 3;, 3655-3658. (a) Tanguay, J.F.; Suib, S.L. Catal. Rev. 1987, _2_2, 1- 40. (b) Suib, S.L.; Carrado, K.A. Inopg, Chem, 1985, 21, 200-202. (c) Suib, S.L.: Kostapapas, A.: McMahon, K.C.; Baxter, J.C.: Winiecki, A.M. W 1985, 2;, 858-863. (d) Suib, S.L.; Kostapapas, A.; Psaras, D. J, Am, Chem, §OQ, 1984, m, 1614-1620. (e) Suib, S.L.; Bordeianu, O.G.; McMahon, K.C.; Psaras, D. hes eympe_ee;& 1982, 111, 225-238. (a) Camara, M.J.: Lunsford, J.H. W 1983, 22, 2498-2501. (b) Quayle, W.H.; Lunsford, J.H. 18229.1. 39. 40. 41. 42. 43. 44. 225 Chem; 1982, 2,1, 97-103. (c) DeWilde, w.; Peeters, 0.: Lunsford, J.H. Qe_£hy§y_gheme 1980, he, 2306-2310. (a) Wilkinson, F.; Willsher, C.J.; Casal, H.L.: Johnston, L.J.: Scaiano, J.C. W 1986, Q1, 539-544. (b) Casal, M.L.; Scaiano, J.C. w 1985, Q, 1308-1314. (c) Scaiano, J.C.: Casal, M.L.; Netto-Ferreira, J.C. AQ§_§ymp&_§ezy 1985, Zlfi, 211-222. (d) Casal, M.L.; Scaiano, J.C. M 1984, _6_2_, 628-629. (a) Verschoor, C.M.; Olken, M.M.: Ellis, A.B. ,1,_;e_s_s; M 1986, 126.. 221-226. (b) Verschoor, C.M.: Ellis, A.B. Solis: ghape Iohies 1986, 23,, 65-69. (c) Olken, M.M.: Verschoor, C.M.; Ellis, A.B. W 1986, 25, 80-82. (d) Olken, M.M.; Ellis, A.B. 11_Am1 Chem, SOC. 1984, M, 7468-7471. (e) Olken, 14.14.; Verschoor, C.M.; Ellis, A.B. 113—LEM 1984, 11:12, 552-554. (f) Olken, M.M.: Biagioni, R.N.: Ellis, A.B. Ingggm_ghemm 1983, 22, 4128-4134. Giannelis, E.P.; Nocera, D.G.; Pinnavaia, T.J. 1.82112 QDEEL 1987, ge, 203-205. Suib, S.L.: Carrado, K.A. m 1985, a, 863- 867. GhOSh, P.K.; Bard, A.J. J, Ehye, Chem, 1984, ee, 5519- 5526. (a) Thomas, J.K. Acc, Chem, Bee. 1988, 1],, 275-280. (b) Nakamura, T.; Thomas, J.K. g, ghye, Chem. 1986, 2Q, 641-644. (c) Nakamura, T.: Thomas, J.K. Lepgmpi; 1985, 45. 46. 47. 48. 49. 50. 51. 52. 226 1, 568-573. (d) Kovar, L.; DellaGuardia, R.; Thomas, J.K. J. Phys. Chem, 1984, ee, 3595-3599. (e) DellaGuardia, R.A.: Thomas, J.K. J, Phye, Chem, 1984, 1Q, 964-970. (f) DellaGuardia, R.A.: Thomas, J.K. L ghye, Qhem. 1983, fl, 3550-3557. (g) DellaGuardia, R.A.; Thomas, J.K. g. Ehye, Qhem, 1983, e1, 990-998. (a) Nijs, H.; Fripiat, J.J.; Van Damme, H. g. Ehye, Chem... 1983, fl, 1279-1282. (b) Nijs, 11.; Cruz, M.I.; Fripiat, J.J.; Van Damme, H. hopy, ,1, Chim, 1982, 5. 551-557. (c) Abdo, S.: Canesson, P.; Cruz, M.: Fripiat, J.J.; Van Damme, H. ,1, Phys. Qhem, 1981, 35, 797-809. (d) Nijs, H.; Cruz, M.: Fripiat, J.: Van Damme, H. J. Chem. Soc,, Chem. Commph. 1981, 1026-1027. Dorman, W.C.; McCarley, R.E. Inopg. ghem. 1974, 1;, 491-493. Sheldon, J.C. J. Chem. 599. 1960, 1007-1014. Mussell, R.D. Ph.D. Dissertation, Michigan State University, East Lansing, MI, 1988. Hamer, A.D.: Smith, T.J.; Walton, R.A. Inozg. Qhem, 1976, 15, 1014-1017. Saito, T.: Nishida, M.; Yamagata, T.; Yamagata, Y.; Yamaguchi, Y. Ihorg, Chem. 1986, 25, 1111-1117. (a) Chisholm, M.H.: Heppert, J.A.: Huffman, J.C. gplyhegmph 1984, 1, 475-478. (b) Nannelli, P.: Block, B.P. Ingmgy_ghemm 1968, 1, 2423-2426. Hogue, R.D.: McCarley, R.E. w 1970, g, 1354- 1360. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 227 Hench, L.L. In "Science of Ceramic Chemical Processing"; Hench, L.L., Ulrich, D.R., Eds.; J. Wiley: New York, 1986: Chapter 4. Beard, J.H.; Casey, J.: Murmann, R.K. 1M 1965, 5, 797-803. Beard, J.H.; Calhoun, C.: Casey, J.; Murmann, R.K. l5 An1_9hem2_§eel 1968. 22. 3389-3394- Murmann, R.K. Inopg, §yhph, 1966, 3, 173-174. Demas, J.N.: Crosby, G.A. gy_£hyey_ghemm 1971, 15, 991- 1024. ‘Wrighton, M.S.; Ginley, D.S.: Morse, D.Iu J. Ehye, Enemy 1974, 15, 2229-2233. Balzani, V.; Moggi, L.: Manfrin, M.F.: Bolletta, F. Cpng, Chem. Rev, 1975, i5, 321-433. Gagne, R.R.: Keval, C.A.: Lisensky, G.C. or . 1980, 52, 2854-2855. Bard, A.J.; Faulkner, L.R. "Electrochemical Methods": J. Wiley: New York, 1980; p 143. Adams, R.N. "Electrochemistry at Solid Electrodes": Marcel Dekker: New York, 1969; p 219. (a) Parker, C.A.; Rees, W.T. Analyse 1960, 55, 587-600. (b) Drushel, H.V.: Sommers, A.L.; Cox, R.C. Ahele_ghemy 1963, 15, 2166-2172. Allen, G.H.: White, R.P.; Rillema, D.P.; Meyer, T.J. gm Am1_Qhem1_§222 1934. 125, 2513'2520- Dye, J.L.: Nicely, V.A. Qy_§hemy_Egy 1971, 55, 443-448. 66. 67. 68. 69. 70. 71. 72. 228 (a) Zeitlow, T.C.; Nocera, D.G.: Gray, H.B. 1m, Qhem, 1986, 25, 1351-1353. (b) Zeitlow, T.C.: Hopkins, M.D.; Gray, H.B. J. 89],, 51;, Qhem, 1985, 51, 112-119. (c) Nocera, D.G.; Gray, H.B. 5, Am, ghem, goe, 1984, 195, 824-825. (d) Maverick, A.W.; Najdzionek, J.S.; MacKenzie, D.: Nocera, D.G.: Gray, H.B. J, Am, Qhem, $95, 1983, 105, 1878-1882. (e) Maverick, A.W.; Gray, H.B. Q, Am, Qhem. S05, 1981, 195, 1298-1300. Hughbanks, T.: Hoffmann, R. 5, Am, Qhem, Sec. 1983, _1_0_5, 1150-1162. Seifert, G.; Groflmann, G.; Muller, H.-J. Mo S c 1980, 54, 93-102. Nocera, D.G. Ph.D. Dissertation, California Institute of Technology, Pasadena, CA, 1984. (a) Ballhausen, C.J. "Molecular Electronic Structures of Transition Metal Complexes"; McGraw-Hill: New York, 1978: Chapter 4. (b) Barrow, G.M. "Introduction to Molecular Spectroscopy"; McGraw-Hill: New York, 1962; Chapter 4. (a) Kober, B.M.; Caspar, J.V.: Lumpkin, R.S.; Meyer, T.J. g. Rhys, ehem. 1986, 2Q, 3722-3734. (b) Vining, W.J.: Caspar, J.V. : Meyer, T.J. J, Ehys, Qhem, 1985, 85, 1095-1099. (c) Caspar, J.V.: Kober, B.M.; Sullivan, B. P. ; Meyer, T.J. J. Am. Chem, Sec. 1982 , 194, 630-632. (a) Sexton, D.A.: Skibsted, L.H.: Magde, D.: Ford, P.C.‘ 1pm 1984, 2;, 4533-4538. (b) Bergkamp, M.A.: 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 229 Brannon, J.; Magde, D.; Watts, R.J.; Ford, P.C. £2.52; ghemy_§pee 1979, 191, 4549-4554. (c) Petersen, J.D.; Watts, R-J-3 Ford, P.C. W 1975: .22: 3188-3194. (a) Rillema, D.P.; Taghdiri, D.G.; Jones, D.S.; Keller, C.D.; Worl, L.A.: Meyer, T.J.: Levy, H.A. Ihepg2_§hemm 1987, '25, 578-585. (b) Caspar, J.V.; Meyer, T.J. Inngy_ghemy 1983, 22, 2444-2453. Caspar, J.V.; Meyer, T.J. J, Ehye, Chem, 1983, 91, 952-957. Englman, R.: Jortner, J. Mol, Ehye, 1970, 22, 145-164. Reference 10: p 68. For a comprehensive list see reference 71a. Pankuch, B.J.: Lacky, D.E.: Crosby, G.A. J. Ehye, Qhem, 1980, 21, 2061-2067. Freed, K.F.; Jortner, J. J. Chem. Phys, 1970, 52, 6272- 6291. Caspar, J.V.; Sullivan, B.P.; Kober, B.M.; Meyer, T.J. Qhem. Phys, peep, 1982, 21, 91—95. Caspar, J.V.; Meyer, T.J. . Am. Chem. S 1983, 195, 5583-5590. Kitamura, N.: Sato, M.: Kim, H.-B.; Obata, R.; Tazuke, S. Ihozg, Chem, 1988, 21, 651-658. Lumpkin, R.S.; Meyer, T.J. J, Ehys. Qhem. 1986, _(_)_, 5307-5312. Winkler, J.R.: Gray, H.B. Inemg. Chem, 1985, 24. 346- 355. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 230 Del Paggio, A.A.; McMillin, D.R. 1mgmgm_gmgmm 1983, 2;, 691-692. (a) Bignozzi, C.A.; Roffia, S.; Scandola, F. L_Am_,_ M 1985, M, 1644-1651. (13) Bignozzi, C.A.; Scandola, F. Ingmg, cggm. 1984, 2;, 1540-1545. Hennig, H.; Rehorek, A.; Rehorek, 0.; Thomas, P. gmgmgm enim, Act; 1984, fig, 41-49. aninski, R.R.; Petersen, J.D. Ingmgm_§ngmm 1982, g;, 3706-3708. Rillema, D.P.; Allen, G.; Meyer, T.J.; Conrad, D. Ingzg,_gggm, 1983, 22, 1617-1622. Braunstein, C.K.; Baker, A.B.; Strekas, T.C.; Gafney, R.D. Emgmg. Chem. 1984, 2;, 857-864. Giordano, P.J.; Bock, C.R.; Wrighton, M.S.; Interrante, V.L.; Williams, R.F.X. J. Am. Chem. Soc. 1977, _9_, 3187-3189. Lay, P.A.; Sasse, W.H.F. M 1984, 2;, 4123- 4125. Shimidzu, T.; Iyoda, T.; Izaki, K. J, Phys. ngm, 1985, 82, 642-645. Hartley, 0.; Ware, M.J. J. C Co 1967, 912-913. Cotton, F.A.: Wing, R.M.; Zimmerman, R.A. Ino . e 1967, 6, 11-15. Yersin, H.; Otto, H.; zink, J.I.; Gliemann, G. J, Am. Qngmm_§ggm 1980, 102, 951-955. 97. 98. 99. 100. 101. 102. 103. 104. 105. 231 Caspar, J.V.; Westmoreland, T.D.; Allen, G.B.; Bradley, P.G.; Meyer, T.J.; Woodruff, W.H. J. Am, enem, See, 1984, meg, 3492-3500. Rice, S.F.: Wilson, R.E.; Solomon, E.I. M 1980, ;2, 3425-3431. (a) Miskowski, V.M.; Gray, H.B. Imezg, Cmem. 1987, lg, 1108-1112. (b) Miskowski, V.M.: Smith, T.P.: Loehr, T.M.; Gray, H.B. J. Am, cmem, $99, 1985, 1,07, 7925- 7934. Estimated from references 94 and 95. Kober, B.M.; Meyer, T.J. Inorg. Chem, 1985, 24, 106- 108. (a) Mussell, R.D.: Nocera, D.G. J, Am, Chem, 502, 1988, 119, 2764-2772. (b) Mussell, R.D.; Nocera, D.G. Polyhegreg 1986, 5, 47-50. Bock, C.R.; Connor, J.A.; Gutierrez, A.R.; Meyer, T.J.; Whitten, D.G.; Sullivan, B.P.: Nagle, J.K. J, Am. Qmem, $99; 1979, 19;, 4815-4824. (a) Creutz, C.; Sutin, N. J, Am. Chem, Sec, 1977, 22, 241-243. (b) Lin, C.-T.; Bottcher, W.; Chou, M.: Creutz, C.: Sutin, N. J. Am. Chem. Soc. 1976, 28, 6536- 6544. (a) Indelli, M.T.; Ballardini, R.; Scandola, F. is w 1984, _8_8, 2547-2551. (1)) Scandola, F.; Balzani, V.; Schuster, G.B. ,1, Am, enem, Soe, 1981, 19;, 2519-2523. (c) Ballardini, R.; Varani, G.: 106. 107. 108. 109. 110. 111. 112 . 232 Indelli, M.T.: Scandola, F.; Balzani, V. J, Am, Chem. £22; 1978, lQQ, 7219-7223. Nagle, J.K.; Dressick, W.J.; Meyer, T.J. J. Am, Chem, $221 1979, 191, 3993-3995. (a) Marshall, J.L.; Stobart, S.R.; Gray, H.B. ,1, M: Chem,_§_ec_,_ 1984, 1526, 3027-3029. (b) Nocera, D.G.: Gray, H.B. J. Am. Chem, §CC. 1981, 121, 7349-7350. Heuer, W.B.; Totten, M.D.: Rodman, G.S.; Hebert, E.J.: Tracy, H.J.: Nagle, J.K. J. fig, Chem, Soc, 1984, m, 1163-1164. (a) Experimental evidence for the inverted regime has recently been presented.1°9b'e (b) Wasielewski, M.P.; Niemczyk, M.P.; Svec, W.A.; Pewitt, E.B. J_,_Am,__Chem_,_ $921 1985, m, 1080-1082 . (C) Closs, G.L.; Calcaterra, L.T.; Green, N.J.; Penfield, K.W.; Miller, J.R. ,1, Phys, Chem, 1986, 9_o, 3673-3683. (d) Miller, J.R.: Beitz, J.V.; Huddleston, R.K. ,1, Am. Chem, See, 1984, 1C6, 5057-5068. (e) Miller, J.R.: Calcaterra, L.T.; Closs, G.L. J. m, Chem, §oc, 1984, 110.5. 3047- 3049. (a) Ballardini, R.; Gandolfi, M.T.; Balzani, V. L M 1988, 22, 56-60. (b) Ballardini, R.: Gandolfi, M.T.: Balzani, V. M. 1987, 2m, 862-867. Rybak, W.; Haim, A.; Netzel, T.L.; Sutin, N. J. Ehye, Chemm 1981, 85, 2856-2860. Frank, R.; Rau, H. J, Ehye, chem, 1983, el, 5181-5184. 233 113. Jamieson, M.A.; Langford, C.K.; Serpone, N.; Hersey, M.W. J, Ehye, Chem, 1983, _81, 1004-1008. 114. White, 11.8.; Becker, W.C.; Bard, A.J. L_P_hys_,__Chem_,_ 1984, _8_§, 1840-1846. 115. Sabbatini, N.; Bonazzi, A.; Ciano, M.; Balzani, V. J_._ Am, Chem, Sec. 1984, 116,, 4055-4056. 116. (a) Deviations from linear SV behavior have been observed which are not a result of ion-pair formation.116b'f' (b) Prasad, D.R.; Mandal, K.: Hoffman, M.z. Coomd. Chem, Rev, 1985, 51. 175-190. (c) Peak, D.; Werner, T.C.; Dennin, R.M.; Baird, J.K. ,L_ Chem_,_£hy§_. 1983, 7_9_, 3328-3335. (d) Chattopadhyay, S.K.; Das, P.K.; Hug, G.L. ,L_Am.__Ch_em_,__§_o_c_,_ 1982, 194, 4507-4514. (e) Nemzek, T.L.; Ware, W.R. J, Chem, Ehye, 1975, _C, 477-489. (f) Lakowicz, J.R.; Weber, G. Bieghemistrx 1973, 12, 4161-4170. 117. (a) Keizer, J. Chem, Rev. 1987, g, 167-180. (b) Keizer, J. Aee, Chem, ges. 1985, me, 235-241. 118. Baird, J.K.: Escott, S.P. J. Chem, Ehye, 1981, fl, 6993-6995. 119. Keizer, J. J, Am, Chem, See, 1983, 195, 1494-1498. 120. Stevens, B. W 1987, 1,14, 519-524. 121. (a) Yang, D.Y.; Cukier, R.I. ,1, Chem. Phys, 1987, CC, 2833-2838. (b) Cukier, R.I. ,1, Chem, Phys. 1985, 8.2. 5457-5469. (c) Cukier, R.I. LAm,__Chem_,__§9_q. 1985, 191, 4115-4117. 122. Adamson, A.W. J. Chem, Ed, 1983, Q, 797-802. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 234 The equilibrium constant is calculated by Keq = 4xNAr3/3000 exp(-w/RT) where w is determined from eq 10. Because 1 = "i + 10, estimates of 4i and 10 are required. 10 was calculated to be 0.79 eV in acetone from eq 9 (a1(pyridinium quencher) = 4.0 A; a2(W5I142') = 5.5 A) and *i = 0.25 eV from reference 48. The area of the electrode (A) was 0.0314 cm2 and the concentration (Co*) of W61142", I, II, and III were 4.44 = 10’7, 2.76 x 10'5, 3.10 x 10‘6, and 7.29 x 10'7 mol/cm3, respectively. Nagle, J.K.; Dressick, W.J.; Meyer, T.J. gm_Amm_szmh_ gee; 1979, 1C1, 3993-3995. Edward, J.T. g, Chem, 3g, 1970, 51, 261-270. Onsager, L. Phys, gev, 1938, 55, 554-557. Bakale, G.; Gregg, R.C.; McCreary, R.D. J, Chem, Ehye, 1977, gl, 5788-5794. Kober, B.M.; Sullivan, B.P.; Meyer, T.J. Ihemgh_Chemy 1984, 21, 2098-2104. Fox, M.A. ACC, Chem, Res, 1983, 1C, 314-321. (a) Paczkowski, J.: Neckers, D.C. neememeleemlee 1985, 1e, 2412-2418. (b) Tamagaki, 8.; Liesner, C.E.: Neckers, D.C. Je_Cmgy_Chemy 1980, 5;, 1573-1576. Urruti, E.B.; Kilp, T. heeremeleemlee 1984, 11, 50-54. Lee, P.C.; Matheson, M.S.; Meisel, D. Ism, ,1, Chem, 1982, 22, 133-137. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 235 Kurimura, Y.; Takato, 12mm 1985. 3.9.. 14.5. W 1986. K.; Takeda, M.; Ohtsuka, N. 11 1023-1026. Wrighton, 231, 32-37. L.: Coche, Moutet, Ju-C. LEW 1987. 221, 111-122. (a) Oyama, P.C. N.; Anson, 1.22: (b) J.M. g: Q. Ellemo figg: 1985, 197' 3442-3450. COQEQ, thfl, 33V, 138‘, 55-226 and references therein. 11.31egtrganal1_9hem1 1986. 467-470. Anson, P.C.; Ni, C.-L.; Saveant, Skorobogaty, 51. Dannhauser, T.; O'Neil, M.; Johansson, K.; Whitten, D.: McLendon, G. J, Rhys, Chem, 1986, 29, J.; g, Rheeeehem, 1985, 23,, ,1, Phye, Chem, 8.; A. ; Smith, T.D. 6074-6076. (a) Rabani, R.E. (b) 1985, e2, Sassoon, 7-26. Sassoon, R.E.; Rabani, J. 5500-5504. R.E.: Gershuni, Rabani, J. J, Ehye. Chem, 1985, fig, 1937-1945. Y.; (c) Sassoon, Nosaka, Kuwabara, A.; Miyama, H. J. s C 1986, 2C, 1465-1470. Kaneko, M.; Yamada, A.; Kurimura, km 1980, 55, L73-L74. Snell, R.D.: Keenan, A.G. Chemy_§eey_3eyy 1979, e, 259- 282. ROY. R- W 1987. 2.3.8.. 1664- 1669. Pope, E.J.A.; Mackenzie, J.D. u3§_emlly 1987, 29-31 and references therein. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 236 Avnir, D.; Levy, D.: Reisfeld, R. J, Ehye, Chem, 1984, CC, 5956-5959. Dunn, B.: Knobbe, E.: McKiernan, J.M.; Pouxviel, J.C.; Zink, J.I.: paper presented at Materials Research Society Meeting, Reno, Nevada, April 1988. Tomoi, M.; Ford, W.T. J. Am, Chem, Sec, 1981, 103, 3821-3828. Buttry, D.A.: Anson, P.C. Jy_Amy_Chemy_§eee 1983, 1C;, 685-689. Majda, M.; Faulkner, L.R. W 1984, 152. 77-95. Sumi, K.; Anson, F.C. J. Phys. Chem, 1986, 19, 3845- 3850. (a) Ballhausen, C.J.; Gray, H.B. Imezge_Chemy 1962, 1, 111-122. (b) Gray, H.B.; Hare, C.R. Inerg1_ghem1 1962, 1, 363-368. (c) Miskowski, V.; Gray, H.B.; Ballhausen, C.J. M91. Bhys, 1974, 1e, 729-745. (d) Winkler, J.R.; Gray, H.B. Cemme_1he;gy_Chemy 1981, 1, 257-263. (a) Collison, D.; Gahan, B.; Garner, C.D.; Mabbs, E.B. J, Chem. $99,, QaLtoh Imehs. 1980, 667-674. (b) Hill, L.M.; Howlader, N.C.; Mabbs, F.E.; Hursthouse, M.B.; Malik, K.M.A. J, Chem, §ee., Deltoh Imems, 1980, 1475- 1481. Sunil, R.K.: Rogers, M.T. IDQEQ- Chem. 1981, 29, 3283- 3287. Rappe, A.K.; Goddard, W°A0 Milk—5.9;; 1982: 19.4.: 448-456. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 237 Antipas, A.; Buchler, J.W.; Gouterman, M.; Smith, P.D. £1_Am1_9hem1_§eg1 1980, 192, 198-207. (a) Tatsumi, K.; Hoffmann, R. Ihem, Chem. 1980, 1e, 2656-2658. (b) Tatsumi, K.; Hoffmann, R. Imemgy_Chemy 1981, 2C, 3771-3784. Hanson, L.K.; Chang, C.K.; Davis, M.S.; Fajer, J. l, Amy_Chemy_§eCy 1981, 1_1, 663-670. Loew, G.B.; Herman. 2.8. MM 1980. 1.0.2. 6173-6174. Winkler, J.R.; Gray, H.B. J. Am, Chem. goe. 1983, 1C5, 1373-1374. Bailey, S.W. In "Crystal Structures of Clay Minerals and Their X-ray Identification"; Brindley, G.W., Brown, G., Eds.; Mineralogical Society: Britain, 1980:\ Chapter 1. Gastuche, M.C.; Brown, G.; Mortland, M.M. WI... 1967, 1, 177-192. Miyata, S. Cleys Clay himem, 1980, CC, 50-56. Taylor, R.M. Clay mihez, 1984, 12, 591-603. Similar intensity variations have been observed in the absorption spectra of other CLO intercalates.4'1'43'169 The factors responsible for these intensity variations have also not been identified. The experimental manifestations of this result are quite striking. Aqueous solutions of Re02(py)4+ yield no luminescence upon visible or ultraviolet irradiation (A < 450 nm). However, with the addition of powdered 168. 169. 170. 171. 238 hectorite to the solution, red luminescence of Re02(py)4+ is observed almost immediately and grows in intensity as the ion-exchange reaction proceeds to completion (~10 min for a 1 wt% suspension of hectorite). (a) Habti, A.; Keravis, D.; Levitz, P.; Van Damme, H. 1, Chem. See,I Eereeey Trans. 2 1984, §_Q, 67-83. (b) Bergaya, F.; ‘Van Damme, H. .J C m. So Iran§1_2 1933. 12. 505-518- Schoonheydt, R.A.; De Pauw, P.; Vliers, D.; De Schrijver, F.C. J, Phye, Chem. 1984, fig, 5113-5118. Despite the apparent electronic similarities of the Re02(en)2+ ion with that of the pyridine and cyanide analogues, the complex does not luminesce.84 It has been suggested that the N-H vibrations of the ethylenediamine ligands provide efficient nonradiative decay channels to the ground electronic state. Stern- Volmer experiments show that Re02(en)2+ does not quench Re02(py)4+ luminescence. ( a) Average dimens ions were calculated from crystallographic: data for’ [Re02(py)4]Cl‘2H20171b and K3[Re02(CN)4]171C and by using the van der Waals radii of the terminal atoms. For Re02(py)4+, the calculated lengths of the C4 (O-Re-O) , C3 (for pseudo-octahedral complex), and C2" axes (bisecting the pyridine-rhenium- pyridine equatorial axes) are 6.32 A, 5.44 A and 9.43 A, respectively. The calculated lengths of the C4, C3, 239 and C2" axes are 6.36 A, 4.66 A, and 6.84 A, respectively, for Reoz (CN) 43' . (b) Calvo, C. ; Krishnamachari, N.; Lock, C.J.L. J . o . t c . 1971, _1, 161-172. (c) Murmann, R.K.; Schlemper, E.0. W 1971, 19, 2352-2354. 172. (a) Miyata, S. Win21;- 1975, 2;, 369-375; 1983, 1;, 305-311. (b) Miyata, S.; Okada, A. Cleye C1ey_h1mem, 1977, Ah, 14-18. (c) Miyata, 5.; Hirose, T. Cleys Clay Miner. 1978, 2C, 441-447. 173. (a) Mann, K.R.; Gray, H.B. Aev. Chem. §er, 1979, 11;, 225-235. (b) Miskowski, V.M.; Sigal, I.S.; Mann, K.R.; Gray, H.B.: Milder, S.J.; Hammond, G.S.; Ryason, P.R. J. Am. Chem. Soc, 1979, 1C1, 4383-4385. (c) Caspar, J.V.; Gray, H.B. J, Am. Chem. §oc. 1984, egg, 3029-3030. (d) Marshall, J.L.; Stiegman, A.B.: Gray, H.B. Agv. Chem. Ser. 1986, 307, 166-176. 174. Roundhill, D.M. J, &, Chem, SOC. 1985, 107, 4354-4356. 175. Caspar, J.V. ,2. fig. Chem. 59C. 1985, 107, 6718-6719. 176. (a) Clearfield, A.; Thakur, D.S. A991. Cetel, 1986, A6, 1-26. (b) Clearfield, A.; Thakur, D.S. .1,__C_a_te]_. 1980 , Q, 185-194 .