LIBRARY Michigan State University This is to certify that the dissertation entitled STRUCTURE, STABILITY, AND REACTIVITY 0F SELF ASSEMBLED MONOLAYERS ON AU (111) presented by L I L I DUAN has been accepted towards fulfillment of the requirements for Ph . D. degreein CHEMISTRY 3b 6 U Major professor Date 08/15/01 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE J DATE DUE DATE DUE NOWI 032ml? .I (.UU/ 6/01 cJClRC/DateDuepBS-p. 15 STRUCTURE, STABILITY, AND REACTIVITY OF SELF-ASSEMBLED MONOLAYERS ON AU(111) By Lili Duan A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 2001 ((1 lane ‘ VU‘ It... Iva M J? UyI m ’1 ABSTRACT STRUCTURE, STABILITY, AND REACTIVITY OF SELF-ASSEMBLED MONOLAYERS ON AU(111) By Lili Duan In order to provide a deeper understanding of the impact of molecular structure on monolayer morphology and properties, and to examine the influence of the monolayer environment on the intrinsic reactivity of unsaturated moieties incorporated into self assembled monolayers (SAMs), four groups of thiol and disulfide SAMs were investigated using surface sensitive techniques, representing a systematic study of aliphatic, aromatic, and aromatic/aliphatic mixed monolayers on Au(111). It was demonstrated that, at room temperature, simple alkanethiols formed densely packed, crystalline-like assemblies on Au(111) with the molecules arranged in a hexagonal pattern. These self- assemblies represented a simple balance between the chemical interaction of the sulfur headgroup/gold surface and the hydrophobic interaction of the hydrocarbon chains. For substituted alkyl thiols, the alkyl chain length, the size of the terminal group together with the terminus interactions played a crucial role in determining the monolayer structure. Shorter chain thiols with bulky endgroups, e.g. phenyl-terminated hexanethiol, lead to lower density coverages and more disordered packing; whereas, small substituents such as olefin groups 111111 ()1 '“f‘ (1') (Li) j L'ra: would have little impact on the monolayer structure. In the terphenyl thiol SAMs, the aromatic interchain rt-n and S-Au interactions appeared to be of similar magnitude. As the result, small domains of order (due to the S-Au bonding) with a high density of defects (due to the structural restrictions imposed by the terphenyl backbone) were observed. Three unsaturated functionalities (olefin, styrene, and cyclobutyl) were incorporated into aliphatic and aromatic thiols to assess the possibility of forming covalent bonding within a single molecular layer ("cross-linking"). Polymerization of the styrene-terminated hexanethiol SAM was achieved by either near-UV irradiation or thermal treatment. In contrast, no polymerization/oligomerization reaction was observed for the olefin-terminated octanethiol SAMs under identical conditions. The different reactivity between styrene- and olefin-terminated monolayers was attributed to the different ' reactivities of a-olefins and styrenes. The cyclobutyl group attached to the terphenyl thiol SAM also appeared to be photostable under the full spectrum UV irradiation. To my precious gift from God-- my daughter Esther Caroline Li ill'fu MW w 0‘3: If 3:1 ACKNOWLEDGEMENTS First and foremost, I wish to acknowledge my .advisor, Simon J. Garrett, for the sound advice and careful guidance he has given me. I will always be grateful for the knowledge, patience, and understanding he has offered to me through the past five years. I would also like to thank the other members of my committee, Dr. Gary Blanchard, Dr. Gregory Baker, and Dr. Mercouri Kanatzidis for the assistance they provided at all levels of the research project. I would also like to acknowledge Dr. Merlin Bruening and Dr. Kathryn Severin, who have generously let me use their lab and equipment whenever I needed. I am very grateful for the daily help and friendship provided by my group members: Todd, Mike, Heather, and Jason. Extra thanks goes to Jong-Bum, Tianqi, and Samantha for helping to synthesize some of the thiol and disulfide molecules used in this study. I would never get this far without the support of my family and friends. To my husband and best friend, Jian, who has accompanied me through the highs and lows of the past five years with his love and encouragement, I acknowledge an appreciation that extends beyond any words at my command. I also owe a special debt of gratitude to my parents for their endless love and confidence in me all along. Without their endurance of one-year separation and my mom’s staying with us and caring for my daughter Esther, the completion of this dissertation would not have been possible. Appreciation also goes out to my friends at Lansing Chinese Christian Ministry for their love, friendship and \l prayers. Finally, I thank God who brought my family to the United States, opened a new world in front of our eyes, saved us through his only son and changed us inside out. A Time for Everything (Ecclesiastes 3:1-8) There is a time for everything, and a season for every activity under heaven: A time to be born, and a time to die, A time to plant, and a time to uproot, A time to kill, and a time to heal, A time to tear down, and a time to build, A time to weep, and a time to laugh, A time to mourn, and a time to dance, A time to throw stones, and a time to gather stones, A time to embrace and a time to refrain, A time to search, and a time to give up, A time to keep, and a time to throw away, A time to tear, and a time to mend, A time to be silent, and a time to speak, A time to love, and a time to hate, A time for war, and a time for peace. And now it’s time for me to graduate-- glory to my Lord! vi "l TABLE OF CONTENTS List of Tables List of Figures List of Schemes Chapter 1: Introduction 1.1. Overview of Self-Assembled Monolayers 1.2. Motivation and Objective of the Present Work 1.3. Outline of the Dissertation 1.4. Literature Cited Chapter 2: Experimental Techniques 2.1. Scanning Tunneling Microscopy 2.2. X-ray Photoelectron Spectroscopy 2.3. Reflection-absorption Infrared Spectroscopy 2.4. Ellipsometry 2.5. Contact Angle Measurements 2.6. Literature Cited Chapter 3: Preparation and Characterization of Gold Substrates 3.1. Introduction 3.2. Experimental 3.3. Results and Discussion 3.4. Conclusion 3.5 Literature Cited Chapter 4: Simple Alkyl Thiol Self-Assembled Monolayers on Au(111) 4.1. Introduction 4.2. Experimental 4.3. Results and Discussion 4.4. Conclusion 4.5. Literature Cited vii Page ix xvii AmhA 17 1 8 33 34 36 38 4O 44 45 46 47 60 61 65 66 68 7O 82 83 (n (y! (n (_J'I $.11 0’15; 1&1; mmmmm Vi Chapter 5: 6-Phenyl-n—hexanethiol and 6-(p-Vinylphenyl)-n-hexanethiol Self-Assembled Monolayers on Au(111) 88 5.1. Introduction 89 5.2. Experimental 92 5.3. Results and Discussion 94 5.4. Conclusion 121 5.5. Literature Cited 122 Chapter 6: Rigid Aromatic p-Methyl Terphenyl Thiol and 2’,3’-Cyclobutyl p- Terphenyl Thiol Self-Assembled Monolayers on Au(111) 127 6.1 . Introduction 128 6.2. Experimental 130 6.3. Results and Discussion 132 6.4. Conclusion 150 6.5. Literature Cited 151 Chapter 7: Olefin-Terminated Di(9-decene) Disulfide Self-Assembled Monolayers on AU(1 11) 155 7.1 . Introduction 156 7.2. Experimental 158 7.3. Results and Discussion 160 7.4. Conclusion 185 7.5. Literature Cited 187 Chapter 8: Conclusions and Outlook 190 8.1. Significance of the Results 190 8.2. Possible Future Directions 196 8.3. Literature Cited 201 viii la la la la la 12 I; Table 1.1 Table 2.1 Table 2.2 Table 2.3 Table 4.1 Table 4.2 Table 4.3 Table 5.1 Table 6.1 Table 7.1 Table 7.2 Table 7.3 Table 7.4 LIST OF TABLES Representative substrates and ligands that form SAMs Comparison of three different etching methods to prepare tungsten and Pto,a/lro,2 tips Comparison of liquid and crystalline state symmetric (vs) and asymmetric (vas) CH2 stretching modes with octadecanethiol SAM on gold Techniques used characterization in the present study for SAMs Thickness of C8, C10, and C18 SAMs Advancing contact angles of water on 08, C10, and C18 SAMs IR assignments of 0-H high frequency stretching modes of neat liquid C10, solid C18 in KBr and C18 SAM on gold Infrared Band Assignments for PHT and VHT Infrared band assignments for MTPT SAM and DMTP solid in KBr Infrared band assignments for DDDS The thickness and water contact angles of DDDS monolayer during annealing The thickness and water contact angles of DDDS monolayer during near UV irradiation in toluene with ~1 mM BME The thickness and water contact angles of DDDS monolayers after 30 min near UV irradiation in toluene with BME followed by anneal Page 2 28 36 4o 71 71 74 97 135 161 169 181 184 “/5 2 fi/.- F.‘ f In Fig. 2.1 Fig. 2.2 Fig. 2.3 Fig. 2.4 Fig. 2.5 Fig. 2.6 Fig. 2.7 Fig. 2.8 Fig. 2.9 Fig. 2.10 Fig. 2.11 Fig. 2.12 LIST OF FIGURES The energy level diagram of a tip and a sample (both are metals) in a vacuum, (a) when the tip and the sample are not connected; (b) when the tip and the sample are sufficiently close and a voltage is applied between them. (adapted from Tersoff and Lang) STM imaging process (adapted from Chen) Image processing on a HOPG surface (44 x 44 A). (a) the original data; (b) after x- and y-slope subtractions Image processing on a HOPG surface (2000 x 2000 A). (a) the original data; (b) after applying median filter Experimental setup for electrochemical etching Sketches of wire preparation for electrochemical etching. (a) bare wire method; (b) epoxy insulation method; (0) tube insulation method SEM Images of a tungsten tip fabricated with the control system shown in Fig. 2.5 and 2.6 (0) SEM Images of a PtQB/Iroz tip fabricated with the control system shown in Fig. 2.5 and 2.6 (b) SEM Images of cut Pto,3/lro,2 tips: (a) a freshly cut tip; (b) a used blunt tip; (c) a crashed tip A series of STM I-V spectra over a gold sample showing the gradual tip degradation STM images of HOPG obtained in air usin cut Pto 3/lr02 tips. (a) 190x190A; (b) 90x90A c) 50x50 ;)(d ).20x20A All images have been slightly filtered The electric field at a metal surface for a high angle of incidence and for s- and p-polarized light. Only the p- polarized component survives at the metal surface, resulting in an anisotropic electric field Ep parallel to the surface normal Page 19 2O 23 24 26 26 28 29 30 31 32 35 "‘0'" (K "J Fig. 2.13 Fig. 2.14 Fig. 3.1 Fig. 3.2 Fig. 3.3 Fig. 3.4 Fig. 3.5 Fig. 3.6 Fig. 3.7 Fig. 3.8 Fig. 3.9 Three-phase parallel layer model (from Collin et al.) Solid-liquid-gas contact angle STM images of the gold single crystal as received, showing single or multiple atomic-height terraces whose step edges meet at 60°. (a) 3000 x 3000 A; (b) 2000 x 2000 A; (c) 1000 x 1000 A XPS surveys of the gold single crystal. (a) as received; (b) after two cycles of sputtering and annealing Experimental apparatus for electrolytic polishing of the gold single crystal STM images of Au single crystal after many cycles of sputtering and annealing, showing the predominant rippled topogrzphy. (a) 2000 x 2000 A; (b) 1000 x 1000 A; (c) 500 x 500 STM images of Au single crystal after many cycles of sputtering and annealing, showing a relatively flat area. The appearance of lines are due to differential sputtering. (a) 3000 x 3000 A; (b) 2000 x 2000 A; (c) 1000 x 1000 A STM images of octanethiol SAM on the gold single crystal. Due to the roughness of the gold substrate, the order of the SAM is limited to small regions. (a) 1500 x 1500 A; (b) 250 x 250 A; (c) 200 x 200 A; (d) 150 x 150 A STM images of Au(111) on mica. The film was comprised of gold grains with atomically flat terraces on top. The gold atoms were arraigned in a hexagonal pattern which is characteristic of the (111) close-packed plane for fcc metals. (a) 3000 x 3000 A; (b) 1300 x 1300 A; (c) 70 x 70 A; (d) 50 x 50 A Schematic representation of the Au(111) 22 X\/3 reconstruction as seen from top and side views. The unit cell is shown b the rectangular which has dimensions of 63.36 A x 4.99 . This figure is taken from Dishner et al. STM images of Au(111) on mica (400 x 400 A), showing the 22x13 surface reconstruction 38 39 48 49 50 51 52 53 55 56 57 Fig. 3.10 Fig. 4.1 Fig. 4.2 Fig. 4.3 Fig. 4.4 Fig. 4.5 Fig. 4.6 Fig. 4.7 Fig. 4.8 Fig. 5.1 A series of time-lapse STM topographic images at room temperature showing a 500 x 500 A area of a gold film on mica. The arrows point out surface details changed with time (a) Average water advancing contact angle for gold substrate is < 20°, (b) average water advancing contact angle for C18 SAM on gold substrate is 109° XPS spectrum of CB SAM on Au(111) IR spectra of CH stretching modes in the high frequency region. (a) transmission spectrum of neat liquid C10; (b) transmission spectrum of solid C18 in KBr; (c) reflection spectrum of C18 SAM on Au(111) STM images of CB SAM at room temperature. (a) 1200x1200 A, (b) 700x700 A, (c) 150x150 A Molecular resolution STM images (50 x 50 A) of CB SAM on Au(111) which exhibits a periodic hexagonal pattern with nearest neighbor distance of 5.0 i0.3 A, consistent with a (V3xl3)R30° adlayer on a Au(111) lattice STM images of C8 SAM on Au(111) after annealed to 75 °C for 30 min, showing large and well-structured low densi wide striped phase. (a) 1500 x 1500 A, (b) 700 x700 , (c) 300 x 300 A, (d) 100 x 100 A RAIRS spectra of C10 SAMs on Au(111) following annealing at the indicated temperature Plot of peak intensity change for C10 monolayers from Fig. 4.7 as a function of annealing temperature in (a) VasCHg and VasCHa modes; (b) VasCHz and vsCHg modes IR spectra of (A) neat VHT; (B) VHT monolayer on Au/Si wafer; (C) neat PHT and (D) PHT monolayer on Au/Si wafer. The arrows indicate the sharp intensity drop of vC=C, 6=CH2, 00=CH-, 00=CH2 and benzene ring 17b mode, respectively, for the VHT monolayer xii 59 72 73 73 76 78 79 81 82 96 ( ‘17 t” ) it Fig. 5.2. Fig. 5.3 Fig. 5.4 Fig. 5.5 Fig. 5.6 Fig. 5.7 Fig. 5.8 Fig. 5.9 Schematic diagram indicating the directions of the various transition dipole moments for the vinyl group of VHT. Both vC=C and 8=CH2 modes are parallel to the vinyl double bond, while vas=CH2 mode is in the H-C-H plane, perpendicular to the vC=C and 8=CH2 modes. The out-of- plane wag modes, 00=CH- and 00=CH2, are orthogonal to the other three modes (A) STM images (950 x 950 A (inset, 200 x 200 A)) of PHT monolayers on Au/mica at room temperature show the 6 stripe phase. (B) Sectional view alone the line in Fig. 5.3 (A) reveals the row spacing of 16 A. (C) A schematic representation of the 5 phase Typical room temperature STM image of VHT monolayers on Au/mica (950 x 950 A) RAIR spectra of VHT monolayers on Au/Si wafer obtained at various UV irradiation time with the sample placed in a load40ck RAIR spectra of VHT monolayers on Au/Si wafer obtained at various UV-Iight exposure time while soaked in the solvent. The band intensities of various vinyl modes all decrease with increasing exposure to UV-light Plot of peak intensity change for VHT monolayers on Au/Si wafers from Fig. 5.6 as a function of UV-light exposure time: (A) changes in the vas=CH2 and vC=C modes and (B) changes in the VasCHz and vsCHg modes STM images of PHT SAMs at 40 °C (A: 1500x1500 A, 8: 500x500 A) and 60 °C (0: 1500x1500 A, D: 450x450 A) (A) STM images (800 x 800 A (inset, 250 x 250 A)) of PHT monolayers on Au/mica at 809C show the x’ stripe phase. The arrows indicate the switching of stripe segments. (B) Sectional view alone the line in Fig. 5.9 (A) reveals the row spacing of either 16 A or 28 A. (C) A schematic representation of the x’ phase xiii 102 103 105 106 108 110 111 113 Fig. (£1? Fig. 5.10 Fig. 5.1 1 Fig. 5.12 Fig. 5.13 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4. Fig. 6.5 Fig. 6.6 (A) STM images (1000 x 1000 A (inset, 250 x 250 A)) of PHT monolayers on Au/mica at 1209C show the B stripe phase. (B) Sectional view alone the line in Fig. 5.10 (A) reveals the row spacing of 28 A. (C) A schematic representation of the [3 phase RAIR spectra of adsorbed VHT monolayers on Au/Si wafer obtained following annealing at the indicated temperature RAIR spectra of pre-photopolymerized VHT monolayers on Au/Si wafer obtained following annealing at the indicated temperature Plot of peak intensity change for unpolymerized and pre- photopolymerized VHT monolayers on Au/Si wafers from Fig. Fig. 5.11 and 5.12 as a function of anneal temperature: (A) changes in the vasCHz and vsCHz modes and (B) changes in the VasCHg mode (shown as the ratio of VaSCHa peak intensity divided by that of initial unpolymerized VasCHz mode) lR spectra of (A) DMTP solid in KBr, and (B) MTPT monolayer on a Au/Si wafer Large scale STM image of MTPT on Au(111). Three terraces separated by monatomic steps are shown. Islands with a height of 26:03 A are observed. Image size 1500 x 1500 A High resolution STM images of MTPT monolayer on Au(111). Image size: (A) and (8) 150x150 A; (C) 40x60 A. (D) and (E) show the corresponding line profiles in (C) Model showing the adsorbed sulfur atoms as black dots and the molecular chains as lines along the molecular axis. The sulfur atoms are ordered while the top of the molecules are slightly displaced from their (V3x13)R30° positions IR spectra of (A) MTPT monolayer on Au(111), and (B) CTPT monolayer on Au(111) STM images of CTPT monolayers on Au(111) at room temperature. Image size: (A) 1500x1500 A; (B) 1000x1000 A; (C) 500x500 A; (D) 500x500 A xiv 115 116 117 120 133 137 140 142 145 146 Hg. E Fig. '1 Fig. it Fig. 6.7 Fig. 6.8 Fig. 7.1 Fig. 7.2 Fig. 7.3 Fig. 7.4 Fig. 7.5 Fig. 7.6 Fig. 7.7 Schematic diagram indicating the “width“ of phenyl and benzocyclobutene units in CTPT molecule STM images of CTPT monolayers on Au(111) after annealing to 45 °C (A,B,C), 75 °C (D,E,F) and 100 °C (G,H,l). Image size: (A) 1300x1300 A; (B) 900x900 A; (C) 600x600 A; (D) 1500x1500 A; (E) 1250x1250 A; (F) 1000x1000 A; (G) 2000x2000 A; (H) 1500x1500 A; (I) 750x750A IR spectra of (A) neat liquid DDDS; (B) DDDS monolayer on Au(111) preparedfrom 2mM solution in decane; (C) DDDS monolayer on Au(111) prepared from 2mM solution in ethanol Schematic diagram indicating the directions of the various vinyl transition dipole moments of DDDS. Both vC=C and v=CH2 modes are parallel to the vinyl double bond, while vas=CH2 mode is in the H-C-H plane, perpendicular to the vC=C and vs=CH2 modes. The out-of-plane wag modes, 00=CH- and 03=CH2, are orthogonal to the other three modes RAIRS spectra of DDDS SAMs on Au(111) following annealing at the indicated temperature Plot of peak intensity change for DDDS monolayers from Fig. 7.3 as a function of anneal temperature in (A) vas=CH2, vC=C and m=CH2 modes; (B) vasCHz and vsCHz modes RAIRS spectra of DDDS monolayer on Au(111) in pure N2(g) environment after 0 and 30 min near-UV (> 300 nm) irradiation. The broad feature at 1000 - 1200 cm'1 is due to polarizer artifact RAIRS spectra of a DDDS monolayer in toluene with ~1 mM AIBN after 0 and 30 min near-UV irradiation (> 300 nm) RAIRS spectra of DDDS SAMs on Au(111) in toluene obtained at various full-UV (250-400 nm) irradiation time 147 149 161 163 166 167 171 172 173 Fig. 1 Fig. ‘ Fig. Hg. Fig Fig. 7.8 Fig. 7.9 Fig. 7.10 Fig. 7.11 Fig. 7.12 Fig. 7.13 Fig. 7.14 Fig. 7.15 Fig. 7.16 Fig. 8.1 Plot of peak intensity change for DDDS monolayers from Fig. 7.7 as a function of full-UV (250-400 nm) irradiation time in (A) vas=CH2, vC=C and m=CH2 modes; (B) VasCHz and VsCHz modes RAIRS spectra (C-H region) of decanethiol SAMs on Au(111) in toluene obtained at various full-UV (250-400 nm) irradiation times Plot of peak intensity change for decanethiol SAMs from Fig. 7.9 as a function of full-UV (250-400 nm) irradiation time in (A) vasCHa and VSCHa modes; (B) VasCHz and vsCHz modes RAIRS spectra of DDDS SAMs on Au(111) in toluene with ~1 mM BME obtained at various near UV (>300 nm) irradiation times Plot of peak intensity change for DDDS SAMs from Fig. 7.11 as a function of UV-light (>300 nm) exposure time with ~1 mM BME in (A) vas=CH2, vC=C and 00=CH2 modes; (B) VasCHz and VsCHz modes RAIRS spectra of decanethiol SAMs on Au(111) in toluene with ~1 mM BME obtained at various near-UV (>300 nm) irradiation times Plot of peak intensity change for decanethiol SAMs from Fig. 7.13 as a function of near UV (>300 nm) exposure time with BME in (A) vasCHa and vsCHa modes; (B) vasCHz and vsCHz modes RAIRS spectra of DDDS SAMs on Au(111) after 30 min UV (>300 rim) irradiation with ~1 mM BME followed by annealing at the indicated temperature Plot of peak intensity change for a 30 min near UV (>300 nm) irradiated DDDS monolayer in toluene with ~1 mM BME initiator from Fig. 7.15 as a function of anneal temperature in vasCHz and VSCHz modes Diagrams illustrating the idea of using a styrene- terminated alkanethiol SAM as a negative resist to transfer the image of a mask into a gold substrate 174 176 177 179 180 182 182 183 184 197 Scher $61161 8010 Scheme 1.1 Scheme 6.1. Scheme 7.1. Scheme 7.2. LIST OF SCHEMES The selected adsorbates systems for investigation in the present study Synthetic scheme for MTPT molecule. a 5 mol % Pd(PPh3)4, 3.18 g Nacha, 15 ml H20 and 40 ml toluene, 48 hours in Ar; b 5 equiv. of NaSCHa in NMP, 100 °C, 48 hours in N2 Synthetic scheme for DDDS molecule. a 2 equiv. of p- toluenesulphonyl chloride and 3 equiv. of pyridine in 30 ml chloroform, 5-8 hours in an ice bath; b 6 equiv. of sodium hydrogen sulfide in absolute ethanol, 42 °C, 22 hours sonication Initial a-cleavage of BME after absorption of a photon in the wavelength range from 313 to 365 nm Page 130 158 178 Chapter 1 Introduction This dissertation presents the findings of the study of aliphatic, aromatic and aliphatic/aromatic mixed thiol monolayers assembled on Au(111) surfaces. The structure, thermal stability and reactivity of these monolayers were examined using several surface sensitive techniques, including scanning tunneling microscopy (STM), reflection-absorption infrared spectroscopy (RAIRS), ellipsometry, contact angle measurements, and X-ray photoelectron spectroscopy (XPS). This chapter aims to provide readers with the context of the present work. It starts with an overview of self-assembled monolayers (SAMs), followed by an outline of the history of the self assembly study and a discussion of the scientific significance of these thin films. The motivation and objective behind the present work is explained. Finally, the chapter concludes with an outline of this dissertation. 1.1. Overview of Self-Assembled Monolayers The study of self assembly began with the work of Irving Langmuir and Katharine Blodgett with monolayers of amphiphilic molecules such as soap or fatty acids on the surface of water};2 Langmuir and Blodgett were able to deposit these films on a solid surface by dipping a solid slab in water that was precovered with a monolayer of the amphiphilic molecules, causing the hydrophobic part of the molecule to adhere to the surface. By dipping a surface throu- 0101 films 56-3” tile L1 .b 1 1') kin Ivias through the monolayer several times, one can deposit multi-layers of alternating orientation.3 Langmuir-Blodgett (LB) films are the first examples of multilayer films with highly ordered layer structure and controlled layer-by-layer growth. However, because of the delicate nature of the van der Waals forces linking each layer and the relatively complex procedures for layer formation, LB films have been limited to mostly studies of layered assemblies where robustness is not an issue.4 The recently developed route to building thin films with molecular orientation is through the use of self-assembled monolayers (SAMs). In contrast to LB films, which are formed in a “mechanical” process described above, SAMs are produced by spontaneous assembly in solution or vapor deposition. This class of thin films is characterized by the presence of a specific headgroup/substrate interaction, and noncolvalent intermolecular forces creating the layer order. To date, a variety of assemblies used in SAMs have been reported45 and a few examples are presented in Table 1.1. Table 1.1 Representative substrates and ligands that form SAMs L Substrate Ligand Bonding RSH Au, Ag, Cu, Fe RSSR’ RS-metal RSR’ SiOz, glass RSICIa Siloxane, Si-O-Si Sl/Si-H (RCOO)2 R-Si L‘ metal oxides RCOOH RCOOH-MO" T 2ro2 RP03H2 RP032---Zr'v j l: i,”- fiat: (it: va The first group of SAMs listed in Table 1.1 are the most widely studied films made of organosulfur compounds. In 1983, Nuzzo and Allara reported that monolayers of n-alkanethiols on gold could be prepared by adsorption of di-n- alkyl disulfides from a dilute solution.6 Since then, SAMs of alkanethiols on gold have been extensively characterized by various surface analytical techniques and studied by theoretical calculations?27 Structural and chemical studies have shown that the monolayers are highly crystalline and are stable in air and water.528 With the goal of producing technologically interesting materials, researchers have explored other derivatized thiols, including dialkyldisulfides39,30 aromatic thiols,31'34 and aromatic/aliphatic mixed thiols.35'38 Another system that has attracted considerable attention is the alkylsiloxane SAMs.39-42 Indeed, the study of ordered, chemisorbed SAMs began with the report of octadecyltrichlorosilane (OTS) adsorbed on silica and alumina.43 It has been shown that the -Cl groups of OTS react with the -OH groups on the hydroxylated silica surface and a network of Si-O-Si bonds forms at the interface.44 The covalent bonded network gives rise to remarkable chemical and thermal stability. A similar system, in which alkyl chains are directly bonded to silicon, is fabricated via a series of free radical reactions of diacylperoxides and H-terminated silicon surfaces.45 In other work it has been demonstrated that long chain n-alkanoic acid molecules spontaneously adsorb on metal oxides (e.g. A920, CuO, and Al203) forming fatty acid SAMs.45v47 The driving force is the formation of a surface salt between the carboxylate anion and a surface metal cation. The last system tiled 51113 tel EWSC :1" Uri 03‘ listed in Table 1.1 is based on alkyl phosphonic acid reactions with ZrOz surfaces. Phosphonic acids form strong, insoluble complexs with Zr'V ions, and the resulting SAMs have robust physical properties with persistent structural anisotropy.“51 Self-assembled monolayers have received a great deal of attention for their fundamental importance in understanding interfacial properties as well as for their potential applications in molecular technologies. For example, SAMs have been used to study important fundamental processes involving interfacial electron transfer, adhesion, surface wetting, lubrication, and catalysis.52'55 Such films have also been used in the design of various interfaces for chemical sensors, nonlinear optical materials, microelectronics, and computer technology.“—58 Another unique character of SAMs is their synthetic flexibility. A wide range of functional groups can be incorporated in the molecular structure; therefore, a variety of surfaces with specific chemical and physical properties can be prepared via self assembly. The ability to vary the chemical and physical properties of materials on the molecular scale, as well as the ability to measure these changes in molecular environment, will become increasingly important for a variety of current and future applications. 1.2. Motivation and Objective of the Present Work SAMs undoubtedly have a promising future in various technological applications. However, it is important to note that the realization of further advancement of these technological applications requires detailed understanding $11th 111 the! 200' ViliJu 81131 ii l int (‘45 N 0A,, in“ 1» a \l‘: "‘1. of several fundamental issues, such as how the molecular structure of the adsorbates alters the morphology of the films, and how the monolayer properties are related with the structure and subsequent reactions of the films. The present study aims to provide some insight into these questions. The first objective of this work is to improve understanding of how the structural diversity of organic molecules affects the morphology of the monolayers. Even though self-assembled monolayers are relatively easy to prepare, they represent a complex system with various interactions involved in the self-assembly process, including the headgroup-substrate interaction, the endgroup-substrate interactions, the chain-chain interactions and the endgroup- endgroup interactions. The final structure is determined by the overall balance of all interactions involved in the assembly process. By changing the molecular structures, different perturbations will be introduced into the system, and as a result, different packing structures may be formed in the SAMs. For example, in (O-SUDSIIIUIGCI alkanethiol SAMs on Au(111), variation in the chemical nature of the endgroup has been shown to change the monolayer morphology completely.59-53 In many environments, alkanethiol SAMs are quite robust. For example, in water or air at room temperature, their packing arrangements and surface coverages do not change over periods of months. But at extreme pH, in many nonaqueous solvents, in the presence of molecules that compete for surface sites (such as Cl', ON, or other thiols), or at elevated temperatures, SAMs reveal 1113 r. {1.21:1 I .5, VI") "II 011 .l. r. rm V l L'AA ‘ .1 PER” PM HA'L. “UN, 1 I II“) IA 0113‘:- FnI Vail] their relatively weak binding interaction with the surface.64'57 This instability has greatly limited SAMs’ potential technological applications. A second theme of the work presented here is the incorporation of unsaturation into the SAMs, which provides the opportunity for subsequent chemistry either through attachment of another species or by intralayer cross- linking reactions (oligomerization/polymerization). After spontaneous assembly of thiols onto a gold substrate, cross-linking of adjacent molecules can be initiated by chemical, photon or thermal means. Increasing the lateral interaction between the chains by covalent bonding can improve the robustness of the SAMs and overcome the weakness of the simple alkanethiol SAMs. At the same time, the film remains chemically bonded to the surface through the specific thiol- gold bonds. There have been previous studies of photoinduced polymerization of adsorbed monomers on metal surfaces. Ford et al. formed polymerizable monolayers by adsorbing 4-(mercaptomethyl)styrene on a roughened silver surface.68 A 514 nm laser light was used to both initiate and probe photoreactivity through surface-enhanced Raman spectroscopy (SERS), and a polymerization reaction was observed. Peanasky and McCarley have studied undec-10-ene-1-thiol/Au SAMs irradiated by y-rays.59 It was proposed that the polymerization reaction was controlled by the distance that the tethered olefin groups were able to move. Several groups have studied UV irradiation of self- assembled monolayers containing diacetylene units.7°'74 The incorporation of conjugated diacetylene groups within thiol or disulfide compounds has permitted 11 se the fabrication of robust monolayer polymers that are more durable and better barriers to electron transfer than the unpolymerized monolayers. The above work has demonstrated that it is possible to polymerize specific molecules based on self-assembled monolayers. In the present study, four groups of adsorbates have been selected, which represent aliphatic, aromatic and aliphatic/aromatic mixed thiol SAMs on Au(111), as shown in Scheme 1.1. C18 MTPT CTPT C8 C10 DDDS PHT VHT __hS Scheme 1.1 The selected adsorbates systems for investigation in the present study. Octanethiol (C8) and octadecanethiol (C18) were Chosen as examples of room-temperature liquid and solid alkanethiols which form a \l3xV3R30° close packing structure on Au(111). The olefin-terminated di(9-decene) disulfide (DDDS) monolayer was chosen to investigate the impact of a small substituent (olefin terminus) on the monolayer structure, and to provide a better understanding of the reactivity of surface-confined olefin groups. Decanethiol (ClO) monolayer was studied as the control in this case. close 110 pl to Sit The phenyl-terminated 6-phenyl-n-hexanethiol (PHT) molecule was used to examine the influence of an aromatic terminus on monolayer packing and order. The aromatic group is much less flexible than an alkyl chain, and the presence of a bulky aromatic group may sterically prevent the formation of the close-packed monolayer structure. The addition of an alkene to the phenyl group (to produce the styrene-terminated 6-vinylphenyl-n-hexanethioI, VHT) allowed us to study the thermal stability and photopolymerization processes. The p-methylterphenyl thiol (MTPT) and 2’,3’-cyclobutyl p-terphenyl thiol (CTPT) molecules were selected to deduce the structure of rigid oligophenyl thiol SAMs. The dominant interactions between aromatic moities include face-to—face or edge-to-face n-bonding contributions in addition to the simple van der Waals’ forces believed to dominate the packing of alkyl chains.75'77 Self-assembled monolayers based on aromatic molecules have been proposed for nanoscale electronic and optoelectronic applications, thus, an understanding of the structural, electronic and optical properties of materials and devices based on oligophenyls is essential. In addition, the benzocyclobutene units in CTPT monolayers can be thermally or photochemically crosslinked in a mechanism similar to a Diels-Alder—type reaction. 1.3. Outline of the Dissertation As a foundation for understanding the experimental findings described in this dissertation, the techniques employed will be summarized in the next chapter. The operation principles of scanning tunneling microscopy (STM) will be presented first, followed by a discussion of issues important to good gene was performance of the STM. The principles of reflection-absorption infrared spectroscopy (RAIRS) and X-ray photoelectron spectroscopy (XPS) will be highlighted. Furthermore, ellipsometry and contact angle measurements will be briefly described. Chapter 3 details the preparation and characterization of a gold single crystal and gold films on mica by STM and XPS. The goal of this work was to generate large, atomically-flat gold surfaces as substrates for later SAM study. It was found that the gold single crystal was dominated by a rippled topography after sputtering and annealing. In contrast, gold films produced by evaporating gold onto mica showed large flat (111) terraces bounded by monatomically high steps (2.5 A). Using time-lapse topography, self-diffusion of gold was observed at room temperature. Chapter 4 reports the study of octanethiol (08) and octadecanethiol (C18) SAMs with multiple surface-sensitive techniques. The formation, decomposition and fundamental properties of these films were assessed using electron spectroscopy, scanning probe microscopy, and optical spectroscopy. These complementary techniques enabled the measurements of critical film properties such as geometric order, molecular orientation, defect sites, chemical structure, and film thickness, providing a solid foundation for later work. In Chapter 5, the results of a comprehensive study of PHT and VHT SAMs on Au(111) are presented. The packing order and structural Changes of the PHT monolayer were investigated at room temperature and following annealing in ultrahigh vacuum. Three different stripe phases (5, x’, and [3), characterized by mole Sill. mole mon- pack 5th (1 Nat! dill: lot 193: clef is: Far“ molecular axes oriented almost parallel to the surface plane, were observed by STM. In contrast, the VHT monolayer had a structure in which the average molecular tilt angle was close to the surface normal. Polymerization of the VHT SAM, as followed by RAIRS, was achieved by either UV-light irradiation or thermal treatment. Chapter 6 describes a molecular scale investigation of MTPT and CTPT monolayers adsorbed on Au(111). At room temperature, MTPT formed densely- packed monolayers on gold with the molecular axes slightly tilted away from the surface normal. Molecular resolution images of the MTPT monolayer revealed a (V3xV3)R30°-like packing with slightly larger lattice vectors than typical alkanethiol monolayers. MTPT monolayers completely desorbed after annealing to 45 °C. In contrast, CTPT formed a disordered, island structure on gold surfaces and the resulting films were more thermally stable than MTPT SAMs. Chapter 7 details the results of a study of monolayers of di(9-decene) disulfide (DDDS) on Au(111). It was found that the film structure depended strongly on the polarity of the solvents used in the film preparation. The thermal stability of ordered DDDS monolayers was demonstrated to be similar to that of decanethiol SAMs. The reactivity of the olefin-terminated DDDS monolayers was tested by exposing them to different wavelengths of UV light. The reactivity of olefin terminated SAMs was further compared to the styrene terminated films discussed in Chapter 5. Finally, Chapter 8 concludes this thesis with a summary of the results and recommendations for future work. 10 N911 1.4. Literature Cited (1) Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848. (2) Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007. (3) Gains, G. L. Insoluble Monolayers at Liquid-Gas lnterfces, lnterscience: New York, 1966. (4) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir- Blodgett to Self-Assembly, Academic Press: San Diego, CA, 1991. (5) Ulman, A. Chem. Rev. 1996, 96, 1533. (6) Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481. (7) Leung, T. Y. B.; Gerstemberg, M. C.; Lavrich, D. J.; Scoles, G.; Schreiber, F.; Poirier, G. E. Langmuir 2000, 16, 549. (8) Eisenberger, P.; Fenter, P.; Liang, K. S. Phys. Rev. Lett. 1993, 70, 2447. (9) Alves, C. A.; Smith, E. L.; Porter, M. D. J. Am. Chem. Soc. 1992, 114, 1222. (10) Delamarche, E.; Michel, B. Thin Solid Films 1996, 273, 54. (11) Poirier, G. E.; Pylant, E. D. Science 1996, 272, 1145. (12) Poirier, G. E. Langmuir1999, 15, 1167. (13) Yamada, R.; Uosaki, K. Langmuir1998, 14, 855. (14) Toerker, M.; Staub, R.; Fritz, T.; Schmitz-Hubsch, T.; Sellam, F.; Leo, K. Surf. Sci. 2000, 445, 100. (15) Hemminger, J. C.; Feher, F. J.; Dishner, M. H. Langmuir1997, 13, 2318. (16) Camillone III, N.; Eisenberger, P.; Leung, T. Y. B.; Schwartz, P.; Scales, G-i Poirier, G. E.; Tarlov, M. J. J. Chem. Phys. 1994, 101, 11031. 11 (17) Camillone, N.; Chidsey, C. E. D.; Liu, G.-y.; Putvinski, T. M.; Scoles, G. J. Chem. Phys. 1991, 94, 8493. (18) Strong, L.; Whitesides, G. M. Langmuir 1988, 4, 546. (19) Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Chem. Phys. 1993, 98, 678. (20) Truong, K. D.; Rowntree, P. A. J. Phys. Chem. 1996, 100, 19917. (21) Revell, D. J.; Chambrier, l.; Cook, M. J.; Russell, D. A. J. Mater. Chem. 2000, 10, 31. (22) Hostetler, M. J.; Stokes, J. J.; Murray, R. W. Langmuir1996, 12, 3604. (23) Chidsey, C. E. D.; Loiacono, D. N. Langmuir1990, 6, 682. (24) Porter, M. 0.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc 1987, 109, 3559. (25) Hutt, D. A.; Leggett, G. J. Langmuir1997, 13, 3055. (26) Heister, K.; Allara, D. L.; Bahnck, K.; Frey, S.; Zhamikov, M.; Grunze, M. Langmuir1999, 15, 5440. (27) Leggett, G. J.; Hutt, D. A. Langmuir1997, 13,3055. (28) Fenter, P. In Thin Films: Self-assembled Monolayers of Thiols; Ulman, A., Ed.; Academic Press: San Diego, 1997; . (29) Troughton, E. B.; Bain, C. D.; Whitesides, G. M.; Nuzzo, R. G.; Allara, D. L; Porter, M. D. Langmuir1988, 4, 365. (30) Mowery, M. D.; Menzel, H.; Cai, M.; Evans, C. E. Langmuir1998, 14, 5594. 12 125, 695‘: 1999 (31) Sabatani, E.; Cohen-Boulakia, J.; Bruening, M.; Rubinstein, l. Langmuir 1993, 9, 2974. (32) Kang, J. F.; Ulman, A.; Liao, 8.; Jordan, R.; Yang, G.; Liu, G.-y. Langmuir 2001, 17, 95. (33) Leung, T. Y. B.; Schwartz, P.; Scoles, G.; Schreiber, F.; Ulman, A. Surf. Sci. 2000, 458, 34. (34) Jin, Q.; Rodriguez, J. A.; Li, C. 2.; Darici, Y.; Tao, N. J. Surf. Sci. 1999, 425, 101. (35) Han, S. W.; Kim, C. H.; Hong, S. H.; Chung, Y. K.; Kim, K. Langmuir 1999, 15, 1579. (36) Tamada, K.; Nagasawa, J.; Nakanishi, F.; Abe, K.; lshita, T.; Hara, M.; Knoll, W. Langmuir1998, 14, 3264. (37) Wang, R.; Iyoda, T.; Jiang, L.; Hashimoto, K.; Fujishima, A. Chem. Lett. 1996, 1005. (38) Caldwell, W. 8.; Campbell, D. J.; Chen, K.; Herr, B. R.; Mirkin, C. A.; Malik, A.; Durbin, M. K.; Dutta, P.; Huang, K. G. J. Am. Chem. Soc. 1995, 117, 6071. (39) Li, D.; Rather, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K. J. Am. Chem. Soc. 1990, 112, 7389. (40) Gun, J.; lscovici, R.; Sagiv, J. J. Collid Interface Sci. 1984, 101, 201. (41) Maoz, R.; Sagiv, J. J. Colloid Interface Sci. 1984, 100, 455. (42) Maoz, R.; Sagiv, J. Thin Solid Films 1985, 132, 135. (43) Sagiv, J. J. Am. Chem. Soc. 1980, 102,92. 13 1145) (16,: 147i let: (44) Wasserrnan, S. R.; Tao, Y. T.; Whitesides, G. M. Langmuir1993, 9, 1749. (45) Linford, M. R.; Chidsey, C. E. D. J. Am. Chem. Soc. 1993, 115, 12631. (46) Allara, D. L.; Nuzzo, R. G. Langmuir1985, 1, 52. (47) Schlotter, N. E.; Porter, M. 0.; Bright, T. B.; Allara, D. L. Chem. Phys. Lefl;1986,132,93. (48) Hong, H. G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3, 521. (49) Vermeulen, L.; Thompson, M. E. Nature 1992, 358, 656. (50) Lee, H.; Kepley, L. J.; Hong, H.; Mallouk, T. E. J. Am. Chem. Soc. 1988, 110,618. (51) Bryd, H.; Snover, J. L.; Thompson, M. E. Langmuir1995, 11, 4449. (52) Chidsey, C. E. D. Science 1991, 251, 919. (53) Ulman, A.; Evans, S. D.; Shnidman, Y.; Sharma, R.; Eilers, J. E.; Chang, J. C. J. Am. Chem. Soc. 1991, 113, 1499. (54) Kim, S.; Choi, G. Y.; Ulman, A.; Fleischer, C. Langmuir1997, 13, 6650. (55) Berman, A.; Steinberg, S.; Campbell, 8.; Ulman, A.; lsraelachvili, J. Tribal. Lett. 1998, 43. (56) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Lstsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078. (57) Berggren, K. K.; Bard, J. L.; Wilbur, J. D.; Gillaspy, A. G.; Helg, J. J.; McClelland, S. L.; Rolston, W. D.; Philis, M.; Prentiss, M.; Whitesides, G. M. Science 1995, 269, 1255. (58) Katz, H. E.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994, 116, 6636. 14 12, ‘ {13,9 WV, 11791 (59) Kawasaki, M.; Sato, T.; Yoshimoto, T. Langmuir2000, 16,5409. (60) Pfaum, J.; Bracco, G.; Scoles, G.; Lee, R.; Kahn, A. Am. Vac. Soc. Int. Symp. Abstr., 2000, 55. (61) Poirier, G. E.; Pylant, E. D.; White, J. M. J. Chem. Phys. 1996, 105, 2089. (62) Boubour, E.; Lennox, R. B. Langmuir2000, 16,7464. (63) Yan, L.; Marzolin, C.; Terfort, A.; Whitesides, G. M. Langmuir1997, 13, 6704. (64) Schneider, T. W.; Buttry, D. A. J. Am. Chem. Soc. 1993, 115, 12391. (65) Chidsey, C. E.; Bertozzi, C. R.; Putvinski, T. M. J. Am. Chem. Soc. 1990, 112,4301. (66) Zamborini, F.; Crooks, R. M. Langmuir1998, 14, 640. (67) Schoer, J. K.; Ross, C. B.; Crooks, R. M. Langmuir1994, 10, 3319. (68) Ford, J. F.; Vickers, T. J.; Mann, C. K.; Schlenoff, J. B. Langmuir1996, 12,1944. (69) Peanasky, J. 8.; McCarley, R. L. Langmuir1998, 14, 113. (70) Mowery, M. D.; Kopta, S.; Ogletree, D. F.; Salmeron, M; Evans, C. E. Langmuir1999, 15,5118. (71) Cal, M.; Mowery, M. D.; Menzel, H.; Evans, C. E. Langmuir1999, 15, 1215. (72) Menzel, H.; Mowery, M. D.; Cai, M.; Evans, C. E. J. Phys. Chem. B1998, 102,9550. (73) Kim, T.; Ye, 0.; Sun, L.; Chan, K. C.; Crooks, R. M. Langmuir1996, 12, 6065. 15 (74) 1995 175) (74) Chan, K. C.; Kim, T.; Schoer, J. K.; Crooks, R. M. J. Am. Chem. Soc. 1995, 117, 5875. (75) Whitten, D. G.; Chen, L.; Geiger, C.; Perlstein, J.; Song, X. J. Phys. Chem. B1998, 102, 10098. (76) Song, X.; Geiger, C.; Farahat, M.; Perlstein, J.; Whitten, D. G. J. Am. Chem. Soc. 1997, 119, 12481. (77) Vadey, S.; Deiger, H. C.; Cleary, B.; Perlstein, J.; Whitten, D. G. J. Phys. Chem. B 1997, 101, 321. 16 Sim: 9105 test XPS Qira Eta Chapter 2 Experimental Techniques Multiple surface sensitive techniques were chosen to elucidate the structure of self-assembled monolayers on Au(111) because no one experimental technique can provide complete characterization. Optical ellipsometry, contact angle measurements, X-ray photoelectron spectroscopy (XPS), and reflection-absorption infrared spectroscopy (RAIRS) were selected to probe the macroscopic structure of the film; scanning tunneling microscopy (STM) was applied to probe the microscopic structure. The value of this multitechnique approach is that each measurement probed the structure of the monolayer by a different physical process, providing complementary information. Optical ellipsometry was used to assess the relative thickness of the monolayer films, whereas the more subtle details of the structure were addressed by other techniques. Contact angles were measured to examine the wettability, while XPS was applied to determine the elemental composition of the film quantitatively and confirm molecular adsorption. RAIRS was used to characterize the chemical identity and average spatial orientation of the monolayer films, whereas STM revealed the atomic scale arrangement of the monolayers, including film morphology, defect densities and chain directions. 17 be pr isolal print unde 2.1 Sill lial ill-re lei rat In this chapter, the operational principles and tunneling current of STM will be presented. Issues important to good performance of STM, including vibration isolation, image processing, and tip preparation, will be discussed. The principles of other structural probes, such as RAIRS and XPS are well understood“3 and only the relevant features will be highlighted here. Ellipsometry and contact angle measurements will be discussed briefly. Readers who are interested in more details are referred to the reviews and textbooks listed in referencesfl‘7 2.1 Scanning Tunneling Microscopy In the two decades since its invention, STM has revolutionized surface science and the way scientists study surface phenomena. The emergence of STM and the techniques it has inspired make it possible to investigate many aspects of material properties at the atomic scale. Principle of operation. The advantage of STM compared with other microscopes, such as scanning electron and transmission electron microscopes (SEM, TEM), is that there is no need for lenses or external electron sources. STM relies solely on electron tunneling between tip and sample, a phenomenon that is based on quantum mechanics. The electrons existing in the sample under investigation serve as the exclusive source of radiation. In a metal or semiconductor, electrons exist within an energy range, designated by the shaded areas in Fig. 2.1. At the interface between that material and an insulator, such as a vacuum, there is an energy barrier. If a second metal or semiconductor is placed near the first one and a voltage is 18 impo there céass 8961) less dista la.) imposed between the two, the slope of the energy barrier will be changed and there will be a driving force for electrons to move across the barrier. According to classical mechanics, the electron cannot travel across the barrier unless its energy is raised above the barrier energy, whereas in a quantum mechanical description, a small numbers of electrons can tunnel through the barrier if the distance 2 is small (Fig. 2.1). Evac -_ I (a) Tunneling Barrier E Tip Sam to F 5F 9 (b) TI EFP Sam la EF P + Fig. 2.1 The energy level diagram of a tip and a sample (both are metals) in a vacuum, (a) when the tip and the sample are not connected; (b) when the tip and the sample are sufficiently close and a voltage is applied between them. (adapted from Tersoff and Langa) Figure 2.1 (a) shows that when tip and sample are not connected, their vacuum levels, which are the reference energy levels, are equal. The Fermi levels E1: of the two materials differ by an amount equal to their work function difference. When tip and sample are sufficiently close and a voltage is applied between them, a tunneling current will occur. In this example, the Fermi level of 19 the li) dorm lipl beta) lnslez local small med the tip is moved upward (negatively biased) while that of the sample is moved downward (Fig. 2.1 (b)). The electrons will tunnel from the occupied states of the tip to the unoccupied states of the sample. STM images should not be interpreted as simple topography of surfaces, because STM senses occupied or unoccupied states near the Fermi level. Instead of a physical topography, STM images represent spatial variations of the local electronic structures and the tunneling barrier at the surface. Tunneling current. The STM imaging process is shown in Fig. 2.2. A sharp metal tip is brought within proximity of a conducting surface while a voltage is applied between them. The barrier separation between tip and sample is so small that the wavefunctions will overlap and electrons can tunnel quantum mechanically through the barrier. i2 0000 500000000 000000000 Fig. 2.2 STM imaging process (adapted from Chen9) 20 CUITE 5110 Sam (,3 1 ( ) T“ I”: i; The probability that an electron will cross the barrier is the tunneling current (I), and it decays exponentially with the barrier width as shown in the simplified expression: I = Cptpsexp(-1.0252"2) (2.1) where z is the barrier width (sample-tip distance), (I) is the average work function of sample and tip, pt and p5 are the densities of states of tip and sample, respectively, and C is a constant.10 The current is exponentially dependent on the sample-tip separation. For typical metal tips and samples (<1) e- 5 eV), a 1 A decrease in the sample-tip separation will Increase the current by one order of magnitude.11 With this level of sensitivity, the tunneling current can be used to control the sample-tip separation with high vertical resolution. STM can be operated in two modes: constant current or constant height mode. In constant current imaging, a feedback mechanism is enabled so that when the tip is scanned across the surface, the tunneling current is maintained constant at a preset value by continuously adjusting the vertical tip position. The height of the tip as a function of the position, which represents a constant Charge density contour of the surface, is read and converted into the surface image. The constant current mode can be used to track surfaces that are not necessarily flat on an atomic scale, such as stepped surfaces. In constant height mode, the tip travels in a horizontal plane above the sample and the tunneling current varies depending on the local surface electronic properties of the sample. The tunneling current is recorded and converted into an image of the surface. The scan rate is faster in constant height mode since 21 no It surfa- al hig post those afeo SUCCI )uocl; micr 93911" no feedback loop is needed. However, this mode is feasible only on flat surfaces; otherwise, the tip might crash into a surface protrusion while scanning at high speed. Vibration isolation. Since the STM is measuring and controlling the tip position to less than an atomic diameter, even the smallest vibrations, such as those caused by sound in air and by people walking around in a building, can affect the stability of the instrument. The design of the STM and ultimately its success depend on damping vibrations. Vibrations that reach the sample-tip junction originate from the building, the table/chamber system on which the microsc0pe sits, and possibly from components of the microscope itself. To minimize the sensitivity of the STM to building vibrations which peak between 10 and 100 Hz in the frequency spectrum, our STM sits on a table that is mounted on air legs. The sputter-ion pump is below table level, helping to lower the center of mass of the system. An additional decrease of the vibration amplitudes was achieved by moving the STM to the sub-basement level. Image processing. To extract information from image data, acquisition artifacts must be removed. There are two common problems in the data. First, it is virtually impossible to bring a tip exactly perpendicular to a sample plane; therefore, all images include a plane containing this deviation. This problem can be solved by using “x slope correct” and “y slope correct” software commands, respectively, which perform an image background subtraction line by line in the x or y direction. Figure 2.3 shows the effect of slope subtraction on highly oriented pyrolytic graphite (HOPG) image data. The original data had slope integrated in 22 (t) al both x and y directions, which obscured the real image information. After slope subtraction, the atomic resolution was clearly shown. Fig. 2.3 Image processing on a HOPG surface (44 x 44 A). (a) the original data; (b) after x- and y-slope subtractions. The second most common problem is high-frequency noise, which originates from digitizing noise, adsorbates on the sample surface or resonance of the piezoelectric components and appears as bright lines in the image data. Since the format of the image data is usually a square array of numbers in which the array position corresponds to x, y coordinates and the number represents the signal intensity at that position, the extreme value pixels (usually corresponding to high frequency noise) can be replaced by the median value of near neighbors. This process is called “median filter". Figure 2.4 (a) shows a two-layer HOPG surface with high-frequency noise appearing mainly near the step edge and originating from electronic noise generated when the tip position was suddenly changed. With "median filter" (Fig. 2.4 (b)), most of the high frequency noise was smoothed out. 23 Fig. 2 data; aria from inlet the d This he! Fig. 2.4 Image processing on a HOPG surface (2000 x 2000 A). (a) the original data; (b) after applying median filter. The slope subtraction and median filter operations have removed only artifacts, not image content. It is also possible to choose a frequency to eliminate from the image, which can be used when interrnediate-frequency electrical noise interferes with the data. This is done by calculating a 2-D Fourier transform of the data, selecting the noise to be removed, and then back transforming the data. This approach is extremely effective for periodic data. However, a caution is that the image content can be removed simultaneously, and artifacts can be introduced into the data. Thus great care must be taken in applying the “Fourier filter” and the details should be reported following the figures. Tip preparation. A frustrating aspect of STM operation is the reliable formation of tunneling tips. The geometry and chemical identity of the tunneling tip influence not only the resolution of a STM scan, but also the measured electronic structure. There are three basic requirements of tips for reliable STM operation: short length, sharp apex, and high purity. Because vibration isolation is a critical factor in microscope design, a tunneling tip should be rather thick (0.2 24 to 1. unit: imag star; ghos reooi COflE BIC): to 1.0 mm) and come to a point rapidly. Short length from the apex to the unetched bulk is necessary to minimize mechanical vibrations that will blur the image. Another critical factor for tip performance is the tip’s microstructure. A sharp apex avoids the problem of multiple tunneling microtips which will produce ghost images. Finally, the chemical purity and homogeneity of the tip surface is required to eliminate tip artifacts which will lead to variations in conductance. Many ways to produce STM tips have been developed. The following contain the most familiar methods: (1) mechanical shaping (grinding,12 cutting,13 etc); (2) electrochemical etching;l4rl7 (3) ion milling;‘8-19 (4) field evaporation;20 (5) electron-beam deposition;21 (6) heat treatment in vacuum?2 Among these methods, electrochemical etching and cutting were chosen as two easier and faster ways to prepare tips. Tungsten and Platinum/Iridium (Pto_g/Irg,2) alloys were used as the materials. Electrochemical Etching. The apparatus used for the electrochemical etching is shown in Fig. 2.5. The etching circuit included the wire, an electrolytic solution of 2 mM NaOH, a stainless steel counter electrode, a variable resistor, an ac power supply and an ammeter. 25 6 1k!) 0 Micromanipulator Stainless Steel 7 Electrode P10 3/"02 r \i or W wire 2 M NaOH .Wr Fig. 2.5 Experimental setup for electrochemical etching Three different methods to prepare the wire before etching are shown in Fig. 2.6. All three ways are based on the "drop-off“ methods developed by Bryant et al.23 The side of the wire will etch more quickly than the bottom and due to the "necking-in" effect, the portion of the wire in the etchant drops off when its weight exceeds the tensile strength of the necking-in region of the wire.24 When the bottom part falls off, the upper part (the tip) separates from the liquid, and the current drops to zero. In this manner, the tip is not damaged even if it is not removed immediately from the preparation cell. (a) M G} G} Epoxy Insulation Epoxy Insulation Fig. 2.6 Sketches of wire preparation for electrochemical etching. (a) bare wire method; (b) epoxy insulation method; (0) tube insulation method. 26 will) 1 Born 0011) " | 3.010 Figure 2.6 (a) shows the bare wire etching method.15 After being washed with ethanol, a piece of tungsten or Ptn,g/Iro,2 wire (0.2 mm) was vertically dipped 8 mm into 2 mM NaOH. The other end of the wire was inserted into a copper connector which was anchored to a micromanipulator. Figure 2.6 (b) shows the epoxy insulation method:16 instead of being inserted into the etchant directly, the end of the wire (~1 mm) was insulated with epoxy. The wire was then immersed in the etchant so that a small (3 0.5 mm) region just above the insulation was brought into contact with the liquid. The exposed region was etched until the insulated lower portion dr0pped, leaving a sharp tip. Another way of wire protection is tube insulation method shown in Fig. 2.6 (c):17 the wire was inserted into a plastic tube (inside diameter 0.2 mm, outside diameter 2 mm), approximately 4 mm long. The end of the tube was sealed with epoxy to protect from etching. This masking procedure ensured that the etching action occurred only at the meniscus position. When the partially covered wire was Immersed into the electrolytic solution, a high etching current was produced. Then the wire was slowly withdrawn with the micromanipulator until a desired current value (ls) was obtained. With the variable resistor, the current was lowered to an etching current (1,.) in order to decrease the large number of bubbles. The exposed region was etched until the insulated portion dropped. The etching conditions and tip results for these three methods are summarized in Table 2.1. 27 Tall and TI nnnqtnn 1"in r30/Ir Tiru Table 2.1 Comparison of three different etching methods to prepare tungsten and Ptogllrog tips. Bare Wire Epoxy Insulation Tube Insulation Method Method Method Volta 3 (\fl 3.5 6 — 7 6 —6.5 .9 Current (mA) 50 - 60 30 -35 '3: 6° ‘ 50 I— I,,: 20 — 25 c . . _— 9 Etching Time _ _ _ 8) min) 20 30 10 15 5 10 S conical shape exponential shape exponential shape 9‘ Tip Shape with curvature with curvature radius with curvature radius radius 3 100 nm s 50nm s 50nm Voltage (V) 20 6.5 6 a Current (mA) 300 - 400 40 :33 23 i: e- E Etching Time (hr) 2 1 1.5 Blunt with Long tall with . . . Tip Shape curvature radius curvature radius of 5 bags: 2;": 1c 36" :2" of 4 p._rn 100 nm ‘ All three methods described above produced reasonably good tungsten tips. Figure 2.7 shows the corresponding SEM images of a tungsten tip obtained by tube-insulation method with Is of 60 mA and I6 of 25 mA. The radius of curvature of the tip was estimated to be less than 50 nm. _40.um _ —300 um _ Fig. 2.7 SEM Images of a tungsten tip fabricated with the control system shown In Fig. 2.5 and 2.6 (0). None of the etching conditions were able to produce good-shape PIQB/Irog tips. probably because Ptgs/lrog wires were inert to oxidation under the same 28 cont imag lip 11 condition and prolonged etching produced long tails. Figure 2.8 shows SEM images of a Pto_3/lrg,2 tip prepared by epoxy insulation etching. The apex of the tip was quite sharp, but the long tail would lead to mechanical vibrations and the tip was not applicable to STM imaging. _ 250nm _ Fig. 2.8 SEM Images of a PIQa/Il'oz tip fabricated with the control system shown in Fig. 2.5 and 2.6 (b). Although most tungsten tips looked almost perfect by SEM, some of them produced unstable tunneling current and consequently poor quality images. The reason could be that the etching dissolution of tungsten wire involves the surface oxidation during the process. As a result, tungsten tips are usually coated with a thick layer of tungsten oxide (WO;;),25.26 which will cause instability in the STM tunnel junction. In contrast, Ptog/lrog tips are more chemically stable. Since etching did not produce good shaped Pto,g/Iro_2 tips, the PIO_3/Il'o_2 tips were manually cut instead. Cutting. Ptog/lrog tips were cut, at almost 90° angle, with a clean pair of cutters (McMaster-Carr, IL, catalog # 5445A13). Cutting tips was simple and fast, with no oxidation/contamination concerns. The major problem was that there was less control of the gross shape of the tip. The formation of two or more 29 micr In or with goor 9105 are: som with coo hail microtips was quite common, which would display multiple imaging of fine details. In order to have a better control, all tips were checked by an optical microscope with 70x magnification before use. By this way, 80% of the tips cut produced good scans. Figure 2.9 (a) shows a freshly cut sharp Pto,3/lrg_2 tip. Although the gross aspect ratio was not as good as that of the etched tungsten tips, the tip apex radius was quite sharp (less than 100 nm). The scanning process sometimes modified the tip end, either due to contamination or accidental contact with the surface. A used tip is shown in Fig. 2.9 (b). The radius of the apex cunrature increased to about 1 pm. Figure 2.9 (c) shows a crashed tip after having contact with the surface. The tip end was bent over. ‘ ‘krz' . _ 2 um ._ 10 um _25 Um— Fig. 2.9 SEM Images of cut PIQa/Irog tips: (a) a freshly cut tip; (b) a used blunt tip; (c) a crashed tip. STM l-V spectra can be used to estimate tip condition. Figure 2.10 shows a series of STM l-V spectra on a gold surface with the tip condition changed from clean to dirty. 30 tori iii a J 1.; lets Tunneling current (nA) Sample Bias (V) Fig. 2.10 A series of STM l-V spectra over a gold sample showing the gradual tip degradation. Figure 2.10 (a) shows a typical behavior in a metal-vacuum-metal tunneling junction. At low biases (-0.5 to +0.5 V) the curve showed a linear ohmic behavior, indicating a continuous density of state exists at the Fermi level. As the voltage became more comparable to the work function voltage, the curve became nonlinear. Deformation of I-V curves indicated the degradation of tips as shown in Fig. 2.10 (b) to (d). As the tip became dirtier, a zero conductance region appeared at low biases (characteristic of semiconductors), and the noise level increased as well. HOPG Imaging. It is difficult to judge the quality of a tip even though the tip may look perfect by SEM. The quality of a good tip is usually evaluated as one which gives stable atomic-resolution images of HOPG in air. HOPG can be easily cleaved to provide atomically flat surfaces. Figure 2.11 presents STM 31 imag unde nor that imag :3" ((2 (C) L_ I) . V ,2: (I) ""V 5995 ill 1- "El images of HOPG obtained in air using cut Ptog/lrgz tips. The current was 1 nA under an applied voltage of 100 mV (tip positive). The images showed the normal periodicity of atomic alignment on a graphite surface, which confirmed that the cut Ptog/Iroz tips were sharp enough to produce atomic resolution images. Fig. 2.11 STM images of HOPG obtained in air using cut PIQe/Irog tips. (a) 190x190 A; (b) 90x90 A; (c) 50x50 A; (d) 20x20 A. All images have been slightly filtered. In summary, STM has several advantages compared with other surface sensitive techniques. Unlike XPS and RAIRS (discussed below) which probe the averaged properties of a large surface area, STM provides three-dimensional real space images and allows spatially localized measurements of structure and properties of surface. The lateral and vertical resolution can reach 1 A and 0.1 A, 32 1950 as i lmm mall Sill inlol 51101 that 22 com anal ha: 01 ( ihol tori nit , Ill Ie' respectively. In addition, STM can be performed in different environments, such as in vacuum or air, at low or high temperatures. Samples can even be immersed in water or other solutions under potential control.27 There are a few limitations of STM: it is not able to image insulating materials, it is blind to the actual chemical nature of the adsorbed species since STM only responds to the electron density, and it cannot provide bonding information or probe the subsurface structure. Furthermore, because of the strong interaction between the tip and the surface, there is always the possibility that the surface properties will be altered by the measurements. 2.2 X-ray Photoelectron Spectroscopy (XPS) XPS provides qualitative and quantitative information about the elemental composition of surfaces with a probing depth of up to ~100 A.1 This technique analyzes photoelectrons emitted from the surface or near-surface regions when irradiated with soft X-rays. The emitted electrons can be detected as a function of energy and angle of emission. Because the kinetic energy of the photoelectron is a “fingerprint” of the elements present and their chemical environment, XPS provides elemental specificity and a measurement of the chemical environment of the elements. In this research, XPS was used to check substrate cleanliness, identify the elemental composition of SAMs, confirm the molecular adsorption, and provide information on layer desorption, dissociation and oxidation. XPS peak intensities and areas can be used as analytical parameters for quantitative determination of elemental compositions; however, extreme care must be taken for 33 meas the tl Sill orien the l hour with is 101 dame measurements with SAMs because the X-ray irradiation can induce damage in the thin films and the quantitative information may be misleading. Several groups have studied X-ray induced damage in alkanethiol SAMs.28'31 It has been demonstrated that irradiation causes the loss of orientational and conformational order, desorption of film fragments, reduction of the thiolate species, and the appearance of new sulfur species. The irradiation- induced new sulfur species are identified as dialkyl sulfide moieties associated with C--S-C links between radicalized alkanethiol SAM fragments. The damage is identical to that caused by electron bombardment, implying that most of the damage is produced by the photoelectrons and secondary electrons.31 In order to reduce X-ray irradiation damage to our films, the X-ray spot was moved to a new position on the sample after every 30 scans (~5 min) for S 2p spectra.” 2.3 Reflection-absorption Infrared Spectroscopy (RAIRS) While XPS affords information on the elemental composition of surface layers, RAIRS can elucidate the structure and orientation of the layers by providing information about average projections of particular dipole moments along the surface normal direction. This selection rule and the high sensitivity of RAIRS are the results of the Fresnel enhancement of the perpendicular field component (p) of the electric field and the cancellation of parallel field component (s) on metal interfaces (Fig. 2.12). 34 Fig. to 5 meta nom com) lite inter relat rode Fig. 2.12 The electric field at a metal surface for a high angle of incidence and for s- and p-polarized light. Only the p-polarized component survives at the metal surface, resulting in an anisotropic electric field Ep parallel to the surface normal. The chemical and structural analysis of RAIR spectra often relies on the comparison with bulk spectra. The binding modes of the adsorbates are easy to determine by the absence or enhancements of certain vibrational bands intensities, relative to the bulk. Deviation in peak locations, line widths, line shapes, and intensities from the behavior of the bulk reflect the local environment of the adsorbate molecules. The infrared spectra of alkanethiols have been studied in detail, and the relationship between peak position and mesoscopic structure is well understood.32-35 In particular, the peak positions and absorbances of the methylene stretching modes provide insights into the local environment and molecular orientation of the alkyl chains. Table 2.2 shows the peak positions for the liquid and solid bulk alkanethiols and ordered octadecanethiol SAM on Au(111). 35 solid simil inter and ' 01 Illl also legit {EEl com; 2.4 ha: ; Table 2.2 Comparison of liquid and crystalline state symmetric (vs) and asymmetric (v85) CH2 stretching modes with octadecanethiol SAM on gold. Modes / (cm'1) Liquid Solid CH3(CH2)178H on Alkanethiol Alkanethiol Au(1 1 1) Va, CH2 2924 2918 2917 vs CH2 2855 2851 2850 The peak positions shift to higher energies in the liquid phase relative to solid due to an increase in the freedom of the hydrocarbon modes.36 The similarity between the solid phase and SAM indicates that a similar distribution of intermolecular environments exists for the CH2 group in both the crystalline bulk and the monolayer phase. Thus these modes can serve as qualitative indicators of the packing of the alkyl chains. There are several limitations of RAIRS. The lower energy limit of detection is on the order of 800 cm", and therefore the information from adsorbate-substrate vibrations, which predominantly occur in the low frequency region of the spectrum, is inaccessible. Electron energy loss spectroscopy (EELS) can provide this vibrational information but, due to experimental complexity, no EELS measurements have been taken. In addition, for a non- metallic substrate, the parallel s-component of the electric field is not completely canceled out and will complicate the orientation information. In this research, gold films on silicon wafers are used to minimize the substrate concern. 2.4 Ellipsometry Ellipsometry is a sensitive surface and thin film measurement technique that uses polarized light. When a polarized light beam reflects from any surface, 36 than PW Ellips meas ratio com; highl) neas olthe Slips. tailor the 91‘ and 9 (’16 3 Wire so; "1%. (9:5: changes occur in both the amplitudes and phases of the oscillating parallel and perpendicular vector components of the electric field associated with the beam. Ellipsometry experiments determine these amplitude and phase changes by measuring the ellipsometric angles delta (A) and psi (‘1’), which are related to the ratio of Fresnel reflection coefficients Rp and Rs for p- and s-polarized light, respectively, as shown in equation 2.2.4 Rp/Rs = tan(\P)exp(iA) (2.2) where RF) and Rs are defined as ratios of the reflected and incident electric fields (including both the amplitude and complex phase factors) for the p and s components. Because ellipsometry measures the ratio of two values it can be highly accurate. In addition, the Woolam M44 ellipsometer used in this research measures A and ‘P at 44 different wavelengths and further improves the reliability of the results. Figure 2.13 shows the monolayer (three-phase) measurement model for ellipsometry. The optical properties of each medium were completely defined by refractive index and extinction coefficient (n., k;). Both quantities are functions of the wavelength of the light used in the measurement. The ellipsometric angles A and ‘P are functions of the angle of incidence (75° for Woolam M44 ellipsometer), the six parameters (n., k;), and the film thickness d. Since the measurements were made in air, na and kg were known. Before the film formation, the gold substrates were cleaned by UV/Oa and the optical properties (n3, ks) were measured. Thus the goal of ellipsometric measurement and data analysis is to determine the unknowns m, k, and d. 37 allow lhlclo ol 0 fella: Sign. ’99“: fl “1. 25 lily- Ambient (n... k.) d (n1! kt) Film (n3. ks) Substrate Fig. 2.13 Three-phase parallel layer model (from Collin et al.4) Since the ellipsometer measures only two parameters (A and ‘1’), it only allows two variables to be fit at a time, i.e., the optical properties (n., k,) and thickness d could not be obtained simultaneously. Furthermore, for a thin film on a gold substrate, ‘1’ does not change with refractive index or thickness, which reduces the number of parameter to one (A) and allows only one variable to be fit. Based on literature results, a refractive index of 1.5 and extinction coefficient of 0 (transparent film) were assumed for the SAMs.37-39 The variation of refractive index from 1.4 to 1.6 did not change the fitted film thickness significantly. Ellipsometry works best for film characterization when the film thickness is comparable to the wavelength of the light used for measurements. The Woollam M44 ellipsometer uses a white light source (414-736 nm) so to characterize a <20 A thick film with a probe of >414 nm is relatively difficult. For thin layer measurements, the instrumental error is estimated to be 1~2 A.40 2.5 Contact Angle Measurements Contact angle measurement is a classical technique that provides information about surface tension or surface free energy of different materials. It 38 is ob it, Ill. and: unit tens inti is observed that in most instances a liquid placed on a solid surface will not wet it, but remains as a drop having a definite contact angle (9) between the liquid and solid (Fig. 2.14). Gas Liquid I Solid Fig. 2.14 Solid-liquid-gas contact angle Surface tension (7) can be thought of as the energy required to create a unit area of an interface. The relationship between contact angle and surface tension is shown in Young’s equation: ysg = 73. + 71,, cos 9 (2.3) where 759, ys. and 719 are the surface free energies of solid-gas, solid-liquid and liquid-gas interfaces, respectively. If the energy required to create the solid- liquid (51) interface is greater than that required for creation of a solid-gas (sg) interface, the contact angle will be greater than 90° and the liquid is said to not wet the solid. The liquid will bead up on the surface to minimize the solid-liquid interfacial area. Contact angle measurement has been widely used as a preliminary tool in the investigation of SAMs, because the wettability (ability of a fluid to cover a surface) of gold substrates covered with monolayers can be correlated with the quality of the monolayers.32»35 In addition, the wettability varies with the polarity of the monolayer terminal groups. For example, for hydrophobic surfaces, the 39 tree this I and corn free energy decreases in the order —CH2 > -CH3 > —CF2 > —CF2H > -CF3.41 In this research, various SAMs with different terminal functional groups were formed and contact angle measurements provided unique information about both the completeness of the monolayer and its degree of order. In conclusion, a combination of surface analysis techniques were used to obtain composition and structural information of SAMs on Au(111). Each measurement probed the structure of the monolayer by a different physical process (Table 2.3), and together providing a complete picture of SAMs. Table 2.3 Techniques used in the present study for SAMs characterization Property of SAMs Techniques structure and order STM, RAIRS stability and reactivity composition XPS wettability contact angle thickness ellipsometry defects STM 2.6 Literature Cited (1) Barr, T. L. Modem ESCA: the Principles and Practice of X-ray Photoelectron Spectroscopy, CRC Press, Inc.: Florida, 1994. (2) Colthup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and Raman Spectroscopy, Academic Press: Boston, 1990. (3) Perry, 3. 3.; Somorjai, G. A. Anal. Chem. 1994, 66, 403. (4) Collin, A. w.; Kim, Y. T. Anal. Chem. 1990, 62, 887. (5) Adamson, A. W.; Gast, A. P. Physical Chemistry of Surfaces, John Wiley & Sons: New York, 1997. (6) Bohn, P. W.; Walls, D. J. Mikrochim. Acta1991, l, 3. (7) Tengvall, P.; Lundstrom, I.; Liedberg, B. Biomaten’als1998, 19, 407. (8) Tersoff, J.; Lang, N. D. In Scanning Tunneling Microscopy, Stroscio, J. A., Ed.; Academic Press, Inc.: San Diego, 1993; Vol. 27; 2. (9) Chen, C. J. Introduction to Scanning Tunneling Microscopy, Oxford University Press: New York, 1993. (10) Bai, C. Scanning Tunneling Microscopy and Its Applications, Springer- Verlag: Berlin, Germany, 2000. (11) Bonnell, D. A.; Huey, B. D. In Scanning Probe Microscopy and Spectroscopy Theory, Techniques, and Applications, second ed.; Bonnel, D. A., Ed.; Wiley-Vch Inc.: New York, 2000; . (12) Binnig, G.; Rohrer, H.; Gerber, C.; Weibel, E. Phys. Rev. Lett. 1982, 49, 57. (13) Stupian, G. W.; Leing, M. S. Rev. Sci. Instrum. 1989, 60, 181. (14) Liu, A.; Hu, X.; Liu, W.; Ji, G. Rev. Sci. Instrum. 1997, 68, 3811. (15) Olive, A. l.; Romero G., A.; Pena, J. L.; Anguiano, E.; Aguilar, M. Rev. Sci. Instrum. 1996, 67, 1917. (16) Weinsterin, V.; Slutzky, M.; Arenshtam, A.; Ben-Jacob, E. Rev. Sci. Instrum. 1995, 66, 3075. (17) Bourque, H.; Leblanc, R. M. Rev. Sci. Instrum. 1995, 66,2695. 41 (18) Sci. (9 Fact (21) 199i (221 (231 so. 1 124} (25} (18) Vasile, M. J.; Grigg, D.; Griffth, J. E.; Fitzgerald, E.; Russell, P. E. J. Vac. Sci. Technol. 1991, B 9, 3569. (19) Hopkins, L. C.; Griffith, J. E.; Harriott, L. R.; Vansile, M. J. J. Vac. Sci. Technol. B 1995, 13,335. (20) Fink, H. W. IBM J. Res. Dev. 1986, 30, 460. (21) Akama, Y.; Nishimura, E.; Sakai, A.; Murakami, H. J. Vac. Sci. Technol. A 1990, 8, 429. (22) Binh, V. T. J. Microsc. 1988, 152, 355. (23) Bryant, P. J.; Kim, H. S.; Zheng, Y. C.; Yang, R. Rev. Sci. Instrum. 1987, 58, 1115. (24) Melmed, A. J. J. Vac. Sci. Technol. 81991, 9, 601. (25) Colton, R. J.; Baker, S. M; Baldeschwieler, J. D.; Kaiser, W. J. Appl. Phys. Lett. 1987, 51, 305. (26) Gamaes, J.; Kragh, F.; Morch, K. A.; Tholen, A. R. J. Vav. Sci. Technol. A 1990, 8, 441. (27) Bard, A.; Fan, F .-R. In Scanning Probe Microscopy and Spectroscopy, 2nd ed.; Bonnell, D., Ed.; Wiley-Vch: New York, 2000; . (28) Heister, K.; Zhamikov, M; Grunze, M.; Johansson, L. S. 0.; Ulman, A. Langmuir2001, 17, 8. (29) Wirde, M.; Gelius, U.; Dunbar, T.; Allara, D. L. Nucl. Instrum. Methods Phys. Res. 81997, 131, 245. (30) Zerulla, D.; Chasse, T. Langmuir1999, 15, 5285. (31) Zhamikov, M; Frey, S.; Heister, K.; Grunze, M. Langmuir2000, 15. 2597- 42 (32) Bain, C. D.; Biebuyck, H. A.; Whitesides, G. M. Langmuir1989, 5, 723. (33) Hostetler, M. J.; Stokes, J. J.; Murray, R. W. Langmuir1996, 12,3604. (34) Hostetler, M. J.; Stokes, J. J.; Greens, S. J.; Murray, R. W. J. Am. Chem. Soc.1996, 118, 4212. (35) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides, G. M; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111,321. (36) Maroncelli, M.; Qi, S. P.; Strauss, H. L.; Snyder, R. G. J. Am. Chem. Soc. 1982, 104, 6237. (37) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559. (38) Porter, M. D.; Walczak, M. W.; Chung, C. J. Am. Chem. Soc. 1991, 113, 2370-2378. (39) Han, S. W.; Kim, C. H.; Hong, S. H.; Chung, Y. K.; Kim, K. Langmuir 1999, 15, 1579. (40) Bruening, M., Personal Communication, 2000. (41) Nishino, T.; Meguro, M.; Nakamae, K.; Matsushita, M.; Ueda, Y. Langmuir 1999, 13, 4321. 43 Gold subs X-ra) suoa Chapter 3 Preparation and Characterization of Gold Substrates Abstract Gold single crystal and gold films grown on mica have been prepared as substrates for monolayer deposition. Scanning tunneling microscopy (STM) and X-ray photoelectron spectrometry (XPS) were used to characterize both surfaces. Atomically flat terraces extending for up to several hundred angstroms can be easily prepared by deposition of gold onto mica. Unsuccessful attempts were made to prepare suitable surfaces from a gold single crystal. In addition, time-lapse topography by STM has shown the effects of surface diffusion of gold atoms at room temperature. 44 3.1 ratios 5th than phen lane of on lean 3.1 Introduction The organization of molecular assemblies at solid surfaces provides a rational approach for fabricating interfaces with a well-defined composition, structure, and thickness. Such assemblies could provide a means to control the chemical and physical properties of interfaces for a variety of heterogeneous phenomena including catalysis,‘ corrosion,2 lubrication3 and adhesion.4 A wide variety of substrates has been investigated to form self-assembling systems, such as gold,57 silverfi-11 metal oxides,12-l5 glassfir18 and semiconductor .surfaces.19'21 For scanning tunneling microscopy (STM) studies, one desirable characteristic of the substrate is that it is atomically flat so that any deposited species can be easily distinguished. In addition, since in our studies deposition of organic species usually takes place in air, the substrate surface should also be relatively inert to oxidation. Furthermore, the substrate needs to be contaminant- free, so that X-ray photoelectron spectroscopy (XPS), reflection-absorption infrared spectroscopy (RAIRS) and other surface chemically sensitive techniques can be used to probe the chemical environment of the self-assembly. For these reasons, one of the most popular substrates has been Au(111) surface. There have been many STM studies of gold surfaces, owing to its inertness in air and its ease of preparation. Some of these studies have focused on gold epitaxially deposited on another substrate, such as mica22r23 or glass,24r 26 while others have investigated gold single crystals”30 and macroscopic gold balls melted from gold wire.“33 These studies have demonstrated that gold surfaces can have atomically flat terraces extending for up to several hundreds of 45 ang COVE 3586 crys alor annl lCEe then Vail filer angstroms, making them ideal for the study of self-assembled monolayers (SAMs). Furthermore, organic thiols and bifunctional disulfides can form S-Au covalent bonds at the adsorbate-substrate interface, and become an ideal self- assembly model for investigation. This chapter reports the preparation and characterization of a gold single crystal and gold films on mica by STM and XPS. The goal is to make large atomically flat gold surfaces as substrates for later SAM study. Sputtering, annealing, mechanical polishing and electrolytic polishing were used to clean and smooth the gold single crystal surface. A contaminant-free surface was produced, but the surface was dominated by a rippled topography. Large atomically flat areas were rarely observed on the gold single crystal. In contrast, gold films produced by evaporating gold onto mica showed large flat (111) terraces bounded by monoatomically high steps (2.5 A). Using time-lapse topography, self-diffusion has been observed at room temperature. 3.2 Experimental Materials. A single crystal of gold (99.99+% pure, 10 mm in diameter, 2 mm thick) with (111) orientation was purchased from Monocrystals Co. (Cleveland, OH). Gold films (approximately 1700 A thick) were prepared by thermally evaporating gold (99.99 °/o) at ~2 A-s" onto freshly cleaved mica in high vacuum (<1045 Torr). To increase the diameter of the Au (111) domains formed, the mica was maintained at 650110 K in vacuum for 12 hours prior to, during and for 2 hours following the deposition. The films were used shortly after preparation. In some instances, the films were flame-annealed for 5-10 5 using a 46 bula incre adve pert: Mg l lake: The Bind butane torch and immediately quenched in methanol. This latter step further increased the size of the Au grains and removed the majority of the observed adventitious carbonaceous contamination. X-ray Photoelectron Spectroscopy. The XPS experiments were performed with a Perkin Elmer PHI Model 1257 spectrometer, equipped with a Mg Ka and Al Koc dual-anode radiation source operated at 300 W. Spectra were taken with a pass energy of 100 eV and at an experimental resolution of 1.0 eV. The base pressure during the measurements was lower than 5 x 10"9 Torr. Binding energies were referenced to the Au 4f 7,2 peak at 84.0 eV. Scanning Tunneling Microscopy. All STM images were obtained in UHV (<1x10'9 Torr) using a RHK Technology Model UHV-300 scanning tunneling microscope. The STM tips were mechanically cut Ptofilroz wires. The STM images were obtained in constant current mode with a typical tunneling current of 3 nA and a bias voltage +300 mV applied to the sample. 3.3 Results and Discussion Au single crystal. Figure 3.1 shows STM images of the untreated gold surface. It was composed of small terraces with single or multiple atomic height steps whose edge met at 60°. This terraced tepography is characteristic of (111) closed packed plane for face-centered-cubic (fcc) metals.34 47 Conle mA, and ; Fig. 3.1 STM images of the gold single crystal as received, showing single or multiple atomic-heiiht terraces whose step edges meet at 60°. (a) 3000 x 3000 A; (b) 2000 x 2000 ; (c) 1000 x 1000 A. The surface cleanliness of the gold single crystal as received was further checked by XPS (Fig. 3.2 (a)). These measurements revealed a substantial carbon and oxygen contamination on the untreated Au(111) surface, which could be attributed to water and airborne organic chemicals. Sputtering and annealing in the vacuum chamber is a very effective method to remove surface contamination. For a dirty surface, hard sputtering (2 keV Ar+ ions, 'sputter ~25 mA, PA, ~20 x 10'3 Pa) was used for 10 min, followed by soft sputtering (500 eV and 200 pA) for another 15 min. For a relatively clean surface, soft sputtering was used solely. The sample was maintained at 200 to 250 °C during sputtering. Subsequently, the sample was annealed at ~360 °C for 30 min. After two cycles of sputtering and annealing, the contamination level was greatly diminished as shown in Fig. 3.2 (b). 48 Au 4f N "O a V a :5 of 3 < . v 3 >1 3 < ’él (a) ‘3 QIL‘“ :‘2‘ r 5i l L- (b) __ l U 1 000 800 600 400 200 0 Binding Energy (eV) Fig. 3.2 XPS surveys of the gold single crystal. (a) as received; (b) after two cycles of sputtering and annealing. Unfortunately, during the ion sputtering and annealing process, the originally mirror-smooth surface became hazy. This sputtering-induced change in optical reflectivity was likely due to microroughening.35 The argon sputtering may have produced structural defects on the surface with dimensions comparable to the wavelength of visible light. Due to our heater limitation, the crystal was not annealed to higher temperature (for example, > 800 °C36) to get rid of this hazy surface. The hazy surface was then sequentially hand polished using 5, 1, and 0.05 micron alumina pastes to regenerate the mirror-like surface. The drawbacks of mechanical polishing were that there were scratches left on the surface and alumina paste became a new contaminant. Electrolytic polishing 49 elec The cont haze imm Fig. illste was further used to remove all traces of alumina paste and produce a scratch- free surface. The apparatus for electrolytic polishing is shown in Fig. 3.3. The electrolyte solution was saturated sodium chloride solution in ultrapure water. The counter electrode was platinum wire. Only the surface to be polished was in contact with the electrolyte. The DC voltage was slowly increased until a yellow haze appeared around the sample (around 3 ~ 6 V, AuCla). The potential was immediately returned to 0 V. The crystal was then withdrawn and rinsed thoroughly with ethanol. Such electrolytic polishing produced scratch-free mirror- like surfaces. The minor trace of contamination was removed by successive cycles of soft sputtering and annealing In UHV. DC Pt counter\ * electrode ,Au single crystal Saturated NaCl Fig. 3.3 Experimental apparatus for electrolytic polishing of the gold single crystal. The surface cleanliness and order were checked by XPS and STM again after these treatments. The crystal was clean based on the absence of carbon and oxygen peaks in XPS spectrum. However, the surface order was quite poor as shown by STM. Figures 3.4 and 3.5 show STM images of the gold single 50 crys of : lop: iii prec How as i surfs. Bis-c l , crystal surface after mechanical polishing, electrolytic polishing and many cycles of sputtering and annealing. The surface was comprised of a variety of topographic features. About 50% of the area examined had a rippled appearance as shown in Fig. 3.4. The rippled topography is also the predominant topographic structure on gold ball surfaces.33 Fig. 3.4 STM images of Au single crystal after many cycles of sputterin and annealing, showing the predominant rippled topography. (a) 2000 x 2000 ; (b) 1000 x 1000 A; (c) 500 x 500 A. About 30% of the area examined was relatively flat at low resolution. However, at high resolution, these regions showed mounds, pits and sharp lines as in Fig. 3.5, which were commonly observed for differentially sputtered surfaces. Rotating samples while sputtering would help minimize the differential effect, however, there was no rotation setup with our instrument. The remaining 20% of the area examined was very rough with small terraces and clusters. Large atomically flat areas were rarely observed. 51 q. . l t Fig. anne dilel cyst alter Fig. 3.5 STM images of Au single crystal after many cycles of sputtering and annealing, showing a relatively flat area. The appearance of lines are due to differential sputtering. (a) 3000 x 3000 A; (b) 2000 x 2000 A; (c) 1000 x 1000 A. A self-assembled monolayer of octanethiol was formed on the gold single crystal, which further demonstrated that the surface was not atomically flat. (For more information about SAM preparation and characterization, please refer to Chapter 4). The molecules did form a monolayer as shown in Fig. 3.6, but the order was limited to small regions (< 50 x 50 A) due to the roughness of the substrate. No vacancy islands were observed, which are characteristic of alkanethiol SAMs on a flat gold substrate.5 Since the large majority of the gold single crystal surface had rippled topography, which would be potentially misleading when used as a substrate for deposition and imaging of organic thiol molecules, we did not use the gold single crystal as the substrate for further experiments. 52 Fig, ( tougl lags if ind i! ”any Wide Fig. 3.6 STM images of octanethiol SAM on the gold single crystal. Due to the roughness of the gold substrate, the order of the SAM is limited to small regions. (a) 1500 x 1500 A; (b) 250 x 250 A; (c) 200 x 200 A; (d) 150 x 150 A. Au (111) on Mica. Due to the unsuccessful attempts to prepare atomically flat surface from the gold single crystal, we changed our focus to make gold films on mica. Gold thin films have several advantages over a gold single crystal. For example, thin films can be prepared directly in vacuum, thereby avoiding the possibility of contamination from polishing or transferring through atmosphere, and the cost is usually a small fraction of the corresponding crystal. There are many procedures published on the subject of gold film growth and there are a wide variety of descriptions for the conditions that produce the highest quality gold films.22-'L’4v26 Our method for the production of gold thin films involved: (1) a 53 higl as e dept quai The 153: The near: the chars ism; that a hen: high vacuum of <10'6 Torr; (2) a clean vacuum chamber; (3) freshly cleaved mica as a smooth, dry substrate; (4) a substrate temperature of 650 K coupled with a deposition rate of 2 A/s. Under these conditions, it was possible to produce high quality thin films of pure (111) terminated gold using relatively simple equipment. Figure 3.7 (a) shows a typical large area STM image of a gold thin film. The film was comprised of many gold grains that have an average size of 1500 x 1500 A. Each grain consisted of flat terraces separated by atomically high steps (Fig. 3.7 (b)). It was relatively easy to find large terraces extending for hundreds of angstroms without a single step. This is in principle the ideal character of a substrate for the study of deposited molecules. The grains were crystalline and (111) terminated with the gold atoms arranged in a hexagonal pattern (shown in Fig. 3.7 (c) and (d)), which is the lowest energy crystallographic plane of gold. The nearest neighbor distance is 29:03 A, which agrees well with the 2.88 A nearest neighbor spacing of the (111) face of gold. It is interesting to note that the gold films on mica did not show any obvious hexagonal faceting characteristic of the (111) close-packed plane for ice metals. The lack of hexagonal faceting at temperatures of 450 °C and below may result from the fact that atoms of gold are unable to achieve sufficient surface mobility to rearrange themselves into a minimum-energy configuration of a flat (111) planar surface densely covered with facets of trigonal (hexagonal) symmetry.37 54 sun's hm. Fig. 3.7 STM images of Au(11 1) on mica. The film was comprised of gold grains with atomically flat terraces on top. The gold atoms were arraigned in a hexagonal pattern which is characteristic of the (111) close-packed plane for fcc metals. (a) 3000 x 3000 A; (b) 1300 x 1300 A; (c) 70 x 70 A; (d) so x 50 A. At large area, surface reconstruction was observed. The reconstructed surface of Au(111) is characterized by a 4.3% uniaxial lateral contraction relative to the bulk layers, with two extra atom per 22X\/3 unit cell.3°»38v39 This contraction causes variations in registry between the surface and subsurface atomic layers such that the stacking arrangement alternates between normal fcc and faulted hexagonal-close-packed (hcp) packing. The fcc and hcp areas are separated by partial surface dislocations in which surface gold atoms are near bridge sites. The gold atoms in bridging sites are vertically displaced by 0.20 A. A model of 55 the inl dis 310 nee the rectangular unit cell for the straight stripes of the reconstruction is illustrated in Fig. 3.8. The short edge of the unit cell is 4.99 A long (~13 times the gold-gold distance of 2.88 A) and lies alone the next-nearest-neighbor direction of gold atoms in a (111) plane. The long edge is 63.36 A long and is aligned along the nearest-neighbor direction of gold atoms in a (111) plane. Fig. 3.8 Schematic representation of the Au(111) 22 x43 reconstruction as seen from top and side views. The unit cell is shown by the rectangular which has dimensions of 63.36 A x 4.99 A. This figure is taken from Dishner et al.33 Figure 3.9 shows a typical unfiltered STM image of 22x43 surface reconstruction on an atomically flat Au(111) terrace deposited on mica. The terrace exhibited a characteristic pattern of painlvise-arranged parallel lines, corresponding to the bridge-site higher ridges. The zig-zag turns of the parallel lines, which are the well-known herringbone turns of reconstructed Au(111) surface, are ways to relieve the surface strains created by the reconstruction.29,4° It has been reported that there exists a regular arrangement of the zig-zag herringbone structure throughout the terrace.30 However, the formation of the 56 zig-z. and ‘ slruc Fig. 2 surlai lo air Au(1- i'llth (i zig-zag herringbone domain was strongly affected by the deposition conditions and temperature distribution on a mica substrate and non-regular herringbone structures were observed in our study. Fig. 3.9 STM images of Au(111) on mica (400 x 400 A), showing the 22X\/3 surface reconstruction. The quality of the gold films generally became poorer after being exposed to air for more than a few hours, which was attributed to the gradual buildup of contaminants on the surface. XPS data showed that the main contamination was C and O, which could be from water and aifoome organic chemicals. The freshly evaporated gold surfaces were hydrophilic with a contact angle < 20° for water droplets. After an exposure to air for 2 ~ 3 days, the surface became more hydrophobic and the average contact angle increased to 55°. Surface Diffusion. Time-lapse topography with STM on a clean, annealed Au (111) on mica showed that at room temperature, the surface details changed with time. Small mounds disappeared, recessed regions tended to fill in, steps 57 move hours 1.5 hr acquii were resell frame moved and initially kinked step edges gradually became straight over a period of hours. The sequence of sixteen topographic images in Fig. 3.10 taken over about 1.5 hours provide direct visual information about this process. Each image was acquired in 25 seconds scanning from the top-left to the bottom-right. All images were 500 x 500 A. At this low resolution, individual gold atoms were not resolved, but features such as steps and islands could be observed. In the first frame, the initial scanning showed an monoatomically high island with a “finger" on top of a flat terrace. There were three small mounds in the left comer and a valley on the right. In subsequent frames, the middle mound gradually disappeared while the two neighboring mounds remained almost unchanged. During the first 24 minutes, the large island’s Au finger grew continuously downwards throughout the images. Subsequently, in the 25th minute frame, a narrow valley formed near the top of the finger. After 37 minutes, the valley became an elongated hole in the large island as the valley entrance became filled in. The hole finally broke through the other side of the island after 58 minutes and the kinked edge became straight after 61 minutes. The valley to the right of the Au finger also changed with time (pointed out from the 42nd minute to the 58th minute frame). The sharpness and the overall appearance of the terrace edges in other parts of the same frame remained unaltered during the sequence, ruling out the possibility of the effect being caused by a gradual tip change. 58 Fig. 3.10 A series of time-lapse STM topographic images at room temperature showing a 500 x 500 A area of a gold film on mica. The arrows pornt out surface details changed with time. 59 To check whether the diffusion was the result of the scanning motion, one area was imaged and then the tip was moved to an adjacent area about 1000 A away. After 30 min, upon returning to the first area, diffusion had taken place, ruling out that scanning itself (i.e., tip-induced diffusion) was responsible for the surface diffusion. Diffusion was observed under a variety of tunneling conditions, regardless of the magnitude or sign of the tunneling current and voltage parameters. These observations showed that there were mobile atoms on the Au(111) surface at room temperature. However, it was not clear from the observations whether atoms were diffusing along steps exclusively or were also crossing the flat terraces. 3.4 Conclusion STM and XPS were used to characterize the surfaces of gold single crystal and gold films grown on mica. Sputtering and annealing were used to clean the gold single crystal surface. Microroughening was observed during sputtering, making the mirror-like surface hazy. It was not possible to anneal the crystal to high enough temperature to get rid of the microroughened surface. Mechanical polishing and electrolytic polishing were subsequently used to generate an optically flat surface. A contamination-free surface was produced, but the surface order was quite poor. STM revealed that rippled topography dominated the surface. Even in a relatively flat area on gold single crystal, pits, mounds and lines were observed, while large atomic flat area was rarely observed. On this surface, the self-assembled monolayer of octanethiol showed very short-range order. 60 ln contrast to the gold single crystal, thin films of gold epitaxially grown on mica consisted of smooth terraces separated by monoatomic steps. Atomically flat terraces extending for up to several hundreds angstroms could be easily observed. Atomic resolution STM images of the Au(111) surface were shown and the 22x13 reconstruction was observed. These gold films are ideal substrates for the study of deposited organic molecules. In addition, time-lapse topography by STM on a clean Au(111) surface on mica showed the effect of surface diffusion of gold atoms at room temperature. Features such as mounds, islands and valleys were seen to change as a result of self-diffusion over the surface. The diffusion did not depend on the magnitude or sign of the tunneling current and voltage parameters, nor was it related to scanning motion or tip condition either. 3.5 Literature Cited (1) Somorjai, G. A. Chemistry in Two Dimensions: Surfaces, Cornell University Press: Ithaca, NY, 1981. (2) Notoya, T.; Poling, G. W. Corrosion 1979, 35, 193. (3) Bowden, F. P.; Tabor, D. The Friction and Lubrication of Solids, Oxford University Press: London, 1968. (4) Kaelble, D. H. Physical Chemistry of Adhesion, Wiley-lnterscience: New York, 1971. (5) Poirier, G. E. Chem. Rev.1997,97,1117. (6) Poirier, G. E.; Heme, T. M.; Miller, C. G.; Tarlov, M. J. Langmuir 1999, 121.9703. 61 (7) Poirier, G. E. Langmuir1999, 15,1167. (8) Harris, A. L.; Rothberg, L.; Dhar, L.; Levinos, N. J.; Dubois, L. H. J. Phys. Chem. 1991, 94, 2438. (9) Dhirani, A.; Hines, M. A.; Fhisher, A. J.; Ismail, O.; Guyot-Sionnest, P. Langmuir1995, 11,2609. (10) Fenter, P.; Eisenberger, R; Li, J.; Chamillone, N.; Bemasek, S.; Scoles, G.; Ramanarayanan, T. A.; Liang, K. S. Langmuir1991, 7, 2013. (11) Gui, J. Y.; Stern, D. A.; Frank, D. G.; F.;, L.; Zapien, D. C.; Hubbard, A. T. Langmuir1991, 7, 955. (12) Smith, E. L.; Porter, M. D. J. Phys. Chem. 1993, 97, 4421. (13) Soundag, A. H. M.; Tol, A. J. W.; Touwslager, F. J. Langmuir 1992, 8, 1127. (14) Tao, Y.-T. J. Am. Chem. 800.1993, 15,4350. (15) Thompson, W. R.; Pemberton, J. E. Langmuir1995, 11,1720. (16) Lin, W.; Yitzchaik, 8.; Lin, W.; Malik, A.; Durbin, M. K.; Richter, A. G.; Wong, G. K.; Dutta, P.; Marks, T. Angew. Chem. Int. Ed. Engl. 1995, 34, 1497. (17) Maoz, FL; Sagiv, J.; Degenhardt, D.; Mohwald, H.; Quint, P. Supramol. Sci.1995, 2, 9. (18) Mino, N.; Ogawa, K.; Hatada, M.; Takastuka, M.; Sha, S.; Morizumi, T. Langmuir 1993, 9, 1280. (19) Hamers, R. J.; Wang, Y. Chem. Rev. 1996, 96, 1261. (20) Schwartz, M. P.; Ellison, M. D.; Coulter, S. K.; Hovis, J. 8.; Hamers, R. J. J. Am. Chem. Soc. 2000, 122, 8529. 62 (21) Hovis, J. S.; Liu, H.; Hamers, R. J. Surf. Sci. 1998, 404, 1. (22) Dishner, M. H.; lvey, M. M.; Gorer, S.; Hemminger, J. C. J. Vac. Sci. Technol. A. 1998, 16, 3295. (23) Porath, D.; Millo, O.; Gersten, J. l. J. Vac. Sci. Technol. 81996, 14,30. (24) Hwang, J. Ph. D. thesis, Michigan State University, 1993. (25) Stamou, D.; Gourdon, D.; Liley, M.; Bumham, N. A.; Kulik, A.; Vogel, H.; Duschl, C. Langmuir1997, 13, 2425. (26) Semaltianos, N. G.; Wilson, E. G. Thin Solid Films 2000, 366, 111. (27) Salmeron, M.; Kaufman, D. S.; Marchon, B.; Ferrer, S. Appl. Surf. Sci. 1987, 28, 279. (28) Salmeron, M.; Marchon, B.; Ferrer, S.; Kaufman, D. S. Phys. Rev. B 1987, 35, 3036. (29) Sandy, A. R.; Mochrie, S. G.; Zehner, D. M.; Huang, K. G.; Gibbs, D. Phys. Rev. B1991, 43, 4667. (30) Chambliss, D. D.; Wilson, R. J. J. Vac. Sci. Technol. 81991, 9,928. (31) Schneir, J.; Sonnenfeld, R.; Marti, O.; Hansma, P. K.; Demuth, J. E.; Hamers, R. J. J. Appl. Phys. 1988, 63, 717. (32) Li, Y. Z.; Vazquez, L.; Piner, R.; Andres, R. P.; Reinfeinberger, R. Appl. Phys. Lett. 1989, 54, 1424. (33) Sommerfeld, D. A.; Cambron, R. T.; Beebe, J. T. P. J. Phys. Chem. 1990, 94,8926. (34) Hegner, M.; Wagner, P.; Semenza, G. Surf. Sci. 1993, 291, 39. (35) Geller, J., Geller MicroAnalytical Lab; Personal Communication, 1998. 63 (36) Leung, T. Y. B. Ph. D. thesis, Princeton University, 1998. (37) DeRose, J. A.; Thundat, T.; Nagahara, L. A.; Lindsay, S. M. Surf. Sci. 1991, 256, 102. (38) Dishner, M. H.; Hemminger, J. C.; Feher, F. J. Langmuir1997, 13,2318. (39) Hara, M.; Sasabe, H.; Knoll, W. Thin Solid Films 1996, 273, 66. (40) Woll, C.; Chiang, 8.; Wilson, R. J.; Lippel, P. H. Phys. Rev. 81989, 39, 7988. Chapter 4 Simple Alkyl Thiol Self-assembled Monolayers on Au(111) Abstract Self-assembled monolayers (SAMs) of n-alkyl thiols (CH3(CH2).,SH, where n=7, 9 and 17), were prepared by adsorption from dilute solutions onto evaporated gold substrates. Their structure and thermal stabilities were characterized by scanning tunneling microscopy (STM), reflection-absorption infrared spectroscopy (RAIR), X-ray photoelectron spectroscopy (XPS), ellipsometry and contact angle measurements. These surface-sensitive techniques probed both macroscopic and microscopic properties of the films, and showed that all three alkyl thiols spontaneously adsorbed on gold and formed well-ordered structures. The alkyl chains, which experienced a crystalline-like environment, were tilted 20~40° from the surface normal and form a 13N3R30° high-density structure at room temperature. After annealing to 75 °C, a low- density striped phase was observed by STM where the molecules were aligned along the substrate surface in double lamellae. The molecular reorientation and film disordering during annealing was further confirmed by RAIRS, ellipsometry, and contact angle measurements. 65 4.1. Introduction Self-assembled monolayers (SAMs) have been studied intensively as model systems for fundamental research areas such as double layer phenomena, biological membrane systems and the nature of organic/inorganic interfaces.“3 They also received considerable attention because of potential technological applications in chemical sensing, corrosion protection, lubrication and nanofabrication of electronic components.‘-45 Whether the focus of research is fundamental or applied science, a firm understanding of structure and thermal properties of the SAMs is an essential requirement. The most widely studied class of monolayers is unquestionably that of films made of organosulfur compounds on metal surfaces. This system has several advantages as a model, including the choice of thiol, sulfide, or disulfide functionalities to study head group/substrate interactions and the variation of the chain length to study the importance of tail interactionsfi‘8 The ability to manipulate the structure and chemical properties of these materials, as well as their stability in both vacuum and ambient environments, make them ideal substrates for surface chemistry studies. Since 1983, after Nuzzo and Allara prepared monolayers of alkanethiols on gold by adsorption of di-alkyl disulfides from a dilute solution,9 SAMs of alkanethiols on gold have been extensively characterized by various surface analytical techniques. The packing order was studied by low energy electron diffraction (LEED),1°»11 He diffractionflfi':13 X-ray diffraction,14-15 and scanning probe microscopy.15-22 The macroscopic structural properties were studied by 66 reflection-absorption infrared spectroscopy (RAIRS),23-25 ellipsometry,26-27 and X-ray photoelectron spectroscopy (XPS).23'30 The wettability and ion-transfer properties were studied by electrochemical methods and contact angle measurements.31'33 It was found that these alkanethiols grow commensurate with the Au(111) surface. The sulfur ends chemisorb on gold surface through S-Au bonds. The external chains of neighboring alkanethiols attract one another through van der Waals interactions and form a \IBXVSRSO" structure with a c(4x2) superlattice.34' 36 There are distinct differences in structure between long and short chain thiols monolayersfi’r‘ifi7 The long chain thiols (C28) form a close-packed, crystalline assembly with fully extended alkyl chains tilted from the surface normal by 20~35°. As the chain length decreases, the surface becomes increasingly disordered with low packing density and coverage. The lack of order in short chain SAMs has been attributed to a greater concentration of gauche defects and to weak interchain interactions.27 This chapter reports the study of octanethiol (CB), decanethiol (C10) and octadecanethiol (C18) SAMs with multiple surface-sensitive techniques. Monolayers of unsubstituted alkanethiols are the natural starting point for the research presented here because a significant amount of structural information is now available for these monolayers. CB and C18 were chosen as examples of room-temperature liquid and solid alkanethiols which can form V3xV3R30° close packing structure on Au(111). In addition, molecular resolution images of CB monolayers were used to calibrate STM scanning parameters. C10 served as a 67 control for the study of the olefin-terminated di(9-decene) disulfide SAMs discussed in Chapter 7. The formation, decomposition and fundamental properties of these films were assessed using electron spectroscopy, scanning probe microscopy, and optical spectroscopy. These complementary techniques enabled the measurements of critical film properties such as geometric order, molecular orientation, defect sites, chemical structure, and film thickness, providing a solid foundation for later work. 4.2. Experimental Materials and Monolayer Preparation. Reagent grade octanethiol, decanethiol and octadecanethiol were bought from Aldrich and were used as received. The gold substrates for infrared spectroscopy and ellipsometry measurements were made by electron-beam evaporation of 200 nm of Au on 20 nm of Ti on Si(100) wafers. The gold substrates for scanning tunneling microscopy (STM) measurements were prepared by thermally evaporating gold onto freshly cleaved mica in high vacuum (details given in Chapter 3). All of the substrates were cleaned with a UV/Oa cleaner for 15 minutes, and then soaked in deionized water for 30 minutes. After this, the substrates were dried in a N2 stream and immediately immersed in 2 mM ethanol solutions of the alkanethiols for more than 5 hours. The resultant SAMs were rinsed with a large amount of ethanol and dried in a N2 stream before characterization. X-ray Photoelectron Spectroscopy. XPS spectra were obtained on a Perkin Elmer PHI Model 1257 spectrometer using Al Kor radiation source operated at 300 W. Survey spectra were taken with a pass energy of 100 eV 68 and at an experimental resolution of 1.0 eV. S 2p spectra were acquired with 45 eV pass energy and at a resolution of 0.1 eV. The X-ray spot was moved to a new position on the sample after every 30 scans to minimize the effects of beam- induced damage. The pressure during the measurements was lower than 5x10“9 Torr. The spectra of monolayers were referenced to the Au 4f7/2 peak at 84.0 eV. Infrared Spectroscopy. RAIRS was performed using a Ng-purged Nicolet 560 Magna IR spectrometer with MCT detector. Spectra were obtained with a PIKE specular reflectance accessory using p-polarized light incident at 80° with respect to the surface normal. All spectra reported are the average of 256 scans obtained at a resolution of 4 cm'1 and referenced against a freshly prepared, clean gold film. Transmission spectra of the pure compounds were recorded on a Mattson 3000 Galaxy Series FT IR spectrometer. Ellipsometry. The thickness of the monolayers were determined by optical ellipsometry in two steps with a rotating analyzer ellipsometer (J.A. Woollam, Model M-44) using a white light source and an incidence angle of 75° from the surface normal. The analyzer and polarizer angles for a reflected light beam from each uncoated substrate were first measured on at least three different spots. The average complex refractive index for each substrate was then calculated with a two-phase parallel layer model from classical electromagnetic theory. After monolayer formation, each sample was again analyzed and the film thickness was calculated from a three-phase parallel layer model, using the average complex refractive index of each substrate and a refractive index of 1.5 69 for the film. At least six different sampling points were considered in order to obtain an average thickness value. Contact Angle Measurements. Advancing contact angles of water were measured in air with an AST model VCA-2500XE goniometer, using the sessile drop technique in which a drop (~ 2 uL) was formed on the end of a hydrophobic, blunt-ended needle.38 The sample was raised until the drop touched the surface, and then the surface was removed and the contact angles were measured. STM measurements. All STM images were obtained in UHV (<1x10'9 Torr) using a RHK Technology Model UHV-300 scanning tunneling microscope. The STM tips were mechanically cut Ptaalrog wires. The STM images were obtained in constant current mode with a typical tunneling current of 100 pA and a bias voltage +1 V applied to the sample. 4.3. Results and Discussion Immersion of a gold film into an ethanol solution of any of the thiols in this study resulted in the formation of a self-assembled monolayer. Elucidating the structure of these films posed a complex problem. A variety of surface-sensitive techniques were chosen to characterize both macroscopic and microscopic structures of these films. Ellipsometry. Optical ellipsometry was applied as a convenient means of determining the average film thickness. The data for C8, C10 and 018 monolayers on Au(111) are shown in Table 4.1. The ellipsometric thickness provided strong evidence that all three thiols formed one-molecular—layer thick films on gold. Furthermore, the apparent tilt angles were calculated through a 70 comparison with model thickness. The apparent tilt angles of CB and C10 are greater than that of C18, indicating the shorter alkyl chains tilted more to achieve the maximum van der Waals interchain interactions. Table 4.1 Thickness of CB, C10, and C18 SAMs CB C10 018 Measured thickness (A) 9.4:o.9 1o.3¢0.s 24.0:03 Calculated thickness (A)‘ 12.8 14.4 25.4 Apparent tilt angle 42°i10° 43°i4° 19°:l:6° TThickness was calculated, from the surface Au atom to the terminal proton based on the minimum-energy extended conformation calculated by molecular mechanics.39 Contact angle measurements. The wettability of surfaces covered with monolayers can be correlated with the quality of the monolayers. Therefore, contact angle measurement is a preliminary tool in the investigation of the alkanethiol films. The advancing contact angle data for 08, C10 and C18 films are listed in Table 4.2 and are all in close agreement with previous measurements of the same or similar adsorbates.38 Figure 4.1 shows the wettability difference between bare gold substrate and C18 SAM. The gold substrate was hydrophilic with contact angle < 20° for a pure water droplet, while the alkanethiol SAMs were hydrophobic with contact angles in the range of 100°~110°, indicating extremely low surface free energies, which is consistent with the hydrophobic character of methyl end groups in the SAMs. Table 4.2 Advancing contact angles of water on CB, C10, and C18 SAMs C8 C10 C18 L__ Contact angles of water 102°12° 107°i2° 109°:l;3° 71 Fig.4.1 (a) Average water advancing contact angle for gold substrate is < 20°, gtggaverage water advancing contact angle for C18 SAM on gold substrate is X-ray Photoelectron Spectroscopy. The adsorption of a close packed alkanethiol monolayer on gold did not result in a significant perturbation of the Au 4f core level spectrum determined by XPS (Fig. 4.2). Though the total integrated spectral intensity decreased, which is an expected result owing to the presence of a relatively thick organic overlayer. Such features as the full width at half- maximum (FWHM) and the Au 4f7/2 (referenced at 84.0 eV) and Au 4f5,2 binding energies remained unchanged. Examination of the spectrum in the S 2p region showed an extremely broad and asymmetric peak with a binding energy maximum at ~162.2 eV (Fig. 4.2). This peak is composed of two components: the S 2pm and 2pm peaks; however, the band widths (0.8 to 0.9 eV) of the source X-rays are too big to resolve them. The C 13 region for this sample showed a single symmetric peak centered at 285.0 eV, which indicates a single environment for the C atoms in the molecules and agrees with other studies.“ The S atom does not produce a measurable chemical shift for the adjacent CH2 group. 72 S 2p 162.2 9 :5. 3 < ,E‘ a 4 ‘ r 5 g\\ 165 160 j E5 1 _ a) 3 I V ( I 2 I F: T L— I fi 1 r 1000 750 500 250 0 Binding Energy (eV) Fig. 4.2 XPS spectrum of C8 SAM on Au(111) Infrared Spectroscopy. Infrared spectroscopy is a valuable probe of monolayer structure. In particular, the CH stretching can indicate the packing environment and average orientation of the adsorbate chains. Figure 4.3 shows IR spectra in the CH stretching region for C18 solid, C10 liquid, and C18 SAM. [0.5 . (a) : x1 Absorbance 3050 2950 2850 2750 Wavenumbers (cm") Fig. 4.3 IR spectra of CH stretching modes in the high frequency region. (a) transmission spectrum of neat liquid C10; (b) transmission spectrum of solid C18 In KBr; (c) reflection spectrum of C18 SAM on Au(111). 73 For the C18 monolayer, the band at 2956 cm‘1 is assigned to the CH3 asymmetric in-plane CH stretching mode (vasCHa). The bands at 2895 and 2877 cm“1 are assigned to the CH3 symmetric CH stretching mode (vsCHa); which splits owing to a Fermi resonance with the lower frequency asymmetric CH3 deformation mode. The asymmetric and symmetric methylene stretching modes absorb at 2918 and 2848 cm", respectively. These modes are qualitative indicators of the packing of the alkyl chains, as shown by the differences in peak positions for the crystalline-like and liquid-like bulk alkanethiols summarized in Table 4.3. Table 4.3 IR assignments of OH high frequency stretching modes of neat liquid C10, solid C18 in KBr and C18 SAM on gold Wavenumbers (cm") Mode assignment Neat liquid of 010 Solid C18 in KBr SAM of C18 2954 2955 2956 v8, CH3 2926 2918 2917 v8, CH2 2897 2895 2895 v3 CH3, FR 2873 2876 2877 vs CH3 2854 2848 2848 vs CH2 The peak positions shift to higher energies in the liquid phase relative to the solid is due to an increase in the freedom of the hydrocarbon modes.41 The similarity between the solid phase and SAM indicates that a similar distribution of intermolecular environments exists for the CH2 group in both the crystalline bulk and the monolayer phase. In particular, the width and the position of the vasCHz mode (2917 cm“) in the SAM suggest that the chains are fully extended and 74 reside in a crystalline-like environment. If there were significant numbers of gauche defects in the monolayer, or if they were not tightly packed, the position of this peak would shift to higher frequency and its width would broaden substantially.“ The relative intensities of the symmetric (polarized perpendicular to, but in-the-plane of, the chain aixs, 2848 cm") and orthogonal asymmetric (polarized perpendicular to the chain aixs, but out-of-the-plane) methylene stretching vibrations imply that the chains must be tilted with respect to the surface normal. Both vas and vsCHa modes are blue shifted compared with the liquid spectrum, indicating a significant perturbation in the methyl group at the monolayer-air interface. We also carefully examined the spectral region between 1600 and 1000 cm". Both the symmetric methyl scissoring (SCHa, 1377 cm“) and the methylene scissoring (SCHZ, 1465 cm") modes were observed. However, the poor signal-to-noise levels precluded the detection of methylene twisting and wagging modes, which are indicative of all-trans methylene segments. Scanning Tunneling Microscopy. Compared with the techniques discussed above which probe macroscopic structure, STM is unique in its ability to characterize structure in direct space with molecular resolution. At large scale (>1000x1000 A), cs, C10 and C18 SAMs looked similar. A typical large area STM image of CB SAM is shown in Fig. 4.4 (a), which exhibited well-defined step edges and uniformly distributed vacancy islands or “pits”. The pits were 2.4 A deep, consistent with Au(111) single-atom step height, suggesting that the pits were defects in the gold layer rather than defects in the alkanethiol layer. 75 Fig. 4.4 STM images of CB SAM at room temperature. (a) 1200x1200 A. (b) 700x700 A, (0) 150x150 A. Edinger et al. have proposed that the vacancy islands were formed by etching of gold in the alkanethiol solutions, a mechanism that was suggested by atomic absorption spectroscopy measurements showing dissolved gold species in the incubation bath.43 This gold etching mechanism was called into question by later experiments showing vacancy islands formation even for assembly in gas-phase transport.18 Poirier suggested that the vacancy islands were created by chemisorption of alkanethiols within the hop regions of the reconstructed gold surfaces.44 Excess gold atoms are forced out of the surface layer by relaxation of the compressed herringbone reconstruction. This creates adatoms on, and vacancies in, the surface layer. The adatoms diffuse rapidly and adsorb at ascending step edges while the vacancies nucleate islands in the terrace. This “lift and reconstruct” theory focused on the changes in the substrates and needs to be supplemented with the consideration of adsorbate intermolecular interactions, because pure aromatic thiol SAMs were observed to have atomically high islands instead of depressions. More discussions on this issue will be presented in Chapter 6. 76 Well-ordered domains of alkanethiol molecules were observed between the vacancy islands (Fig. 4.4 (c)). Atomically resolved images were found for CB and C10 films, which exhibited a periodic hexagonal pattern with nearest neighbor distance of 5.0 :l:0.3 A, consistent with a (V3xV3)R30° adlayer on a Au(111) lattice (Fig. 4.5). The c(4x2) superlattice structure was not observed, probably because the tunneling gap impedance we employed (<10 G0) was not high enough. A large tunneling resistance corresponds to a large tip-substrate separation and the c(4x2) superlattice was reported to appear only at high tunneling gap impedance (>100 GO).34.45 Similarly, no molecularly resolved images were obtained for C18 SAMs, because the tunneling gap impedance was not high enough to resolve the long alkyl tail. Thermal Properties. The thermal pr0perties of alkanethiol SAMs were studied by annealing films in a vacuum. After annealing to 75 °C, as shown in Fig. 4.6, the vacancy islands disappeared and the surface was covered by parallel rows of features oriented at approximately 120° with respect to each other. 77 0 51015 20 25 30 35(A) (Pm) 100 (b) 80 33 /\/\/ NMVA 20 0 0 5 10 15 20 25 30 35 40010 Fig. 4.5 Molecular resolution STM images (50 x 50 A) of 08 SAM on Au(111) which exhibits a periodic hexagonal pattern with nearest neighbor distance of 5.0 10.3 A, consistent with a (VSXVS)R30° adlayer on a Au(111) lattice. 78 Fig. 4.6 STM images of C8 SAM on Au(111) after annealed to 75 °C for 30 min, showing large and well-structured low density wide striped phase. (a) 1500 x 1500 A, (b) 700 x700 A, (c) 300 x 300 A, (d) 100 x 100 A. 79 According to the literature, this phase was called the wide striped phase (B phase).‘9‘21r46 The distance between the stripes was approximately twice the length of the methylene chain (20 A for C8, 22 A for C10 and 31 A for C18), suggesting the gray areas between the alternate stripes were caused by the methylene chains lying down tail to tail on the gold surface. Subsequently, interactions between the alkyl chains and the gold surface were able to stabilize the wide striped structure. Annealing these films to 100 °C resulted in further molecular desorption with only a disordered, liquidlike phase left on the surface. Two other stripe phases have been reported with an intermediate molecular surface density between (13xV3)R30° structure and the [3 phase.”- 21:45 A narrow striped phase (6 phase) forms in coexistence with the (‘13X‘/3)R30° high density structure after the first annealing step. ln the 8 phase, the axes of one row of molecules are oriented parallel to the substrate surface, while a pairing row is tilted out of the surface plane, forming gauche defects near the sulfur terminus. Further annealing of these samples results in the next intermediate phase: a mixed narrow striped and wide striped phase (x phase). After the formation of x state, the continued annealing of alkanethiol SAMs results in surfaces covered with the pure wide striped phase ([3 phase). Obviously, the arrangement of molecules in the different phases is governed by the change in surface coverage due to annealing process. Therefore, different adsorbate structures are induced by consecutive annealing to varying temperatures. 80 The annealing process was also followed by RAIRS. Figure 4.7 shows a representative series of RAIRS spectra of the C10 monolayer on Au(111) as a function of annealing temperature. The splitting of the asymmetric methylene mode at 2920 cm'1 is due to the Fermi resonance with the methyl mode at 2936 cm'1 . 3050 2950 2050 2750 Wavenumbers (cm") Fig. 4.7 RAIRS spectra of C10 SAMs on Au(111) following annealing at the indicated temperature. As the temperature increased, the intensity of both v330H3 and vsCHa modes decreased (Fig. 4.8 (a)), indicating that molecular desorption occurred on the surface. At the same time, the position of both VasCHz and vsCHz modes shifted to higher frequency (~3 cm") during annealing and their peak widths broadened, suggesting that a significant number of gauche defects were created in the C10 monolayer during the thermal treatment and the film was no longer tightly packed. The intensity increase before 75 °C in both methylene stretching modes (Fig. 4.8 (b)) further confirmed that annealing induced orientation change 81 in the C10 films. The rapid reduction in both methyl and methylene modes after 75 °C signaled complete monolayer decomposition. O vas CH3, 2963 cm'1 100 - . 3 90‘ .v, CH3, 2878 cm1 e g 30, G g 70-( 0 '3 501 g 40-( .E 30 -i i 20 , a ‘0 fi‘ fi r JT—"_ 20 40 60 80 100 Annealing temperature (°C) I vas CH2, 2921 cm'1 9 vs CH2, 2850 cm" gee 3,451 c _ Base 0 . £25100 0 :15f ‘5 l i 5i 3 51__-_1..1_._..__ .- - _ A -, _ _ 20 40 so so 100 Annealing temperature (°C) Fig. 4.8 Plot of peak intensity change for C10 monolayers from Fig. 4.7 as a function of annealing temperature in (a) vasCHa and vasCHa modes; (b) vasCHz and vsCHz modes. The average film thickness decreased by ~5 A after annealing to 75 °C for 30 min and the surface became more hydrophilic (61.5° water contact angle), which further confirmed the changes in film morphology and surface coverage discussed above. 82 4.4. Conclusion This chapter has presented the results of a multitechnique study of the structure and thermal stability of simple alkanethiol SAMs (C8, C10 and C18) on Au(111). The IR spectroscopic and ellipsometric data indicated that all three thiols formed a densely packed, crystalline-like assembly with fully extended alkyl chains at room temperature. Atomically resolved images were found for CB and 010 films at room temperature. The images exhibited a periodic hexagonal pattern with nearest neighbor distance of 5.0 10.3 A, consistent with a (V3x1l3)R30° adlayer on Au(111). No molecular resolution images were observed for C18 SAM due to the insulating properties of the long alkyl tail. The structure of SAMs depends strongly on the surface coverage. The structural changes with decreasing molecular surface density were studied by annealing densely packed films in a vacuum. STM showed that the vacancy islands disappeared and a stripe phase was formed after annealing. The molecular reorientation and film disordering during annealing was further confirmed by RAIRS, ellipsometry, and contact angle measurements. The thermal fragility of SAMs becomes a serious deficiency for practical applications. This leads us to investigate the possibility of increasing their thermal and temporal stability by incorporating unsaturated functionalities into the SAM. After spontaneous assembly of thiols onto a gold substrate, cross-linking of adjacent molecules can be initiated by chemical, photo or thermal means. Increasing the lateral interaction between the chains by covalent bonding can, in principle, improve the robustness of the SAMs. 83 4.5. Literature Cited (1) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir- Blodgett to Self-Assembly, Academic Press: San Diego, CA, 1991. (2) Whitesides, G. M.; Laibinis, P. E. Langmuir1990, 6, 87. (3) Dubois, L. H.; NUzzo, R. G. Annu. Rev. Phys. Chem. 1992, 43, 437. (4) Ulman, A. Chem. Rev. 1996, 96, 1533. (5) Berggren, K. K.; Bard, J. L.; Wilbur, J. D.; Gillaspy, A. G.; Helg, J. J.; McClelland, S. L.; Rolston, W. D.; Philis, M.; Prentiss, M.; Whitesides, G. M. Science 1995, 269, 1255. (6) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558. (7) Fenter, P. In Thin Films: Self-assembIed Monolayers of Thiols; Ulman, A., Ed.; Academic Press: San Diego, 1997. (8) Poirier, G. E. Chem. Rev. 1997, 97, 1117. (9) Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481. (10) Strong, L.; Whitesides, G. M. Langmuir1988, 4, 546. (11) Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Chem. Phys. 1993, 98, 678. (12) Camillone III, N.; Eisenberger, P.; Leung, T. Y. 8.; Schwartz, P.; Scoles, G.; Polrier, G. E.; Tarlov, M. J. J. Chem. Phys. 1994, 101, 11031. (13) Camillone, N.; Chidsey, C. E. D.; Liu, G.-y.; Putvinski, T. M.; Scoles, G. J. Chem. Phys. 1991 , 94, 8493. 84 (14) Leung, T. Y. B.; Gerstemberg, M. G.; Lavrich, D. J.; Scoles, G.; Schreiber, F.; Polrier, G. E. Langmuir 2000, 16, 549. (15) Eisenberger, P.; Fenter, P.; Liang, K. 8. Phys. Rev. Lett. 1993, 70, 2447. (16) Alves, C. A.; Smith, E. L.; Porter, M. D. J. Am. Chem. Soc. 1992, 114, 1222. (17) Delamarche, E.; Michel, B. Thin Solid Films 1996, 273, 54. (18) Poirier, G. E.; Pylant, E. D. Science 1996, 272, 1145. (19) Polrier, G. E. Langmuir1999, 15,1167. (20) Yamada, H.; Uosaki, K. Langmuir1998, 14,855. (21) Toerker, M.; Staub, R.; Fritz, T.; Schmitz-Hubsch, T.; Sellam, F.; Leo, K. Surf. Sci. 2000, 445, 100. (22) Hemminger, J. C.; Feher, F. J.; Dishner, M. H. Langmuir1997, 13,2318. (23) Truong, K. D.; Rowntree, P. A. J. Phys. Chem. 1996, 100,19917. (24) Revell, D. J.; Chambrier, l.; Cook, M. J.; Russell, D. A. J. Mater. Chem. 2000, 10,31. (25) Hostetler, M. J.; Stokes, J. J.; Murray, R. W. Langmuir1996, 12,3604. (26) Chidsey, C. E. D.; Loiacono, D. N. Langmuir1990, 6, 682. (27) Porter, M. 0.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc 1987, 109, 3559. (28) Hutt, D. A.; Leggett, G. J. Langmuir1997, 13,3055. (29) Heister, K.; Allara, D. L.; Bahnck, K.; Frey, S.; Zhamikov, M.; Grunze, M. Langmuir 1999, 15, 5440. (30) Leggett, G. J.; Hutt, o. A. Langmuir1997, 13, 3055. 85 (31) Bruening, M.; Cohen, R.; Guillemoles, J. F.; Moav, T.; Libman, J.; Shanzer, A.; Cahen, D. J. Am. Chem. Soc. 1997, 119.5720. (32) Whitesides, G. M.; Hickman, J. J.; Ofer, D. J. Am Chem. Soc1991, 113, 1128. (33) Yu, H. 2.; Ye, 8.; Zhang, H. L.; Uosaki, K.; Liu, 2. F. Langmuir2000, 16, 6948. (34) Poirier, G. E.; Tarlov, M. J. Langmuir1994, 10, 2853. (35) Fenter, P.; Eberhardt, A.; Eisenberger, P. Scinece1993, 1994. (36) Camillone III, N.; Chidsey, C. E. D.; Liu, G.-y.; Scoles, G. J. Chem. Phys. 1993, 98,3503. (37) Poirier, G. E.; Tarlov, M. J.; 'Rushmeier, H. E. Langmuir1994, 10, 3383. (38) Bain, C. D.; Biebuyck, H. A.; Whitesides, G. M. Langmuir1989, 5,723. (39) Molecular mechanics calculations were performed by using SPARTAN 5.0 molecular modeling program (Wavefunction Inc., CA), based on empirical Merck force fields. (40) lshita, T.; Hara, M.; Kojima, l.; Tsuneda, S.; Nishida, N.; Sasabe, H.; Knoll, W. Langmuir1998, 14,2029. (41) Maroncelli, M.; Oi, S. P.; Strauss, H. L.; Snyder, R. G. J. Am. Chem. Soc. 1982, 104, 6237. (42) Hostetler, M. J.; Stokes, J. J.; Greens, S. J.; Murray, R. W. J. Am. Chem. Soc. 1996, 118, 4212. (43) Edinger, K.; Golzhauser, A.; Demota, K.; Woll, C.; Grunze, M. Langmuir 1993, 9, 4. 86 (44) Polrier, G. E. Langmuir1997, 13,2019. (45) Schonenberger, C.; Sondag—Huethorst, J. A. M.; Jorritsma, J.; Fokkink, L. G. Langmuir1994, 10, 611. (46) Kondoh, H.; Kodama, C.; Sumida, H.; Nozoye, H. J. Chem. Phys. 1999, 11,1175. 87 Chapter 5 6-Phenyl-n-hexanethiol and 6-(p-Vinylphenyl)-n-hexanethiol Self- Assembled Monolayers on Au(111) Abstract Self-assembled monolayers (SAMs) of 6-phenyl-n-hexanethiol (PHT) and 6-(p-vinylphenyl)-n-hexanethiol (VHT) on Au(111) have been investigated by reflection-absorption infrared spectroscopy (RAIRS), ellipsometry, and scanning tunneling microscopy (STM). Both molecules chemisorbed as thiolates. The packing order and structural changes of the PHT monolayer were investigated at room temperature and following annealing in ultrahigh vacuum. Three different stripe phases (8, x’, and (3), characterized by molecular axes oriented almost parallel to the surface plane, were observed by STM. In contrast, the VHT monolayer had a structure in which the average molecular tilt angle was close to the surface normal. Polymerization of the VHT SAM, as followed by RAIRS, was achieved by either UV-light irradiation or thermal treatment. Ultraviolet irradiation produced longer chain polymers with a maximum of ~70% conversion, whereas annealing produced shorter chain polymers with CH3 as the end group. The UV- light polymerized film was more robust than the thermally polymerized film. 88 5.1. Introduction Self-assembled monolayers (SAMs) have received a great deal of attention because of their fundamental importance in understanding interfacial properties and for their potential application in molecular—based technologiesJ-2 One of the most popular systems has been monolayers of (lo-substituted thiols of the form HS-(CH2)n-X chemisorbed on Au(111). Such substituted SAMs modify both the chemical and physical properties of the substrate surfaces to which they bond and can be used to provide a variety of chemically tailored surfaces. Moreover, it has long been recognized that the (ii-terminus can provide a convenient attachment point for subsequent chemistry. In this way, derivatized surfaces or multilayer assemblies may be produced. The dependence of the monolayer morphology on the chemical nature of the endgroup has been studied for a range of terminal groups including X = CH3, COOH, CH=CH2, CF3, OH and NH3"CI‘.3'12 The first three endgroups result in similar hexagonal (\13X\/3)R30°-based structures, whereas CF3-terminated SAMs show only short-range order. Both OH- and NH3*CI“-terminated SAMs adopt unique packing structures, which are fundamentally different from the hexagonal lattice. It can be concluded that, in addition to surface-adsorbate interactions, the packing arrangement of (D-SUbStitUted thiols is influenced by the interactions between both alkyl chains and endgroups. We are interested in studying alkanethiol SAMs terminated by aromatic groups. The aromatic group is much less flexible than an alkyl chain and the dominant interactions between aromatic moieties may include face-to-face or edge-to-face rt-bonding contributions in 89 addition to the simple van der Waals’ forces believed to dominate the packing of alkyl chains.13'15 Furthermore, the presence of a bulky aromatic group may sterically prevent the formation of the well-documented (13x13)R30° close- packed monolayer structure. Azobenzene derivatives have been studied as model SAMs in order to determine the magnitudes of the aromatic and aliphatic interactions and their effect on monolayer structureflfi-2o In general, the alkyl portions of the chains are somewhat disordered, and the packing is dominated by attractive interactions between the aromatic portions of the molecules. Interestingly, minor molecular modification produces major differences in the morphologies of the SAM, suggesting that in these systems there is a delicate balance between aromatic, aliphatic and substrate interactions. Simple alkanethiols can form dense, well-packed monolayers on some metal surfaces; however, a serious deficiency for practical applications is their thermal and mechanical fragility. The incorporation of unsaturation into the SAM provides the opportunity for subsequent chemistry either through attachment of another species or by intralayer cross-linking reactions (oligomerization/polymerization). Increasing the lateral interaction between the chains by covalent bonding can improve the robustness of the SAMs.21 After Spontaneous assembly of thiols onto a gold substrate, cross-linking of adjacent molecules can be initiated by chemical, photon, or thermal means. However, in highly ordered monolayers, such as SAMs, geometric structure, molecular motion, and diffusion are likely to play important roles in determining reaction 90 probability and specificity. As in solid state polymerization, structural control of potential reaction geometry is expected, and in appropriately organized systems, there may be minimal disruption of the monolayer structure by the interconnection process. In principle, polymerization reactions in SAMs can be used to produce crystalline macromolecular thin films. There have been previous studies of photoinduced polymerization of adsorbed monomers on metal surfaces. Ford et al. formed polymerizable monolayers by adsorbing 4-(mercaptomethyl)styrene on a roughened silver surface.22 The orientation of the benzene ring plane in the styrene moiety was deduced from band intensities in surface-enhanced Raman spectroscopy (SERS) to be slightly inclined from the surface normal. Subsequently, 514 nm laser light was used to both initiate and probe photoreactivity through SERS, and a polymerization reaction was observed. Peanasky and McCarley have studied undec—tO-ene-1-thiol/Au SAMs irradiated by ways-'33 The polymerization of the monolayers during y-ray exposures was indicated by the decrease in the intensities of the infrared bands associated with the olefin functionality. Some disordering of the monolayer occurred during the reaction. It was proposed that the polymerization reaction was controlled by the distance that the tethered olefin groups were able to move. Several groups have studied UV irradiation of self- assembled monolayers containing diacetylene units.?—""28 The incorporation of conjugated diacetylene groups within thiol or disulfide compounds has permitted the fabrication of robust monolayer polymers that are more durable and better barriers to electron transfer than the unpolymerized monolayers. It was 91 suggested that the diacetylene groups were able to undergo topochemical (structure-controlled) reaction in these monolayers. In this chapter, we present our study of phenyl- and styrene-terminated hexanethiol SAMs on Au(111). The phenyl-tenninated molecule was used to examine the influence of a simple aromatic terminus on monolayer packing order. The addition of an alkene to the phenyl group (to produce a styrene- tenninated SAM) allowed us to study the thermal stability and photopolymerization processes. In both cases, the monolayer thickness and the molecular orientation were determined by ellipsometry and reflection-absorption infrared spectroscopy (RAIRS), respectively. Film morphologies were characterized by scanning tunneling microscopy (STM). Chemical and structural changes induced by thermal or UV irradiation of the SAMs were followed by RAIRS. 5.2. Experimental Synthesis. The synthesis of 6-(p-vinylphenyl)-n-hexanethiol (VHT) was carried out by sequential alkylation and thiolization reactions as follows: (4- Vinylphenyl)magnesium bromide was prepared from 4-bromostyrene (2.32 g, 12.7 mmol) and magnesium metal tumings activated with a few crystals of l2 in 30 mL of dry THF.29 Once the reaction was initiated, dry ice cooling was applied to minimize potential polymerization. This Grignard solution was added to a solution of 1,6-dibromohexane (2.78 g, 11.4 mmol) and L12CUCI4 (0.184 g, 0.72 mmol) in 30 mL of THF.30 The reaction temperature was kept at 0 °C in an ice- water bath for 20 min and then allowed to warm to room temperature. After 3 h, 92 the reaction mixture was filtered through silica gel and the solvent was removed under reduced pressure. The crude product was purified by silica gel chromatography. 4-(6—Bromohexyl)styrene was obtained as an oily product at 36% yield. Next, to a dried and argon-filled Schlenk flask, sodium bisulfide dihydrate (0.92 g, 10.0 mmol) and the solution of 4-(6—bromohexyl)styrene (0.89 g, 3.3 mmol) in 50 mL of acetone were added. The reaction mixture was stirred at room temperature for 6 h and then filtered through silica gel. Solvent was removed under reduced pressure without air exposure. The flask containing an oily residue was moved into adrybox filled with helium. The crude product was separated on a silica gel column (28x23 cm, 230-400 mesh) using hexane- chloroform (2:1) giving VHT as a clear oil in 45% yield. The structure of VHT was confirmed from 1H NMR spectra taken in CDCI3 media; 6 values were 8 7.35-7.05 (2d, aryl 4H), 6.66 (quartet, vinyl 1H), 5.75-5.15 (2d, vinyl 2H), 2.57 (t, 2H), 2.49 (quartet, 2H), 1.65-1.52 (m, 4H), and 1.44-1.27 (m, 4H + SH). 6-Phenyl-n— hexanethiol (PHT) was synthesized in 84% yield from 1-bromo-6-phenylhexane (Lancaster Synthesis Inc.) under the same thiolization reaction condition as described above. Purity was confirmed by 1H NMR and GC-MS. Substrate and Monolayer Preparation. The gold substrates for infrared spectroscopy and ellipsometry measurements were made by electron-beam evaporation of 200 nm of Au on 20 nm of Ti on Si(100) wafers. The gold substrates for STM measurements were prepared by thermally evaporating gold onto freshly cleaved mica in high vacuum. In both cases, STM revealed the clean surfaces to be composed of Au grains with (111)-textured terraces (1000- 93 2500 A) separated by monoatomic steps. All of the substrates were cleaned by a UV/Oa cleaner for 15 minutes, followed by soaking in deionized water for 30 minutes. After this, the substrates were dried in flowing N2 and immediately ' transferred into a N2-filled glovebag where they were immersed in 2 mM CH2CI2 solutions of the adsorbates for more than 5 hours. The resultant SAMs were sonicated in CH2CI2 to remove excess adsorbate from the surface and dried in a N2 stream before characterization. Surface Characterization. Ellipsometry, reflection-absorption infrared spectroscopy (RAIRS), and scanning tunneling microscopy (STM) were applied to characterize the structure, stability, and reactivity of PHT and VHT SAMs. For detailed descriptions of these techniques and their working conditions, please ' refer to Chapter 2 and 4. 5.3. Results and Discussion Ellipsometry Measurements. Optical ellipsometry was used to determine the average thickness of the various films. After soaking for >5 h, the measured thicknesses of the PHT and VHT SAMs were 612 and 1412 A, respectively. These results were obtained reproducibly on multiple samples. Molecular mechanics calculations31 were used to calculate the minimized energy geometry for PHT and VHT and in both cases, the plane of the phenyl ring was approximately orthogonal to the plane of the all-trans methylene chains. Based on these calculations, the distance from the gold surface to the (o-terrninal hydrogen was estimated to be 14.2 A for PHT and 16.3 A for VHT. This immediately indicated that in the PHT SAM, the molecules were inclined with an 94 apparent average tilt angle of about 65° from the surface normal. It should be noted that this tilt angle is misleading since STM data (discussed below) revealed that PHT molecules formed a complex monolayer structure. In contrast, theSAM thickness values obtained for VHT indicated a more perpendicular molecular orientation (tilt angle of ~30°, similar to alkanethiols”) with respect to the underlying Au(111) surface plane. The surprisingly large difference between PHT and VHT molecular orientations suggested that the vinyl group attached to the para position of the benzene ring in VHT contributed significantly to the interchain interactions in these SAMs. IR Measurements. Parts (B) and (D) of Fig. 5.1 show the RAIR spectra of VHT and PHT monolayers after extended (>8 h) self-assembly on gold, respectively. For comparison, the neat VHT and PHT liquids' infrared spectra are shown in Fig. 5.1 (A) and (C). All peaks in these spectra can be assigned by consulting the data in the literature,33'35 and the results are summarized in Table 5.1. 95 Absorbance é X1 lgLn l a_l_a l a l__.|_.|_.|_|._|_/ 3200 3000 2800 r1600 1400 1200 1000 800 wavenumbers (cm") Fig. 5.1 IR spectra of (A) neat VHT; (B) VHT monolayer on Au/Si wafer; (C) neat PHT and (D) PHT monolayer on Au/Si wafer. The arrows indicate the sharp intensity drop of vC=C, 8=CH2, w=CH-, m=CH2 and benzene ring 17b mode, respectively, for the VHT monolayer. 96 Table 5.1 Infrared Band Assignments for PHT and VHT Wavenumbers (cm") PHT PHT/Au VHT VHT/Au Assignment 11212112: Neat liquid SAM Neat liquid SAM 3084 3089 \r,,,=CH2 3084 3088 3047 3051 vCH(arom) 20a 3061 3068 2 3025 3032 301 8 3020 20b 3006 3009 v=CH- 2977 2981 V3=CH2 2928 2927 2928 2928 v,,,,cH2 2855 2857 2855 2856 VSCHz 1 629 1 630 vC=C 1604 1 606 1 609 1608 vCC(ring 8a stretch) 1584 o o 0 8b 1496 1498 1511 1511 198 1453 1450 O 1 422 1 9b 1 465 1 465 1 463 1461 80112 1407 1409 6=CH2 1030 1032 1017 1018 6CH(arom) 18a 990 990 w=CH- 905 905 w=CH2 844 844 wCH(arom) 17b 826 826 vCC(arom) 1 o-overlapped band, v-stretching vibration, 6—in-plane vibration, w-out-of-plane vibration. 97 The RAIR spectra taken from films prepared by a 30 min self-assembly were almost identical to those taken from films prepared by up to 24 h self- assembly. This suggests that in common with conventional alkanethiols, a rapid initial adsorption occurred for these mixed aromatic/aliphatic thiols. A weak vS-H feature observed at 2572 cm'1 in the liquid RAIR spectra but not in the monolayer spectra, confirmed that both PHT and VHT dissociatively chemisorbed as thiolates. Modes Associated with the Methylene Backbone. It is well-known that the peak positions of the symmetric and the asymmetric CH2 stretching vibrations can be used as a sensitive indicator of the alkyl chain ordering.3-32 Lower wavenumbers indicate highly ordered conformations with preferential all-trans characteristics (pseudo crystalline), and for well ordered alkanethiol SAMs. the vsCH2 and vasCH2 modes are usually observed below 2850 and 2920 cm", respectively. As shown in Fig. 5.1, the vsCH2 and vasCH2 modes for PHT and VHT neat liquids were observed at 2855 and 2928 cm“, respectively. Upon monolayer formation, although there was some narrowing of these bands, they remained almost completely unshifted with respect to the liquid phase (at vsCH2 ~2857 cm'1 and vasCH2 ~2927 cm"). This implies that in the adsorbed state on Au(111), the aliphatic chains for PHT and VHT were disordered (liquidlike), in marked contrast to alkanethiols with similar overall molecular length (octanethiol) on Au(111). Modes Associated with the PhenyI Ring. There are five infrared-active aromatic C-H modes in the 3100-3000 cm‘1region: 20a, 2. 7b, 20b and 7a modes 98 in approximate order of decreasing energy.34 Of these, only the 20a, 2 and 20b modes are commonly observed in substituted benzenes. For the PHT neat liquid, the bands at 3084, 3061 and 3025 cm‘1 were assigned to benzene ring modes 20a, 2 and 20b, respectively. They were in the intensity ratio of approximately 1/3.2/5.5. For the PHT SAM, the bands at 3088, 3068 and 3032 cm'1 were similarly assigned but the intensity ratio changed to 1/7.7/5.2. Clearly, the intensity ratio of modes 20a and 20b remained constant but both band intensities decreased markedly with respect to mode 2 after formation of the monolayer on Au(111). According to the infrared surface selection rule, only vibrations with transition dipole moments oriented perpendicular to the metal substrate are observed. Therefore, peak intensities are directly related to the component of each transition moment that is perpendicular to the metal surface (affected by both molecular tilt and twist angles). It should be noted that changes in the intrinsic oscillator strength, for example by a perturbation of the vibrational potential energy surface induced by changes in layer packing, will also affect the intensity of an IR absorption band. This effect is ignored in the present discussion. The dipole moment changes associated with modes 2, 20a and 20b are all in the plane of the phenyl ring. Simplifying the symmetry of PHT to C2,, the transition dipole moment of mode 2 (a1 symmetry) is aligned parallel and modes 20a and 20b (b2 symmetry) are aligned perpendicular to the 1.4-axis of the benzene ring.36-38 Since mode 2 was clearly visible in the spectrum of the PHT SAM, the phenyl 1,4-axis was not parallel to the surface plane. 99 Furthermore, the decrease in mode 20a and 20b intensities relative to mode 2 indicated that the in-plane phenyl direction, perpendicular to the 1,4-axis, was 1 almost parallel to the surface plane. This conclusion was also supported by intensity changes in the aromatic C-C stretch region between 1400 and 1610 cm". Figure 5.1 shows that upon formation of the PHT monolayer from the PHT solution, mode 19b (1453 cm", b2 symmetry) decreased relative to mode 198 (1496 cm“, a1 symmetry).35 It Should be noted that modes 8b (b2 symmetry) and 8a (a1 symmetry) should also show the same effect, but interpretation was complicated since these two bands were overlapped in the spectra presented in Fig. 5.1 (D). Similar orientational arguments can be made for the VHT monolayer, but measurement of the intensities of modes 2, 20a and 20b was complicated by the proximity of vibrational bands associated with the vinyl terminus of this molecule. The aromatic C-H stretching vibrations shifted to lower energy by 10-20 cm'1 for VHT relative to PHT because these modes involve some motion of the 1,4 position C atoms and so are sensitive to substitution at the para position.33 Consequently, the bands at 3047 and 3018 cm'1 were assigned to benzene ring modes 20a and 20b in solution phase VHT. Mode 2 is strictly IR inactive for identically p-disubstituted benzenes,33 so this mode was expected to be weak in all VHT spectra. The aromatic C-C stretching modes for the VHT neat liquid were observed at approximately 1609 (8a) and 1511 cm‘1 (19a). Unfortunately, the complementary modes 8b and 19b could not be clearly distinguished from a 100 broad to con based distinc counle 5.1 (B) an ode out-of- were e vinyl g termini 3084 2 Sim (\’=CH Plane - Charac 905 C11 belwei with less in leaks on'enla broad envelope of methylene modes in the 1400-1600 cm'1 region. It is difficult to comment confidently on the relative orientation of the phenyl group in VHT based on these bands. The out-of-plane benzene ring 17b mode appeared very distinctly at 844 cm'1 in the neat liquid spectrum of VHT (Fig. 5.1 (A)), but its counterpart was greatly diminished in the RAIR spectrum of the monolayer (Fig. 5.1 (B)). This observation suggests that the plane of the benzene ring adopted an orientation almost perpendicular to the gold substrate. Unfortunately, other out-of-plane modes (for example, mode 17a) that might support this hypothesis were either very weak or were obscured by features associated with the alkyl or vinyl groups of the VHT molecule. Modes Associated with the Vinyl Group of VHT. For the VHT liquid, the terminal vinyl group gave rise to seven readily observable IR bands. The band at 3084 and 2977 cm'1 can be assigned to the =CH2 asymmetric (vas=CH2) and symmetric (vs=CH2) modes, and that at 3006 cm’1 is due to the vinyl C-H stretch (v=CH-). The C=C stretching vibration (vC=C) occurs at 1629 cm", and the in- plane CH2 deformation (8=CH2) is at 1407 cm". The vinyl group also has two characteristic hydrogen wag (out-of-plane) vibrations at 990 cm'1 (m=CH-) and 905 cm'1 ((0=CH2). Of significance is the difference in the ratio of peak intensities between the RAIR spectrum of the VHT monolayer and the absorbance spectrum of the liquid. The vC=C, 5=CH2, w=CH- and w=CH2 peaks were much less intense in the VHT monolayer spectrum whereas the vas=CH2 and v=CH- peaks were more intense. Since the intensities of these modes are related to the orientation of the vinyl group through the IR transition dipole moment (shown 101 schematically in Fig. 5.2), the approximate orientation of the vinyl moiety can be deduced. 1412 A (measured) Fig. 5.2. Schematic diagram indicating the directions of the various transition dipole moments for the vinyl group of VHT. Both vC=C and 8=CH2 modes are parallel to the vinyl double bond, while vas=CH2 mode is in the H-C-H plane, perpendicular to the vC=C and 6=CH2 modes. The out-of-plane wag modes, co=CH- and w=CH2, are orthogonal to the other three modes. The transition dipole moments for the vC=C and 5=CH2 modes are parallel to the vinyl double bond, whereas that for the vas=CH2 mode is in the H-C-H plane, perpendicular to the vC=C and v=CH2 modes. The direction of the transition dipole moments for the out-of-plane wag modes, (0=CH- and 03=CH2, is orthogonal to the other three modes. The band intensity changes observed upon monolayer formation were consistent with an orientation in which the vinyl group double-bond was almost parallel to the Au(111) plane, with the vinyl terminal H- C-H plane almost parallel to the surface normal. STM Measurements. To obtain further information on the structure of the PHT and VHT films, STM measurements were performed. The STM images of the PHT/Au(111) monolayer, acquired at room temperature after soaking in the PHT solution for at least 24 h, are shown in Fig. 5.3. 102 PnhtBl 40 ml i O 0 20 40 60A (C) 6 phase Facial Room T Fig. 5.3 (A) STM images (950 x 950 A (inset, 200 x 200 A)) of PHT monolayers on Au/mica at room temperature show the 8 stripe phase. (B) Sectional view alone the line in Fig. 5.3 (A) reveals the row spacing of 16 A. (C) A schematic representation of the 8 phase. 103 Figure 5.3 (A) shows many 2.5102 A deep “vacancy islands”. These monatomic depressions appeared to be similar in number density, size and shape to those formed in simple alkanethiol SAMsF‘m-‘f0 Between the vacancy islands, the monolayer was composed of domains of molecular rows oriented 12015° with respect to each other. The domains appeared to be at least partially bounded by the vacancy islands. The rows in each domain had a 1611 A corrugation period. This value is larger than the calculated single molecule length (12.6 A) but much less than twice the molecular length. By analogy with annealed or gas-phase deposited alkanethiol films, we believe that the packing pattern adopted by the PHT film resembled the so-called 8 phase of alkanethiol SAMs.41 '43 In the 5 phase, the axes of one row of molecules are oriented parallel to the substrate surface, while a pairing row is tilted out of the surface plane, forming gauche defects near the sulfur terminus. A schematic representation is shown in Fig. 5.3 (C). In this way, both S-S and alkyl-alkyl interactions are maximized. Unfortunately, at higher resolution, shown in Fig. 5.3 (B), no molecular resolution was observed and the exact geometric arrangement of the phenyl-phenyl and S-S interactions cannot be determined. A room temperature large area STM image of a VHT monolayer is shown in Fig. 5.4. In common with the room temperature PHT monolayer, it displayed monatomic deep depressions on the surface. However, no molecular-scale information was obtained from VHT monolayers. This failure probably resulted from disorder in the initial monolayer. As discussed previously, IR measurements suggested that at least part of the VHT molecule (the hexyl chain) 104 was in a liquidlike environment. It was also likely that the vinyl terminated SAM was sensitive to the tunneling current used to image the layer. As such, electron- induced reactions on the surface might contribute to the apparent poor order observed. A range of experimental tunneling conditions (tip bias from 0.7 to 3.0 V and currents down to <50 pA) were investigated but none produced significantly improved images. Fig. 5.4 Typical room temperature STM image of VHT monolayers on Au/mica (950 x 950 A). Photoreactivity of PHT and VHT Monolayers. The photoreactivity of PHT and VHT monolayers was examined by exposing them to UV light from a low pressure Hg arc lamp (A~250-400 nm). Initially the VHT SAMs on gold were placed in a quartz tube which was purged continuously with pure N2 during UV irradiation. However, RAIR data showed a strong carbonyl stretch at 1737 cm"1 in the films after irradiation, indicating that photo-oxidation reactions occurred in the films (probably due to the trace amount of 02 left in the tube).44.45 105 To avoid the photo-oxidation, the SAMs were then placed in a load-lock which was pumped down to 1x10'6 Torr. The UV light shined through a quartz window right above the sample. No carbonyl stretches were observed in the irradiated films under this photopolymerization condition; however, dramatic increases in C-H stretching modes in the VHT SAMs were noticed after only 2min UV irradiation (Fig. 5.5). As the control experiment, the PHT SAMs were irradiated under identical conditions. Similar results were observed in the C-H high frequency region in the irradiated PHT SAMs. suggesting that the abnormal increases of C-H modes were probably due to the contamination of pump oil vapon Absorbance 0min ‘ g l e l a l a l a l a 3200 3100 3000 2900 2800 2700 2600 Wavenumbers (cm*‘) Fig. 5.5 RAIR spectra of VHT monolayers on Au/Si wafer obtained at various UV irradiation time with the sample placed in a load-lock. In order to check the film photoreactivity in an O2- and contamination-free environment, we soaked the SAMs in pure CH2CI2 (solvent) in vials and purged 106 N2 continuously before and during UV irradiation. Under these conditions, no oxidation or contamination occurred in either the PHT or VHT SAMs during irradiation. Reactivity (consumption of monomer) in the VHT monolayer was observed by noting a decrease in integrated IR band intensity for all modes associated with the vinyl group, consistent with vinyl polymerization. For the PHT monolayer, neither band intensities nor positions for the PHT methylene backbone and benzene ring modes changed upon UV irradiation for up to 45 min, which ensured that the spectral changes noted for VHT monolayers were due to reactions involving only the vinyl functionality. A representative series of RAIR spectra of VHT monolayers are shown in Fig. 5.6 as a function of UV exposure. The peak intensities of the vinyl modes all decreased as expected for cross-linking oligomerization reactions. During UV irradiation, intensity changes of the benzene ring modes pointed to molecular orientation changes in the VHT monolayer. The in-plane skeletal stretch at 1511 cm“1 (mode 19a) decreased markedly, implying that the phenyl ring plane became more parallel to the surface plane during VHT oligomerization (as mentioned above, no changes in phenyl mode intensities were observed during PHT monolayer irradiation). Unfortunately, additional supporting information provided by VHT mode 19b/88/8b intensity ratios was not obtainable: mode 19b was strongly overlapped by the =CH2 deformation and modes 8a and 8b could not be resolved as separate features. 107 Absorbance 0min l I 1A4 11 a4 . 1_A L_a 3200 3100 3000 2900 2800 2700 2600 Wavenumbers (cm ") L 'vccriing strI—et) ’5 ' 7" 737 vC=C j1x104 5:sz 0)=CH2 I ‘ I I I l . 1 12W ‘ 1 Absorbance s ‘27 I I 0min I I 1 I I I I J 4114 n II_L lu# ”Li 1700160015001400 71001000 900 800 Wavenumbers (cm '1) Fig. 5.6 RAIR spectra of VHT monolayers on Au/Si wafer obtained at various UV-light exposure time while soaked in the solvent. The band intensities of various vinyl modes all decrease with increasing exposure to UV-light. 108 Confirmation that the average molecular tilt angle increased during photopolymerization of VHT monolayers was provided by pre- and post-UV exposure ellipsometry. These measurements indicated that the thickness of the film decreased from 1412 to 1111 A following >15 min of UV irradiation. Combined with the IR intensity changes noted above, this suggests that there was some reorientation occurring within the film involving an increase in the average molecular tilt angle from ~30° to ~48° during polymerization. Similar reorientation was also observed by Peanasky and McCarley following photopolymerization of undec-10-ene-1-thiol SAMs on Au?-3 The absolute peak intensity change of the vas=CH2 and vC=C modes against UV exposure time was quantified in Fig. 5.7 (A). Both peak intensities decreased and reached a minimum after approximately 15 min, at which time about 30% of the vinyl bonds remained unreacted on the surface, presumably due to steric or geometric effects. No new spectral features appeared during UV irradiation. In contrast to the decreased vinyl group peak intensities, Fig. 5.7 (B) shows that the peak intensity for methylene (vasCH2 and vsCH2) modes increased during UV irradiation. Peanasky and McCarley noted similar effects in their study of the polymerization of undec-10-ene-1-thiol SAMs on Au by y-ray irradiation.23 In their work, the decrease in vinyl intensity and increase in vasCH2 and vsCH2 intensity was attributed to reaction accompanied by reorientation of the alkyl chains needed to accommodate coupling of the vinyl monomer groups. However, alkyl reorientation was not the only contribution to increased methylene 109 intent new i also I as ob with f mono photo chain contril Fit. 5. Fe 5. 1"0:0 intensity after oligomerization. During irradiation, the vinyl bonds reacted and new cross-linking methylene units were created. These new CH2 groups would also contribute to an apparent increase in the intensity of the methylene bands, as observed during photopolymerization of styrene on Ag(110).45 In conjunction with the intensity changes noted in the phenyl ring modes and the changes in monolayer thickness measured by ellipsometry, we concluded that during photopolymerization, both the new methylene units in the growing poly(styrene) chain and the orientation change of the hexyl portion of the VHT molecule contributed to the increased methylene intensity. % Q) g 0 (A) g -10 —A—v”=CH2, 3089 cm“ -20 _ .z-fiso +vc=c,1830cm‘ 2 Q «to .5 .50 a 0 -60 0- -70 -90. 2 1 g L L UV light exposure time (min) a, —A— vaCHz, 2928 cm" 80 r -e— V.CH2' 2856 cm'1 (B) Peak intensity change 0 0‘3110‘13‘23‘723 UV light exposure time (min) Fig. 5.7 Plot of peak intensity change for VHT monolayers on Au/Si wafers from Fig. 5.6 as a function of UV-light exposure time: (A) changes in the Vas=CH2 and vC=C modes and (8) changes in the vasCH2 and vsCH2 modes. 110 Thermal Stability of PHT and VHT Monolayers. The thermal stability of PHT and VHT monolayers was also tested by controlled thermal treatment in a pure N2 environment. Upon annealing the PHT/Au(111) SAM to elevated temperatures, the vacancy islands noted at room temperature became larger and less frequent. After heating to 60 °C the average diameter of the depressions was 100-200 A and contained an identical row or striped 8 phase (Fig. 5.8). Fig. 5.8 STM images of PHT SAMs at 40 °C (A: 1500x1500 A, B: 500x500 A) and 60 °C (0: 1500x1500 A, D: 450x450 A). After heating to 80 °C, as shown in Fig. 5.9, the vacancy islands disappeared and the surface was completely covered by parallel rows of features 111 oriented at approximately 120° with respect to each other. The domains were much larger than at room temperature and were terminated by sharp boundaries. Again, by analogy with the phases observed for annealed alkanethiols on Au(111), we assigned this structure as the x’ phase,41 shown schematically in Fig. 5.9 (C). In our PHT SAM, the x’ phase was characterized by parallel stripes, randomly separated by either 1611 or 2811 A. We attributed the 16 A stripe spacing as a similar structure to the 8 phase. We believed that the 28 A spacing (which is almost exactly twice the calculated molecular length plus a S-S distance of 4 A41) was due to alternating head-to-head and tail-to-tail molecules arranged with their molecular axes almost parallel to the surface plane. An interesting feature of this phase was that the distance between rows was observed to switch from 28 to 16 A, or vice versa, at random places along the row, as shown in Fig. 5.9 (B). A plausible explanation for this phenomenon involves the expansion of a 16 A row or compression of a 28 A row, followed by the lateral shift of the entire stripe segment as proposed by Toerker et al. for decanethiol phases on Au(111).41 112 (B) 50 100 150 200A (C) x’ phase E16 $2871T5| 80°C Fig. 5.9 (A) STM images (800 x 800 A (inset, 250 x 250 A)) of PHT monolayers on Au/mica at 80 °C show the x’ stripe phase. The arrows indicate the switching of stripe segments. (B) Sectional view alone the line in Fig. 5.9 (A) reveals the row spacing of either 16 A or 28 A. (C) A schematic representation of the x’ phase. 113 The x’ phase appeared to be an intermediate state between the 8 phase and the [3 phase observed at higher annealing temperatures. After the formation of the x’ state, the continued annealing of PHT monolayers on Au(111) resulted in surfaces covered with the [3 phase in which all the molecular axes were parallel to substrate surface plane with sulfur atoms paired and chains packed head-to-head and tail-to-tail (Fig. 5.10). After annealing to 120 °C, as shown in Fig. 5.10 (A), large single domains of striped [3 phase characterized by a row separation of 2811 A were visible. The ordered regions of the monolayer contained a few short sections of the narrow 8 phase rows as illustrated in Fig. 5.10 (B). Additionally, significant fractions of the surface were covered by a disordered or mobile liquidlike phase, visible in the lower right-hand comer of Fig. 5.10 (A). We also acquired STM images of VHT monolayers annealed up to 120 °C. The images were qualitatively very similar to Fig. 5.4 at all temperatures. Significantly, at no time did we observe evidence of row or stripe phases typically observed for annealed conventional alkanethiols and indicative of a reduced surface coverage.“ The increased stability of the VHT monolayer versus the PHT monolayer probably resulted from the thermal polymerization discussed below. 114 0 50 100 150 200 A (C) B phase ls—28 AA—Agl L 120°C Fig. 5.10 (A) STM images (1000 x 1000 A (inset, 250 x 250 A)) of PHT monolayers on Au/mica at 120 °C show the B stripe phase. (B) Sectional view alone the line in Fig. 5.10 (A) reveals the row spacing of 28 A. (C) A schematic representation of the 0 phase. 115 The annealing process of VHT SAMs was also followed by RAIRS. Figure 5.11 shows the 2700-3200 cm'1 region RAIR spectra of an unpolymerized VHT monolayer as a function of annealing temperature.47 200°C 123 . E 8 130°C .0 < 100°C 60°C 40°C 20°C 3200 3000 . 2800 Wavenumbers (cm") Fig. 5.11 RAIR spectra of adsorbed VHT monolayers on Au/Si wafer obtained following annealing at the indicated temperature. The peak intensity of all modes associated with the vinyl bond decreased with increasing temperature: the vas=CH2 mode had effectively disappeared after annealing to 130 °C. Concurrently, a new feature at 2964 crn’1 attributed to VasCHa grew in as the annealing temperature increased. The appearance of the methyl mode was evidence that a fraction of the vinyl bonds opened during heating became terminated with CH3 and groups. It should be stressed that no methyl features appeared in the case of VHT photopolymerization (shown in Fig. 116 5.6 (A)), even after extended UV irradiation times. Although we cannot quantitate the degree of polymerization for the thermal and photopolymerized monolayers, it seems clear that themial activation produced shorter oligomers with a larger number of terminal methyl groups. Figure 5.12 shows the effect of annealing on a prephotopolymerized VHT monolayer (15 min UV irradiation). v CH 1x10'4 as 2 vaSCH3 N VSCHZ r l or C) 5 0c e , 200 g 160°C < l 130°C 100°C 60°C 20°C l l j 40°C I l I Ag 1* a l 3200 3000 2800 Wavenumbers (cm'i) Fig. 5.12 RAIR spectra of pro-photopolymerized VHT monolayers on Au/Si wafer obtained following annealing at the indicated temperature. The data are similar to those in Fig. 5.11 but the initial intensity of the vinyl group vas=CH2 mode at 3089 cm'1 was much reduced (approximately 30% of the vinyl bonds remained intact after photopolymerization). Annealing of the pre- phot0polymerized monolayer also resulted in the appearance of a feature assignable to methyl (vasCHa at 2964 cm") but at a slower rate than for the unpolymerized VHT film. 117 The integrated peak areas for the methylene vs and vas modes of the pre- photopolymerized and unpolymerized VHT monolayers are shown in Fig. 5.13 (A) as a function of annealing temperature. For the unpolymerized monolayer, both vasCH2 and vsCH2 modes increased up to 60 °C, but the vsCH2 mode increased faster than the vaSCH2 mode. Such behavior was also observed during the photopolymerization experiments, as discussed previously, and attributed to growing poly(styrene) oligomers together with orientational effects in the underlying molecule. We believe the growth of the vasCH2 and vsCH2 modes during thermal treatment was also related to oligomerization reactions within the monolayer. However, the absolute increase of both modes was less than that of the photopolymerized film. Ellipsometry indicated that the film thickness also decreased by 311 A during annealing, identical to the decrease observed during photopolymerization. If we assumed that the hexyl chain orientation change was similar for both thermal and photopolymerization, the less rapid increase of the vasCH2 and vsCH2 modes during thermal polymerization indicated a lower average degree of polymerization. As discussed above, the increased vasCHa mode intensity with annealing supported the idea that the polymers formed by thermal annealing were short-chain oligomers with CH3 end groups. Decomposition of the thermally polymerized monolayer above 60 °C was signaled by rapid decreases in both methylene modes and a continuous increase in the vasCHa mode up to 250 °C, as shown in Fig. 5.13 (B). The appearance of methyl vibrational modes implied cleavage of the poly(styrene) chains. 118 For the pre-photopolymerized monolayer, as shown in Fig. 5.13 (A), both vaSCH2 and vsCH2 modes decreased monotonically with increasing annealing temperature. The rate of decrease of these modes was much less rapid than for the unpolymerized film (above 60 °C), suggesting that the prephotopolymerized film had increased thermal stability. It will be recalled that the prephotopolymerized film contained about 30 % unreacted vinyl groups at the start of the annealing process. These bonds appeared to be thermally activated to produce methyl groups upon annealing up to 100 °C, as signaled by a rapid rise in the vasCHalvasCH2 intensity ratio shown in Fig. 5.13 (B). For all temperatures above 100 °C, this ratio remained constant, likely due to the consumption of all reactive vinyl groups. The compositional differences between the thermally polymerized and photopolymerized VHT SAMs are striking. Thermal polymerization produced a significant number of CH3 groups whereas these functionalities were almost completely absent for the photopolymerized monolayer. We believe that the most probable cause of these differences was related to additional molecular motion present in the VHT monolayer at elevated temperatures. Increased motion created a wide range of potential reaction geometries and, in turn, a reduced specificity for successful cross-linking. There appeared to be no significant change in the degree of order or gross structure of the VHT SAM during annealing as evidenced by the constant width and position of the methylene vs or V” bands and by the ellipsometric thickness changes (an identical decrease in film thickness was observed for both therrno- and 119 photopolymerized VHT SAMs). We can therefore tentatively discount the formation of a discrete high temperature phase with a structure not conducive to cross-linking as the cause of a reduced degree of polymerization during thermal versus photochemical polymerization. l— unpoly-v,CH2 % e— un 01 -v CH 1 Am prepon-\r,CH2 go 201 *...prepoly-quH2 8 0 j 4. >5 3: l 6:“ ~ M .‘é ‘20 ‘ “““““ ‘13 . a g -401 - ' 3071050 ' 1507 260 52er '300 Anneal temperature (°C) Am prepoly-vafiH3 0— unpoly-vasCI-I3 % .9 E 8:” U a > \M I U a > Anneal temperature (°C) Fig. 5.13 Plot of peak intensity change for unpolymerized and pre- photopolymerized VHT monolayers on Au/Si wafers from Fig. Fig. 5.11 and 5.12 as a function of anneal temperature: (A) changes in the vasCH2 and vsCH2 modes and (B) changes in the VasCHa mode (shown as the ratio of VasCHa peak intensity divided by that of initial unpolymerized vasCH2 mode). 120 5.4. Conclusion The self-assembly of mixed aliphatic/aromatic thiols, 6-phenyl-n- hexanethiol and 6-(p-vinylphenyl)-n-hexanethiol, on Au(111) has been investigated. We have characterized both SAMs by RAIR spectroscopy, ellipsometry and STM. The RAIR spectral data indicated that both molecules chemisorbed on gold as thiolates. however, the alkyl chains were disordered for both SAMs. Ellipsometry and infrared measurements suggested that the PHT monolayer had an average molecular axis close to parallel to the surface plane. Scanning tunneling microscopy revealed that the PHT SAM was composed of small domains of molecules arranged in rows. The average corrugation of the rows was 1611 A, consistent with a so-called 8 phase by analogy with the structures of simple alkanethiol SAMs. We investigated the structural changes of the PHT monolayer with annealing in ultrahigh vacuum and three different stripe phases (8, x’, and [3) were observed by STM. All three phases were characterized by alignment of the molecular axes with the surface plane. In contrast, the VHT monolayer had a structure in which the average molecular tilt angle was close to the surface normal. As with the azobenzene derivative SAMs,16-20 a small modification of the molecular structure (from PHT to VHT) changes the monolayer morphology completely. In our case, this effect could be simply related to the presence of an additional two C atoms in the case of VHT stabilizing the "standing up" orientation (as observed for hexanethiol versus octanethiol, on Au(111), for example). However, the disordered nature of the alkyl tethers of both PHT and VHT monolayers implied that aliphatic 121 stabilizing interactions did not completely dominate film morphology. The involvement of styrene-styrene contributions to the molecular orientation in styrene-containing SAMs can also be inferred from the standing up orientation deduced for ' much shorter p-alkyl styrene monolayers ((mercaptomethyl)styrene).22 Like the PHT SAM, the room temperature VHT SAM exhibited monatomic depressions (vacancy islands) as imaged by STM, but further details could not be resolved. Polymerization of the VHT SAM, as followed by RAIRS, was achieved by either UV-Iight irradiation or thermal treatment. We speculated that UV irradiation produced longer chain polymers up to a maximum of ~70% conversion. In contrast, thermal annealing produced shorter chain polymers with a large proportion of CH3 end groups. The UV-light polymerized film was more robust to degradation than the thermally polymerized film. 5.5. Literature Cited (1) Ulman, A. An Introduction to Ultra thin Organic Films from Langmuir- Blodgeft to Self-Assembly, Academic Press: San Diego, CA, 1991. (2) and ( 3) 2358. ( 4) 558. Whitesides, G.; M.; Gorman, C. B. In 7719 Handbook of Surface Imaging Visualization; Hubbard, A. T., Ed.; CRC Press: Boca Raton, FL, 1995; 713. Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987, 109, Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, (5) Leung, T. Y. B. Ph.D. thesis, Princeton University, 1998. 122 (6) Li, J.; Liang, K. S.; Scoles, G.; Ulman, A. Langmuir1995, 11,4418. (7) Hutt, D. A.; Leggett, G.J. Langmuir1997, 13, 2740. (8) Yan, L.; Marzolin, C.; Terfort, A.; Whitesides, G. M. Langmuir1997, 13, 6704. (9) Boubour, E.; Lennox, R. B. Langmuir2000, 16,7464. (10) Polrier, G. E.; Pylant, E. D.; White, J. M. J. Chem. Phys. 1996, 105, 2089. (11) Pfaum, J.; Bracco, G.; Scoles, G.; Lee, R.; Kahn, A. Am. Vac. Soc. Int. Symp. Abstr., 2000, 55. ( 12) Kawasaki, M.; Sato, T.; Yoshimoto, T. Langmuir 2000, 16,5409. (13) Whitten, D. G.; Chen, L; Geiger, C.; Perlstein, J.; Song, X. J. Phys. Chem. B 1998, 102, 10098. (14) Song, X.; Geiger, H. C.; Farahat, M.; Perlstein, J.; Whitten, D. G. J. Am. Chem. Soc. 1997, 11.9, 12481. ( 15) Vadey, S.; Deiger, H. C.; Cleary, 8.; Perlstein, J.; Whitten, D. G. J. Phys. Chem. 81997, 101,321. (16) Caldwell, W. B.; Csmpbell, D. J.; Chen, K.; Herr, B. R.; Mirkin, C. A.; Malik, A.; K.;, D. M.; Dutta, P.; Huang, K. G. J. Am. Chem. Soc. 1995, 117, 6071. (17) Wang, R.; Iyoda, T.; Jiang, L.; Hashimoto, K.; Fujishima, A. Chem. Lett, 1995, 1005. (18) Tamada, K.; Nagasawa, J.; Nakanishi, F.; Abe, K.; lshita, T.; Hara, m; Knoll, W. Langmuir 1 998, 14, 3264. 123 (19) Han, S. W.; Kim, C. H.; Hong, S. H.; Chung, Y. K.; Kim, K. Langmuir 1999, 15, 1579. (20) Yu, H. 2.; Ye, S.; Zhang, H. L.; Uosaki, K.; Liu, 2. F. Langmuir 2000, 16, 6948. (21) Kim, T.; Chan, K. C.; Crooks, R. M. J. Am. Chem. Soc. 1997, 119, 189. (22) Ford, J. F.; Vickers, T. J.; Mann, C. K.; Schlenoff, J. B. Langmuir1996, 12, 1944. (23) Peanasky, J. S.; McCarley, R. L. Langmuir1998, 14, 113. (24) Mowery, M. D.; Kopta, S.; Ogletree, D. F.; Salmeron, M.; Evans, C. E. Langmuir1999, 15,5118. (25) Cal, M.; Mowery, M. D.; Menzel, H.; Evans, C. E. Langmuir 1999, 15, 1215. (26) Menzel, H.; Mowery, M. 0.; Cal, M; Evans, C. E. J. Phys. Chem. B1998, 102, 9550. (27) Kim, T.; Ye, 0.; Sun, L.; Chan, K. C.; Crooks, R. M. Langmuir1996, 12, 6065. (28) Chan, K. C.; Kim, T.; Schoer, J. K.; Crooks, R. M. J. Am. Chem. Soc. 1995, 717, 5875. (29) Sekiya, A.; 811116, J. K. J. Am. Chem. Soc 1981, 103, 5096. (30) Bums, D. H.; Miller, J. 0.; Chan, H. K.; Delaney, M. O. J. Am. Chem. Soc, 1997, 779, 2125. 124 (31) Molecular mechanics calculations were performed by using SPARTAN 5.0 molecular modeling program (Wavefunction Inc., CA), based on empirical Merck force fields. (32) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. 800.1987, 109, 3559. (33) Colthup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and Raman Spectroscopy, Academic Press: Boston, 1990. (34) Varsanyi, G. Vibrational Spectra of Benzene Derivatives, Academic Press: New York and London, 1969. (35) Varsanyi, G. Assignments for Vibrational Spectra of Seven Hundred Benzene Derivatives, John Wiley and Sons: New York, 1974; Vol. 1. (36) Carron, K. T.; Hurley, L. G. J. Phys. Chem. 1991, 95, 9979. (37) Tour, J. M.; Jones ll, L.; Pearson, D. L.; Lamba, J. J. S.; Burgin, T. P.; Whitesides, G. M.; Allara, D. L.; Parkh, A. N.; Atre, S. V. J. Am. Chem. Soc. 1995, 117, 9529. (38) Wan, L. J.; Terashima, M.; Noda, H.; Osawa, M. J. Phys. Chem. B 2000, 104, 3563. (39) Schbnenberger, C.; Sondag-Huethorst, J. A. M.; Jorritsma, J.; Fokkink, L. G. J. Langmuir1994, 10, 611. (40) Poirier, G. E. Langmuir1997, 13, 2019. (41) Toerker, M.; Staub, R.; Fritz, T.; Schmitz-Hubsch, T.; Sellam, F.; Leo, K. Surf. Sci. 2000, 445, 100. (42) Yamada, R.; Uosaki, K. Langmuir1998, 14,855. 125 (43) (44) (45) (46) (47) Polrier, G. E. Langmuir1999, 15,1167. Lewis, M.; M., T. J. Am. Chem. Soc. 1995, 117, 9574. Huang, J.; Hemminger, J. C. J. Am. Chem. Soc. 1993, 115, 3342. Carlo, S. R.; Grasslan, V. H. Langmuir1997, 13, 2307. During annealing, the sample was heated at approximately 3 °C/min to the indicated temperature and immediately cooled to room temperature. The average cooling rate was approximately 5-8 °C/min. 126 Chapter 6 Rigid Aromatic.p-Methylterphenyl Thiol and 2’,3’-Cyclobutyl p- Terphenyl Thiol Self-Assembled Monolayers on Au(111) Abstract Self-assembled monolayers (SAMs) of rigid p-methylterphenyl thiol (MTPT) and 2’,3’-cyclobutyl p-terphenyl thiol (CTPT) on Au(111) have been prepared. According to ellipsometry and reflection-absorption infrared spectroscopy (RAIRS), MTPT formed densely-packed monolayers on gold with the molecular axes slightly tilted away from the surface normal (~17°). Molecular resolution STM images of the MTPT monolayer reveal a (113x13)R30°-like packing with slightly larger lattice vectors (a=5.3:1:0.3 A, b=5.3:1:0.4 A, y:6012°) than typical alkanethiol monolayers. In contrast, CTPT molecules formed a completely different film structure on Au(111). The molecular axes were tilted far away from the surface normal (~72°), and only monatomically high islands were observed with no molecular order visible within the islands. The structural differences between MTPT and CTPT monolayers were attributed to the physical sizes of the molecules: the bulky cylcobutene unit attached to the phenyl ring in CTPT prevented the formation of an ordered commensurate CTPT film. Compared with MTPT films, which completely desorb after being annealed to 45°C, CTPT monolayers were more thermally stable. 127 6.1. Introduction The spontaneous formation of organized organic monolayers on solid surfaces continues to generate broad academic interest. Many studies have been motivated by the potential technological applications of self-assembled monolayers (SAMs). Monolayers of n-alkanethiols on Au(111) are the most thoroughly studied SAMs to date, and these systems are now relatively well characterized.“ In recent years, the structural diversity of organic molecules investigated for SAM formation has expanded significantly. In particular, aromatic thiol SAMs have received attention due to their high electronic conductivity and nonlinear optical properties.5-ll A number of studies have been carried out to deduce the structure and kinetics of film formation of aromatic thiols. It has been demonstrated that new packing structures may form when the relatively flexible hydrocarbon chain of the conventional alkanethiol SAM is replaced by a rigid and bulky aromatic moiety.”20 The nature of the stabilizing intermolecular forces has been significantly modified: the dispersion forces responsible for chain ordering in alkyl thiols have been augmented by stronger and anisotropic noncovalent 1w: interactions.21 For molecules containing only a single phenyl unit, such as benzenethiol on Au(111), consensus on the detailed structure of the SAMs has not yet been achieved. Well-ordered monolayers with either an upright or strongly inclined adsorption geometry, or poorly organized films have been found after immersing Au(111) surfaces into benzenethiol.”-19.202223 This structural variation could arise from the different sample preparation conditions, and may indicate that the 128 interactions between single phenyl rings are only marginally strong enough to order commensurate films. In addition to studies on thiols containing a single phenyl unit, oligophenylthiols have been studied. The pioneering work done by Sabatani et al. showed that 4-mercaptobiphenyl and 4-mecaptoterphenyl can form stable, ordered monolayers with similar surface concentrations to conventional alkanethiols.24 They predicted a hexagonal packing structure based on bulk crystal structures for these films. Other investigations include 4'-substituted biphenylthiolsfi-IZ-W-25 biphenyl-4-yl alkanethiols,26 p-terphenyl derivatized thiols9-“vl4 and oIigo(phenylethynyl)benzenethioI3-10-18-19 adsorbed on gold surfaces. For phenyl-substituted alkanethiols, the stability of the corresponding SAMs depends on the location of the benzene ring in the alkyl chain, as well as on the length of the alkyl chain.15 With methylene “spacers” inserted between the biphenyl rings and the thiol headgroup, close packed and ordered biphenyl- derivatized monolayers were fon'rled.15'26 When the thiol headgroup is directly connected with the oligophenyl rings, such as in 4'-substituted biphenylthiols, both incommensurate and hexagonal close-packed structures were reportedJZ-i7 Conjugated arenethiols, such as 4-[4'-(phenylethynyl)-phenylethynyl] benzenethiol, form ordered and incommensurate SAMs on gold surfaces.”-19 This chapter describes a molecular scale investigation of p-terphenyl derivatized thiol monolayers adsorbed on Au(111). The purpose is to achieve a better understanding of the intermolecular interactions and their effect on the stability and morphology of the films. Such a study is essential in understanding 129 the electronic and optical properties of materials and devices based on oligophenyls. In addition, the cyclobutene unit attached to the phenyl ring in CTPT molecules allowed the photo and thermal reactivity study of these films. The monolayer thickness and the molecular orientation were determined by ellipsometry and reflection-absorption infrared spectroscopy (RAIRS). Film morphologies were characterized by scanning tunneling microscopy (STM). 6.2. Experimental Synthesis. The preparation of p-methylterphenyl thiol was carried out according to the route shown in scheme 6.1. CH3" Bl“ (1) + CH3$‘@'B(OH)2 (2) a CH3..." SCH3 (3) vlvb CH3 SH (4) Scheme 6.1. Synthetic scheme for MTPT molecule. a 5 mol °/o Pd(PPh3)4, 3.18 g Na2COa, 15 mL H20 and 40 mL toluene, 48 hours in Ar; b 5 equiv. of NaSCHa in NMP, 100 °C, 48 hours in N2. Para-methylbiphenyl bromide (1) was prepared using published procedures.27 The (p-methylterphenyl)methyl sulfide (3) was synthesized via Suzuki coupling28 using approximately 0.5 g of 1 refluxed with 0.34 g of 4- (methylthio)phenylboronic acid (2) in the presence of 5 mol °/o Pd(PPh3)4, 3.18 g Na2003, 15 mL H20 and 40 mL of toluene under argon for two days. The resulting solution was extracted with CH2CI2, dried over Na2SO4, filtered through 130 silica gel and concentrated to yield crude 3. The crude product was recrystallized from CH2CI2 to give a white solid. Carbon-13 NMR and mass spectrometry were used to verify the structure and purity of 3. The p methylterphenyl thiol (MTPT) (4) was prepared by reaction of 3 and sodium thiomethoxide (5 equiv.) in 1-methyl-pyrrolidin-2-one (NMP) under nitrogen.29 The reaction ran at 100 °C for two days and the solution was then extracted with ether and filtered through silica gel using CH2CI2 as the eluting solvent. The light yellow solution of 4 was then concentrated until saturation (~1 mM). Mass spectra of the solution give the appropriate molecular peak at MW. 276. The CTPT molecule was synthesized from 1-bromo-2,3-cyclobutyl blphenyl3° under the same Suzuki coupling and thiolization conditions as described above. Purity was confirmed by 1H NMR and GC-MS. Substrate and Monolayer Preparation. The gold substrates for infrared spectroscopy and ellipsometry measurements were made by electron-beam evaporation of 200 nm of Au on 20 nm of Ti on Si(100) wafers. The gold substrates for STM measurements were prepared by thermally evaporating gold onto freshly cleaved mica in high vacuum. All of the substrates were cleaned by a UVI03 cleaner for 15 minutes, followed by soaking in deionized water for 30 minutes. After this, the substrates were dried in flowing N2 and immediately transferred into a saturated solution of MTPT or CTPT in CH2CI2 for 3-18 hours, followed by rinsing with copious quantities of CH2CI2 and drying in a N2(g) stream. 131 Surface Characterization. Ellipsometry, reflection-absorption infrared spectroscopy (RAIRS), and scanning tunneling microscopy (STM) were applied to characterize MTPT and CTPT film structure, stability, and reactivity. For detailed descriptions of these techniques and their working conditions, please refer to Chapter 2 and 4. 6.3. Results and Discussion Characterization of MTPT monolayers. Self-assembled monolayers of MTPT on Au(111) were initially characterized by ellipsometric and RAIRS measurements. Ellipsometry showed that the average film thickness was 1711 A, indicating the presence of a densely packed monolayer with the molecular axes slightly tilted away from the surface normal. The length of one MTPT molecule was calculated to be 15.4 A31 and, based on the S-Au distance of 2.36 A,3 a tilt angle of 1718° was determined. Similar results were also observed for other terphenyl thiol derivatized SAMs on gold, with tilt angles ranging from 5° to 270 reported] l,I4,I8,21,24 Reflection-absorption infrared spectroscopy confirmed the adsorption of MTPT on gold. Comparison of the reflectance spectrum of the MTPT SAM with the transmission spectrum of 4,4”-dimethyl terphenyl (DMTP)32 solid in KBr, supported the conclusion that the molecular axes in the SAM were near- perpendicular to the substrate surface plane. Due to the limited quantities of MTPT generated, DMTP was chosen for RAIRS comparison because of the structural similarity to the MTPT molecule. Variation in para-substituents may 132 affect band positions and intensities;33 however, to change the 4-substituent from methyl (in DMTP) to thiol (in MTPT) should not alter the charge distribution in the aromatic ring significantly because thiol is a weak electron donor.34 Therefore, we believe the IR spectrum of MTPT solid will be similar to that of the DMTP solid and it is valid to compare MTPT SAM RAIRS data with DMTP transmission data in order to obtain orientational information in the monolayer. Figure 6.1 (A) and (B) shows the transmission spectrum of DMTP solid in KBr, and the RAIR spectrum of the MTPT SAMs on Au(111), respectively. All peaks in these spectra can be assigned by consulting the data in the literature,12-26-34r36 and the results are summarized in Table 1. L Wfilfi—fi'jfilfi '01 Absorbance grass... 1 1 I ' I ' I V r V I V I T V 3100 3000 2900 1600 1400 1200 1000 800 Wavenumbers (cm") Figure 6.1 IR spectra of (A) DMTP solid in KBr, and (B) MTPT monolayer on a Au/Si wafer. The DMTP bulk spectrum is dominated by two aromatic vibrations: the ring C-C stretch (1490 cm", 19a) and the out-of-plane C-H bend (808 cm", 17b), whose transition dipoles are oriented parallel to the 4,4”-axis and orthogonal to 133 the ring plane of the terphenyl moiety, respectively. Four aromatic C-H in-plane deformation bands occur in the region 1187-970 cm'1 in the DMTP bulk spectrum, and among these bands, only mode 18a at 1003 cm'1 has its dynamic dipole aligned parallel to the 4,4”-axis. In the range of aromatic C-H stretching modes, one strong band shows up at 3025 cm" (20b). The terminal CH3 has two prominent bands at 2914 and 2856 cm“, both assigned to symmetric CH3 stretch. The asymmetric CH3 stretching vibrations are not degenerate and are seen as two weak bands at 2974 and 2940 cm". The asymmetric and symmetric CH3 bending vibrations are located at 1466 and 1389 cm", respectively, with the first band strongly overlapped by the aromatic ring stretch. The CH3 rocking vibration appears as a weak band at 1040 cm“. The RAIR spectrum of the MTPT SAM is much simpler than the bulk DMTP spectrum. The simplification of the spectrum in the SAM environment reflects the near-perpendicular orientation of the terphenyl moiety to the substrate surface. Most striking is the intensity drop of the aromatic out-of-plane bending mode at 806 cm'1 (17b). In contrast to the out-of-plane vibration, the ring C-C stretching mode (1481cm'1, 19a), and in-plane C-H bending mode (1003 cm“, 18a) whose transition dipoles are parallel to the 4,4”-axis, have significant intensity. According to the infrared surface selection rule, only vibrations with transition dipole moments oriented perpendicular to the metal substrate are observed. Therefore, peak intensities are directly related to the component of each transition moment that is perpendicular to the metal surface. 134 Table 6.1 Infrared band assignments for MTPT SAM and DMTP solid in KBr Peak Position (cm‘1) Solid DMTP in KBr MTPT SAM Mode Assignment Direction of Transition Dipole§ 3026 3029 vCHamm (20b)* .L 4,4”-axis, ip 2974 - vasCH3 .L 4,4”-axis 2940 - vasCH3 J. 4,4"-axis 2914 2920 vsCH3 ll 4,4”-axis 2856 2865 vsCHa ll 4,4"-axis 1490 1481 vCCarom(19a)* ll 4,4”-axis 1466(sh) - 8asCH3 .L 4,4”-axis 1388 1379 8sCH3 ll 4,4”-axis 1 187 - 8CHarom .L 4,4”-axis 1 1 15 - 8CHarom .L 4,4”-axis 1041 - 8asCH3 .L 4,4”-axis 1003 1002 8CHarom (18a)' ll 4,4”-axis 970 - 8CHarom .l. 4,4"-axis 825 826 vCCamm (1 )* J. 4,4”-axis, op 808 806 OJCHarom (17b)* 1. 4,4"-axis, op §, based on C3 symmetry *, Wilson’s notation; -, not observed; sh, shoulder peak; ip, in-plane; op, out-of-plane; v, stretching vibration; 8, in-plane bending vibration; to, out-of-plane bending vibration. 135 It should be noted that changes in the intrinsic oscillator strength, for example by a perturbation of the vibrational potential energy surface induced by changes in layer packing, will also affect the intensity of an IR absorption band. This effect is ignored in the present discussion. The dramatic diminution of mode 17b and the strong absorption of modes 19a and 18a suggest that the aromatic rings in the film are roughly perpendicular to the plane of the gold surface. Similar conclusions were derived from the RAIRS data of other oligophenylthiol SAMs-12.21.26 The small molecular tilt argument is also supported by intensity changes in the CH3 vibrational modes. Both symmetric stretching (2920 and 2865 cm") and in-plane bending (1379 cm") modes, whose transition dipoles are parallel to the 4,4”-axis, appear distinctively in the spectrum of MTPT SAM; whereas, the asymmetric stretching and in-plane bending modes were absent. A typical large area STM image of the MTPT monolayer on Au(111) after >10 hours soaking is shown in Fig. 6.2. The image shows three terraces separated by monatomic height steps in the Au substrate. Also visible are small, irregularly shaped islands approximately 50~150 A in diameter distributed on the terraces. The islands do not appear to be correlated with the position of step edges in the substrate. The height of the individual islands was measured as 2.6103 A, consistent with the height of a single atomic layer of Au (measured in our instrument as 2.5102 A).37 136 Figure 6.2 Large scale STM image of MTPT on Au(111). Three terraces separated by monatomic steps are shown. Islands with a height of 2610.3 A are observed. Image size 1500 x 1500 A. Compared with aliphatic thiol‘i-37 or aliphatic/aromatic mixed thiol38 SAMs, the presence of islands rather than depressions in the monolayer is rather surprising. Conventional n-alkanethiols typically produce monatomic deep depressions in the monolayer (termed “vacancy islands").39 The vacancy islands form during the adsorption process of alkanethiols on Au(111), and at high surface coverage, the (113x113)R30° lattice structure is observed within the depressions.40 The origin of vacancy islands is still controversial but appears to be related to lifting of the 22x13 compressed herringbone-reconstruction of the bare Au surface followed by expulsion of additional atoms from the relaxed surface.39 Under our monolayer preparation conditions the islands appeared to be stable for at least several hours without observable changes in size, shape or structure (very rapid self diffusion of gold atoms on Au(111) surfaces has been observed in the same timescale). In addition, identical molecular resolution 137 images were obtained both on top of the islands and on the surrounding terraces, suggesting that the islands were not associated with different structural phases but rather formed in a homogenous layer. Similar island structures were also observed on other aromatic thiol SAMsJ-i-41 Upon adsorption of 4- mercaptopyridine (4MPY) or 4-hydroxythiophenol (4HTP) molecules on Au(111), monatomic high islands appeared immediately on the gold terraces. The islands initially increased both in size and number, and then decreased slowly after reaching maximum values (the coverage of the islands reached 50 % in 15 minutes). The formation of the islands was attributed to gold atoms on the surface that became highly mobile due to strong binding to the thiols. It was suggested that islands are formed on aromatic thiols rather than depressions because the strong attractive intermolecular interactions between the phenyl moieties favors aggregation of the adsorbed aromatic molecules (each with a gold atom attached) into islands. Although we were not able to observe the formation of the islands on flat Au(111) regions by STM (for example, during the adsorption process), based on the dimensions, density, and surface coverage of the islands, we believe the monatomic high islands on our MTPT SAMs were probably formed in the same manner as those in 4MPY and 4HTP SAMs during the initial adsorption under conditions of high Au mobility. On the terraces we were able to observe the molecular arrangement of MTPT molecules on Au(111). Figure 6.3 (A) shows a typical high resolution STM image of the MTPT monolayer on a single Au(111) terrace. Molecules appeared as single circular or oval features arranged in a hexagonal pattern, and this same 138 packing was consistently observed on the terraces and the monatomic high islands. A close inspection of MTPT SAMs revealed that there were many defects and disordered regions in the monolayer (some are pointed out in Fig. 6.3 (B)). Although rows of molecules can be observed easily, the degree of order is significantly less than that of a typical alkanethiol SAM. The image appears to be composed of many small regions of order (approximately 30-60 A in diameter containing about 25-100 molecules) surrounded by disorder and defects. The origin of the defects will be discussed below. A similar hexagonal packing structure was also reported for 4-methyl-4‘- mercaptobiphenyl (MMB) SAM on Au(111) as characterized by grazing incidence X-ray diffraction (GIXD) and low-energy atomic beam diffraction (LEAD).17 The MMB molecule has one iess phenyl ring than the MTPT molecule of our study but has the same or, (0 substituents. In addition to the high-density hexagonal phase, a low-density striped phase was observed following gas phase deposition. After conventional liquid-phase deposition, no phase other than the hexagonal phase was seen and, based on the diffraction peak widths, the domain size was estimated to be ~65 A. Clearly, our STM data for the terphenyl SAM supports the conclusions of Leung et al. for the biphenyl system.17 These aromatic SAMs appear to be composed of small domains of ordered material when deposition is carried out in solution. 139 0-80 5 10 15 20 25 30 35(4) (A) 1.3 E 1.0/\qflm \ 0.7 0 5 10 15 20 25 30(4) Figure 6.3 High resolution STM images of MTPT monolayer on Au(111). Image 3129; (A) and (B) 150x150 A; (C) 40x60 A. (D) and (E) show the corresponding line profiles in (C). 140 In small ordered domains as shown in Fig. 6.3 (C), the measured lattice vectors were a=5.3:1:0.3 A, b=5.3:1:0.4 A, 7:605:23 These values were calculated from at least 20 different domains, and typical cross-section data are presented in Fig. 6.3 (D) and (E). While generally consistent with the (13x13)R30° structure observed for many alkanethiols (a=b=4.99 A, 7:60°), the MTPT primitive unit cell is ~11 % larger in area. Indeed, calibration experiments on octanethiol SAMs were performed under identical conditions and produced lattice vectors of a=b=5.0:l:0.2 A. An examination of the molecular crystal of terphenyl42 shows that the area per molecule in the (001) plane of the crystal is about 6 % larger than that of the molecule in the (1/3x13)R30° lattice.24 It appears from our STM images that commensurate monolayers of MTPT cannot be formed with long- range two-dimensional order. Instead, local (V3xx/3)R30° Au-S bonding probably occurs; however, the aryl part of the molecule attempts to adopt the larger separation dictated by 1t-rrIvan der Waals’ interactions as in the molecular crystal. The MTPT molecules may accommodate the strain by tilting of the molecular axis (and likely some twisting). This tilt most likely increases progressively with distance from the center of the ordered area, but eventually restricts (\/3X~l3)R30°-Iike Au-S bonding as shown schematically in Fig. 6.4. Such a situation would account for the small areas of order surrounded by disorder observed in our STM images. 141 Figure 6.4. Model showing the adsorbed sulfur atoms as black dots and the molecular chains as lines along the molecular axis. The sulfur atoms are ordered while the top of the molecules are slightly displaced from their (13x13)R30° positions. In the biphenyl-based MMB monolayer, GIXD results showed the underlying sulfur atoms were in an ordered (~13x13)R30°-Iike structure.17 However, some surface disorder was detected by LEAD, a technique that is only sensitive to the position of the topmost layer of atoms. This disorder was attributed to mismatch between the lowest energy configuration of the biphenyl backbone and that of the underlying (13x13)R30° structure of the sulfur headgroups. Small ordered domains and high defect density were also reported in other aromatic thiol SAMs. 12118-1920 For 4-[4’(phenylethynyl)-phenylethynyl]-benzenthiol and 4’-chloro-4- mercaptobiphenyl SAMs,12-18 an ordered but incommensurate structure has been reported. In both monolayers, six equivalent domains were present and superstructures were observed. A rectangular unit cell containing two molecules was revealed in STM images and was similar to the bulk terphenyl crystal structure. Compared with these monolayers, our MTPT SAMs did not show evidence of large dimensional superstructures or periodic Moire fringes. Our unit cells were hexagonal and exhibited the three rotational domains characteristic of a commensurate lattice. The precise cause of the structural differences between 142 MTPT and 4-[4’(phenylethynyl)-phenylethynyl]-benzenthiol or 4’-chloro-4- mercaptobiphenyl monolayers may be related to steric factors or electronic effects due to different a) substitution. It has been already demonstrated that different (lo-position substituents on the aromatic ring can affect the molecular structure of the self-assembled films significantly.“-15 We annealed the MTPT monolayers in UHV at 45 °C for 30 minutes to investigate whether “lying-down” phases, as observed for octadecane thiol (C13H37SH), could be generated. Octadecanethiol, which has a similar mass and number of carbon atoms as MTPT, exhibited a variety of px13 monolayer structures after 30 minutes of annealing at 45 °C.43 However, STM revealed that the MTPT monolayer had completely desorbed after annealing: we found no evidence for the intermediate pxx/3 phases. This clearly suggests that the overall intermolecular interchain interactions are weaker in aromatic versus aliphatic thiols. Unfortunately, we cannot independently assess the role of the domain size in the thermal stability, although we might expect the small domain/moderately disordered terphenyl thiol SAMs to exhibit reduced thermal stability compared with the large domain/fully ordered alkanethiol SAMs. Characterization of CTPT monolayers. The CTPT monolayers on Au(111) were also investigated by ellipsometry, RAIRS and STM. The results are dramatically different from the MTPT films. After soaking for >5 h, the measured thickness of the CTPT SAMs was only 511 A. The distance from the gold substrate to the 0) terminal hydrogen of a CTPT molecule was calculated to be 16.0 A, thus, the CTPT molecules were inclined from the surface normal with 143 an apparent molecular tilt angle of 72°. This result is in great contrast to the MTPT molecules which were oriented with their molecular axes almost perpendicular to the Au(111) plane. The large difference between CTPT and MTPT monolayers suggests that the cyclobutene group attached to the phenyl ring in CTPT influenced the packing order of the self-assembly significantly. The RAIRS data further demonstrated the difference in molecular orientation between CTPT and MTPT SAMs. Figure 6.5 (A) and (B) show the RAIRS spectra for MTPT and CTPT monolayers, respectively. It should be noted that due to the structural variation, there is disparity in vibrational modes and peak positions between CTPT and MTPT films. For example, the symmetric and asymmetric methylene modes were observed at 2856 and 2927 cm", respectively, in the CTPT monolayers, while no other aliphatic vibrational modes but the terminal methyl vibrations were detected in the MTPT films. In addition, the phenyl substitution-sensitive modes, such as the phenyl C-H and C-C stretch modes, had different frequencies in the MTPT and CTPT spectra. 144 ll ' l fi— fir r T r n ‘1— '— T j r -l vC=C .4 / +1 2'0 X 10 8CH(arom) it) 1 (1)CH(arom) 8 .. a . 31 "NJ A . L1 8 1 -° 1 < m1 B 3000 2800 1600 1200 1000 800 Wavenumbers 400(cm") Fig. 6.5 IR spectra of (A) MTPT monolayer on Au(111), and (B) CTPT monolayer on Au(1 1 1). The peak intensity changes observed for the phenyl modes in the CTPT film relative to those in MTPT provide useful information about the oligophenyl backbone orientation. Both the ring C-C stretching mode (1471cm'1, 19a), and in-plane C-H bending mode (1010 cm", 18a), whose transition dipoles are parallel to the 4,4”-axis, dropped in intensity by 40~60 % in CTPT monolayers compared with MTPT films. The aromatic out-of—plane bending mode at 809 cm'1 (17b) in the CTPT monolayers, whose transition dipole is perpendicular to the terphenyl plane, increased by ~40 % relative to the MTPT SAMs. The above observations further confirmed the postulate that CTPT molecular axes were more inclined towards the substrate plane than the MTPT molecules on Au(111). Large area STM images of CTPT monolayers are shown in Fig. 6.6. Monatomically high terraces in the gold substrate were observed (Fig. 6.6 (A) 145 and (B)). Unlike the MTPT molecules, which self-assembled into a close-packed pattern on Au(111), the CTPT molecules formed irregularly shaped islands. Some texture was visible in the islands due to the presence of individual molecules; however, no periodic structures were ever observed despite the use of a variety of imaging conditions (Fig. 6.6 (C) and (0)). This observation immediately suggests that CTPT did not form an ordered SAM. Interestingly, the height of the islands in CTPT monolayers was about 2.5 A, consistent with the height of a single layer of gold. Fig. 6.6 STM images of CTPT monolayers on Au(111) at room temperature. Image size: (A) 1500x1500 A; (B) 1000x1000 A; (C) 500x500 A; (D) 500x500 A. As discussed previously, monatomically high islands approximately 50~150 A in diameter were also observed on MTPT films, however, the density of islands was much lower than that in CTPT films. Identical (13x13)R30°-Iike 146 structure was imaged both on top of the islands and on the surrounding terraces in MTPT monolayers. On the contrary, no flat terraces were observed in CTPT films: only monatomically high islands with no apparent order were imaged. Considering that similar island structures were also reported in other aromatic thiol SAMs13119r“, our results suggest that aromatic thiols probably form islands upon initial adsorption on gold and larger, ordered domains grow later. The reason that no domains of order were formed in CTPT films was likely due to the bulky size of the molecule. The van der Waal’s "molecular width" of CTPT is 5.3 A (Fig. 6.7). Considering the fact that MTPT molecules, whose "molecular width“ is 0.9 A less than CTPT, formed a (13x13)R30°-like structure on Au(111) with 11% larger unit cell (in area) and had many defects in the film due to the packing (Fig. 6.4), the geometry of CTPT molecules would prevent the formation of a similar structure. Fig. 6.7 Schematic diagram indicating the "width” of phenyl and benzocyclobutene units in CTPT molecule. 147 The thermal stability of the CTPT monolayers was tested by controlled thermal treatment in UHV. Upon annealing the CTPT/Au(111) SAM to elevated temperatures, the density and size of the islands noted at room temperature by STM decreased (Fig. 6.8). In addition, the step edges became jagged indicating that reconstruction occurred in the gold substrate during annealing. After heating to 100 °C, most CTPT molecules had desorbed from the surface. The thermal behavior of the CTPT SAM was quite different from MTPT films: it will be recalled that the MTPT SAM had completely desorbed after annealing to only 45 °C. This disparity is likely related to the different molecular orientations in CTPT and MTPT films. Another reason that may contribute to the better then'nal stability of the CTPT films is the possibility of thermal polymerization of the benzocylcobutene units in the SAM. Such a reaction may link individual molecules into a rigid oliomeric film with enhanced thermal properties. However, according to the literature, the possibility that benzocylcobutene (liquid) polymerization at 100 °C or lower temperatures is small (less than 10 %)."°4 In conjunction with the fact that most CTPT molecules desorbed from the surface at 100 °C, we concluded that under our annealing conditions, polymer formation was negligible. 148 Fig. 6.8 STM images of CTPT monolayers on Au(111) after annealing to 45 °C (A,B,C), 75 °C (D,E,F) and 100 °C (G,H,l). Image size: (A) 1300x1300 A; (B) 900x900 A; (C) 600x600 A; (D) 1500x1500 A; (E) 1250x1250 A; (F) 1000x1000 A; (G) 2000x2000 A; (H) 1500x1500 A; (I) 750x750 A. The probability of polymerization reactions in CTPT SAMs was probed by investigating their photochemical stability. The photoreactivity of CTPT monolayers was tested by exposing the films to the UV light emitted from a low pressure Hg arc lamp (1~250-400 nm) in a pure N2 environment. After 45 min irradiation, there was no peak intensity change nor peak position shift observed in the RAIRS spectrum of the CTPT monolayer. It has been reported that direct UV irradiation of cyclobutene in the gas phase or solution will initiate the ring 149 opening reaction.45:46 However, compared with gas or solution phase reactions, it appears that the reaction probability in the film was significantly reduced. Reaction on a two-dimensional surface will be determined by geometric structure, molecular motion and diffusion. Kurth et al. have shown that a classical solution-phase S~2 reaction failed to occur when one of the reactants was immobilized on a silica surface.“ Thus, we speculated that the reduced photoreactivity of cyclobutene in CTPT monolayers was probably due to the immobilization of the monomers on the surface by formation of the strong Au-S bond or an unfavorable reaction geometry. 6.4. Conclusion Self-assembled monolayers (SAMs) of rigid p-methylterphenyl thiol (MTPT) and 2’,3’-cyclobutyl p-terphenyl thiol (CTPT) on Au(111) were investigated by ellipsometry, RAIRS and STM. The aromatic nature and rigidity of the molecular backbone influenced the final monolayer structure, and different film properties were observed compared with conventional alkanethiol SAMs. According to ellipsometry and RAIRS, MTPT formed densely packed monolayers with the molecular axes slightly tilted away from the surface normal (~17°). At room temperature, monatomically high islands were observed by STM instead of the typical monatomically deep holes in alkanethiol SAMs. Molecular resolution images of the MTPT monolayer revealed a (13x13)R30°-like packing with slightly larger lattice vectors than typical alkanethiol monolayers. 150 The addition of a cyclobutene unit to the system changed the film properties completely. The CTPT molecules were tilted far away from the surface normal (~72°). Only irregularly shaped islands were imaged in CTPT SAMs with no apparent order in the islands. The structural variation between MTPT and CTPT monolayers was attributed to the geometric difference between the two molecules: the cylcobutene unit attached to the central phenyl ring in CTPT prevented the formation of an ordered commensurate CTPT monolayer on Au(111). Unlike MTPT films, which completely desorbed after annealed to 45°C, CTPT monolayers were more thermally stable. The different thermal behavior was attributed to the various molecular orientations in the films. 6.5. Literature Cited (1) 'UI man, A. An Introduction to Ultrathin Organic Films from Langmuir- BIodgett to Self-Assembly, Academic Press: San Diego, CA, 1991. (2) Ulman, A. Chem. Rev. 1996, 96, 1533. (3) Polrier, G. E. Chem. Rev. 1997, 97, 1117. (4) Polrier, G. E.; Fitts, W. P.; White, J. M. Langmuir2001, 17, 1176. (5) Buckel, F.; Effenberger, F.; Yan, C.; Gblzhauser, A.; Grunze, M. Adv. Mater. 2000, 12,901. (6) Kang, J. F.; Ulman, A.; Jordan, R.; G., K. D. Langmuir1999, 15,5555. (7) Eck, W.; Stadler, V.; Geyer, W.; Zhamicov, M.; Gélzhauser, A.; Grunze, M. Adv. Mater. 2000, 12, 805. 151 (8) Tour, J. M.; Jones ll, L.; Pearson, D. L.; Lamba, J. J. S.; Burgin, T. P.; Whitesides, G. M.; Allara, D. L.; Parkh, A. N.; Atre, S. V. J. Am. Chem. Soc. 1995, 117, 9529. (9) lshida, T.; Mizutani, W.; Akiba, U.; Umemura, K.; lnoue, A.; Choi, N.; Fujihira, M.; Tokumoto, H. J. Phys. Chem. B1999, 103, 1686. (10) Zehner, R. W.; Sita, L. R. Langmuir1997, 13,2973. (11) Himmel, H.-J.; Terfort, A.; Wbll, C. J. Am. Chem. Soc. 1998, 120,12069. (12) Kang, J. F.; Ulman, A.; Liao, 8.; Jordan, R.; Yang, G.; Liu, G.-y. Langmuir 2001, 17,95. (13) Jin, G.; Rodriguez, J. A.; Li, C. 2.; Darici, Y.; Tao, N. J. Surf. Sci. 1999, 425, 101. (14) Frey, S.; Stadler, V.; Heister, K.; Eck, W.; Zhamikov, M.; Grunze, M.; Zeysing, B.; Terfort, A. Langmuir2001, 17, 2408. (15) Tao, Y.-T.; Wu, C.-C.; Eu, J.-Y.; Lin, W.-L. Langmuir1997, 13,4018. (16) Ulman, A.; Scaringe, R. P. Langmuir1992, 8, 894. (17) Leung, T. Y. B.; Schwartz, P.; Scoles, G.; Schreiber, F.; Ulman, A. Surf. Sci. 2000, 458, 34. (18) Yang, G.; Qian, Y.; Engtrakul, C.; Sita, L. R.; Liu, G.-y. J. Phys. Chem. B 2000, 104, 9059. (19) Dhirani, A.; Zehner, R. W.; Hsung, R. P.; Guyot-Sionnest, P.; Sita, L. R. J. Am. Chem. Soc. 1996, 118, 3319. (20) Wan, L. J.; Terashima, M.; Noda, H.; Osawa, M. J. Phys. Chem. B 2000, 104, 3563. 152 (21) Reese, 8.; Fox, M. A. J. Phys. Chem. B1998, 102, 9820. (22) Szafranski, C. A.; Tanner, W.; Laibinis, P. E.; Garrell, R. L. Langmuir 1998, 14,3570. (23) Whelan, C. M.; Smyth, M. R.; Barnes, C. J. Langmuir1999, 15, 116. (24) Sabatani, E.; Cohen-Boulakia, J.; Bruening, M.; Rubinstein, l. Langmuir 1993, 9, 2974. (25) Kang, J. F.; Ulman, A.; Liao, S.; Jordan, R. Langmuir1999, 15,2095. (26) Rong, H.-T.; Frey, S.; yang, Y.-J.; Zhamikov, M.; Buck, M.; Wl'ihn, M.; Wbll, C.; G., H. Langmuir2001, 17, 1582. (27) Bumagin, N. A.; Luzikova, E. V.; Beletskaya, I. P. Russ. J. Org. Chem. 1995, 31, 1480. (28) Miyaura, N.; Yanagi, T.; Suzuki, A. Syn. Comm. 1981, 11,513. (29) Shaw, J. J. Org. Chem. 1991, 56, 3728. (30) Liu, T.; Baker, G. . (31) Molecular mechanics calculations were performed by using SPARTAN 5.0 molecular modeling program (Wavefunction Inc., CA), based on empirical Merck force fields. (32) DMTP was synthesized using Suzuki coupling of para-methylbiphenyl bromide and 4-methylphenylboronic acid. (33) Avram, M.; Mateescu, G. H. D. Infrared Spectroscopy Application in Organic Chemistry, Wiley-lnterscience: New York, 1966. (34) Colthup, N. B.; Daly, L. H.; Wiberley, s. E. Introduction to Infrared and Raman Spectroscopy, Academic Press: Boston, 1990. 153 (35) Varsanyi, G. Vibrational Spectra of Benzene Derivatives, Academic Press: New York and London, 1969. (36) Varsanyi, G. Assignments for Vibrational Spectra of Seven Hundred Benzene Derivatives, John Wiley and Sons: New York, 1974; Vol. 1. (37) Polrier, G. E.; Pylant, E. D. Science 1996, 272, 1145. (38) Duan, L.; Garrett, S. J. Langmuir 2001 , 17,2986. (39) Poirier, G. E. Langmuir1997, 13, 2019. (40) Schbnenberger, C.; Sondag-Huethorst, J. A. M.; Jorritsma, J.; Fokkink, L. G. Langmuir1994, 10,611. (41) Hara, M.; Sasabe, H.; Knoll, W. Thin Solid Films 1996, 273, 66. (42) Rietveld, H. M.; Maslen, E. N.; Clews, C. J. B. Acta Cryst. B 1970, 26, 693. (43) More details have been discussed in Chapter 4. (44) Niklaus, F.; Enoksson, P.; Kalvesten, E.; Stemme, G. 13th IEEE Int. Symp. Abstr., 2000, 247. (45) Leigh, W. J.; Zheng, K.; Clark, K. B. Can. J. Chem. 1990, 68, 1988. (46) Leigh, W. J.; Postigo, J. A. J. Am. Chem. Soc. 1994, 117, 1688. (47) Kurth, G. G.; Bein, T. Langmuir1993, 9,2965. 154 Chapter 7 Olefin-Terminated Di(9-decene) Disulfide Self-Assembled Monolayers on Au(111) Abstract Self-assembled monolayers (SAMs) of di(9-decene) disulfide (DDDS) on Au(111) were investigated by ellipsometry, RAIRS and contact angle measurements. The film structure depended strongly on the polarity of the solvents used in the soaking solutions. In a nonpolar solvent, ordered crystalline- like DDDS monolayers were formed on the gold substrate with an apparent tilt angle of 48°; whereas in a polar solvent, the films were disordered and liquid-like with large amounts of gauche defects present. The thermal stability of ordered DDDS monolayers was studied following annealing in decane and was demonstrated to be similar to that of decanethiol SAMs. The reactivity of the olefin-terminated DDDS monolayers was tested by exposing them to different wavelengths of UV light. Irradiating the SAMs with light of 71<300 nm caused loss of orientational and conformational order in the films, while near-UV irradiation (>300 nm) induced no discemable change in the monolayer structure. The radical initiator, benzoin methyl ether (BME), could be used to initiate some degree of vinyl polymerization/oligomerization in DDDS monolayers under near UV exposure. 155 7.1 Introduction In the past few decades, spontaneously organized monolayers formed from thiol or disulfide compounds on gold surfaces have been extensively studied.‘-2 The most exhaustive studies have been centered on n-alkyl and c0- functionalized assemblies.3'1o By suitable molecular design, these systems provide a way to tailor surface properties such as wetting, adhesion, lubrication, and corrosion“,12 We are interested in the photochemical and thermal reactions of self-assembled monolayers (SAMs) that have been modified with unsaturated functionalities at external positions. Two critical issues to be addressed are the role of the substituent on the structure of the self-assembled monolayer and the influence of the monolayer environment on the intrinsic reactivity of the unsaturated moiety. This chapter will report a study of olefin- : terminated di(9-decene) disulfide (DDDS) self-assembled monolayers on Au(111) and their subsequent therrnal- and photo-reactivity. There have been previous studies of olefin-terminated monolayers on metal or metal oxide surfaces. Allara and Nuzzo characterized olefin-terminated carboxylic acids adsorbed on oxidized aluminum substrates using reflection- absorption infrared spectroscopy (RAIRS).13 In the close-packed highly-oriented monolayer assemblies observed in that work, the CH stretching modes of the terminal vinyl groups were blue-shifted compared with the liquid phase, implying the environment of the ambient-monolayer interface exhibited very diminished intermolecular interactions for the terminal vinyl groups compared to the bulk. Leung et al. have studied a series of olefin-terminated thiol SAMs on Au(111) 156 using grazing incidence X-ray diffraction (GIXD) and low energy atom diffraction (LEAD) techniques.14 It was demonstrated that the olefin-terminated thiol monolayers had the same packing order as that of n-alkanethiol films: a c(4x2) superlattice structure on Au(111). As the impact of the terminal olefin groups on the structure of the monolayers has been demonstrated to be minimal, this well ordered layer of olefin groups opens up new opportunities for subsequent chemistry either through attachment of another species or by intralayer cross-linking reactions (oligomerization/polymerization). For example, Peanasky and McCarley have studied undec-10-ene-1-thiol SAMs on gold irradiated by y—rays.‘5 The polymerization of the monolayers during y-ray exposures was indicated by a decrease in the intensities of the infrared bands associated with the olefin functionality. Some disordering of the monolayer occurred during the reaction. It was pr0posed that the polymerization reaction was controlled by the distance that the tethered olefin groups were able to move in the monolayer. The optical and spectroscopic data describing the structure, thermal stability and photoreactivity of olefin-terminated DDDS monolayers on Au(111) are presented in this chapter. The goal is to have a better understanding of the reactivity of surface-confined olefin groups and fabricate robust monolayer polymers with structural control in three dimensions. Several surface sensitive techniques, including ellipsometry, RAIRS, STM, and contact angle measurements, were applied to characterize DDDS monolayers. In all cases, direct comparison was made with decanethiol monolayers on Au(111). 157 7.2 Experimental Synthesis. The di(9-decene) disulfide (DDDS) was prepared according to the procedures described in the literature."5'18 The reaction route is shown in scheme 1. o . a CH2=CH-(CH2)8-OH —-> CHFCH‘10H2I8‘O'I'@’CH3 l: . CH2=CH-(CH2)8-S-S-(CH2)8-CH=CH2 Scheme 7.1. Synthetic scheme for DDDS molecule. a 2 equiv. of p- toluenesulphonyl chloride and 3 equiv. of pyridine in 30 mL chloroform, 5-8 hours in an ice bath; b 6 equiv. of sodium hydrogen sulfide in absolute ethanol, 42 °C, 22 hours sonication. The starting material, 9-decene-1-ol, was purchased from Aldrich and used without further purification. The terminal alcohol was first tosylated (step a) in chloroform, using a 1:2:3 ratio of alcoholzp-toluenesulfonyl chloridezpyridine. The solution was stirred in an ice bath for 5-8 hours. The reaction was monitored for completion by thin-layer chromatography (TLC). Upon completion, the solution was added to 30 mL ether and 7 mL water. The organic layer was washed successively with 2 M hydrochloride acid, 5% sodium bicarbonate, and water. The organic phase was then dried over magnesium sulfate and the solvent was evaporated. After recrystalization from hexane, the tosylated decene was collected in ~80 % yield. Proton NMR and mass spectrometry were used to verify the structure and purity of the product. Step b involved nucle0philic substitution of the tosylate with the -SH. Initially, the tosylated decene was 158 dissolved in a minimum amount of absolute ethanol. Six equivalents of sodium sulfide were added and the mixture was sonicated at 42 °C for 22 hours. Proton NMR indicated that the product was a mixture of 9-decene-1-thiol and di(9- decene) disulfide. Prolonged sonication at the elevated temperature drove the reaction toward disulfide formation. The crude disulfide was then purified by flash silica gel column chromatography using n-hexane as eluent. The final product was a slightly yellow liquid, whose purity was checked by proton NMR and mass spectrometry. Substrate and Monolayer Preparation. The gold substrates for infrared spectroscopy, ellipsometry, and contact angle measurements were made by electron-beam evaporation of 200 nm of Au on 20 nm of Ti on Si(100) wafers. The gold substrates for STM measurements were prepared by thermally evaporating gold onto freshly cleaved mica in high vacuum. All of the substrates were cleaned by a UV/Oa cleaner for 15 minutes, followed by soaking in deionized water for 30 minutes. After this, the substrates were dried in a N2(g) stream and transferred into a 2 mM solution of DDDS in different solvents as detailed below. After more than 10 hours soaking, the samples were rinsed with copious quantities of the corresponding solvent and dried again in the N2 stream. Surface Chracterization. Ellipsometry, contact angle measurements, reflection-absorption infrared spectroscopy (RAIRS), and scanning tunneling microscopy (STM) were applied to characterize the DDDS film structure, stability, and reactivity. For detail descriptions of these techniques and their working conditions, please refer to Chapter 2 and 4. 159 7.3 Results and Discussion Monolayer structure. Self-assembled monolayers of DDDS on Au(111) were initially characterized by ellipsometric measurements. Ellipsometry showed that the film thickness depended , strongly on the solvent used in the self- assembly (soaking). With polar solvents, such as ethanol and methylene chloride, the average film thickness was 512 A; whereas in non-polar solvents, such as decane and octane, the average film thickness was 912 A. Molecular mechanics calculations showed the distance from the gold surface to the (o- terrninal hydrogen of 9-decene-1-thiolate was 14.0 A (DDDS dissociated upon adsorption onto gold and formed 9-decene gold thiolate on the surface).19 Based on these calculations, it was estimated that the molecules were tilted from the surface normal by ~48° and ~69° on the films prepared from nonpolar solvents and polar solvents, respectively. Decanethiol, which has the same number of carbon atoms as 9-decene-1-thiol and a similar calculated film thickness (14.4 A), formed a 1111 A thick monolayer on Au(111) regardless of whether the film was prepared from polar or nonpolar solvents.20 The solvent effect in preparing DDDS films will be discussed further in the following section. Reflection-absorption infrared spectroscopy data confirmed the adsorption of DDDS on gold and revealed differences in film structure due to different preparation solvents (Fig. 7.1). Figures 7.1 (B) and (C) show the RAIRS spectra of DDDS monolayers after >10 hours self-assembly on gold in decane and ethanol solutions, respectively. For comparison, the neat DDDS liquid infrared spectrum is shown in Fig. 7.1 (A). All the peaks in these spectra can be 160 assigned by consulting the data in the Iiterature,13115v21 and the results are summarized in Table 7.1. J! ' j—fi‘j IFTfi— rfifi ' I—'_ 1 l0.5 x1 I r A A 1 3200 3000 2800 1600 1400 1200 1000 800 Wavenumbers (cm 1) Absorbance a_ k _a_ _L . I_ Fig. 7.1 IR spectra of (A) neat liquid DDDS; (B) DDDS monolayer on Au(111) prepared from 2mM solution in decane; (C) DDDS monolayer on Au(111) prepared from 2mM solution in ethanol. Table 7.1 Infrared band assignments for DDDS Band frequency (cm'1) liquid DDDS DDDS SAM prepared DDDS SAM prepared mode from decane solvent from ethanol solvent 3076 3084 3079 vas=CH2 2998 3003 - v=CH- 2976 2983 - vs=CH2 2927 2922 2927 vaSCH2 2854 2851 2857 vsCH2 1641 1645 1641 vC=C 1 462 - - 5CH2 1414 - - 8=CH2 994 - - w=CH- g 910 914 912 m=CH2 v-stretching vibration, 8-in-plane vibration, (o-out-of-plane vibration. 161 Modes associated with the methylene backbone. In bulk compounds, the symmetric and asymmetric methylene C-H stretching vibrations are highly sensitive to the structural environment of the alkyl chains. The V3CH2 and vasCH2 modes occur at 2850 and 2920 cm“, respectively, for highly crystalline systems. Both vibrations increase in frequency systematically with decreasing order.3v22 Likewise, in single-layer assemblies, a monolayer of highly ordered alkyl chains in an all-trans conformation should exhibit methylene stretching vibrations near those of crystalline systems, with disorder induced by gauche defect sites resulting in a systematic increase in vibrational frequencies. As shown in Fig. 7.1, the vsCH2 and vasCH2 modes for DDDS neat liquid were observed at 2854 and 2927 cm", respectively, indicative of a very disordered chain environment. In the film prepared from a decane solution, both bands exhibited much narrower line widths than those in the liquid spectrum, and were shifted to lower frequencies (vsCH2 at 2851 cm'1 and vasCH2 at 2922 cm"), reflecting a more ordered, crystalline-like molecular environment in the monolayer. For the monolayer prepared from a ethanol solution, although there was some narrowing of these bands, they remained at relatively high frequencies: vsCH2 at 2857 cm'1 and vaSCH2 at 2927 cm". This implies that the aliphatic chains in the monolayer prepared from ethanol were disordered and liquid-like, in marked contrast to DDDS monolayer prepared from decane. Modes associated with the (tr-olefin group. The terminal olefin group gave rise to seven IR bands in the IR region of 800-3200 cm“.13.15.21 In the liquid spectrum, the band at 3075 and 2976 cm‘1 can be assigned to the vinyl C-H 162 asymmetric (vas=CH2) and symmetric (vs=CH2) modes; whereas that at 29980m'1 is due to the B C-H stretch (v=CH-). The C=C stretching vibration (vC=C) occurs at 1641 cm", and the in-plane CH2 deformation (8=CH2) is at 1414 cm". The vinyl group also has two characteristic hydrogen wag (out-of-plane) vibrations at 994 cm‘1 ((o=CH-) and 910 cm'1 ((o=CH2). The orientation of these modes is shown schematically in Fig. 7.2. 912 A A I I I I I (measured) I I I I I I Fig. 7.2 Schematic diagram indicating the directions of the various vinyl transition dipole moments of DDDS. Both vC=C and v=CH2 modes are parallel to the vinyl double bond, while vas=CH2 mode is in the H-C-H plane, perpendicular to the vC=C and vs=CH2 modes. The out-of-plane wag modes, w=CH- and ro=CH2, are orthogonal to the other three modes. In the DDDS monolayer prepared from a decane solution, the three vinyl C-H stretches (vas=CH2, vs=CH2, and v=CH-) were well resolved and blue shifted by 5~8 cm‘1 with respect to the liquid spectrum. In addition, the vas=CH2 mode exhibited a much narrower line width. These phenomena reflect the change in environment of the vinyl group upon assembly. Similar results have been observed for other olefin-terminated mnolayers and were attributed to decreased hydrogen bonding of the terminal vinyl group.13-15 The vC=C and w=CH2 modes 163 appeared distinctively in the monolayer spectrum and exhibited narrower line widths with respect to the liquid results. The hydrogen wag vibration at 994 cm'1 did not appear in the monolayer spectrum due to an overlap with a strong (negative) Si-O absorption in this region. Based on IR peak intensities, we can make some qualitative arguments concerning the orientation of the vinyl bond. Figure 7.2 shows the minimum energy configuration of 9-decene-1-thiolate on gold substrates assuming a tilt angle of 48° from the surface normal. Since both vas=CH2 and vC=C modes, whose dipole moments are perpendicular to each other, exhibited strong absorption in the reflection spectrum, the vinyl bond must be neither perpendicular nor parallel to the surface normal. In addition, the strong absorption of the oo=CH2 bending mode indicated the H-C-H plane must be canted away from the surface normal. For the DDDS monolayer prepared from ethanol, a completely different film structure was observed. The peak intensities of all vinyl modes were greatly diminished with respect to the monolayer prepared from decane discussed above. In addition, the vinyl band positions were similar to those in the liquid sample, indicating that the molecular environment of the terminal vinyl group was similar to that of the neat liquid. This result is not surprising, considering that the methylene backbone frequencies for this film also suggested a liquid-like environment, and variation in structural order of the methylene backbone would directly modify the range of orientations of the terminal olefin groups. 164 It is interesting to note that nonpolar solvents facilitated the formation of ordered and crystalline-like DDDS monolayers on Au(111) while polar solvents did not. No such solvent effect was observed for decanethiol20 or didecane disulfide monolayers,23 indicating that the design and fabrication of o)- functionalized SAMs may not be a simple modification of n-alkyl structure in all cases. Similar solvent effects have been reported for mixed SAMs of 4’- substituted 4-mercaptobiphenyls on Au(111).24-25 The composition of the mixed SAMs in equilibrium depended on the polarity of the solvent from which they were assembled and the kinetics of SAM formation depended strongly on intermolecular dipolar interactions. Because DDDS monolayers prepared from polar solvents were disordered and liquid-like, only the DDDS films prepared from nonpolar solvents were used in the following studies of film properties. Scanning tunneling microscopy measurements were performed to obtain further information on the structure of DDDS monolayers. In common with decanethiol SAM, DDDS monolayer displayed monatomic deep depressions on the surface; however, the STM images gradually degraded after a few scans and no molecular-resolution information could be obtained. This is probably because the olefin terminus in DDDS SAM was sensitive to the electron-induced reactions caused by the tunneling process, which apparently destroyed the film order after prolonged scanning. A range of experimental tunneling conditions was investigated but none produced significantly improved images. Thermal stability. The thermal stability of DDDS monolayer was tested by placing the sample in a vial filled with decane, and immersing the vial in a 165 water bath for 20 min at the target temperature. Pure N2 gas was continuously purged through the vial before and during the thermal treatment. The annealing process was followed by RAIRS, ellipsometry, and contact angle measurements. Figure 7.3 shows a representative series of RAIRS spectra of the DDDS monolayer on Au(111) as a function of annealing temperature. fifi—fifi '1' fifi— A q I 5x104 23°C 45°C 75°C W *1 3000 2900 2800 2700 Absorbance fl— 3200 3l00 Wavenumbers (cm") I B 4 45 :AJ a_l A L_1A Absorbance a l 4 L \J \ LII O 0 % ._l* 8 O n 1700 I600 950 850 Wavenumbers (cm") Fig. 7.3 RAIRS spectra of DDDS SAMs on Au(111) following annealing at the indicated temperature. As the temperature increased, the peak intensity of all modes associated with the vinyl bond decreased. Closer inspection of the spectra in Fig. 7.3 (A) also revealed peak intensity changes and blue shifts of the vasCH2 and vsCH2 166 modes during the annealing process. The integrated peak areas for these modes are depicted in Fig. 7.4. A (0:012, 914 cm-l 100 . I vC=C, 1645 cm" sol 3 C a c 801 e vu=CH2, 3083 cm" 2 70- 0 5‘ 604 2 50'1 3 40- .5 301 '3 20 A 8 101 20 4o 60 80 100 Annealing temperature (°C) A vas CH2, 2922 cm" . v, €112,285] cm“ 01501 010101 e: . A. J . d 01 01 01 f N .31 Ol Peak Intenelty change (96) N GI 40 60 80 IOO Anneellng temperature (°C) Fig. 7.4 Plot of peak intensity change for DDDS monolayers from Fig. 7.3 as a function of anneal temperature in (A) vas=CH2, vC=C and 01=CH2 modes; (B) VasCHz and VsCH2 modes. Three processes can contribute to the decrease in the integrated IR band intensity of the olefin modes during annealing: orientation change, molecular desorption, and polymerization/oiigomerization. Bulk DDDS liquid was annealed to 100 °C for 1 hour in N2 environment. Neither polymers nor oligomers were detected by transmission IR or mass spectrometry. It has been speculated that the formation of unsaturated monolayers on surfaces would further reduce the reactivity of the terminal groups due to the immobilization of the monomers”.15 167 Kurth et al. have shown that a classical solution-phase 8N2 reaction failed to occur when one of the reactants was immobilized on a silica surface.26 Thus, the diminution in vinyl modes was probably not due to polymerization/oligomerization of the olefin group, but a result of reorientation and/or desorption. The position of both vasCH2 and vsCH2 modes shifted to higher frequency (~3 cm“) during annealing and their peak widths broadened, indicating that a significant number of gauche defects were created in the DDDS monolayer during the thermal treatment and the film was no longer tightly packed. These results are similar to the annealing behavior of alkanethiol monolayers?o Decanethiol SAM formed a mixture of close-packed hexagonal phase and less dense “lying-down" pxxl3 phase after annealed to 45 °C. Upon heating to 75 °C, the surface was completely covered by the pX\/3 phase. At 100 °C, most of the decanethiol molecules desorbed, and a disordered, liquid-like phase was left on the surface. In conjunction with the DDDS contact angle measurements and the changes in monolayer thickness (Table 7.2), we concluded that the DDDS monolayer behaved very similarly to a decanethiol film during annealing. At room temperature, the average DDDS monolayer thickness was ~2 A less than that of decanethiol, indicating the DDDS molecules were more tilted on the substrate surface. The water contact angle was ~ 10° lower than decanethiol SAM, compatible with the higher polarity of the olefin terminus compared to the methyl terminus of decanethiol. It is speculated that, upon annealing to 45 °C, some of the DDDS molecules desorbed, and the monolayer left on the surface 168 was a mixture of the hexagonal phase and the px13 phase. Consequently, the average monolayer thickness decreased by 1.7 A compared to the room temperature value. The water contact angle dropped by 5° probably because more methylene backbones were exposed to the surfaced-27-28 In the RAIRS data, the vinyl modes decreased while vasCH2 mode increased due to orientational changes, film disorder and molecular desorption.29 After annealing at 75 °C, the surface was covered by the px13 phases, leading to the dramatic diminution in film thickness (A = -5.5 A) and water contact angle (A = -22°). At the same time, the vinyl mode intensity decreased by ~70 % and the vasCH2 mode increased by ~55 % of their room temperature values. After annealed to 100 °C, most of the DDDS molecules desorbed from the surface, as evidenced by a reduction in all modes, and the liquid-like disordered film left on the substrate was only 3.0 A thick. The surface became more hydrophilic and had a water contact angle of 68°. Table 7.2 The thickness and water contact angles of DDDS monolayer during annealing Temperature Thickness \ A thickness“I Contact angle \ A contact anglebW 23 °c 8.8 A l 0 A 96° 0° 1 l \ \ 45°C 7.1A \ -1.7A T 91° X -5° \ l 75 °c 3.3 A I -5.5 A 1 74° 22° W 100 °c 3.0 A J -5.8 A j 68° 1 -28° fi\ ° A thickness = thickness at 23 °C - thickness at corresponding temperatures; b A contact angle = contact angle at 23 °C - contact angle at corresponding temperatures. 169 Similar thermal behavior has been observed for 18-octadecene-1-thiol SAMs on Au(111).14 Grazing incidence X-ray diffraction data showed that after the film was annealed to 55 °C, the average domain size decreased from ~500 A to ~300 A and the Bragg peak intensity dr0pped by ~75 %. Further annealing resulted in greater degradation of the 18-octadecene-1-thiol monolayer and an irreversible structural phase transition occurred at 90 °C. These results further supported our conclusions of the thermal behavior of DDDS SAMs. Reactivity of DDDS monolayers on Au(111). The reactivity of the olefin- tenninated DDDS monolayers was tested by exposing them to the broad spectrum emission UV light from a low pressure l-lg arc lamp (A~250—400 nm). Initially the DDDS SAMs on gold were placed in a glass vial (transparent > 300 nm) during UV irradiation and the vial was continuously purged with pure N2 to avoid any photo-oxidation reaction occurring in the films.30-31 After 30 min irradiation, there was no significant change in the RAIRS spectrum of the DDDS monolayer (Fig. 7.5). The intensity of all vinyl modes remained constant. The width of the vaSCHz mode broadened slightly, indicating that a small degree of alkene-chain disordering occurred during the irradiation procedure.29 This result is in great contrast to the styrene-terminated thiols which can react and polymerize under the irradiation of 514 nm laser light.32’33 This difference is probably originated from the different photo-reactivities between or-olefins and styrenes. It is know that alkenes absorb strongly at 177 nm (emax ~ 10,000), while styrenes absorb strongly at 244 nm (8",," ~ 12,000) and 282 nm (8max ~ 450) due to the conjugated 1: system.34 170 11 fl I—r I 1rfi ' l fir' I ' 5x104 ._4‘ Absorbance 0min lit/Vt 3...... Jl 3200 3000 2800 1600 1400 1200 1000 Wavenumbers (cm") Fig. 7.5 RAIRS spectra of DDDS monolayer on Au(111) in pure N2(g) environment after 0 and 30 min near-UV (> 300 nm) irradiation. The broad feature at 1000 - 1200 cm'1 is due to polarizer artifact. Since the DDDS films appeared photostable under near-UV (>300 nm) irradiation, it was decided that a radical initiator should be added to the soaking solvent to facilitate the vinyl polymerization. The first initiator chosen was 2,2’- azobisisobutyronitrile (AlBN) which is commonly used in free radical vinyl polymerization reactions.35 About 2 mg of AIBN was dissolved in 20 mL of pure toluene to make a solution of ~1 mM concentration. The DDDS films were then placed in the toluene solution in a glass vial. Pure N2 was purged through the vial for 30 min before the UV irradiation procedures were repeated. Surprisingly, no reaction occurred in the films even after AIBN addition (Fig. 7.6). All IR peaks remained constant (position and intensity) after 30 min of near-UV irradiation with AIBN. According to the literature, the photochemical decomposition of AIBN has the largest quantum yield at short UV wavelengths (<300 nm).36 Since our reaction conditions were limited to near-UV (>300 nm) irradiation, the 171 photochemical decomposition of AIBN might be limited and as the result, no reaction occurred on the DDDS monolayers. Absorbance 1441 1‘1—1_L4 l_n l I V ' Y ' Y ' ‘r I ' firfiil fi r 3200 3000 2800 1600 1400 1200 1000 Wavenumbers (cm") Fig. 7.6 RAIRS spectra of a DDDS monolayer in toluene with ~1 mM AIBN after 0 and 30 min near-UV irradiation (> 300 nm). In order to have the full spectrum of UV light emitted from the lamp (A~250-400 nm) irradiating on the films, the sample container material was changed from glass to quartz (transparent >235 nm). Before adding the AIBN initiator, we checked the DDDS monolayer structure as a function of full UV (250- 400 nm) irradiation by RAIRS (Fig. 7.7). The RAIRS data revealed structural changes occurring in the DDDS monolayers without the initiator: the peak intensities of all vinyl modes decreased, while the intensity of methylene stretch modes increased and blue shifted (Fig. 7.8). 172 _l S9 (.11 X p—A O L E 6min 10min Absorbance A I . 15min 20 min r fi w— 3200 31Too 3300 ' 29'00 - 2810—0 2700 Wavenumbers (cm") ‘II —I . , _._ r , '— ap/AMVV I2.0)(10‘4 . 0mm .1 3min 81 . a 6mm “l '8‘ 10min 84 -° 15min «W ‘MW 1700 1600 1000 900 Wavenumbers (cm") Fig. 7.7 RAIRS spectra of DDDS SAMs on Au(111) in toluene obtained at various full-UV (250-400 nm) irradiation time. 173 A m=CH2, 914 cm" I vC=C, 1645 cm'1 9 vas=CH2, 3083 cm'1 100 sol 80J 701 so - 50 ~ 40 ~ 30 - 20 Peak intensity (%) 0 5 10 15 20 Time (min) A vas CH2, 2922 cm'1 I vs CH2, 2851 cm'1 Peak intensity change (%) 30 j 10+ -109- .__- . e - -9. o 5 to 15 20 Time (min) Fig. 7.8 Plot of peak intensity change for DDDS monolayers from Fig. 7.7 as a function of full-UV (250-400 nm) irradiation time in (A) vas=CH2, vC=C and m=CH2 modes; (B) vasCHz and VSCHz modes. 174 Figure 7.8 (A) quantifies the absolute peak intensity change of the three vinyl modes plotted against UV irradiation time. The peak intensities of veg-=CH2 and vC=C stretches, whose dipole moments are in the vinyl plane and perpendicular to each other, decreased at a similar rate with increasing exposure; whereas the w=CH2 mode, whose dipole moment is perpendicular to the vinyl plane, diminished at a slower rate. No new spectral features appeared during the irradiation. In contrast to the decreased vinyl group peak intensities, Figure 7.8 (B) shows that the peak intensity of vasCH2 mode increased almost 100% after 6 min irradiation but the vsCH2 mode increased only ~10°/o. The substantial increase in vasCH2 peak intensity and peak width suggests that the methylene backbones were no longer fully extended or resided in a crystalline- like environment, instead there were significant numbers of gauche defects created in the monolayer during the full spectrum UV irradiation. As discussed previously in the thermal stability section, three processes can contribute to the intensity decrease of the vinyl modes during irradiation: orientation change, molecular desorption, and polymerization/oligomerization. Since no polymerization/oligomerization reactions occurred for pure DDDS liquid under the same UV irradiation in N2 environment for >45 min, the peak intensity changes observed in the RAIRS data of DDDS monolayers were likely not due to reactions, but were the results of reorientation or molecular desorption during irradiation. As the control experiment, decanethiol SAMs were irradiated under identical full spectrum UV conditions (Fig. 7.9). The intensity of the vsCHa mode, 175 whose dipole moment is aligned parallel to 09-010 bond, decreased by ~60% after 20 min full range UV (250-400 nm) exposure (Fig. 7.10 (A)). At the same time, the vasCHa mode, whose dipole moment is perpendicular to the 09—010 bond, increased by ~10°/o. This result suggests that the terminal methyl group inclined further away from the surface normal during irradiation and the film morphology became more disordered. The intensity increase in the vasCH2 and vsCH2 stretching modes (Fig. 7.10 (B)) and their high frequency shifts further confirmed that a full range UV irradiation-induced orientation change occurred in the decanethiol films. 5x104 1 0min O 81 “El 10min 8 .D < 20min 4 3200 3100 3000 2900 2800 2700 Wavenumbers (cm") Fig. 7.9 RAIRS spectra (C-H region) of decanethiol SAMs on Au(111) in toluene obtained at various full-UV (250-400 nm) irradiation times. 176 A 120 .1 g v v 100 Q v;5 CH3, 2963 cm'1 >» g ,1 I vs CH3, 2878 cm‘1 8 80 .E 5 °°1 On 40 'U r 1 j 0 5 10 15 20 Time (min) I v” CH2. 2921 cm'l 9 vs CH2, 2850 cm'1 Peak intensity change (%) 0 5 1‘0 — 15 20 Time (min) Fig. 7.10 Plot of peak intensity change for decanethiol SAMs from Fig. 7.9 as a function of full-UV (250-400 nm) irradiation time in (A) vasCHa and vsCHa modes; (B) vasCH2 and vsCH2 modes. Similar irradiation-induced damage has been reported in the study of investigating the potential of utilizing SAMs as templates for patteming.37'39 Ultraviolet light, electron or X-rays irradiation-induced film modification can be used as lithographic approach to pattern surfaces, with SAMs serving as either a positive or a negative resist or as a chemical template for a lateral structuring.4°:41 It has been demonstrated that irradiation can cause the loss of the orientational and conformational order in the assembly, the desorption of the film fragments, the reduction of the thiolate headgroup species, and the appearance of new sulfur species.3°:31:42v43 Our RAIRS data demonstrate that both alkanethiol and thiol derivatized SAMs are sensitive to short wavelength UV 177 light (A<300 nm) and the original close-packed crystalline-like monolayer structure becomes disordered upon such irradiation. Since both DDDS and decanethiol SAMs remained unchanged under near UV (>300 nm) irradiation in solvent, another radical initiator, benzoin methyl ether (BME), was tested for its ability to initiate the vinyl polymerization. Benzoin and its alkyl ether derivatives undergo the so-called a-cleavage after absorbing photons in the wavelength region from 313 to 365 nm4445 leading rapidly to benzoyl and substituted benzyl radicals (Scheme 7.2). OCH OCH ©—(C)— $—© —h—>©—c. + .i—@ :x_@ "CH3 Scheme 7.2. Initial rat-cleavage of BME after absorption of a photon in the wavelength range from 313 to 365 nm. Figure 7.11 contains RAIRS spectra acquired from a DDDS monolayer in toluene with ~1 mM BME obtained at different near UV (>300 nm) irradiation times. Similar to the full spectrum UV irradiation (Fig. 7.9), all the modes associated with vinyl bonds lost intensity while methylene modes gained intensity. No new spectral features appeared during the irradiation. It should be noted that without the BME initiator, there was no structural change in the DDDS monolayer (Fig. 7.6) under the same UV exposure; thus, all the peak intensity changes observed in Fig. 7.11 were related to the photochemical reaction of the initiator. 178 2.0x10’4 1 Absorbance 3200 3 100 3000 2900 2800 2700 Wavenumbers (cm") 15min Haw 3Oflun§//\CA 1700 1600 1000 900 Wavenumbers (cm") Absorbance ‘ I2.5x10“ Fig. 7.11 RAIRS spectra of DDDS SAMs on Au(1 1.1) in toluene with ~1 mM BME obtained at various near UV (>300 nm) irradiation times. The integrated peak areas for the vinyl and methylene modes are shown in Fig. 7.12. Compared with the RAIRS data of full spectrum UV irradiation (Fig. 7.8), the increase of vasCH2 mode was much slower (Fig. 7.12): only ~15% increase after 15 min irradiation in contrast to the 90% increase observed for full spectrum UV irradiation. In addition, the peak width of vasCH2 mode almost remained the same during the UV exposure, indicating that the order of the original crystalline-like methylene backbones was not destroyed. The rate of 179 peak intensity changes for three vinyl modes and vsCH2 mode was similar to that in full spectrum UV irradiation (Fig. 7.8). A m=CH2, 914 cm'1 I vC=C, 1645 cm" 0 vas=CH2, 3083 cm-1 A Peak intensity change (%) 5‘ 60 .. w I 40 I I a 0 10 20 ‘ 30 time (min) A vas CH2, 2922 cm-1 B I vs CH2, 2851 cm’1 Peak lntenslty change (%) tlme (min) Fig. 7.12 Plot of peak intensity change for DDDS SAMs from .Fig. 7.11 as a function of UV-light (>300 nm) exposure time with ~1 mM BME In (A) vas=CH2, vC=C and m=CH2 modes; (B) vasCH2 and vsCH2 modes. In addition to RAIRS, the near UV (>300 nm) irradiation of DDDS monolayers with BME was also followed by ellipsometry and contact angle measurements. The film thickness and wettability changes during irradiation are listed in Table 7.3. 180 Table 7.3 The thickness and water contact angles of DDDS monolayer during near UV irradiation in toluene with ~1 mM BME. Irradiation Thickness A thicknessa Contact angle A contact angleb Time 0 min 8.5 A o A 97° 0° 15 min 8.2 A -2.3 A 92° -5o 30 min 5.0 A -2.5 A 91° -50 " A thickness = thickness at 0 min irradiation - thickness at corresponding exposure times; b A contact angle = contact angle 0 min irradiation - contact angle at corresponding exposure times. After 15 min near-UV irradiation of the DDDS monolayer in toluene with BME, the avaerage ellipsometric thickness dropped by 2.3 A while the surface water contact angles decreased by 5°. Another 15 min irradiation barely changed the thickness or wettability, indicating that the photochemical reaction was almost complete during the first 15 min. The peak intensity changes in RAIRS spectra also agreed with this conclusion: after the first 15 min irradiation, both vasCH2 and vsCH2 modes reached maximum values, and the three vinyl mode intensities remained almost constant during the second 15 min UV irradiation. As a control, decanethiol SAMs were irradiated under identical conditions (Fig. 7.13). Surprisingly, peak intensity changes were also observed (Fig. 7.14), indicating that reorientation and alkyl chain disorder occurred in the decanethiol assembly during the near-UV irradiation when BME was present in the soaking solution. 181 I2.5x10“‘ 0 min in o d 5 . '8 15 mm o (D .t: - < 30 min 3200 3100 3000 2900 2800 2700 Wavenumbers (cm' I) Fig. 7.13 RAIRS spectra of decanethiol SAMs on Au(111) in toluene with ~1 mM BME obtained at various near-UV (>300 nm) irradiation times. § .53 (I) g 80 g .5. 'fi 9 vas CH3, 2963 cm'1 2‘3 I vs CH3, 2878 am1 so k-_h _ 2 _ .__.fi 0 10 20 30 Time (min) 100 i , I vas CH2, 2921 cm'1 9 vs CH2, 2850 cm'1 Peak intensity change (%) O 10 20 30 Time (min) Fig. 7.14 Plot of peak intensity change for decanethiol SAMs from Fig. 7.13 as a function of near UV (>300 nm) exposure time with BME in (A) vasCHa and VSCH3 modes; (B) vasCH2 and vsCH2 modes. 182 Thermal Stability of the irradiated DDDS Monolayers. Since the free radicals produced by BME induced damage to decanethiol SAMs, it is difficult to judge whether the decrease in vinyl mode intensities observed in DDDS monolayers during the same UV exposure was the result of a disorder effect, or a combination of reorientation and vinyl reaction. One method employed to probe for any polymer/oligomer formed on the DDDS film, was to investigate the monolayer thermal stability. Polymerized/oligomerized SAMs have demonstrated better thermal stability and robustness than the unreacted monomer films.33v45 Thus we annealed the 30 min near-UV irradiated DDDS monolayer to 45, 75 and 100 °C. The annealing process was followed by RAIRS (Fig. 7.15), ellipsometry and contact angle measurements (Table 7.4). Absorbance l 75°C + 100°C 3200 3 100 3000 2900 2800 2700 Wavenumbers (cm") Fig. 7.15 RAIRS spectra of DDDS SAMs on Au(111) after 30 min UV (>300 nm) irradiation with ~1 mM BME followed by annealing at the indicated temperature. 183 88 _l__l Peak intensity (%) 8 3 A v” CH2, 2922 cm" I v: CH2, 2851 cm" 20 4o 60 80 100 Annealing temperature (°C) Fig. 7.16 Plot of peak intensity change for a 30 min near UV (>300 nm) irradiated DDDS monolayer in toluene with ~1 mM BME initiator from Fig. 7.15 as a function of anneal temperature in vasCH2 and vsCH2 modes. Table 7.4 The thickness and water contact angles of DDDS monolayers after 30 min near UV irradiation in toluene with BME followed by anneal. Temperature Thickness A thicknessa Contact angle A contact angleb 23 °c 6.0 A o A 91 ° 0° 45 °C 6.5 A +0.5 A 33° -30 75 °c 5.3 A +0.3 A 79° -120 100 °c 4.5 A -1.5 A 75° -150 a A thickness = thickness at 0 min irradiation - thickness at corresponding exposure times; b A contact angle = contact angle 0 min irradiation - contact angle at corresponding exposure times. In comparison with the unirradiated DDDS monolayers (Fig. 7.3), the near- UV irradiated DDDS film at room temperature has only ~45% of the original vas=CH2 peak intensity. Both the vasCH2 and vsCH2 modes decreased monotonically with increasing annealing temperature (Fig. 7.16), in contrast to the unirradiated DDDS films, in which the vasCH2 mode increased rapidly until 75 °C and started to decline at higher annealing temperatures. Another pronounced difference between pre-irradiated and unirradiated DDDS films is the thickness 184 changes during annealing. For the unirradiated film, the thickness decreased continuously and dropped to 3.3 A at 75 °C; whereas, the irradiated film maintained an almost constant 6 A thickness up to 75 °C, indicating a better thermal stability. The pre-irradiated film was more hydrophobic than the unirradiated film during annealing, suggesting that less methylene backbones were exposed to the surface. Based on the above results, we believed that some degree of polymerization/oligomerization did happen on the DDDS monolayers in addition to the reorientation and disorder effect during near-UV irradiation with BME. During annealing, the unreacted molecules (~45 %, probably due to geometric reaction considerations) desorbed from the surface resulting in a decrease of methylene stretching modes. The polymerized DDDS molecules were more robust and the average film thickness did not change substantially upon heating to 75 °C. At higher temperatures, the polymerized molecules started to degrade, as indicated by a rapid decrease in vasCH2 and vsCH2 mode intensities. 7.4 Conclusion Self-assembled monolayers (SAMs) of di(9-decene) disulfide (DDDS) on Au(111) were investigated by ellipsometry, RAIRS and contact angle measurements. Monolayers prepared from polar solutions were disordered with large amounts of gauche defects present, and the RAIRS spectra resembled free DDDS in the liquid state. In contrast, DDDS films prepared from nonpolar solvents were well-ordered with the methylene backbones in a crystalline environment, similar to decanethiol monolayers on Au(111). The solvent effect 185 revealed that the monolayer formation depended strongly on intermolecular and adsorbate-solvent interactions for thiol derivatized SAMs. In addition to structural investigations, the thermal stability of DDDS monolayers was also studied. It was demonstrated that DDDS monolayers behaved very similarly to decanethiol films during thermal treatment. After annealing to 45 °C, DDDS SAMs was speculated to form a mixture of close- packed hexagonal phase and less dense pX1/3 phase. At 75 °C, the surface was completely covered by the px113 phase. When heated to 100 °C, most of the DDDS molecules desorbed, and a disordered, liquid-like phase was left on the surface. The reactivity of the olefin—terminated DDDS monolayers was tested by exposing them to different wavelength ranges of UV light. It was found that both DDDS and decanethiol SAMs were sensitive to UV light with A<300 nm and the original close-packed crystalline-like monolayer structure became disordered upon such irradiation in solvent. On the contrary, no irradiation-induced damage was observed on SAMs under near-UV (>300 nrn) exposure in solvent. Benzoin methyl ether (BME), which undergoes photodissociation after absorbing photons in the wavelength region from 313 to 365 nm, was selected to facilitate the vinyl reaction. It was demonstrated that polymerization/oligomerization as well as reorientation/disorder occurred on DDDS monolayers after near-UV irradiation in the presence of BME. 186 7.5 (1) Literature Cited Ulman, A. An Introduction to Ultra thin Organic Films from Langmuir- Blodgett to Self-Assembly, Academic Press: San Diego, CA, 1991. (2) Whitesides, G.; M.; Gorrnan, C. B. In The Handbook of Surface Imaging and Visualization; Hubbard, A. T., Ed.; CRC Press: Boca Raton, FL, 1995; 713. (3) 2358. (4) 558. (5) (6) (7) 6704. (8) (9) (10) (11) (12) Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987, 109, Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, Li, J.; Liang, K. S.; Scoles, G.; Ulman, A. Langmuir1995, 11, 4418. Hutt, D. A.; Leggett, G. J. Langmuir1997, 13,2740. Yan, L.; Marzolin, C.; Terfort, A.; Whitesides, G. M. Langmuir1997, 13, Boubour, E.; Lennox, R. B. Langmuir2000, 16,7464. Poirier, G. E.; Tarlov, M. J. J. Chem. Phys. 1996, 105, 2089. Kawasaki, M.; Sato, T.; Yoshimoto, T. Langmuir2000, 16,5409. Ulman, A. Chem. Fiev. 1996, 96, 1533. Bumagin, N. A.; Luzikova, E. V.; Beletskaya, l. P. Russ. J. Org. Chem. 1995, 31, 1480. (13) (14) (15) (16) Allara, D. L.; Nuzzo, R. G. Langmuir1985, 1,52. Leung, T. Y. B. Ph.D. thesis, Princeton University, 1998. Peanasky, J. S.; McCarley, R. L. Langmuir1998, 14, 113. Kabalka, G. W.; Varrna, M.; Varma, R. S. J. Org. Chem. 1986, 51, 2386. 187 (17) Kim, T.; Crooks, R. M. Tetrahedron Lett. 1994, 35, 9501. (18) Mowery, M. D.; Evans, C. E. Tetrahedron Lett. 1997, 38, 11. (19) Molecular mechanics calculations were performed by using SPARTAN 5.0 molecular modeling program (Wavefunction Inc., CA), based on empirical Merck force fields. (20) More details have been discussed in Chapter 4. (21) Colthup, N. B.; Daly, L. H.; Wiberley, S. E. Introduction to Infrared and Raman Spectroscopy, Academic Press: Boston, 1990. (22) Porter, M. 0.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559. (23) Bain, C. D.; Biebuyck, H. A.; Whitesides, G. M. Langmuir1989, 5,723. (24) Kang, J. F.; Ulman, A.; Jordan, R.; G., K. D. Langmuir1999, 15, 5555. (25) Kang, J. F.; Ulman, A.; Liao, 8.; Jordan, R. Langmuir1999, 15,2095. (26) Kurth, G. G.; Bein, T. Langmuir1993, 9, 2965. (27) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111,321. (28) Troughton, E. B.; Bain, C. D.; Whitesides, G. M.; Nuzzo, R. G.; Allara, D. L.; Porter, M. D. Langmuir1988, 4, 365. (29) Hostetler, M. J.; Stokes, J. J.; Murray, R. W. Langmuir1996, 12,3604. (30) Huang, J.; Hemminger, J. C. J. Am. Chem. Soc. 1993, 115, 3342. (31) Lewis, M.; Tarlov, M. J. Am. Chem. Soc.1995, 117, 9574. (32) Ford, J. F.; Vickers, T. J.; Mann, C. K.; Schlenoff, J. B. Langmuiri 996, 12, 1944. 188 (33) Duan, L.; Garrett, S. J. Langmuir2001, 17, 2986. (34) Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instrumental Analysis, Saunders College Publishing: Philadelphia, 1998. (35) Motherwell, W. B. Free radical chain reactions in organic synthesis, Academic Press: London, 1992. (36) Simoes, J. A. M.; Greenberg, A.; Liebman, J. F. Energetics of organic free radicals, Blackie Academic & Professional: London, UK, 1996. (37) Laibinis, P. E.; Graham, R. L.; Biebuyck, H. A.; Whitesides, G. M. Science 1991, 254, 981. (38) Seshadri, K.; Froyd, K.; Parikh, A. N.; Allara, D. L.; Lercel, M. J.; Craighead, H. G. J. Phys. Chem. 1996, 100, 15900. (39) Heister, K.; Zhamikov, M.; Grunze, M.; Johansson, L. S. 0.; Ulman, A. Langmuir2001, 17,8. (40) Huang, J.; Dahlgren, D. A.; Hemminger, J. C. Langmuir1994, 10, 626. (41) Xia, Y.; Whitesides, G. M. Angew. Chem. Int. Ed. 1998, 37, 550. (42) Zhamikov, M.; Frey, 8.; Heister, K.; Grunze, M. Langmuir2000, 16,2697. (43) Zhamikov, M.; Frey, S.; Golzhauser, A.; Geyer, W.; Grunze, M. Phy. Chem. Chem. Phys. 1999, 1, 3163. (44) Viklund, C.; Ponten, E.; B., G.; lrgum, K.; Horstedt, P.; Svec, F. Chem. Mater. 1997, 9, 463. (45) Pokhrel, M. R.; Janik, K.; Bossmann, S. H. Macromolecules 2000, 33, 3577. (46) Kim,T.; Chan, K. C.; Crooks, R. M. J. Am. Chem. Soc. 1997, 119,189. 189 Chapter 8 Conclusions and Outlook The objectives of this work were to provide a deeper understanding of the impact of molecular structure on monolayer morphology and properties, and to examine the influence of the monolayer environment on the intrinsic reactivity of the unsaturated moiety incorporated in the self assembled monolayers (SAMs). Four groups of thiol and disulfide SAMs have been investigated, representing a systematic study of aliphatic, aromatic, and aromatic/aliphatic mixed monolayers on Au(111). In this chapter, a summary of the experimental results will be reported. Possible future work will be presented at the end of the chapter. 8.1. Significance of the Results Impact of molecular structure on monola yer structure and properties. The self-assembly of simple alkanethiols or derivatized thiols on Au(111) is initiated by strong chemical interactions between the sulfur headgroup and the gold surface which result in chemisorption of the molecules as thiolates. In addition to the headgroup-substrate interaction, several other interactions are involved in the self assembly process, such as endgroup-substrate, chain-chain and endgroup- endgroup interactions. The overall balance of all interactions determines the final monolayer structure and the film pr0perties. 190 In the studies presented in this dissertation, octanethiol (CB), decanethiol (C10) and octadecanethiol (C18) formed densely packed, crystalline-like assemblies with fully extended alkyl chains tilted from the surface normal by 20°~40°. Molecular resolution STM images revealed that the n-alkanethiol molecules adopted the hexagonal and commensurate \/3X\/3R30° overlayer structure on Au(111). These self-assembling systems demonstrated the simple balance between the chemical interaction of the sulfur headgroup/gold surface and the hydrophobic interaction of the hydrocarbon chains. Annealing to elevated temperatures resulted in cleavage of chemical bonds in the SAMs: cleavage of the S-Au bond led to molecular desorption, and cleavage of the S-C/C-C bond led to monolayer decomposition. STM data suggested that, at a temperature <75 °C, molecular desorption occurred in the monolayer and low density "striped" phases were formed. In the low density striped phases, the molecular axes were near-parallel to the surface plane in order to maximize the molecular/substrate interaction. After annealing to 100 °C, most of the molecules had desorbed, and a disordered liquid-like phase was left on the Au(111) surface, indicating monolayer decomposition. Di(9-decene) disulfide (DDDS) SAMs on Au(111) were investigated to study the impact of a small substituent (olefin terminus) on the monolayer structure. The olefin terminal was nonbulky; therefore, the V3x1/3R30° lattice should be sterically allowed. It was found that the film structure depended strongly on the polarity of the solvents used in the soaking solutions. In a nonpolar solvent, ordered crystalline-like DDDS monolayers similar to n- 191 alkanethiol SAMs were formed on the gold substrate; whereas in a polar solvent, the films were disordered and liquid-like with large amounts of gauche defects present. This solvent effect suggested that the design and fabrication of 0)- functionalized SAMs was not a simple modification of n-alkyl structure. Instead, for thiol derivatized SAMs, the monolayer formation depended strongly on intermolecular and adsorbate-solvent interactions. The thermal stability of ordered DDDS monolayers was studied following annealing in decane and was demonstrated to be similar to that of decanethiol SAMs. These findings suggested that (lo-substitution by an ethylene group had a small impact on the monolayer structure if prepared from a nonpolar solvent, and a well ordered array of ethylene termini could be prepared via the self-assembly strategy. The phenyl-terminated 6-phenyl-n—hexanethiol (PHT) and styrene- terrninated 6-(p-vinylphenyl)-n-hexanethiol (VHT) were investigated to examine the influence of aromatic termini on monolayer packing and order. Both molecules chemisorbed on gold as thiolates, but the alkyl chains were disordered for both SAMs. Three different stripe phases, characterized by alignment of the molecular axes with the surface plane, were observed in PHT monolayers following annealing in ultrahigh vacuum. In contrast, the VHT monolayer had a structure in which the average molecular tilt angle was close to the surface normal. The presence of an additional two C atoms in the case of VHT stabilized the "standing up“ orientation. Such a minor molecular modification produced major differences in the morphologies of the SAMs, suggesting that in these systems there was a delicate balance between the aromatic and aliphatic 192 interactions. Similar effects were also observed in azobenzene derivatized thiol SAMs: the alkyl-terminated azobenzene thiols formed a complete different packing structure on Au(111) in comparison with azobenzene-terminated alkanethiols.“5 In general, for substituted alkanethiols, the alkyl chain length, the size of the terminal group together with the terminus interactions appear to play a crucial role in determining the resulting monolayer structure. Shorter chain thiols with bulky endgroups, e.g. PHT, will lead to lower density coverages and more distorted packing; whereas, small substitutes such as olefin groups will have little impact on the monolayer structure. Furthermore, to study the impact of the nature of the molecular backbone on the self assembly system, two terphenyl derivatized thiol SAMs assembled on Au(111) were characterized. Compared to n-alkanethiol molecules which had rather flexible aliphatic chains, the terphenyl units in p-methylterphenyl thiol (MTPT) and 2’,3’-cyclobutyl p-terphenyl thiol (CTPT) molecules were much more rigid and. bulkier. The nature of the intermolecular forces was significantly modified: the dispersion forces responsible for chain ordering in n-alkanethiols have been augmented by anisotropic noncovalent 1t-Tl: interactions.6 It was demonstrated that MTPT formed a densely-packed monolayer on Au(111) with the molecular axes slightly tilted away from the surface normal (~17°). Molecular resolution STM images of the MTPT monolayer revealed a (1/3xx/3)R30°-like packing with slightly larger lattice vectors than typical alkanethiol monolayers, which led to the formation of small domains of order and 193 large density of defects in the film. As discussed earlier, several energetic interactions contribute to the self-assembly and stabilization of thiol derivatized monolayers on Au(111). In general, when S-Au bonding was dominant, the usual (\lSXJB)R30° packing resulted. However, in aromatic thiols there were additional intermolecular forces derived from the it systems of neighbors7 and it was possible to form new room temperature monolayer phases. When interchain interactions were dominant, an ordered but incommensurate layer may be formed as observed for 4-[4’ (phenylethynyl)-phenylethynyl]-benzenthiol and 4’- chloro-4-mercaptobiphenyl monolayersfi.9 In the MTPT system examined here, the interchain 1H: and S-Au interactions appeared to be of similar magnitude. The result was that small domains of (43x43)R30°-like packing were observed by STM, but the structural restrictions imposed by the terphenyl portion of the MTPT created strain in the domain. This inhibited the formation of long range two- dimensional order. The CTPT molecules formed a completely different structure on Au(111) with molecules tilted far away from the surface normal (~72°). Additonally, only monatomically high islands were observed with no molecular order visible within the islands. The structural variance between MTPT and CTPT monolayers was attributed to steric factors: with the cyclobutyl unit attached to the phenyl ring, the van der Waals cross section of CTPT was too large to accommodate a commensurate VSN3R30°~Iike packing arrangement. Reactivity of SAMs with unsaturated functionalities. In this study, three unsaturated functionalities (olefin, styrene, and cyclobutyl) were incorporated into 194 both aliphatic and aromatic thiols to assess the possibility of forming covalent bonding within a single molecular layer. The ability to form monolayers that can be photopolymerized is of potential interest for creating pattemed surfaces by photolithography, and in the fundamental understanding of factors affecting polymerization in two dimensions. Polymerization of the styrene-terminated hexanethiol SAM was achieved by either near-UV irradiation or thermal treatment. It was demonstrated that the UV irradiation produced polymers up to a maximum of ~70 % conversion. In contrast, thermal annealing produced shorter chain polymers with a large pr0portion of CH3 and groups. The UV-light polymerized film was more robust to subsequent thermal degradation than the thermally polymerized film. The compositional differences between the thermally polymerized and photopolymerized VHT SAMs were attributed to the different molecular motion present in the VHT monolayer at various temperatures. At elevated temperatures, increased molecular motion created a wide range of potential reaction geometries and, in turn, a reduced specificity for successful cross- linking. The reactivity of the olefin-terminated DDDS monolayers was tested under the same conditions as those for VHT SAMs. Neither the thermal treatment ($100 °C) nor the near-UV irradiation initiated the polymerization/oligomerization reaction of the ethylene groups in DDDS SAMs. The different reactivity between VHT and DDDS monolayers probably originated from the different photo- reactivities of a-olefins and styrenes. it is know that upon irradiation with 514 nm 195 green light or UV light, styrenes can be excited and undergo polymerization“),11 Whereas, olefins only undergo photodimerization with irradiation by polarized short wavelength UV light, and the cross-linking is predominantly along the direction of the exciting polarizationfiz»13 Benzoin methyl ether (BME), which undergoes photodissociation under near-UV irradiation, was selected to facilitate the vinyl reaction. It was demonstrated that polymerization/0|igomerization as well as reorientation/disorder occurred on DDDS monolayers after near-UV irradiation in the presence of BME. The CTPT SAM which had cyclobutene units within the monolayer appeared photostable under the full spectrum UV irradiation. This result is in great contrast to cyclobutene, which undergoes ring opening reaction in the gas phase or solution state under direct UV irradiation.14115 The reduced photoreactivity of cyclobutene in CTPT monolayers was probably due to the immobilization of the monomers on the surface by formation of the strong Au-S bond or an unfavorable reaction geometry. 8.2. Possible Future Directions The information and understanding gained through the research detailed in this dissertation represents a starting point for engineering the structure and properties of SAMs for future technological applications. A number of interesting experiments can be carried out to expand both the basic understanding and applications of these films. We have demonstrated the feasibility of fabricating polystyrene terminated monolayers by UV light. One possible future direction is to take advantage of 196 this reactivity in surface patterning. The radiation-induced chemistry allows for select areas of the substrate to be polymerized, opening the way for the production of positive and negative resists. There has been considerable interest in using ultrathin SAM resist for patterning surfacesJG-20 Several reports have appeared illustrating the viability of SAMs as resist materials?“23 One possibility of employing the styrene-terminated SAM as a negative photo resist to transfer the image of a mask into a gold substrate is illustrated in Fig. 8.1. UV irradiation I I I I I I I " I] I] l Electrochemical desorption 8w l Etching CW Fig. 8.1 Diagrams illustrating the idea of using a styrene-terminated alkanethiol SAM as a negative resist to transfer the image of a mask into a gold substrate. mask Au Initially, the mask can be placed in contact with a styrene-terminated SAM (Fig. 8.1 A). The entire assembly can then be exposed to UV light, which will induce polymerization in the unmasked regions of the styrene-terminated SAM. Next, the unpolymerized portion of the resist can be selectively desorbed using reductive stripping voltammetry in a single compartment cell.24 It has been demonstrated that this electrochemical procedure completely removed the 197 monomeric resist, but removed little of the polymerized resist.23 The monomeric- resist remove will result in a negative image of the mask (Fig. 8.1 B). It has been demonstrated in this proposal that simple thermal treatment causes short oligerrnerization reactions to occur within the SAM. Hence removing the unexposed portions of a lithographically-pattemed styrene based SAMmay not be feasible. Finally, the mask image can be etched into the gold surface using an etching solution such as an aqueous KOH/KCN solution (Fig. 8.1 C).23 The exposed bare gold region, where the CN' can make intimate contact with, will be dissolved, while the thiol monolayer will retard this process and the monolayer- covered region will survive the CN’ etching.25-27 Since such resists consist of single-layered, small molecules, the theoretical resolution of lithographically defined features can be as small as a few nanometers. In addition, the substrate can be extended to other materials, such as Si”, Al”, and GaAs30, and nanometer-scale pattering will be straightforward. Scanning tunneling microscopy would be a useful tool to examine the precision with which structures can be lithographically patterned in this manner. Clearly, once polymerization has been initiated by UV exposure if chain lengths are long the chain may propagate into a region outside the illuminated area (into an area that was initially masked). Such "bleeding" of the pattem near the edges of the mask is in addition to any diffraction-related issues that currently limit the size of features that can be produced by UV and deep-UV lithography. Oligophenyl thiol SAMs on Au(111) have proven to be excellent model systems in which factors governing the rate of electron transfer/tunneling across 198 interfacial barriers can be probed.31'33 An understanding of these factors is of considerable importance to a number of scientific and technological problems, extending from long-range electron transfer in biological systems to the design and construction of molecular-based electronic devices ("moltronics").34 We have discovered that MTPT, which has a terphenyl backbone, formed a densely- packed, (VSxV3)R30°-like overlayer on Au(111) with the molecular axes slightly tilted away from the surface normal. It will be interesting to investigate the possibility of using self-assembled MTPT monolayer as an efficient tunneling conductor, which may find utility in thin-film-based electronic devices. The conductance of MTPT molecules can be measured after forming nanoscale domains in alkanethiol SAMs using STM.35 It has been demonstrated that the apparent height of oligophenyl thiol domains in alkanethiol SAMs increased as the lateral size of the domains grew, reflecting the increase in the vertical conductance of the domains due to the lateral intermolecular interaction.33 Using a model of conducting disks with various radii, the resistance of a single molecule and the effective lateral conductivity within a large domain (> 100 A2) can be estimated.3335”?6 ' We have demonstrated that the olefin-terminated disulfide (DDDS) formed densely packed, ordered monolayers on Au(111), and had similar film properties as n—alkanethiol SAMs. Modification of the terminal group to butadiene or diacetylene will lead to better reactivity of these films, while still keep the ordered assembly structure. Butadiene and diacetylene absorb strongly at 217 nm (Emax~20000 L/molocm) and 250 nm (Emax~21000 Umol-cm), respectively.37 Both 199 absorption bands are displaced to longer wavelengths compared to the a-olefin due to conjugation effects. The photo- and thermal-polymerization of butadiene on Au(111) is currently under investigation in our research group. The reactivity of diacetylenic thiol SAMs, which has the diacetylene unit inserted within the alkyl chains, has been investigated by several research groups.“40 Little rearrangement was required to yield linearly polymerized films if the diacetylenic thiols adopted a similar geometry on gold to that of rt-alkanethiols.41"42 The polymerized diacetylenic SAMs are extremely durable compared to either the n- alkanethiol SAMs or the unpolymerized diacetylenic SAMs. Further investigations of these high performance materials will yield applications in nonlinear optical materials, photo-conductors, photo-lithographic resists and chemical sensors.23o43v44 In the future, new and innovative self-assembled monolayer systems are expected to be devel0ped. As has been demonstrated in the work presented here, the superb spatial resolution offered by scanning probe techniques when coupled with other surface sensitive analytical instrumentation and computational studies, is essential to understand film morphology and structure. The effects of molecular modification, for example by incorporating bulky aromatic or unsaturated groups, on film properties are beginning to be understood. It is hoped that the scientific community will soon be able to apply such fundamental knowledge to the development of practical materials incorporating unique self- assembled monolayer architecture. 200 8.3. Literature Cited (1) Caldwell, W. B.; Campbell, D. J.; Chen, K.; Herr, B. R.; Mirkin, C. A.; Malik, A.; Durbin, M. K.; Dutta, P.; Huang, K. G. J. Am. Chem. Soc. 1995, 117, 6071. (2) Wang, R.; Iyoda, T.; Jiang, L.; Hashimoto, K.; Fujishima, A. Chem. Lett. 1996, 1005. (3) Tamada, K.; Nagasawa, J.; Nakanishi, F.; Abe, K.; lshita, T.; Hara, M.; Knoll, W. Langmuir1998, 14, 3264. (4) Han, S. W.; Kim, C. H.; Hong, S. H.; Chung, Y. K.; Kim, K. Langmuir 1999, 15, 1579. (5) Yu, H. Z.; Ye, S.; Zhang, H. L.; Uosaki, K.; Liu, 2. F. Langmuir2000, 16, 6948. (6) Reese, 8.; Fox, M. A. J. Phys. Chem. 31998, 102, 9820. (7) Lil, J.-H.; Allinger, N. L. J. Am. Chem. Soc. 1989, 111, 8576. (8) Yang, G.; Qian, Y.; Engtrakul, C.; Sita, L. R.; Liu, G.-y. J. Phys. Chem. B 2000, 104, 9059. (9) Kang, J. F.; Ulman, A.; Liao, 8.; Jordan, H.; Yang, G.; Liu, G.-y. Langmuir 2001, 17,95. (10) Ford, J. F.; Vickers, T. J.; Mann, C. K.; Schlenoff, J. B. Langmuir1996, 12, 1944. (11) Carlo, S. R.; Grassian, V. H. Langmuir1997, 13,2307. (12) Shashidhar, R.; Gruneneberg, K.; Shenoy, D. B.; Peek, B.; Kakala, U. K.; Naciri, J. SID Digest 1997, 315. 201 (13) 1998, (14) (15) (16) (17) 5305. (18) 1070. (19) (20) 1995, (21) (22) 5495. (23) 1995, (24) 335. (25) (26) (27) (28) Shenoy, D.; Grueneberg, K.; Naciri, J.; Shashidhar, R. Jpn. J. Appl. Phys. 37, L1326. Leigh, W. J.; Zheng, K.; Clark, K. B. Can. J. Chem. 1990, 68, 1988. Leigh, W. J.; Postigo, J. A. J. Am. Chem. Soc. 1994, 117, 1688. Huang, J.; Hemminger, J. C. J. Am. Chem. 800.1993, 115, 3342. Tarlov, M. J.; Burgess, D. F.; Gillen, G. J. Am. Chem. 800.1993, 115, Yang, X.; Kim, E.; Whitesides, G. M. J. Electrochem. Soc. 1996, 143, Xia, Y.; Whitesides, G. M. Angew. Chem. Int. Ed. 1998, 37, 550. Prichard, D. J.; Morgan, H.; Cooper, J. M. Angew. Chem. Int. Ed. Engl. 34, 91. Huang, J.; Dahlgren, D. A.; Hemminger, J. C. Langmuir1994, 10,626. Hutt, D. A.; Cooper, E.; Parker, L.; Leggett, G. J. Langmuir1996, 12, Chan, K. C.; Kim, T.; Schoer, J. K.; Crooks, R. M. J. Am. Chem. Soc. 1 17, 5875. Widrig, C. A.; Chung, C.; Porter, M. D. J. Electroanal. Chem. 1991, 310, Sun, L.; Crooks, R. M. Langmuir1993, 9, 1951. Laibinis, P. E.; Whitesides, G. M. J. Am. Chem. Soc. 1992, 114, 9022. Abbott, N. L.; Rolison, D. R.; Whitesides, G. M. Langmuir1994, 10, 2672. Fang, J.; Chen, M.; Shashidhar, R. Langmuir2001, 17, 1549. 202 (29) Allara, D. L.; Nuzzo, R. G. Langmuir1985, 1, 52. (30) Sheen, C. W.; Shi, J.; Martensson, J.; Parkih, A. N.; Allara, D. L. J. Am. Chem. Soc. 1992, 114, 1514. (31) Zehner, R. W.; Sita, L. R. Langmuir1997, 13,2973. (32) Zehner, R. W.; Parsons, B. F.; Hsung, R. P.; Sita, L. R. Langmuir1999, 15, 1121. (33) lshida, T.; Mizutani, W.; Akiba, U.; Umemura, K.; lnoue, A.; Choi, N.; Fujihira, M.; Tokumoto, H. J. Phys. Chem. B 1999, 103, 1686. (34) Waldeck, D. H.; Beratan, D. N. Science 1993, 261, 576. (35) Bumm, L. A.; Arnold, J. J.; Cygan, M. T.; Dunbar, T. D.; Burgin, T. P.; Jones, L.; Allara, D. L.; Tour, J. M.; Weiss, P. S. Science 1996, 271, 1705. (36) lnoue, A.; Mizutani, W.; lshita, T.; Tokumoto, H. Appl. Phys. 1996, A66, S1241. (37) Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instrumental Analysis, Saunders College Publishing: Philadelphia, 1998. (38) Batchelder, D. N.; Evans, 8. D.; Freeman, T. L.; Haussling, L.; Ringsdorf, H.; Wolf, H. J. Am. Chem. Soc. 1994, 116, 1050. (39) Kim, T.; Chan, K. C.; Crooks, R. M. J. Am. Chem. Soc. 1997, 119, 189. (40) Menzel, H.; Mowery, M. D.; Cai, M.; Evans, C. E. J. Phys. Chem. B 1998, 102, 9550. (41) Mowery, M. D.; Kopta, S.; Ogletree, D. F.; Salmeron, M.; Evans, C. E. Langmuir 1999, 15, 5118. . 203 (42) Kim, T.; Ye, 0.; Sun, L.; Chan, K. C.; Crooks, R. M. Langmuir1996, 12, 6065. (43) Berman, A.; Ahn, D. J.; Lio, A.; Salmeron, M.; Reichert, A.; Charych, D. H. Science 1995, 269, 515. (44) Axon, T. L.; Bloor, D.; Molyneux, S.; Kar, A. K.; Wherrett, B. 8. Proc. SPIE 1993, 2025, 374. 204 '“ll‘l .» I “I MIL HIGAN STATE L‘AIBR AEFlI 31293 02177 069M2