Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electrochemical and spectroscopic studies of 2-(2’-thienyl)pyridine on Au(111) Chung, Emily Susanne 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-931110.pdf [ 10.18MB ]
Metadata
JSON: 831-1.0059339.json
JSON-LD: 831-1.0059339-ld.json
RDF/XML (Pretty): 831-1.0059339-rdf.xml
RDF/JSON: 831-1.0059339-rdf.json
Turtle: 831-1.0059339-turtle.txt
N-Triples: 831-1.0059339-rdf-ntriples.txt
Original Record: 831-1.0059339-source.json
Full Text
831-1.0059339-fulltext.txt
Citation
831-1.0059339.ris

Full Text

E L E C T R O C H E M I C A L A N D S P E C T R O S C O P I C S T U D I E S O F 2- (2 ' -THIENYL)PYRIDINE O N AU(111) by E M I L Y S U S A N N E C H U N G B . S c , The University of Western Ontario, 1999 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F BRITISH C O L U M B I A August 2004 © Emily Susanne Chung, 2004 c Abstract The adsorption behaviour of the bifunctional molecule 2-(2'-thienyl)pyridine (TP) on Au(l 11) was characterized under potential control in neutral and basic aqueous solution. Initial characterization of adsorbed TP was by cyclic voltammetry (CV) and ac voltammetry in aqueous solution. C V was also performed in organic media for comparison purposes. The electrochemical behaviour of TP was compared to that of pyridine, which has been well-characterized in the literature. C V and ac voltammetric characterizations of thiophene in aqueous solution were also performed within the same potential range for comparison to TP. It was found that TP underwent both capacitive changes similar to those observed in the presence of pyridine, and redox processes similar to those observed in the presence of thiophene. Chronocoulometric characterization of the capacitive changes showed that TP initially adsorbs at negative potentials as a low-coverage layer in which interaction with the Au surface is through the aromatic TC orbitals and the plane of the rings is parallel to the surface. Scanning to more positive potentials induces a pyridine-like two-dimensional phase transition to a more condensed phase in which the molecules are bonded to Au via the lone pairs of the N and/or S (likely both). The redox processes of TP were hypothesized to be an electrochemical oc,a'-coupling of the thiophene rings, followed by an oxidative charging process; where subsequent reductive discharging gives the neutral dimer, 5,5'-bis(2-pyridyl)-2,2'-bithienyl (PTTP). PTTP was synthesized chemically for comparison to the product of TP electrochemical oxidation on the electrode. The photophysical properties of TP and PTTP in solution were used to interpret in situ fluorescence results obtained at a Au(l 11) electrode under potential control. Spectra obtained after holding the electrode at positive potentials in the presence of TP were identical to that of chemically synthesized PTTP, confirming the dimerization hypothesis. In addition, fluorescence microscopy images were consistent with the electrochemical results regarding other states of TP. This system demonstrates that multifunctional adsorbates can create chemically-tuned surfaces capable of switching between multiple states, and allow subtle comparisons of the influence of surface-adsorbate interactions between different functional groups. Table of Contents Abstract ii Table of Contents iv List of Tables vi List of Schemes vii List of Figures viii Acknowledgements x List of symbols and abbreviations xii 1 Introduction 1 1.1 Motivation 1 1.2 Scope 4 1.3 References 5 2 Theory and Literature Review 9 2.1 Organic Adsorption at Metal Electrodes 9 2.1.1 The Chemical Nature of Metal-Adsorbate Bonding 9 2.1.2 Effect of Potential on Electrode-Chemisorbate Bonding 12 2.1.3 Literature Review of Adsorption of Relevant Molecules 17 2.1.3.1 Pyridine and Bipyridine Adsorption and Electrochemistry 17 2.1.3.2 Thiophene and B ithiophene Adsorption 18 2.1.3.3 Multifunctional Adsorbates 20 2.2 Electrochemistry of Thiophene 21 2.2.1 Theory and Mechanism of Electropolymerization 21 2.2.2 Electrochemistry of Thiophene in Aqueous Media 25 2.2.3 Effect of Pyridine Substituents on Thiophene Electrochemistry 27 2.3 Fluorescence Near an Electrode 28 2.3.1 Effect of the Metal 28 2.3.2 Fluorescence of Adsorbates 29 2.3.3 Effect of the Electric Potential 30 2.4 References 30 3 Experimental 37 3.1 Introduction 37 3.2 Synthesis 37 3.2.1 Synthesis of 2-(2'-Thienyl)pyridine (TP) 37 3.2.2 Synthesis of 5,5'-Bis(2-pyridyl)-2,2'-bithienyl (PTTP) 38 3.3 Electrochemistry 39 3.3.1 Materials and Instrumentation for Aqueous Electrochemistry 39 3.3.2 Experimental Set-Up for Aqueous Electrochemistry 40 3.3.3 Cyclic Voltammetry and Capacitance in Aqueous Electrolyte 43 3.3.4 Chronocoulometry in Aqueous Electrolyte 44 3.3.4.1 Chronocoulometry Procedures 44 3.3.4.2 Chronocoulometry Error Analysis 47 iv 3.3.5 Organic Electrochemistry 48 3.4 Photophysical measurements 49 3.4.1 UV/Vis Absorbance Measurements 49 3.4.2 Solution and In Situ Fluorescence Spectroscopy and Microscopy .'. 50 3.5 References 52 4 Cyclic Voltammetry and Capacitance of 2-(2'-Thienyl)pyridine, Pyridine and Thiophene 54 4.1 Introduction 54 4.2 Au( 111) in KC10 4 Electrolyte at pH 7 and pH 11 55 4.3 T P o n A u ( l l l ) 58 4.4 Pyridine on Au( 111) 65 4.5 Thiophene on Au(l 11) 68 4.6 Comparison of Electrochemistry of TP, Pyridine and Thiophene 77 4.7 Electrochemistry of TP on Polycrystalline Au in Organic Electrolyte 82 4.8 Summary 85 4.9 References 86 5 Characterization of Capacitive Changes in Adsorbed TP by Chronocoulometry 88 5.1 Introduction 88 5.2 Charge Density 88 5.3 Film Pressure and Surface Pressure 97 5.4 Adsorption Isotherm and Gibbs Energy of Adsorption 105 5.5 Summary 108 5.6 References 109 6 Characterization of Faradaic Processes of Adsorbed TP by In Situ Fluorescence and Comparison to PTTP I l l 6.1 Introduction I l l 6.2 Electrochemistry of PTTP 112 6.3 Solution and Solid State Photophysical Characterization of TP and PTTP... . 113 6.4 In Situ Fluorescence Spectroscopy of TP and PTTP 118 6.5 In Situ Fluorescence Microscopic Imaging of TP and PTTP 127 6.6 Summary 135 6.7 References 136 7 Conclusions 138 7.1 A Model for TP Adsorption on Au(l 11) 138 7.2 Future Work 141 7.3 Concluding Remarks 144 Appendix 146 Examples of Chronocoulometry Current Transients 146 A note regarding the calculation of A G from the surface pressure curves 147 v List of Tables Table 4-1 Dielectric constants of selected pure substances 67 Table 5-1 Selected values of electric dipole moments for molecules in the gas phase.... 95 Table 5-2 Surface pressure equations for various adsorption isotherms 106 Table 6-1 Solvent effects on the wavelengths of PTTP excitation and emission 117 Table 6-2 Solvent effects on the wavelength of the TP emission maximum 118 vi List of Schemes Scheme 2-1 Mechanism for thiophene, pyrrole and furan electropolymerization 23 Scheme 3-1 Reaction scheme for the synthesis of 2-(2'- thienyl)pyridine 38 Scheme 3-2 Reaction scheme for synthesis of 5,5'-bis(2-pyridyl)-2,2'-bithienyl (PTTP) 39 Scheme 4-1 Numbering scheme for TP carbon atoms 81 Scheme 7-1 A model for behaviour of a TP monolayer on Au(l 11) as a function of potential on the first potential scan 138 vii List of Figures Figure 2-1 A diagram of the electrical double layer 13 Figure 2-2 The effect of the increase of bulk activity of the surfactant on the interfacial tension, charge density and capacitance 16 Figure 3-1 Electrochemical set-up for cyclic voltammetry, capacitance and chronocoulometry 41 Figure 3-2 A representation of the electrochemical cell as a series R C circuit 43 Figure 3-3 Potential profile for a chronocoulometric measurement 45 Figure 3-4 Set-up for in situ fluorescence measurements 50 Figure 4-1 A cyclic voltammogram of a Au(l 11) electrode 55 Figure 4-2 Cyclic voltammograms of Au(l 11) in 0.05 M KCIO4 within the double layer potential range at pH 7 and pH 11, at a scan rate of 20 mV/s 57 Figure 4-3 Capacitance and C V of 1 m M TP on Au(l 11), showing the effect of opening the positive potential limit on the reversibility of the C V 58 Figure 4-4 Effect of bulk TP concentration on cyclic voltammograms of a Au(l 11) electrode 60 Figure 4-5 A comparison of 1 m M TP C V and capacitance curves at pH 7 and pH 11... 61 Figure 4-6 Changes in the C V of a Au(l 11) electrode in the presence of 1 m M TP with successive scanning 62 Figure 4-7 CVs of a Au(l 11) electrode in the presence of 1 m M TP at pH 7 after holding at 0.4 V for 5 min and gradually increasing the negative potential scan limit on successive scans 63 Figure 4-8 C V and capacitance curves of a Au(l 11) electrode in the presence of 1 m M pyridine at pH 7 and pH 11 66 Figure 4-9 C V and capacitance plots of Au(l 11) in the presence of 1 m M thiophene at pH 7 69 Figure 4-10 A comparison of the first and steady state CVs for at Au(l 11) electrode in the presence of 10 m M thiophene at pH 7 71 Figure 4-11 C V and capacitance curves of 10 mM thiophene on Au (111) at pH 7 and pH 11 72 Figure 4-12 C V and capacitance plots of a Au(l 11) electrode in the presence of 1 m M TP and 1 m M pyridine 78 Figure 4-13 C V and capacitance plots in the presence of 1 m M TP and 1 m M thiophene atpH7 80 Figure 4-14 CVs of a polycrystalline Au bead in the presence of 0.1 M tetrabutylammonium hexafluorophosphate/MeCN electrolyte only (blank); DMFc and electrolyte; and 15 m M TP, DMFc, and electrolyte 84 Figure 5-1 Relative charge density as a function of potential, E , for a Au(l 11) in the presence of various bulk concentrations of TP 89 Figure 5-2 Capacitance plots of 1 mM TP from differentiation of the charge density curve (in Figure 5-1) and from ac voltammetry measurements 90 Figure 5-3 Relative charge density for Au(l 11) in the presence of 1 m M TP to a positive potential limit of +200 mV 96 viii Figure 5-4 Film pressure vs. potential on Au(l 11) in the presence of various bulk concentrations of TP 98 Figure 5-5 Surface pressure as a function of charge density for various concentrations of TP 100 Figure 5-6 Surface coverage vs. In [TP] from differentiation of film pressure curves... 102 Figure 5-7 Surface coverage of TP from differentiation of surface pressure vs. In [TP] (smoothed) 103 Figure 5-8 Fit of the natural log of surface pressure vs. In [TP] + f(a) to various adsorption isotherms 107 Figure 6-1 CVs and capacitance plots of PTTP adsorbed on Au(l 11) from the air/solution interface at pH 7 and p H l l 112 Figure 6-2 Normalized fluorescence excitation and emission; and UV/Vis absorption spectra of a) 7.1 x 10"5 M TP and b) 2.2 x 10"5 M PTTP in CH 2 C1 2 114 Figure 6-3 A comparison of solution and solid-state absorption spectra of TP and PTTP. 115 Figure 6-4 Solvent effects on the excitation (solid line) and emission (dotted line) of a) 1.5 x 10"5 M TP at excitation and emission maxima of 303 nm and 360 nm respectively and b) 7.6 x 10"6 PTTP at excitation and emission maxima of 390 nm and 460 nm respectively...: 117 Figure 6-5 Difference in fluorescence on desorption, AI, for a layer of PTTP adsorbed on Au(l 11) at after holding at various potentials, Ei, for 15 minutes 119 Figure 6-6 Background subtracted fluorescence upon desorption, AI, for a layer of TP adsorbed on Au(l 11) at after holding at various potentials, Ej for either 15 or 30 minutes 121 Figure 6-7 Change in background-subtracted fluorescence intensity upon desorption as a function of time spent at Ej = +0.200 V vs. S C E 122 Figure 6-8 A comparison of normalized spectra of (a) a solution spectrum of PTTP in CHCI3, (b) a background-subtracted in situ spectrum of PTTP adsorbed on Au(l 11) at Ej = -0.600 V and (c) a background-subtracted in situ spectrum of TP at Ei = -0.600 V and E = +0.200 V 124 Figure 6-9 Images of the Au(l 11) electrode in the presence of 1 m M TP: (a) brightfield and with U V filters (b) at -0.200 V and (c) -1.000 V vs. S C E 128 Figure 6-10 The first in situ image after desorption (with background subtracted i.e. AI=I(Edes)-I(Ej)) of the electrode with U V filters in place after holding for 10 minutes at various potentials in the presence of 1 m M TP solution 129 Figure 6-11 A plot of average image intensity time (with background subtraction i.e. AI=I(EdesH(Ei)) for different values of E;. Ed e s was -1.00 V 132 Figure 6-12 Changes in background-subtracted fluorescence images (AI=I(Edes)-I(Ei)) of a Au(l 11) electrode in the presence of 1 mM TP after varying amounts of time following desorption 134 ix Acknowledgements First and foremost, I would like to thank my supervisors, Prof. Michael Wolf and Prof. Dan Bizzotto for allowing me to pursue this project even though I had no clue what I was getting into when I started it; for always making time to give me helpful suggestions, encouragement and hands-on help in the lab; and for all the other ways in which they "made it happen". I could not have asked for better mentors, nor for a more rewarding graduate school experience. Thanks also to the other members of my supervisory committee, Prof. Keith Mitchell, (who kindly offered to give the thesis a pre-reading) and Prof. Alan Storr; and to my examiners, Prof. Elod Gyenge, Prof. Mark Thachuk and Prof. Nongjian Tao. I would also like to thank the students and post-docs of the Bizzotto and Wolf research groups with whom I have had the pleasure of sharing lab and office space for the past five years; and with whom I have enjoyed many pitchers of beer and many games of hacky-sack. I would especially like to acknowledge Jeff Shepherd of the Bizzotto research group, who went far beyond the call of duty helping me with the fluorescence ' microscope, including modifying many of the programs to my specifications on short notice, and who also taught me many of the electrochemical procedures; Yanguo Yang, formerly of the Bizzotto group, who also taught me many of the electrochemical procedures; Dr. Cerrie Rogers, formerly of the Wolf group, who offered synthetic advice and synthesized the catalyst that I used; and Tracey Stott of the Wolf group who recrystallized the organic electrolyte and first suggested the dimerization hypothesis; but also the rest of the Bizzotto group (John Odiko Agak, Anisa Akhtar, Dr. Ed Guerra, Eewei Goh, Mike Lathulliere, Irina Manisali, Amanda Musgrove, Aya Sode, and Robin x Stoodley) and the Wolf Group (Pierre-Louis Brunner, Dr. Olivier Clot, Tamara Kunz, Glen Kurokawa, Keri Kwong, Carolyn Moorlag, Kristen Matkovich, Dr. Andras Pattanyus-Abraham, Agostino Piettrangelo, Dr. Katja Rademacher, Dr. Ruth Mueller, Dr. Madiba Saidy, Bryan Sih, and Lisa Thorne.) Thanks also to Dr. Carl Scott and Stephen Reid, formerly in the group of Prof. Steve Withers, who helped me with the G C and the fluorimeter respectively. I received a great deal of assistance from chemistry and A M P E L support staff, including Ben Clifford, with whom I have had helpful discussions and who gave me the sample of quinine sulphate; Brian Ditchburn of the glass shop; Brian Snapkausis, Ken Love, and the rest of the mechanical shop; Judy Wrinskelle, Sheri Harbour, and Lani Collins and everyone else in the chemistry front office; Jason Gozjolko and the rest of the electronics shop; Scotty in the C-Horse lab; and all the helpful people at chemistry stores. I would also like to thank all the other people who have helped me with other aspects of my research that will unfortunately be reserved for future documents other than this one including Robin Coope, Dr. Alina Kulpa, Erwin Lin, Prof. Sylvie Morin, Jim Mackenzie, Dr. Jody Mandeville, Dr. Katja Rademacher, A l Schmalz, Dr. Ken Wong, and Dr. Philip Wong. Thanks to everyone who paid me, including NSERC, the U B C U G F fund, and the Dept. of Chemistry Gladys Estella Laird endowment. And, of course thanks to all my friends and family for making very sure I didn't spend all my time in the lab. Goldschlager has been the official drink of Emily Chung's thesis. List of symbols and abbreviations Symbol/abbreviation Meaning, units r surface excess, mol/cm r m a x maximum surface excess T elastic surface stress p equilibrium constant e dielectric constant Ee elastic surface strain eo permittivity of free space <p surface pressure (J) potential drop y surface tension X wavelength ju chemical potential 7C film pressure 0 relative coverage o charge density, pC/cm 2 a M charge density on the metal co angular frequency, Hza £ Parson's function ° degree, unit of angle a activity A area of electrode A angstrom, unit of distance a.u. arbitrary units ac alternating current (aq) aqueous A F M atomic force microscopy b Volmer coefficient B virial coefficient Bu butyl c concentration C capacitance, pF C differential capacitance, pF/cm 2 C coulomb, unit of charge C C D charge coupled device Cai double-layer capacitance, pF/cm 2 C E counter electrode C V cyclic voltammetry; cyclic voltammogram d distance between capacitor plates d distance from electrode d.f. degrees of freedom xii Symbol/abbreviation Meaning, units D A Q data acquisition board D C M dichloromethane DMFc decamethylferrocene or Bis(pentamethylcyclopentadienyl)iron D M F N,N-dimethylformamide dppp Ph2P(CH2)3PPh2 E potential, V Em half-wave potential, V Ebase base or resting potential E d e s desorption potential E F Fermi level Ej potential of interest E N shift in pzc E Q C M electrochemical quartz crystal microbalance ESR electron spin resonance E t 2 0 diethylether EtOAc ethyl acetate eV electron volt, unit of energy F Faraday's constant, 9.64846 x 104 C F Farad, unit of capacitance f frequency FFT fast Fourier transform A G 0 standard free energy of adsorption A G Gibbs free energy AGads Gibbs free energy of adsorption G C gas chromatography - A H 0 standard enthalpy of adsorption h hour, unit of time H P L C high performance liquid chromatography HSAB hard-acid soft-base Hz hertz, unit of frequency I current I intensity iimag quadrature component of current i r e a i in-phase component of current AI difference in fluorescence between desorbed and adsorbed spectrum or image IR infrared IHP inner Helmholtz plane rrO indium tin oxide J number of intervals j current density, pA/cm 2 k Boltzmann constant, 1.38065 x 10 2 3 J/K kJ kilojoule, unit of energy (/) liquid M molar, unit of concentration Symbol/abbreviation Me MeCN MeOH MHz min n n N A NIR N M R OCP OHP O L E D Ph PTTP pzc q $ R RC R E R H E rms S (s) S A M S C E SEIRAS SERS SSCE S T M T t T H F TP U V V Vis wd W E X x XPS Meaning, units methyl acetonitrile methanol megahertz, unit of frequency minute, unit of time number of carbons in an alkyl chain number of electrons numerical aperture near infrared nuclear magnetic resonance open cell potential outer Helmholtz plane organic light-emitting diode phenyl 5,5'-bis(2-pyridyl)-2,2'-bithienyl potential of zero charge charge charge on a metal charge in solution gas constant, S.SMJK'moi" 1 resistor capacitor reference electrode reversible hydrogen electrode root mean squared spin solid self-assembled monolayer saturated calomel electrode surface enhanced infrared absorption spectroscopy surface enhanced Raman scattering sodium saturated calomel electrode scanning tunnelling microscopy temperature time tetrahydrofuran 2-(2'-thienyl)pyridine ultraviolet volt, unit of electric potential visible working distance working electrode mole fraction distance x-ray photoelectron spectroscopy 1 Introduction 1.1 Motivation interface On&r-fisjl- *1 n. 1. A surface forming a common boundary between adjacent regions, bodies, substances, or phases. 2. A point at which independent systems or diverse groups interact. Scientific knowledge grows like a living thing; not linearly, but dendritically as a tree does. From the trunk, each branch divides into boughs, some of which continue on a long journey away, towards the sky or open air. Those that took an earlier fork continue to grow and branch off into smaller twigs, filling the space between other branches. Eventually, there comes a place where leaves at the ends of two boughs that took off from the trunk in separate directions long ago may find themselves side by side, at an interface. In fact, in the growth of a tree, it isn't long before the number of such boundaries and interfaces in the canopy outnumber by far those between the tree and the open air. Such boundaries and interfaces within science are numerous and much can be learned by crossing back and forth between them. This thesis is primarily the study of one interface, one that can be described simply as that between a Au(l 11) electrode and a solution of 2-(2'-thienyl)pyridine. But there are many other interfaces in supporting roles - that between surface chemistry and organometallic chemistry; that between electrochemistry and polymer chemistry; and other combinations thereof, to name a few. This thesis tells a story that takes place somewhere in the canopy. 1 The modification of surfaces through the adsorption of organic molecules has been extensively studied as a way of chemically tuning surfaces to engineer specific physical and chemical properties for applications such as corrosion inhibition, nanolithography, microsensors, biosensors, catalysis and electronic and optoelectronic devices.!2"6] In particular, many organic adsorbates are chosen for their ability to self-organize or self-assemble into organized two-dimensional structures that are conducive to cooperative and homogeneous behaviour. t 5i On electrode surfaces, the growth and organization of such structures can be further controlled via the applied potential, f3-7 - 9] In some cases, chemical reactions can also be induced by the application of an electric potential and electrosynthesis can be conducted at the interface, t7] In addition to providing a means of control, electrochemistry also provides a means for analysis, and has been used for the thorough characterization of model organic adsorbates such as pyridine.HO] The strong and stable gold-sulphur interaction enjoyed by thiols has made thiol-on-gold systems a popular choice for the chemical modification of surfaces.!11-13] Thiophenes, in which the interaction with gold is weaker, have recently been shown to produce stable monolayers also.!6'1 4"1 8] Thiophene polymerizes to polythiophene, a 7t-conjugated organic polymer that is unusually stable in both neutral and oxidized (conductive) forms, !fi] Polythiophenes and other conductive polymers also have interesting optical properties and have been widely studied for possible applications in electronic devices such as transistors and batteries, as well as optoelectronic devices such as organic light emitting diodes (OLEDs).! 6 - 1 9 - 2 0 ] Recently, polymer-based OLEDs have been incorporated into some commercial cellular phone and digital camera displays.!21] 2 Polythiophene is difficult to characterize due to its insolubility, thus thiophene oligomers have been used as model systems in order to gain a better understanding of polymer properties such as charge storage. I6'22! In addition, derivatives of thiophene have been studied, as novel properties can be introduced by functionalizing the monomer precursor. 12 31 Pyridine has been incorporated both into the backbone and as a side-chain of polythiophene in order to induce third order non-linear optical properties due to intramolecular charge transfer, 12 41 in order to stabilize the reduced form of the polymer, 12 31 or to allow binding of transition metal ions.! 2 5! Pyridine has been a popular model compound for investigating the adsorption of neutral organic molecules to metal surfaces, as it can take on multiple surface orientations, interacting with the metal via either the aromatic % orbitals or through the lone pair of the nitrogen heteroatomi 2 6' 2 7! In contrast to the Au-thiol systems, pyridine on Au(l 11) is a system where the ligand-metal interaction is described as "moderate", or weak chemisorption.!3] Thus far, there has been little research focused on the heterodentate binding of adsorbates capable of interacting with a metal surface through more than one functional group simultaneously. Such a system would expand the functionality of the modified surface, and provide new insights into the interplay between intramolecular interactions of organic adsorbates and electronic effects in adsorbate-metal binding, and, by analogy, ligand-metal binding. 2-(2'-thienyl)pyridine (TP) is known to form complexes with many transition metals including gold, in several different binding modes.l28-32] _ contains pyridine and thiophene groups whose geometry relative to one another makes simultaneous interaction 3 of both moieties with the surface possible. Its rigid aromatic structure should allow for multiple binding modes and long-range packing. Therefore, TP is an interesting model for examining how linking two surface-active functional groups affects their very different individual interactions with the electrodes. Specifically, will one group dominate? If so, which one? Will the molecule take on the individual characteristics of both groups or a blend, or behaviour that is completely unlike either group? Finally, this study bridges those of weakly chemisorbed species such as pyridine and those of more strongly adsorbed surfactants, and is a starting point for the investigation of heterodentate adsorbates on surfaces. O o CH3 X S S N = / pyridine thiophene 2-(2'-thienyl)pyridine (TP) 1.2 Scope This thesis brings together a broad range of concepts and topics from many areas of chemistry. With this in mind, it begins in Chapter 2 with a review of theory regarding adsorption and fluorescence on metal surfaces, the effect of a potential applied across the metal-solution interface on these processes, and the electrochemistry of thiophene, topics that may be helpful in the interpretation and understanding of the experimental results. In addition, the literature on adsorption of similar molecules such as pyridine and thiophene will be thoroughly reviewed in this section. The experimental methods, including materials used, synthesis of relevant compounds, characterization by a variety of electrochemical and spectroscopic methods, and treatment of errors, have been collected 4 in Chapter 3 for easy reference. These will reviewed briefly in subsequent chapters where relevant. The experimental results begin in Chapter 4, which contains a mostly qualitative electrochemical characterization of TP adsorption on Au(l 11) by cyclic voltammetry and capacitance measurements, including TP adsorption behaviour as a function of potential, concentration and pH. Here, stable adsorbed phases and the capacitive and faradaic changes during transitions from one stable phase to another are characterized. Based on these results, a hypothesis regarding the different phases and transitions will be presented. Chapter 5 will contain a more detailed, quantitative examination of the purely capacitive changes by chronocoulometry, providing details about the thermodynamics of different adsorbed states, including surface coverages and adsorption energies. In Chapter 6, a hypothesis regarding the nature of an oxidative process in the adsorbed TP layer will be tested using fluorescence spectroscopy to monitor in situ the electrochemical generation of a new species. Fluorescence microscopy will be used to further characterize the morphology of desorbed TP after adsorption in different potentially-controlled states on the electrode. Finally, Chapter 7 will sum up the findings of previous chapters, present a complete model of TP adsorption behaviour on Au(l 11) as a function of potential, discuss its implications, and briefly outline possible future directions for this research. 1.3 References (1) The American Heritage® Dictionary of the English Language; 4th ed.; Houghton Mifflin: Boston, 2000. (2) Xu, J.; L i , H. -L. J. Colloid Interface Sci. 1995,176, 138-149. 5 (3) Bizzotto, D.; Zamlynny, V.; Burgess, I.; Jeffrey, C ; L i , H.-Q.; Rubinstein, J.; Galus, Z.; Nelson, A.; Pettinger, B.; Merril, R.; Lipkowski, J. In Interfacial Electrochemistry; Wieckowski, A. , Ed.; Marcel Dekker: New York, 1999, pp 405-425. (4) Su, J.; Mrksich, M . Langmuir 2003,19, 4867-4870. (5) Mandler, D.; Turyan, I. Electroanalysis 1996, 8, 207-213. (6) Matsuura, T.; Nakajima, M . ; Shimoyama, Y. Jpn. J. Appl. Phys. 2001, 40, 6945-6950. (7) Buess-Herman, C ; Bare, S.; Poelman, M . ; van Krieken, M . In Interfacial Electrochemistry; Wieckowski, A. , Ed.; Marcel Dekker: New York, 1999, pp 427-446. (8) Buess-Herman, C. In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P. N. , Ed.; V C H : New York, 1992, pp 77-118. (9) Lipkowski, J.; Stolberg, L . In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P., Ed.; V C H , 1992, pp 171-238. (10) Lipkowski, J.; Stolberg, L . ; Yang, D. F.; Pettinger, B.; Mirwald, S.; Henglein, F.; Kolb, D. M . Electrochim. Acta 1994, 39, 1045-1056. (11) Lavrich, D. J.; Wetterer, S. M . ; Bernasek, S. L . ; Scoles, G. I. Phys. Chem. B 1998,102, 3456-3465. (12) Jin, Q.; Rodriguez, J. A.; L i , C. Z.; Darici, Y.; Tao, N. J. Surf. Sci. 1999, 425, 101-111. (13) Zharnikov, M . ; Grunze, M . I. Phys.: Condens. Matter 2001,13, 11333-11365. 6 (14) Dishner, M . H. ; Taborek, P.; Hemminger, J. C ; Feher, F. J. Langmuir 1998,14, 6676-6680. (15) Dishner, M . H.; Hemminger, J. C ; Feher, F. J. Langmuir 1996,12, 6176-6178. (16) Noh, J.; Ito, E . ; Araki, T.; Hara, M . Surf. Sci. 2003, 532-535, 1116-1120. (17) Liu, G.; Rodriguez J. A.; Dvorak J.; Hrbek J.; T., J. Surf. Sci. 2002, 505, 295-307. (18) Ishida, T.; Choi N.; Mizutani W.; Tokumoto H. ; Kojima I.; Azehara H. ; Hokari H.; Akiba U.; M . , F. Langmuir 1999,15, 6799-6806. (19) Casado, J.; Pappenfus, T. M . ; Miller, L. L. ; Mann, K. R.; Orti, E . ; Viruela, P. M . ; Pou-Amerigo, R.; Hernandez, V.; Lopez Navarrete, J. T. J. Am. Chem. Soc. 2003, 125, 2524-2534. (20) Noh, J.; Kobayashi, K.; Lee, H.; Hara, M . Chem. Lett. 2000, 630-631. (21) Souza, C. Electronic Buyer's News 2003, 3. (22) Guay, J.; Kasai, P.; Diaz, A.; Wu, R.; Tour, J. M . ; Dao, L . H . Chem. Mater. 1992, 4, 1097-1105. (23) Wang, J. X.; Keene, F. R. J. Electroanal. Chem. 1996, 405, 59-70. (24) Lu, H.-F.; Chan, H. S. O.; Ng, S.-C. Macromolecules 2003, 36, 1543-1552. (25) Jenkins, I. H. ; Salzner, U.; Pickup, P. G. Chem. Mater. 1996, 8, 2444-2450. (26) Stolberg, L . ; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1987, 238, 333-353. (27) Yang, D. F.; Stolberg, L . ; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1992, 329, 259-278. (28) Balashev, K. P.; Puzyk, M . V.; Kotlyar, V . S.; Kulikova, M . V . Coord. Chem. Rev. 1997,159, 109-120. (29) Constable, E . C ; Sousa, L . R. / . Organomet. Chem. 1992, 427, 125-139. 7 (30) Fuchita, Y.; Ieda, H.; Wada, S.; Kameda, S.; Mikuriya, M . J. Chem. Soc, Dalton Trans. 1999, 4431-4435. (31) Neve, F.; Ghedini M . ; A. , C . J. Organomet. Chem. 1994, 466, 259. (32) Sigel, H. ; Wynberg, H.; Van Bergen, T. J.; Kahmann, K. Helv. Chim. Acta 1972, 55, 610-613. 8 2 Theory and Literature Review 2.1 Organic Adsorption at Metal Electrodes 2.1.1 The Chemical Nature of Metal-Adsorbate Bonding The strength of adsorbate-surface interactions varies from physical adsorption, involving only van der Waals and/or dipole-dipole interactions (standard enthalpy of adsorption, - A H 0 , typically < 35 kJ/mol)!1! to strong chemisorption (-AH 0 typically > 35 kJ/mol), in which the adsorbate-surface interaction can resemble the covalent bonds within molecules. Adsorption at a solid can occur from the gas phase or from the liquid or solution phase. In the latter case, which is relevant to subsequent chapters, adsorption is driven by competition between adsorbate-solvent interactions and surface-adsorbate interactions. Adsorbates that interact not only with the surface and the solution but with one another may form organized structures on the surface called self-assembled monolayers (SAMs). Self-assembly involves a fast adsorption step, driven by surface-adsorbate interactions, and a slower organizational assembly process, driven by adsorbate-adsorbate interactions such as hydrophobic and %-% interactions.I2-3! Models of adsorption to solids can be described from a chemical perspective, an electronic perspective, or a combination thereof. In the first case, the surface is considered to be a collection of surface sites, atoms, and clusters that can react chemically with adsorbing species, while in the second it is described by surface states, according to the band model. In the band model, the bulk metal consists of metal atoms and electrons, where a portion of the valence electrons are not confined to a particular metal atom. The electronic levels combine to form a continuous, partially filled band. The upper limit of electron occupation for this band at absolute zero is defined by the Fermi 9 level, E F . Within this model, surface states are localized electronic states found at the interface between the metal and another phase. t 4l Muetterties first noted in 1975 that many gas-surface interactions have organometallic analogues, and suggested that the interaction between an adsorbate and a metal surface can be thought of as similar to a ligand-metal interaction. 151 This way of viewing adsorption from a chemical perspective, the surface-cluster analogy, is mostly used to describe interfacial interactions between the gas phase and the solid phase. However, in cases where adsorbate-solvent interactions are small, it may also be relevant to liquid-solid interfacial interactions. 161 In general, the adsorbate-metal bond can be described as a covalent-like bond between the dangling d-orbitals of the metal atoms at the interface and the orbitals of the adsorbate. As such, ligand field effects and crystal field effects may come into play, favouring geometries and bonding modes where the interaction between the adsorbate orbitals and the metal d-orbitals are maximized. There are two analogies to ligand-metal bonding in coordination complexes in which o-bonding and electronic effects can be described. The first is an analogy to Dewar-Chatt-Duncanson bonding, involving o-donation of electrons from the ligand to the metal and rc-backbonding from the metal to the ligand, although the opposite case has also been known.!7! Higher electron density around the metal will favour the bonding of a better rc-electron-accepting (back-bonding) atom or functional group on the ligand. This type of bonding accounts for second-order effects such as inductive and trans effects from nearby ligands. 10 Electronic effects can also be accounted for by another coordination chemistry theory, hard-soft acid-base (HSAB) theory. Here, the metal acts as a Lewis acid acceptor for the donor non-bonding electrons of atoms such as N. Polarizable or soft acids will interact more favourably with soft bases, while hard acids will interact more favourably with hard bases. Au is considered a soft acid, while pyridine is a "borderline" base.!6! Thiophene, bearing a larger and more polarizable sulphur heteroatom, is a softer base than pyridine, and thus would be expected to interact more favourably with Au. When there is competition between different types of interactions, n interactions with the metal tend to dominate for harder bases, while interactions via nonbonding molecular orbitals dominate for softer bases. While the surface-cluster analogy can describe many aspects of metal-adsorbate interactions, some ligand-surface interactions are unlike those seen in complexes. There are a number of differences between metal clusters and surfaces, as highlighted by the electronic description of metal surface-adsorbate interactions. Clusters and complexes have discrete orbitals, whereas in surfaces, the orbitals form bands. The surface-adsorbate bond strength therefore depends on the Fermi level of the metal, E F . Additional factors also play a role in surface-adsorbate bonding, such as dipole-dipole interactions. Charged adsorbates may be stabilized by the image charge induced on the opposite side of the interface. Adsorbate-adsorbate interactions may also be quite significant, and the nature of adsorbate-surface bonding may be affected by the competition between adsorbate-adsorbate and adsorbate-surface interactions. 11 2.1.2 Effect of Potential on Electrode-Chemisorbate Bonding The above chemical description of adsorbate-metal bonding is useful in many ways, but is insufficient to completely describe the special case of adsorption on a metal surface that will be dealt with in subsequent chapters. Herein, the surface is an interface between a metal and a solution, and is capable of accumulating charge. In this case, the potential across the interface, and therefore the interfacial conditions, can be varied and controlled. The simplest case of such an interface is the IPE or ideal polarized electrode, in which no charge transfer takes place when an electrical potential is applied. In this case, the electrode-solution interface acts like a capacitor. Charge, q M , accumulates at one "plate", the surface of the electrode, and an opposite, balancing charge, q s = - q M , accumulates on the solution side of the interface, which conceptually makes up the other plate. Even in cases where charge transfer does occur at the electrode-solution interface, capacitive charging will precede electron transfer. Except at the potential of zero charge (pzc), charge on the metal, q M , accumulates as a thin layer on the electrode surface, while q s is distributed through a region of the solution called the electrical double layer, made up of charged species and oriented dipoles in solution at the interface (Figure 2-1). The two layers of the double layer are known as the inner layer (also called the compact, Helmholtz, or Stern layer) and the diffuse layer. The inner layer is made up of solvent and specifically adsorbed species, unsolvated species in direct contact with the electrode. The plane in which the electrical centres of specifically adsorbed ions are found is called the inner Helmholtz plane (IHP). The plane further away from the electrode in which the electrical centres of the nearest solvated ions are found is called the outer Helmholtz plane (OHP). The diffuse layer is defined as the 12 Figure 2-1 A diagram of the electrical double layer, where IHP is the inner Helmholtz plane and OHP is the outer Helmholtz plane, located at distances and x 2 from the metal respectively; «))M, <[>i and <j>2 are potential drops across the metal, inner layer, and diffuse layer; and a 1 and <rd are excess charge density in the inner and diffuse layers, respectively.!**! region from the OHP to the bulk, and its thickness is inversely related to the concentration of ions in solution. The formation of the double layer produces an associated double layer capacitance, Cai. A distinguishing feature of Cai is that unlike the capacitance of an ideal capacitor, it is usually a function of potential. I8! Processes occurring in the double layer, including changes in the diffuse part of the double layer and the interaction of the inner layer with the adsorbate, can have significant effects on adsorption. For a thermodynamic understanding of adsorbate processes within the double layer, it is useful to revert to an even simpler model of the interface, an ideal two-component system. Here, the interface region can be thought of as a single dividing plane, with all of one component residing on one side of the plane and all of the other component residing 13 on the other side of the plane. In real systems, the interfacial region is larger, however, a dividing plane, often called the Gibbs dividing plane, can still be designated as a convenient reference. The degree of adsorption at the Gibbs dividing plane is known as the surface excess, T, and is defined as T = — (2.1) A where n is the amount of adsorbate at the surface in excess of the amount in the bulk, and A is the surface or interfacial area. Using thermodynamic relationships, T can be related to the surface tension at the interface, y, and the chemical potential, ju, by Gibbs adsorption equation dr = -^Xidfi, (2.2) This holds also for an electrochemical cell, where the components, i, include the solvent, the electrodes and the electrons within them, the ions in solution, and possibly other adsorbates. From the Gibbs adsorption equation, it is possible, via thermodynamic relationships, to derive an equation describing the effect of adsorbates on the surface tension at the electrode-surface interface, the electrocapillary equation. For a mercury surface in contact with an aqueous solution, the electrocapillary equation is:!9! -dy = o~MdE + 5 ] r l w ^ (2.3) i where G M is the charge density on the metal, E is the potential of the electrode with respect to the reference, r,.(W 0 ) i s the relative surface excess of a given component with respect to H2O. For example, for component M : 14 r - r X m r (2-4) LM(H20) ~ L M Y L H20 XH20 where X w is the mole fraction of M and XHi0 is the mole fraction of H 2 0 . The electrocapillary equation can be modified to apply to a solid electrode by including terms accounting for elastic stress and strain of the solid surface. The elastic surface stress, T, is the reversible work required to form a unit area of surface by stretching, and the elastic surface strain, Be, is the change in y as a result of changes in metal atom spacing in response to variations in electron density. However, the values of the stress and strain terms are very small compared to measurable changes in the charge density, and therefore can be neglected when calculating changes in y from integration of GM as a function of E.16,10] The electrocapillary equation is extremely useful for understanding changes in capacitance at the interface due to changes in ordering and orientation of molecules at the electrode. Potential-induced transitions to a more condensed phase are common for rigid molecules with some dipolar character.!11] Stable condensed phases are characterized by a region of low capacitance independent of E , T and concentration within a limited potential range. As the bulk concentration of adsorbate increases, the lowering of y for the condensed phase tends to be more pronounced relative to the less condensed phase, and the potential region of stability for the condensed phase increases (Figure 2-2). Adsorption to an electrode from the solution phase can be described as a function of potential or charge and bulk solution activity, ab, using adsorption isotherms, much as adsorption from the gas phase can be described as a function of pressure. 15 1 ~~a~ a |g ~cT -E Figure 2-2 The effect of the increase (a-d) of bulk activity of the surfactant on the interfacial tension, charge density and capacitance, where a and p are less condensed and more condensed adsorbed phases respectively. The maximum in y corresponds to the pzc.t* ^ The general form for an adsorption isotherm isf9] aA = ab exp r AG 0^ RT (2.5) where AG is the standard free energy of adsorption, which is a function of electrode potential, and aA is the activity of the adsorbed species. This can be simplified by the use of an equilibrium constant, p, where fl = exp f AG 0^ RT (2.6) For an ideal or ideally dilute solution, a f cis equivalent to the mole fraction or concentration, c. Similarly, can be defined in terms of surface excess, T. The simplest case of adsorption is described by the Henry's isotherm, to which most more complex isotherms reduce at low coverage. f 6l It is valid for low values of T and assumes no interaction between adsorbates. Therefore T is directly proportional to c and 16 (2.7) where r m a x is the maximum or saturation coverage, the limit imposed by the number of surface sites available for adsorption. The Henry's isotherm can be slightly modified to acknowledge the effect of the declining availability of vacant sites for adsorption as the coverage increases, giving the Langmuir isotherm! 1,6,12] (2.8) Isotherms based on more complex models of adsorption, such as the Frumkin, Freundlich, Volmer and virial isotherms, include additional parameters to account for attractive or repulsive forces between adsorbates and/or the inequivalence of different adsorption sites.! 6' 1 2' 1 3! 2.1.3 Literature Review of Adsorption of Relevant Molecules 2.1.3.1 Pyridine and Bipyridine Adsorption and Electrochemistry Pyridine is a rigid molecule with a benzene-like ring structure, a large dipole moment along the C2 axis, and the ability to interact with a metal through either the aromatic rc orbitals or the lone pairs of the N heteroatom.!14! These characteristics have allowed it to be studied by a variety of methods including surface enhanced Raman scattering (SERS) and IR spectroscopy, and have made it an attractive model compound for the study of how orientation and adsorption energetics are influenced by geometry and density of coordination sites, and hence crystallographic orientation of the gold surface. ! 1 5 ! Pyridine adsorbs to Au(l 11) in a flat-lying, rc-bonded orientation at negative potentials, and undergoes a 2-dimensional phase transition near the pzc of Au to a more vertical, N -17 bonded orientation at positive potentials, as shown by chronocoulometry^16] and S T M J 1 7 ' 1 8 ] S T M studies show that at positive potentials, hexagonal ordered domains indicative of strong pyridine-pyridine interactions coexist with disordered domains. The interaction of pyridine with gold is considered to be moderate (a weak chemisorption) with a A G a d s on Au(l 11) of -37 kJ/molJ 1 6 - 1 9 ] 2,2'-bipyridine on Au(l 11) shows similar behaviour to pyridine, but the 7t-bonded region is narrower compared to pyridine and the phase transition is less steep. PO] By S T M , it has been found that 1 m M 2,2'-bipyridine makes a transition from an aromatic n-bonded to a vertical orientation (cis-bonded through both nitrogens) between -0.5 V and -0.3 V vs. S C E J 2 1 ] At yet more positive potentials, there are some reports that it undergoes a second phase transition through a disordered phase to a more compact, densely-packed configuration, f21-22] At positive charge densities, bipyridine is stacked in "polymer-like" chains,f22] in which 71-71 interactions are hypothesized to play a roleJ 2 3] 2.1.3.2 Thiophene and Bithiophene Adsorption Until recently, whether thiophene adsorbed to gold was somewhat controversial, as theoretical studies showed that the conjugation of one of the lone pairs on the S into the aromatic TC system would result in no chemical interaction of the S atom with A u J 2 4 ] and even some recent theoretical25] and experimental!26] studies have found no evidence for bonding of the thiophene S atom to Au. However, as early as 1987, studies of thiophene electropolymerization on gold reported a depression in the capacitance at potentials below the onset of electropolymerization, 18 suggesting significant adsorption.!27] Thiophene adsorption from ethanolic solutions in the absence of potential control has been reported by a number of groups and characterized by S T M , ! 2 8 " 3 0 ] IR,! 3 1] and X P S . ! 3 0 - 3 2 - 3 3 ] S T M images show rows of features proposed to correspond to re-stacked rows of thiophene molecules. Like alkanethiols, thiophene interacts strongly enough with Au(l 11) to relax the 23 x V3 reconstruction and stabilize large vacancy islands. However, there is ample evidence that the thiophene interaction with gold is weaker than that of alkanethiols. Thiophene can be displaced by molecules that interact more strongly with Au, such as methanethiolJ29] Its temperature stability on A u is also lower. Under U H V conditions, thiophene desorbs from the surface at temperatures above 250 K. l 3 2 ] At temperatures of 35-40°C under ambient conditions, the thiophene layer can be annealed to remove the vacancy islands. 12 91 XPS studies have found that thiophene adsorbed to gold shows no unbound S peaks at 164 eV, but does show S(2p) binding peaks at 161-163 eV,! 3 0> 3 2] indicating a chemical interaction between Au and S. In addition, they show that thiophenes, alkanethiols, sulphides, and dialkyl disulphides all have S(2p) peaks at the similar binding energies.!33] A calculation of AGatis for thiophene estimated it to be about 9 kT (22.3 kJ/mol), of the same order of magnitude as for alkanethiol, terthiophene and benzenedithiol S A M s . ! 3 1 ! IR studies show that thiophene adsorbs initially with the plane of the ring parallel to the gold surface. However, intermolecular interactions between the adsorbed thiophenes lead to a change to an orientation normal to the surface. This change has very slow kinetics, taking 15 hours to reach equilibrium, in contrast to alkanethiols, which are reported to reach equilibrium in a few seconds,!311 although there is evidence that the ordering step 19 following adsorption for thiols may be slower than previously believed. I 3 4' 3 5! Slow kinetics of adsorption are also reported for 3-octylthiophene, which shows a more gradual change in coverage with time initially compared to the corresponding thiol of similar alkyl chain length, suggesting a lower affinity for the gold surface. It was suggested that this might be because of the strong interactions between the TC orbitals of thiophene and the Au surface.!36! Besides the reports of thiophene adsorption to gold from ethanolic solutions, there is also a recent report of thiophene adsorption onto Au from 1 m M thiophene in 0.1 M H C I O 4 J 3 7 ] Cyclic voltammograms showed no peaks in the double layer region (0 to 1.2 V vs. RHE), however, a depression in the capacitance was presented as evidence of adsorption. S T M images showed at +0.1 V vs. R H E (-0.1 V vs. SCE), the molecules were completely desorbed. At +0.3 vs. R H E (+0.1 V vs. SCE), the molecules were flat-lying, spaced by 0.88 nm, and at +0.6 V vs. R H E (+0.4 V vs. SCE), they were in a vertical orientation, spaced by 0.5 nm. The kinetics of the change between the two states was described as very slow, consistent with studies in ethanolic solution. I3 !1 2.1.3.3 Multifunctional Adsorbates Most adsorption studies of multidentate or chelating ligands onto metal surfaces have tended to look at adsorbates with identical donor groups, such as terthiophene,!38! bipyridine,[20,21,23,39] a n d dithiols.^ 4!] There are a few studies of adsorbates theoretically capable of adsorption through more than one different functional group, most of which have at least one thiol group that dominates the interaction with Au. It was found that for a thiophene thiol monolayer where the thiophene and thiol were on opposite ends of an alkyl chain, interaction with 20 Au occurred only through the thiol,! 4 2! while no comment was made on thiophene-Au interaction in a study where the thiol group was bonded to the oc-carbon of the thiophene ring.[43] There are also a small number of adsorption studies of molecules with both sulphur and nitrogen donor atoms available, such as 2-mercaptobenzthiazolet44! and 2-mercaptobenzoimidazolesJ45! Typically, the interaction through the thiol was confirmed, while a secondary interaction through the N was proposed but not supported by any evidence. 2-mercaptopyrimidine was studied by S T M , and it was found that it adsorbed to gold as flat-lying, H-bonded dimers i.e. interacting with the metal through the aromatic % orbitals rather than S or N J 4 6 1 The only literature examples of molecules for which there is evidence for interaction with Au via more than one functional group simultaneously are studies of 2-mercaptopyridine by S E I R A S j 4 7 ! S T M , and electrochemistry.f48i They found that the pyridine is not protonated even in very acidic solution, and that the reductive desorption peak is very broad compared to that of 4-mercaptopyridine and other thiols, suggesting that for 2-mercaptopyridine, both N and the S are coordinated to the gold. 2.2 Electrochemistry of Thiophene 2.2.1 Theory and Mechanism of Electropolymerization Polythiophene can be formed by chemical or electrochemical polymerization from monomers, dimers, or short-chain oligomers, t 4 9! While chemical polymerization is also possible, electropolymerization has the advantage of speed, control of the thickness via deposition charge, no need for a catalyst, and the production of the polymer directly in its 21 conducting form.!4 9] Little input of electricity is required and the resulting polymers have favourable mechanical properties. 15°1 Many polythiophene derivatives can be charged or "doped" by oxidation (p-doping) or reduction (n-doping) to give conductive polymers. I5'1 The mechanism for thiophene polymerization from a monomer solution ( Scheme 2-1) involves the loss of an electron to give a radical cation. It is generally accepted that this is followed by the coupling of two radical cations and the loss of two hydrogens. However, it has also been suggested that the dependence of the reaction on monomer concentration points to the mechanism being an electrophilic aromatic substitution reaction between a cation radical and a neutral monomer species.1521 IR studies show that a, a'-coupling is typically favoured.!51] The initial dimer formed is immediately oxidized, as the oxidation potential of the dimer is lower than that of the monomer, due to its longer conjugation length. Further coupling is possible and the polymerization continues. It should be noted that it is not a chain polymerization. The short chain oligomers initially formed are soluble in organic solvents, but their solubility decreases with length, and they tend to deposit on the surface of the electrode once reaching a threshold level of polymerization. I5 ^ The oxidation peak observed in the C V during thiophene polymerization encompasses both the formation of radical cations that couple as well as a charging or "doping" step. The polymerization itself consumes 2 electrons per thiophene unit, while the charging consumes a further 0.25 to 0.50 electrons per thiophene uni tJ 4 9 - 5 1 ! A reduction peak is associated with the discharging of the stored charge. There is typically a large separation between the cathodic and anodic peaks during thiophene oligomer and polymer charging, 22 for which several explanations have been offered, based on different proposed mechanisms of charge storage. Originally, it was believed that charge was stored during the formation of a diionic bipolaron state. An intermediate polaron state with S = 1/2 has been observed by ESR and optical spectra, however, there is some evidence disagreeing with the formation of a Scheme 2-1 Mechanism for thiophene, pyrrole and furan electropolymerization. [49] n e u t r a l u n d o p e d p o l y m e r o x i d i z e d d o p e d p o l y m e r bipolaron state. Based on this model, the large separation between peaks was attributed to local geometric relaxations to a more planar structure with greater double bond character between monomeric units during polaron and bipolaron formation.t 5 °] More recent explanations include phase changes within the polymer!5 1] or a stepwise nonfaradaic process following charge transfer involving the desolvation of the polymer and the insertion of anions, in which conformational changes may occur.!5 3] The latter process is evidenced by the larger apparent reduction charge relative to the oxidation charge, which is attributed to capacitive charge storage through the ionic/solvent 23 movements mentioned. This happens more quickly at more positive potentials, and it is believed that the electric field enhances anion movement. There has also been evidence the charging/discharging process for longer polymer chains involves a reversible o-dimerization (without proton loss) of shorter chains, implying that the large anodic and cathodic peak separation is due to the fact that the peaks belong to two different species, and is the result of the lower oxidation potential for the dimer than for the monomer. I5 J l Polymerization is affected by a number of factors such as chain length, solvent, pH and substituents on the monomer. Typically, the reactivity of oligomers decreases with chain length!54! and protonated hexathiophene and octathiophene cations are stable at modest levels of charging. I51! There has been a large volume of literature devoted to the electropolymerization of thiophene in organic solvents such as acetonitrile. The working electrode is usually Pt or rrO, and thiophene electropolymerization is reported to occur at 1.6 V vs. S C E on a Pt electrode in acetonitrile. 15 51 Au is a less popular choice of substrate as generally it dissolves into solution at 1.35 V vs. SCE; however the dissolution of gold is suppressed in the presence of thiophene.!27! Chain lengths and morphology of polythiophene are solvent-dependent. ! 5 6 ! It has been reported that nucleophilic solvents may react more quickly with the thiophene monomers than they can react with one another and thus prevent growth of oligomers.!511 The onset of thiophene electropolymerization happens at more negative potentials at higher thiophene concentrations.!271 Typically, concentrations of between 0.01 M and 0.5 M thiophene are used. 24 Electron donating substituents reduce the oxidation potential of thiophene, while electron withdrawing substituents typically increase the oxidation potential of the monomers or oligomers i 5 1 ] The effect of pH on thiophene electropolymerization in organic solvents has been described by Can et a l J 5 7 ' 5 8 ! In general, the authors found that H + is required for chain growth, which is enhanced in the presence of some acid. However, high acid concentrations inhibited proton loss from the radical dications and therefore resulted in lower conductivity. On the other hand, in the presence of base, such as pyridine, there was a loss of electrochemical activity, which was attributed to deprotonation and degradation of the polymer. 2.2.2 Electrochemistry of Thiophene in Aqueous Media Only a small number of studies have addressed the electrochemistry of thiophene in aqueous media. Likely, earlier such studies were discouraged by a paper by Downard and Pletcherf59] on thiophene electropolymerization in C H 3 C N + B U 4 N B E 4 , where the addition of water resulted in a shift to more positive values for deposition of the polymer film as the water content was increased. "Certainly the presence of water is catastrophic," they wrote. They suggested that the presence of water led to a non-conducting and passivating layer on the Pt or polythiophene surface. IR studies of polythiophene even in organic solvents have detected C=0 stretches, suggesting that reactions with H2O or O2 during polymerization can occur. t 5 ° ] Generally, nucleophilic solvents often negatively affect electropolymerization, as the radical cation formed in the first step is very reactive, and must react more quickly with other monomers than with the solvent, f5 1] 25 More recent studies have been significantly more encouraging. M u and Parkl 5 6 ! found that while typically, in H2O, oxygen evolution takes place at potentials prior to the oxidation potential of thiophene, the polythiophene film hinders the oxidation of H 2 0 . However, they found that the chain length of the polythiophene depended on the dielectric constant of the solvent, and that chain lengths were shorter for polythiophene generated in water than in organic solvents, although the oxidation potential was found to be lowered slightly to +1.35 V vs. Ag/AgCl. A number of studies have used strong acidic media such as HCIO4, H2SO4 and H B F 4 as solvents.[55,56,60] Under these conditions, electropolymerization occurs at potentials below 1 V . The theory put forward by these papers is that the acids form a % complex with the thiophene, reducing the resonance stabilization of the aromatic ring, causing a negative shift in the oxidation potential. The authors of these papers also suggest that the acidic media may stabilize the cation radical. Hu et al. 1611 observed that the oxidation potential of bithiophene was lowered in acetonitrile solution upon addition of perchloric acid, but attributed this to enhanced solution conductivity in the aqueous medium. However, electropolymerization of thiophene at 0.8 V to 1.0 V vs. S C E has also been reported in neutral aqueous media, 1621 suggesting that the lowering of the oxidation potential has to do with the use of aqueous media and not the pH. Generally, it has been found that overoxidation of polythiophene (above 0.9 V vs. SSCE) in the presence of solvent or electrolyte anions results in degradation to oxides and oligomers. I631 However, compact polythiophene films are significantly more stable 26 than porous ones to overoxidationJ64] In addition, it has been proposed that some anions such as CIO4" are able to shield the doped polythiophene from nucleophilic attack, t6 3] 2.2.3 Effect of Pyridine Substituents on Thiophene Electrochemistry Pyridine side-chains and backbone substituents have been added to thiophene in order to engineer donor-acceptor systems and n-dopable polymers^6 5"6 7] permit the binding of transition metal ions via the N donor atoms,!68] or to produce conjugated polymers with alternating 7C-rich and 7t-deficient aromatic units, which can lead to third order non-linear optical properties and a decrease in the optical band gap due to intramolecular charge transfer. t 6 9 i Wang et alT6 5l found that the incorporation of pyridyl groups linked to the polymer via an alkyl side-chain significantly increases the oxidation potential required for electropolymerization of the thiophene backbone, due to the electron-withdrawing nature of the pyridine. However, it does stabilize the reduced (n-doped) polymer, in which the doping is believed to involve the reduction of the pyridyl groups themselves. Jenkins et a l J 6 8 ! found that incorporating pyridyl groups into the polymer backbone resulted in polymers with highly-conductive but unstable p-doped states, and n-doped states that showed poor conductivity. Lu et alJ 6 9 ] produced similar polymers chemically rather than electrochemically and showed that the pyridine groups resulted in polymers with good electron transport and electron injection properties that were both n- and p-doped easily, and had a low reduction potential. Polymers based on 2-(2'-thienyl)pyridine show a set of thiophene doping and undoping peaks at positive potentials (+0.4 V and -1.7 V vs. Ag/Ag + ) and a set of pyridine doping and undoping peaks at negative potentials (-2.2 V and -2.0 V vs. Ag/Ag +). There is a notably large separation (about 1.3 V) between 27 the doping and de-doping peaks for thiophene, proposed to be due to an additional associated step, such as the shift of an electron from the pyridine ring to the thiophene ring, accompanied by an anion. [66,67] 2.3 Fluorescence Near an Electrode 2.3.1 Ef fect o f t h e Meta l Many conjugated aromatic organic molecules, including thiophene and pyridine derivatives^6 7'7 0] are fluorescent, providing a sensitive and convenient method for their characterization. Because fluorescence involves electronic transitions, it is a source of information about the electronic structure of molecules, such as degree of conjugation. However, it can be strongly affected by electronic effects of the local environment. The total radiated power of fluorophores is observed to decrease dramatically near metal blocks, films and islands. Fluorescence is affected by a metal surface in three ways. At fairly large metal-fluorophore separations (d > 500 nm), there can be wave interference between directly emitted light and that reflected by the surface. At closer separations (d < 500 nm), some of the energy of the excited state can be non-radiatively transferred to propagating surface plasmons or to electron-hole pairs (excitons) and converted to heat. The classical model for energy transfer to a metal describes the excited state of the fluorophore as a point dipole and the metal as a medium with a frequency-dependent dielectric constant. Energy transfer from the fluorophore excited state results from coupling between the near field (the component of the radiation that decays exponentially away from the dipole) of the fluorophore and the surface plasmon modes of a metal-28 dielectric interface. The standard Forster model for energy transfer between two dipoles predicts that the rate of energy transfer between two dipoles should be proportional to d"6. Considering a metal block or metal electrode as a "volume of point dipoles" gives a d"3 dependence, which has been confirmed experimentally in many systems.I71"73! Fluorescence near a metal can also be affected by factors other than average separation between the metal and the fluorophore. The orientation of the excited state dipole determines whether the induced image dipole on the other side of the metal interface will interfere with or cancel the emitted electric field. 1 7 4 _ 7 6 1 On rough surfaces, surface plasmons can have the effect of enhancing the emission of the fluorophore, increasing observed fluorescence intensity. 175,77,78] 2.3.2 Fluorescence of Adsorbates For the most part, fluorescence is quenched for molecules adsorbed to metal surfaces. However, some fluorescence has been observed for fluorophores tethered to Au surfaces via long spacers.! 7 5- 7 7' 7 9' 8 0] Broadening! 7 5' 7 7- 8 1] in the emission spectra of species adsorbed on metal surfaces or "tailing" at the red edge of the peak!80] is often seen relative to the bulk. This has been attributed to the effects of adsorbate-surface interactions or adsorbate-adsorbate interactions on a surface that may lead to a less planar molecular structure!81] or ground state re-re stacking of aromatic molecules.!80] In addition, red-shifting of the emission is also frequently seen, another effect of adsorbate-adsorbate interactions such as stacking or aggregation. !7 7'7 9] 29 2.3.3 Effect of the Electric Potential Molecular species that experience a change in the electric dipole moment with potential (including the double layer potential near an electrode) may show spectral shifts with changes in the local electric field due to interactions between the ground and excited state dipoles and the electric field. This is known as electrochromism or the Stark effect. 17 51 The effective polarity/interfacial dielectric constant of the local environment may also be affected by the potential of the electrode. 17 51 However, the most pronounced effects on the electric potential on the fluorescence of fluorophores near the surface are the result of potential-induced changes in the fluorophore-metal separation. A study of Alexa 488 fluorophore attached via a cysteamine linkage was found to vary with potential, which was attributed to reorientation and desorption. [ 8 2 > 8 3 1 Little change in fluorescence intensity with potential was observed for Alexa 488 attached via a longer tether (18 carbons), suggesting that electrostatic interactions between the electrode and the fluorophore were relatively unimportant. Shepherd et al.l 8 4< 8 5l found that in monolayers of octadecanol mixed with 3 mol% of a fluorescent carbocyanine dye, no fluorescence was observed at potentials where the molecules were adsorbed. However, there was strong fluorescence on desorption. Aggregates were observed on desorption whose sizes were dependent on the type of potential perturbation. 2.4 References (1) Attard, G.; Barnes, C. Surfaces; Oxford University Press: Oxford, 1998, pp 9-10. (2) Mandler, D.; Turyan, I. Electroanalysis 1996, 8, 207-213. 30 (3) Xu, J.; L i , H. -L . /. Colloid Interface Sci. 1995,176, 138-149. (4) Morrison, S. R. The Chemical Physics of Surfaces; 2nd ed.; Plenum Press: New York, 1990, p 6. (5) Muetterties, E . L . Bull. Soc. Chim. Belg. 1975, 84, 959-986. (6) Lipkowski, J.; Stolberg, L . In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P., Ed.; V C H , 1992, pp 171-238. (7) Albert, M . R.; Yates, J. T., Jr. The Surface Scientist's Guide to Organometallic Chemistry; American Chemical Society: Washington, 1987, pp 19-23. (8) Bard, A . J.; Faulkner, L . R. In Electrochemical Methods: Fundamentals and Applications; 2nd ed.; John Wiley: New York, 2001, pp 12-13. (9) Bard, A . J.; Faulkner, L . R. In Electrochemical Methods: Fundamentals and Applications; 2nd ed.; John Wiley: New York, 2001, p 538. (10) Lipkowski, J.; Schmickler, W.; Kolb, D. M . ; Parsons, R. J. Electroanal. Chem. 1998,452, 193-197. (11) Buess-Herman, C. In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P. N. , Ed.; V C H : New York, 1992, pp 77-118. (12) Atkins, P. Physical Chemistry; 6th ed.; Freeman: New York, 1998, pp 858-862. (13) Parsons, R. Proc. Roy. Soc. (London) 1961, A261, 79-90. (14) Yang, D. F.; Stolberg, L . ; Lipkowski, J.; Irish, D. E . / . Electroanal. Chem. 1992, 329, 259-278. (15) Stolberg, L . ; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1987, 238, 333-353. (16) Stolberg, L . ; Morin, S.; Lipkowski, J.; Irish, D. E. J. Electroanal. Chem. 1991, 307, 241-262. 31 (17) Andreasen, G.; Vela, M . E . ; Salvarezza, R. C ; Arvia, A . J. J. Electroanal. Chem. 1999, 467, 230-237. (18) Andreasen, G.; Vela, M . E . ; Salvarezza, R. C.; Arvia, A . J. Langmuir 1997,13, 6814-6819. (19) Bizzotto, D.; Zamlynny, V.; Burgess, I.; Jeffrey, C ; L i , H.-Q.; Rubinstein, J.; Galus, Z.; Nelson, A.; Pettinger, B.; Merril, R.; Lipkowski, J. In Interfacial Electrochemistry; Wieckowski, A. , Ed.; Marcel Dekker: New York, 1999, pp 405-425. (20) Yang, D.; Bizzotto D.; Lipkowski J.; Pettinger B.; S., M . / . Phys. Chem. 1994, 98, 7083-7089. (21) Noda, H.; Minoha, T.; Wan, L . J.; Osawa, M . J. Electroanal. Chem. 2000, 481, 62-68. (22) Cunha, F.; Tao, N. J.; Wang, X. W.; Jin, Q.; Duong, B.; D'Agnese, J. Langmuir 1996,12, 6410-6418. (23) Dretschkow, T.; Lampner D.; T., W. J. Electroanal. Chem. 1998, 458, 121-138. (24) Elfeninat, F.; Fredriksson, C ; Sacher, E . ; Selmani, A. / . Chem. Phys. 1995, 702, 6153-6158. (25) Majumder, C.; Mizuseki, H.; Kawazoe, Y . J. Chem. Phys. 2003,118, 9809-9813. (26) Lavrich, D. J.; Wetterer, S. M . ; Bernasek, S. L . ; Scoles, G. J. Phys. Chem. B 1998,102, 3456-3465. (27) Hillman, A . R.; Mallen, E . F. J. Electroanal. Chem. 1987, 220, 351-367. (28) Dishner, M . H.; Hemminger, J. C.; Feher, F. J. Langmuir 1996,12, 6176-6178. 32 (29) Dishner, M . H. ; Taborek, P.; Hemminger, J. C ; Feher, F. J. Langmuir 1998,14, 6676-6680. (30) Noh, J.; Ito, E . ; Araki, T.; Hara, M . Surf. Sci. 2003, 532-535, 1116-1120. (31) Matsuura, T.; Nakajima, M . ; Shimoyama, Y . Jpn. J. Appl. Phys. 2001, 40, 6945-6950. (32) Liu, G.; Rodriguez J. A.; Dvorak J.; Hrbek J.; T., J. Surf. Sci. 2002, 505, 295-307. (33) Ishida, T.; Choi N.; Mizutani W.; Tokumoto H.; Kojima I.; Azehara H.; Hokari H.; Akiba U.; M . , F. Langmuir 1999,15, 6799-6806. (34) Garg, N.; Friedman, J. M . ; Lee, T. R. Langmuir 2000,16, 4266-4271. (35) Diao, P.; Jiang, D.; Cui, X.; Gu, D.; Tong, R.; Zhong, B. / . Electroanal. Chem. 1999, 464, 61-67. (36) Noh, J.; Kobayashi, K.; Lee, H.; Hara, M . Chem. Lett. 2000, 630-631. (37) Su, G.-J.; Zhang, H . -M. ; Wan, L.-J . ; Bai, C . -L . Surf. Sci. 2003, 531, L363-L368. (38) Matsuura, T.; Takamura T . ; Y. , S. Jpn. J. Appl. Phys., Part 1 1999, 38, 2874. (39) Hoon-Khosla, M . ; Fawcett, W. R.; Goddard, J. D.; Tian, W.-Q.; Lipkowski, J. Langmuir 2000,16, 2356-2362. (40) Cavallini, M . ; Bracali, M . ; Aloisi, G.; Guidelli, R. Langmuir 1999,15, 3003-3006. (41) Yang, J.-S.; Lee, C . - C ; Yau, S.-L.; Chang, C . - C ; Lee, C . - C ; Leu, J . -M. J. Org. Chem. 2000, 65, 871-877. (42) Ito, E . ; Yamamoto, M . ; Kajikawa, K.; Yamashita, D.; Ishii, H.; Ouchi, Y.; Seki, K.; Okawa, H.; Hashimoto, K. Langmuir 2001,17, 4282-4286. (43) Peng, Z.; Dong, S. Langmuir 2001,17, 4904-4909. 33 (44) Sandhyarani, N.; Skanth, G.; Berchmans, S.; Yegnaraman, V.; Pradeep, T. / . Colloid Interface Sci. 1999, 209, 154-161. (45) Whelan, C. M . ; Smyth, M . R.; Barnes, C. J.; Brown, N. M . D.; Anderson, C. A . Appl. Surf. Sci. 1998,134, 144-158. (46) Pinheiro, L . S.; Temperini, M . L . A . Surf. Sci. 1999,441, 45. (47) Taniguchi, I.; Yoshimoto, S.; Sunatsuki, Y.; Nishiyama, K. Electrochemistry 1999, 67, 1197-1199. (48) Sawaguchi, T.; Mizutani, F.; Yoshimoto, S.; Taniguchi, I. Electrochim. Acta 2000, 45, 2861-2867. (49) Tourillon, G. In Handbook of Conducting Polymers Vol. 1; Skotheim, T. A. , Ed.; Marcel Dekker: New York, 1986, pp 301-350. (50) Heinze, J. In Electrochemistry IV; Steckhan, E . , Ed.; Springer-Verlag: Berlin, 1990; Vol. 152, pp 1-47. (51) Heinze, J. In Organic Electrochemistry; 4th ed.; Lund, H . and Hammerich, O., Ed.; Marcel Dekker: New York, 2001, pp 1309-1339. (52) Satoh, M . ; Imanishi, K.; Yoshino, K. J. Electroanal. Chem. 1991, 317, 139-151. (53) Visy, C.; Kankare, J. Electrochim. Acta 2000, 45, 1811-1820. (54) Roncali, J.; Giffard, M . ; Jubault, M . ; Gorgues, A . J. Electroanal. Chem. 1993, 361, 185-191. (55) Bazzaoui, E . A.; Aeiyach, S.; Lacaze, P. C. J. Electroanal. Chem. 1994, 364, 63-69. (56) Mu, S.; Park, S.-M. Synth. Met. 1995, 69, 309-312. 34 (57) Can, M . ; Pekmez, K.; Pekmez, N.; Yildiz, A . J. Appl. Polym. Sci. 2000, 77, 312-322. (58) Can, M . ; Sevin, F.; Yildiz, A . Appl. Surf. Sci. 2003, 210, 338-345. (59) Downard, A . J.; Pletcher, D. / . Electroanal. Chem. 1986, 206, 147-152. (60) Jin, S.; Xue, G. Macromolecules 1997, 30, 5753-5757. (61) Hu, X.; Wang, G.; Wong, T. K. S. Synth. Met. 1999,106, 145-150. (62) Fujita, W.; Teramae, N.; Haraguchi, H . Chem. Lett. 1994, 511-514. (63) Pud, A. A. Synth. Met. 1994, 66, 1-18. (64) Christensen, P. A.; Hamnett, A.; Hillman, A . R. / . Electroanal. Chem. 1988, 242, 47-62. (65) Wang, J. X.; Keene, F. R. J. Electroanal. Chem. 1996, 405, 59-70. (66) Zhou, Z. H. ; Maruyama, T.; Kanbara, T.; Ikeda, T.; Ichimura, K.; Yamamoto, T.; Tokuda, K. J. Chem. Soc, Chem. Commun. 1991, 1210-1212. (67) Yamamoto, T.; Zhou, Z. H. ; Maruyama, T.; Kanbara, T. Synth. Met. 1993, 55, 1209-1213. (68) Jenkins, I. H.; Salzner, U.; Pickup, P. G. Chem. Mater. 1996, 8, 2444-2450. (69) Lu, H.-F.; Chan, H. S. O.; Ng, S.-C. Macromolecules 2003, 36, 1543-1552. (70) Becker, R. S.; de Melo, S.; Macanita, A . L . ; Elisei, F. Pure Appl. Chem. 1995, 67, 9-16. (71) Chance, R. R.; Prock, A.; Silbey, R. In Advances in Chemical Physics; Prigogine, I., Ed.; Interscience: New York, 1978; Vol. 37, pp 1-65. (72) Barnes, W. L . J. Mod. Opt. 1998, 45, 661-699. (73) Waldeck, D. H. ; Alivisatos, A . P.; Harris, C. B. Surf. Sci. 1985,158, 103-125. 35 (74) Axelrod, D.; Hellen, E . H.; Fulbright, R. M . In Topics in Fluorescence Spectroscopy; Lakowicz, J. R., Ed.; Plenum Press: New York, 1992; Vol. 3, pp 289-343. (75) Pope, J. M . ; Buttry, D. A . J. Electroanal. Chem. 2001, 498, 75-86. (76) Comi, S.; Tomasi, J. /. Chem. Phys. 2003,118, 6481-6494. (77) Haas, U.; Thalacker, C ; Adams, J.; Fuhrmann, J.; Riethmueller, S.; Beginn, U.; Ziener, U.; Moeller, M . ; Dobrawa, R.; Wuerthner, F. J. Mater. Chem. 2003,13, 161-112. (78) Gersten, J. I.; Nitzan, A. Surf. Sci. 1985,158, 165-189. (79) Kittredge, K. W.; Fox, M . A.; Whitesell, J. K. J. Phys. Chem. B 2001,105, 10594-10599. (80) Fox, M . A.; Wooten, M . D. Langmuir 1997,13, 7099-7105. (81) Belletete, M . ; Mazerolle, L . ; Desrosiers, N.; Leclerc, M . ; Durocher, G. Macromolecules 1995, 28, 8587-8597. (82) Gaigalas, A . K.; L i , L . ; Ruzgas, T. Curr. Top. Colloid Interface Sci. 1999, 3, 83-101. (83) L i , L . ; Meuse, C.; Silin, V.; Gaigalas, A . K.; Zhang, Y.-z. Langmuir 2000,16, 4672-4677. (84) Shepherd, J.; Yang, Y.; Bizzotto, D. / . Electroanal. Chem. 2002, 524-525, 54-61. (85) Shepherd, J. L . ; Bizzotto, D. J. Phys. Chem. B 2003,107, 8524-8531. 36 3 Experimental 3.1 Introduction In this section, experimental details regarding materials, instrumentation and procedures used in subsequent chapters are collected for easy reference. Those important for the understanding of the results will be noted again briefly where relevant. Overall, this section describes the synthesis of compounds studied, the methods involved in their electrochemical characterization, and the methods for their photophysical characterization both under and in the absence of potential control. Though some procedures used show only minor variations from the literature, they are nevertheless summarized for completeness. 3.2 Synthesis 3.2.1 Synthesis of 2-(2'-Thienyl)pyridine (TP) 2-(2'-thienyl)pyridine (TP) is available commercially (Lancaster), however, its synthesis is relatively simple and economical. With the exception of the catalyst, all chemicals in the synthesis of 2-(2'-thienyl)pyridine (TP) were reagent grade and from Aldrich, and were used without further purification. The catalyst was synthesized in the Wolf research group by Dr. Cerrie Rogers according to a literature procedure. I1! The Et20 solvent was dried over Na and benzophenone before use. TP was synthesized by a Kumada coupling (Scheme 3-1) of 2-bromopyridine and 2-bromothiophene, using NiCl2(dppp) as a catalyst (dppp = Ph2P(CH2)3PPh2), based on a general procedure by Tamao et a U 2 l The organic layer was washed with dilute HC1 37 Scheme 3-1 Reaction scheme for the synthesis of 2-(2'- thienyl)pyridine k N A B r f \ M g > f \ NiCI2(dppp) ^ S ^ B r Et2o ^ S ^ M 9 B r H2O > reflux 20 h and saturated sodium bicarbonate solution. The crude solid was then purified by column chromatography, using 2:1 hexanes:ethyl acetate as an eluant, to give a pale yellow solid. Its melting point and ' H N M R peaks were in close agreement with literature values. 12>31 The TP was further purified by recrystallization in petroleum ether and twice subliming to give colourless crystals. The product was >99% pure by G C . 3.2.2 Synthesis of 5,5'-Bis(2-pyridyl)-2s2,-bithienyl (PTTP) The synthesis of PTTP, based on a literature procedure,!3-4! is shown in Scheme 3-2. TP was dissolved in T H F and lithiated by dropwise addition of f-butyllithium at -50 °C. After stirring for 1 h, the solution was allowed to warm to room temperature. CuCl was added and 0 2 was bubbled through the solution, which subsequently turned black. The solution was then heated to reflux under O2 for 3 h. At this point, the solution, which was red and contained a pale precipitate, was removed from heat. 10% HC1 was added to quench any remaining metallated TP, and saturated aqueous NaHC03 was added until the reaction mixture was neutralized and a precipitate formed. The crude product was purified by column chromatography using 2:1 hexanes:EtOAc as the eluent, then recrystallized from CH 2 Cl2 to give yellow crystals. To ensure high purity, these crystals were recrystallized from 38 Scheme 3-2 Reaction scheme for synthesis of 5,5'-bis(2-pyridyl)-2,2'-bithienyl (PTTP) PTTP acetone. The melting point and *H N M R spectrum of the product agreed with literature values. C3 >4] 3.3 Electrochemistry 3.3.1 Materials and Instrumentation for Aqueous Electrochemistry All aqueous measurements were performed in a 3-electrode cell with a A u ( l l l ) single crystal working electrode, Au coil counter electrode and saturated calomel electrode (SCE) reference. The electrolyte was 0.05 M aqueous KCIO4 and the pH was altered by addition of NaOH solution, primarily to suppress H 2 evolution at the negative potentials required for adsorbate desorption in the chronocoulometry and fluorescence experiments. NaOH (Fluka, puriss. grade and AnalR from E M Science) was calcined overnight at 290°C before use. Pyridine (HPLC grade, 99.9%+) and thiophene (99+%) were both used as received from Aldrich. Potassium perchlorate was used as the electrolyte for all aqueous electrochemistry experiments because it does not adsorb specifically to Au(l 11) and is stable throughout the potential range of interest. Al l H2O was purified using a Millipore Gradient H 2 0 purification system, which electrodeionizes the H2O and removes residual organic material via U V photolysis. The resulting purified H 2 0 has a resistivity >18.2 Mil-cm. KCIO4 (Fluka, puriss. grade, 99%+) was recrystallized twice in H2O before use. Ar (Praxair) was filtered through a Supelco charcoal purifier before use. 39 A single crystal Au(l 11) electrode with a cross-sectional area of 0.2854 cm 2 , cut and polished by Mr. Yanguo Yang, was used as the working electrode. It was electropolished by applying a voltage of 2.5 V vs. S C E for 30 s in a solution of 1 M HCIO4, then rinsing with 10% HC1 to remove the gold oxide formed. The electrochemical results were obtained using a potentiostat (FHIELAB) connected via a data acquisition board (National Instruments PCI-MIO-16E-1, 12-bit, 12.5 MHz) to a computer. Capacitance measurements employed a frequency generator (HP 204C Oscillator) and lock-in analyzer (EG & G Princeton 5204). The software used to run all electrochemistry experiments was written in-house by Prof. Dan Bizzotto using Lab VIEW 6.0. The pH was measured using an Accumet Model 810 pH meter. 3.3.2 Experimental Set-Up for Aqueous Electrochemistry All experiments were performed at 22 ± 2 °C. A l l solutions were made up with Millipore H2O in volumetric flasks. The set-up procedure described herein for the electrochemical measurements was originally developed by Prof. Jacek Lipkowski 151 but will be summarized here for completeness. Before the experiment, all glassware was cleaned using the following procedure: After rinsing with Millipore H 2 0 , it was soaked for at least 2.5 hours in a 1:1 mixture of H2S04:HN03 that had been heated until it began to steam. After removal from the acid bath, all parts were rinsed thoroughly with H2O. Vessels were stored filled with H2O. All Teflon-coated magnetic stir bars were cleaned in a 1:1 mixture of H2S04:HN03 at room temperature, rinsed and stored in a water-filled cell. 40 Computer] andDAQ I Potentiostat & Lock-in Amplifier RE i WE Ar Figure 3-1 Electrochemical set-up for cyclic voltammetry, capacitance and chronocoulometry, showing connections to the working electrode(WE), counter electrode(CE), and reference electrode(RE). KCIO4 solutions were made without sonication, as it was found that sonication resulted in decomposition and the subsequent electrochemical detection of Cl". A diagram of the cell and its connections is shown in Figure 3-1. The cell and all parts were first rinsed thoroughly with H 2 0 , then rinsed with a small amount of electrolyte solution, and then the cell was filled with electrolyte solution. Ar was bubbled through the solution for at least 10 minutes in order to remove 0 2 . Throughout the experiment, a blanket of Ar was maintained over the surface of solution to exclude 0 2 . In order to humidify this Ar and reduce H 2 0 evaporation over the course of a long experiment, the 41 surface Ar flow was passed through a H 2 0 bubbler before being directed towards the cell. The exit port of the cell was equipped with another bubbler, which also served to minimize evaporation and allowed the Ar flow to be monitored easily. Once the solution was deaerated, electrical contact with the saturated calomel reference electrode was made via a salt bridge. The reference side of the salt bridge was filled with saturated KCI solution. Both the working and counter electrodes were flame annealed using a butane torch and quenched with water. The counter-electrode was immersed in the electrolyte solution, while contact with the working electrode was made via a hanging meniscus in order to maintain a constant surface area and in order to expose only the polished <111> surface of the electrode to the solution. Cyclic voltammograms were first performed in the absence of surfactant, and the curves were compared to the literature. Once it was confirmed that there were no contaminants in the system, NaOH solution was added to adjust the pH as necessary. The maximum solubility of TP was slightly more than 1 m M in H 2 0 at 25 °C. Solutions were made by heating the TP until it melted and then sonicating and vigorously shaking the mixture until the TP was completely dissolved. Enough KCIO4 was then added to the solution to bring the electrolyte concentration to 0.05 M . When stored at room temperature, all solutes remained in solution indefinitely provided [TP] < 1 mM. Electrochemical measurements of 1 m M TP were performed directly with a 1 m M TP/0.05M KCIO4 stock solution. For 3 x 10"5 M < [TP] < 1 mM, solutions were made by adding an electrolyte solution with 1 mM/0.05 M TP. For [TP] < 1 x 10"4M, solutions were made by adding an electrolyte solution with 1 x l O ^ M TP. 42 c + 1 -1 Figure 3-2 A representation of the electrochemical cell as a series RC circuit, where Rsoln is the solution resistance and C is the capacitance of the electrode-solution interface. The electrochemical characterization of PTTP required dilution in CHCI3 (1-2 mg/mL) because PTTP is poorly soluble in H2O. A drop of this solution was introduced to the cell via microlitre syringe, touched to the surface of the electrolyte solution, and allowed to spread across the surface. 161 Additional drops were added until saturation was reached (usually after 5-7 uL), as indicated by the appearance of crystals on the surface of the solution. The CHCI3 was then allowed to evaporate, and the working electrode was then touched to the surface, forming a hanging meniscus. The electrochemical characterization was then begun. 3.3.3 Cyclic Voltammetry and Capacitance in Aqueous Electrolyte For all cyclic voltammetry (CV) experiments, the voltage (E) vs. S C E was scanned linearly at a rate of 20 mV/s as the current (i) was measured. The current density, j, was obtained by dividing the current by the electrode area. The capacitance was measured using the phase-sensitive ac voltammetry or differential capacitance technique. An ac sine wave with an amplitude, E a c , of 5 mV rms and a frequency, f, of 25 Hz (f = (£I/2K) was superimposed on the 5 mV/s voltage ramp. The current response of the working electrode was measured. The in-phase (ireai) and quadrature (iimag) components of the 43 current were resolved by a lock-in amplifier and were used to calculate the capacitance, C, assuming a series RC circuit (); that is, C = - ^ ( l + (-^L)2) « E a c i i m a g (3.1) Values of C were divided by the electrode area to obtain the capacitance per unit area, C. Charge density, o, under C V peaks was calculated by integrating from j = 0 to the measured value of j over the E range of the peak and subtracting the measured charge due to the charging current, f7-8] The charge due to the charging current was calculated from the average capacitance C over the E range of the peak using the following equation: a = CAE ( 3 2 ) 3.3.4 Chronocoulometry in Aqueous Electrolyte 3.3.4.1 Chronocoulometry Procedures Al l chronocoulometry experiments were preceded by C V and ac voltammetry to confirm the absence of contaminants. The instrumentation used was identical. The potential profile for a single chronocoulometric measurement is shown in Figure 3-3.The procedure can be summarized as follows: The electrode was held at a resting potential, EbaSe, for a fixed length of time, with stirring. It was then stepped to a potential of interest, Ej, and held there, with stirring, until equilibrium was reached between adsorbed and solution species (as determined from capacitance-time transients). The electrode continued to be held without stirring at Ej for 60 s to allow for equilibrium to be re-established in the absence of stirring, and was then stepped briefly (0.2 s) to a very 44 Ei •^ base If Edes Ebase time Figure 3-3 Potential profile for a chronocoulometric measurement. negative potential, Ed e s , at which all species desorbed from the electrode, as the current was measured vs. time (see Appendix for sample current transients). Finally, the potential was returned to EbaSe- This was repeated multiple times, varying E; at 25 mV increments from the desorption potential to a positive limit. Ebase was chosen to be -0.950 V , which was negative enough to prevent adsorption of TP but positive enough to avoid significant hydrogen evolution. Ed e s varied between -1.000 V for the lowest concentration of TP (9.8 x 10"7 M) to -1.175 V for the highest concentration of TP (1 x 10" M). Despite the fact that TP desorbs at around -0.950 V, a large overpotential was necessary to ensure desorption within a short time scale (the current transient was measured over 150 ms). Ej was varied from Ed e s to -0.250 V . The appropriate amount of time spent holding the electrode at Ej was determined by recording capacitance vs. time transients for a given applied potential. In the absence of TP, this time was quite short, and the electrode was held at E; for 40 s. In the presence of TP, the processes being studied were kinetically quite slow, therefore, the electrode was held at 45 Ei for 200 s for E < 0.950 mV and for 400 s for E > -0.950 V (the point Ei = -0.950 V was measured twice). The current-time transients recorded were integrated over 150 ms immediately following desorption. The integrated curves were corrected for instrumental offsets and interference from faradaic reactions (such as hydrogen evolution) by extrapolating the linear part of the curve to t = O j 9 ! and taking the intercept as the charge, and dividing by the electrode area to obtain the relative charge density on the metal, AoM(Ei). Before the addition of TP, blank chronocoulometric measurements were performed on the Au(l 11) electrode in the presence of electrolyte only, both before and after the adjustment to basic pH by addition of NaOH solution. Al l measurements in the presence of TP were conducted at a basic pH to allow for a very negative Ed e s , ensuring complete desorption with minimal interference from hydrogen evolution. Charge densities on the metal relative to the potential of zero charge (pzc) were plotted assuming that the pzc of the Au(l 11) electrode was +0.255 V at neutral pH. This value was measured previously by Mr. Yanguo Yang in the Bizzotto research group, based on the position of the diffuse layer minimum of the capacitance curve in dilute KCIO4 solution.no] x h e capacitance for the bare electrode was assumed to be independent of the desorption potential over this limited potential range. Following the chronocoulometric measurements, C V and capacitance plots were obtained and compared to those taken prior to the chronocoulometry experiment to re-confirm the absence of chemical reactions or contamination. 46 3.3.4.2 C h r o n o c o u l o m e t r y E r r o r A n a l y s i s Since there were no faradaic reactions in the potential range studied in these experiments, errors due to pH were assumed to be negligible. The electrolyte concentration used was consistently 0.05 M , and varied by <1%. As Ar was passed through a water bubbler before entering the cell and any gases exiting the cell passed through an exit port water bubbler, errors due to evaporation were also assumed to be negligible. Errors in the concentrations of the solutions were estimated based on accepted errors in the glassware, pipettes and syringes used. Generally, they increased with decreasing solution concentration, and were never greater than 8 %, the error in the lowest concentration of 9.8 x 10"7 M . Errors in the measurement of a were due to the height of the meniscus, the position of the stir bar and the speed of the stir bar, which affected the surface area of the electrode in contact with the solution and the amount of area (on the sides of the electrode) in contact with the solution that was not crystallographically <111>. These errors contributed to random error between measurements. Therefore, an estimate of these errors was made by calculating the standard error from a comparison of the 3-point test curves obtained before each chronocoulometric run with the CT values at those values of E in the full experiment that followed. As only two points at each potential were used, this estimate is likely high, and represents a worst-case scenario. While the error might be expected to be larger in the steep part of the curve, this was not found to be the case, and the error was assumed to be the same for all values of E . On average, the error increased as [TP] decreased. However, the error was largest for the highest concentration, 1x10" 47 M , where the average error for the 3 points was 0.8 pC (about 3% of the total charge). This concentration is close to the limit of TP solubility in H2O, therefore the large error in the concentration may be an indication of the onset of solubility problems. Overall, in most cases, the error in a was < 0.3 pC and was consistently < 1 pC. Errors due to the instrumentation and software calculations were small in comparison to errors due to the experimental set-up. For example, the error in the slope after integration was found to be about 0.08 %. The error in a vs. E for the blank curve, in the absence of TP, was 0.08 pC on average. The lower value may be attributed to the fact that this curve was itself an average of 6 curves. The error in E was ± 0.0003 V , based on the standard deviation of 6 blank curves. The error in y, by, was calculated using average values of the error in a, based on the formula^ 1 1! 1 - -Sy = (Sa)A(J - -)2 «(Sa)AJ2 (3.3) where So is the error in o, A is the x interval (in this case, 0.025 V), and J is the number of intervals. The error calculations for 7C and T were derived from these previous calculations and were straightforward. 3.3.5 Organic Electrochemistry The gold bead working electrode was fabricated using 0.5 m M wire (Goodfellow, 99.9%). Acetonitrile (Fisher, H P L C grade) was dried for at least 24 hours over molecular 48 sieves (Fisher, type 3A, grade 562). Dichloromethane (Fisher, H P L C grade) was dried in solvent towers. Pt mesh was used as a counterelectrode and an A g wire was used as a quasi-reference electrode. N 2 was obtained from Praxair. The electrolyte used was tetrabutylammonium hexafluorophosphate [CH3(CH2)3]4NPF6 (Aldrich, 98%), was recrystallized three times from EtOH before use, and was kept under N 2 . Bis(pentamethylcyclopentadienyl)iron, also known as decamethylferrocene (DMFc; Strem, 99%), was used as received as an internal potential reference. The Au bead was fabricated by heating a gold wire with a butane torch until it melted and the molten Au started to move up the wire. To ensure that the drop of molten Au did not fall off, the bead was immediately cooled as soon as it began to move noticeably up the wire. Before each experiment, the electrodes were sonicated in CH 2 C1 2 , flamed with a butane torch, then allowed to cool. Al l electrochemistry experiments were performed in dry solvents under N 2 . Cyclic voltammetry was performed at a typical scan rate of 100 mV/s in a 3-electrode cell under N 2 , using a potentiostat (Pine, AFCBP1). DMFc was added as a solid, while TP and PTTP were added pre-dissolved in M e C N and M e C N / C H 2 C l 2 respectively. 3.4 Photophysical measurements 3.4.1 UV/Vis Absorbance Measurements U V / Vis spectra were obtained using a Varian Cary 5000 UV-Vis-NIR spectrophotometer. Solution spectra were performed in CH 2 C1 2 (Fisher, H P L C grade). Solid state spectra were performed on films dropcast from CH 2 C1 2 solutions onto quartz 49 microscope cover glass (Fisher). Cuvettes were quartz and had a path length of 10 mm (Hellma). They were cleaned with concentrated H N O 3 overnight and rinsed before use. 3.4.2 Solution and In Situ Fluorescence Spectroscopy and Microscopy Fluorescence spectra were obtained with a Varian Cary Eclipse spectrometer using special optical glass (Fisher) or quartz (Hellma) cuvettes. CE WE RE XeArc Lamp Potenl &Lo Amp iostat ck-in lifier Com] and I puter )AQ Spot RT CCD S™7 Ocean Optics S2000 Figure 3-4 Set-up for in situ fluorescence measurements, showing connections to the working electrode (WE), counter electrode (CE) and reference electrode (RE), and light paths to the Spot RT camera and Ocean Optics S2000 spectrometer. Drawing by Jeff Shepherd. 50 In situ measurements were performed in an electrochemical cell set-up similar to that used by Shepherd et al . ! 6 - 1 2 ! , as shown in Figure 3-4. Light from a Xenon arc lamp was passed through a 11000v2 filter set from Chroma Technology Corp. (excitation 340-380 nm, dichroic 400 nm, emission >430 nm) and the selected excitation wavelengths were focused on the electrode from below by the lOx objective (NA = 0.3, wd = 10 mm) of an inverted microscope (Olympus 1X70). The emitted and reflected light passed through the dichroic and emission filters, and was collected through the same microscope objective. The emitted light was directed to either a fibre optic spectrometer (Ocean Optics S2000) or a camera (Spot CCD). In the case of the spectrometer, integration time was 150 ms and the resulting spectrum was corrected for the wavelength-dependence of the spectrometer response using quinine sulphate as a reference, which has a similar spectrum to PTTP. The electrode was held at a potential of interest, Ei, typically for 15 minutes. During this time, a background spectrum, I(Ej), was taken of the adsorbed monolayer to account for leakage through filters, solubilized fluorophores (e.g. TP) in the optical path between the microscope objective and the electrode, etc. After collection of the background spectrum, the potential was stepped to Ed e s (which was -1.00 V) and another spectrum was collected. The intensity of background fluorescence from species not in direct contact with the electrode should vary little between Ei and Ed e s compared to the changes in fluorescence intensity for adsorbate species. Fluorescence intensity from species adsorbed to the electrode at E; should be almost completely quenched at Ej, since non-radiative energy transfer to the metal significantly decreases the lifetime of an excited state within close proximity to a metal surface.! 1 3- 1 4! However, organic molecules 51 adsorbed on the electrode at Ej desorb into solution at E d e s . Upon desorption, the molecules separate from the metal surface, significantly decreasing the efficiency of metal-mediated quenching. This results in a large increase in the fluorescence intensity of the previously adsorbed species at E d e s relative to their fluorescence intensity at Ej , since the rate of energy transfer to the metal surface is inversely related to the distance from the electrode (Section 2.3). Therefore, AI, the spectrum of only the species that had been adsorbed at Ej , without background from non-adsorbing species, scattering, etc. is obtained by subtracting the background spectrum measured at E ; from the spectrum measured at E d e s , that is AI = I ( E d B 1 ) - I ( E i ) (3-4) 3.5 References (1) Van Hecke, G. R.; Horrocks, W. D., Jr. Inorg. Chem. 1966, 5, 1968-1974. (2) Tamao, K.; Kodama, S.; Nakajima, I.; Kumada, M . Tetrahedron 1982, 38, 3347-3354. (3) Constable, E . C ; Sousa, L . R. J. Organomet. Chem. 1992, 427, 125-139. (4) Kauffmann, T.; Wienhoefer, E . ; Woltermann, A. Angew. Chem. Int. Ed. 1971,10, 741-743. (5) Bizzotto, D. Ph.D. Thesis, University of Guelph, 1996, p 20. (6) Shepherd, J.; Yang, Y.; Bizzotto, D. J. Electroanal. Chem. 2002, 524-525, 54-61. (7) Yang, D. F.; Wilde, C. P.; Morin, M . Langmuir 1996,12, 6570-6577. (8) Yang, D. R; Wilde, C. P.; Morin, M . Langmuir 1997,13, 243-249. (9) Stolberg, L . ; Lipkowski, J.; Irish, D. E . / . Electroanal. Chem. 1987, 238, 333-353. 52 (10) Yang, Y.; Bizzotto, D. J. Electroanal. Chem. 2001, 500, 408-417. (11) Oldham, K.B.J. Electroanal. Chem. 1986, 208, 1-12. (12) Shepherd, J. L . ; Bizzotto, D.J. Phys. Chem. B 2003,107, 8524-8531. (13) Axelrod, D.; Hellen, E . H.; Fulbright, R. M . In Topics in Fluorescence Spectroscopy; Lakowicz, J. R., Ed.; Plenum Press: New York, 1992; Vol. 3, pp 289-343. (14) Chance, R. R.; Prock, A.; Silbey, R. In Advances in Chemical Physics; Prigogine, I., Ed.; Interscience: New York, 1978; Vol. 37, pp 1-65. 53 4 Cyclic Voltammetry and Capacitance of 2-(2'-Thienyl)pyridine, Pyridine and Thiophene 4.1 Introduction 2-(2'-thienyl)pyridine (TP) was first characterized by cyclic voltammetry (CV) and capacitance measurements to gather basic information about its adsorption on Au(l 11) and the effect of potential on its behaviour. The fact that TP is made up of two separate rings bearing two different surface active functional groups adds a level of complexity to the interpretation of its behaviour. On the other hand, it can serve as a starting point for characterization. The constituents of TP, thiophene and pyridine, are each known to adsorb to Au(l 11) and their individual surface chemistries on gold have been well-studied. The electrochemistry of pyridine on Au(l 11) has been very well characterized by the group of Prof. Jacek Lipkowski at the University of Guelph, as described in section 2.1.3.1. It is reproduced herein at neutral pH and performed at a more basic pH. The electrochemistry of thiophene at potentials negative of the onset of polymerization has been less well-characterized. However, it was clear that studies of thiophene within this potential window might be a great help in understanding the behaviour of TP, and so a study of thiophene electrochemistry was also undertaken. In this chapter, the electrochemical behaviour of TP on Au(l 11) will be introduced and compared to that of thiophene and of pyridine on the same substrate. 54 4.2 Au(111) in KCI04 Electrolyte at pH 7 and pH 11 Before each electrochemical experiment (with the exception of the ones involving the highest concentration of TP), the electrode was first characterized by C V and ac voltammetry in the presence of 0.05 M KCIO4 electrolyte, but in the absence of adsorbates, to confirm that the system was free of contaminants and to serve as a blank for comparison purposes. A typical C V showing current density, j, as a function of E from -0.8 V to +1.25 V at pH 6 is shown in Figure 4-1 and agrees well with the literature, confirming the <111> -80-| 1 1 1 1 1 1 1 1 1 1 -1.0 -0.5 0.0 0.5 1.0 1.5 E (V vs. SCE) Figure 4-1 A cyclic voltammogram of a Au(l 11) electrode in the presence of 0.05 M KC10 4 at pH 7, using a scan rateof 20 mV/s. 55 crystallography and the absence of contaminants.! 11 Gold oxidation peaks can be seen at +0.83 V and +1.1 V on the positive scan, while reduction peaks can be seen at +0.68 V, +0.27 V and +0.17 V on the negative scan. Looking only at the potential region negative of the onset of gold oxidation (the double layer region) the current is very small and is capacitive in nature. The cyclic voltammogram and capacitance curves for this potential region at pH 7 and pH 11 are shown in Figure 4-2. When the Au (111) is flame annealed, the surface reconstructs to form the (23 x V3) overlayer structure. A small peak at +0.43 V (also visible in Figure 4-1) is associated with the lifting of the reconstruction and the reversion of the surface back to the (1 x 1) structure.I1-2! Near -0.8 V , the negative current increases due to the onset of hydrogen evolution from the electroreduction of H2O. Defects in the surface of the electrode acting as nucleation sites for the reaction decrease the overpotential required for this reaction. Also shown in Figure 4-2 are C V and capacitance measurements of the same Au electrode in the presence of OH" (pH 11). In these plots, a sharp peak is observed at +0.2 V. According to a detailed study of OH" adsorption on Au(l 11), this peak is associated with Au surface reconstruction in the presence of hydroxide adsorption. I31 It shifts to more negative potentials as the pH is increased. Positive of this sharp peak, evidence of oxide formation was found. I31 For studies at pH 11 herein, the potential was kept below +0.2 V in order to avoid Au oxidation. At negative potentials, the curve overlaps with the pH 7 curve. 56 Figure 4-2 Cyclic voltammograms of Au(l 11) in 0.05 M KC10 4 within the double layer potential range at pH 7 and pH 11, at a scan rate of 20 mV/s. indicating that the OH" is desorbed. Hydrogen evolution is also shifted to more negative potentials in the presence of OH." 57 4.3 TPonAu(111) TP was characterized by C V and capacitance in order to determine its behaviour on Au(l 11) as a function of potential, concentration and pH. For clarity, only 80-, 60 H E 40-LL O 20-1 0 6 4-2 blank "}1 mM TP i — 1 — r I ° |-2. -4 -6 T P 4 T P 1 B I T P 1 A T P 2 T P 1 A T P 2 I ' \ •: ' \ :• ' \'.' / • i \ / \ ' i ' t / \ / \ / T P 3 T P 1 B ^ > i > i > i • i > i • i •1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 E(V vs. SCE) Figure 4-3 Capacitance and CV of 1 mM TP on A u ( l l l ) , showing the effect of opening the positive potential limit on the reversibility of the CV. Scans were performed in 0.05 M KCIO4 at pH 11. The scan direction is indicated by the arrows. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blank scans of the electrode under identical conditions in the absence of TP are shown with a dotted line. 58 the positive sweep capacitance scans will be shown for experiments performed in the absence of surfactant in this section. Steady state curves in the presence of 1 m M TP at pH 11 are shown in Figure 4-3, and the plots in basic solution are shown first, as it will become apparent that the two sets of peaks are better separated at this pH. Generally, experiments were conducted at pH 11 because at neutral pH, the desorption of TP overlaps with hydrogen evolution at negative potentials. Corresponding features in the C V and capacitance are marked with arabie numerals. Two main peaks are visible in the positive scan of both the C V and capacitance curves. A very sharp peak TP1A, appears near -0.42 V and a broader peak TP2 appears at +0.13 V in the capacitance curves. Peak TP1A is reversible, giving a very similar peak TP1B on the negative scan, provided the positive potential limit is kept negative of the onset of TP2. However, after scanning past TP2 on the positive scan, significant hysteresis is observed. This is particularly noticeable in the capacitance plot. Here TP1B disappears, and a single broad peak TP3 appears at -0.67 V . In addition, after scanning to a positive potential limit of +0.2 V , on subsequent scans TP1A and TP1B each split into two peaks that can be resolved at concentrations greater than 3 x 10"4 M . In 1 m M TP, one is at the original potential of -0.42 V and the second is at -0.46 V . The original peak decreases in size, while the new peak increases in size during subsequent cycles until a steady state is reached. (Note that both TP1B peaks only appear when the positive limit is kept negative of the onset of TP2.) In addition, a small peak TP4 near -0.9 V is often observed in the full range steady state C V . It increases significantly in size on holding the potential at +0.2 V for longer than 10 minutes. 59 no TP i — | — i — | — i — | — i — | — i — | — i — | — i — | — i — | -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 E (V vs. SCE) Figure 4-4 Effect of bulk TP concentration on cyclic voltammograms of a Au(l 11) electrode, performed in 0.05 M KC10 4 at a scan rate of 20 mV/s. The shape and peak potential of TP1 and TP2 vary with concentration. With increasing concentration, TP1 shifts to more negative potentials and sharpens dramatically (Figure 4-4). TP2 also sharpens with increasing concentration, but less dramatically, and shifts to more positive potentials with increasing concentration. Interestingly, the charge beneath TP2 does not appear to be concentration dependent. TP3 varies little in potential, size or shape with solution concentration. The electrochemistry of TP also shows a dependence on the pH of the supporting electrolyte. In Figure 4-5, capacitance plots and CVs of the full potential range are compared in neutral and basic solution at pH 7 and pH 11. Corresponding peaks at the 60 pH 11 E (V vs. SCE) E (V vs. SCE) Figure 4-5 A comparison of 1 mM TP CV and capacitance curves at pH 7 (a and c) and pH 11 (b and d), performed on Au(l 11) in 0.05 M KC10 4. CVs were performed at a scan rate of 20 mV/s, and the arrows indicate the scan direction. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blanks scans of the electrode under identical conditions in the absence of TP are indicated by dotted lines. Peaks are labelled with Arabic numerals in the CVs, while different states are indicated by roman numerals in the capacitance plots. two pHs are numbered in the CVs using the same numbering system as in Figure 4-3. There are a number of differences between the plots of TP at pH 7 and pH 11. It is clear that the potential of TP1 peaks is pH independent. However, the ratio between the peaks appears to favour the more negative peak at pH 7 and the more positive peak at pH 11. TP2 and TP3 both shift to different potentials with a change in pH. At pH 7, TP2 appears at +0.35 V in the C V . This results in a shift of close to 60 mV/decade towards 61 negative potentials as the pH increases, suggesting the occurrence of a faradaic process involving 1 proton per TP molecule. TP3 appears at +0.50 V at pH 7, giving it a potential shift with pH that is less than 60 mV/decade. The potential shift of TP3 with pH is slightly less than 60 mV/decade in this figure. It is possible that TP3 does not involve H + , as will be discussed in later chapters. In addition to these shifts, TP2 and TP3 are broader at pH 7 than at pH 11, and show some shoulders in neutral solution that are not present in basic solution. These have been labelled in Figure 4-6. T P 3 B is only visible when the positive potential limit is above +0.4 V. Also shown in this figure is that at pH 7, there is a more noticeable evolution - 8 - | — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E (V vs. SCE) Figure 4-6 Changes in the CV of a Au(l 11) electrode in the presence of 1 mM TP with successive scanning at pH 7, in 0.05 M KC10 4 at a scan rate of 20 mV/s. 62 with time of the shape of TP2 and TP3 (along with the change in the relative sizes of the TP1 peaks seen at all pHs, vide infra) from the first full potential range scan to the time at which a steady state is reached. With repeated scanning, the current of the more positive peak, TP2B, decreases and the onset of TP2A shifts to more negative potentials. The nature of the shoulder peaks can be further resolved by holding at positive potentials and gradually opening the negative scan limit (Figure 4-7). As can be seen, the product of TP2 is stable to a negative potential limit of 0 V. TP3A appears to be a reduction, with TP2A comprising most of the corresponding oxidation. The peak at -0.5 V appears to be the reverse peak of TP 1A (or specifically the more negative of the two E o 1-1 3-2-1 -0--2--3--4-TP2B TP2A TP3A r TP3C V ( 2nd scan 4th scan 8th scan " i — • — r -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 E (V vs. SCE) Figure 4-7 CVs of a Au(l 11) electrode in the presence of 1 mM TP at pH 7 after holding at 0.4 V for 5 min and gradually increasing the negative potential scan limit on successive scans. Selected peaks are labelled. Scans were performed in 0.05 M KC10 4 at a scan rate of 20 mV/s. 63 sub-peaks). A summary of the important features of TP electrochemistry, based on Figure 4-5, follows. Generally, for 1 m M TP at pH 11, if a C V scan is started at -1.0 V vs. SCE, where TP is desorbed, and the potential is scanned initially in a positive direction, several potential regions in which there is adsorption to the electrode can be distinguished. At potentials negative of peak TP1, the adsorbed TP molecules are in State I, where the capacitance is similar to that of the bare electrode. At potentials between peak TP1 and peak TP2, the adsorbed TP results in a low capacitance region corresponding to State II, which can be converted reversibly back to state TP1 by reversing the scan direction at any potential negative of peak TP2. On scanning the potential positive of TP2, the adsorbed TP undergoes a transition to State HI, which is characterized by an even lower capacitance than State n, and cannot be reversibly converted back to State II. When the potential is reversed and scanned in a negative direction, State III is stable until the onset of peak TP3, a process that results in State IV, along with significant desorption. At lower concentrations of TP, because peak TP1 is shifted to more positive potentials and peak TP3 is shifted to more negative potentials, the potential region of stability for State II increases with TP concentration. The molecules are completely desorbed at -1.0 V, as demonstrated by the equivalence of the capacitance with and without the TP molecule at this potential. The surface coverage of TP after scanning positive of TP2 was estimated using CVs of 1 x 10~3 M TP in basic solution. Assuming the molecules are adsorbed as a monolayer, the coverage T can be calculated using 64 (4.1) where q is the charge, A is the electrode area, n is the number of electrons, and F is Faraday's constant. Assuming that n = 1, integrating under TP3 at equilibrium and accounting for the capacitive charge gives a coverage of 4.6 (± 0.4) x 10"10mol/cm2. As will be explained later (section 4.6), assuming 1 electron per TP molecule may be an overestimate by up to a factor of 6 and therefore an underestimate of surface coverage. On the other hand, this may be mitigated by the possibility that a significant portion of the apparent reductive current is actually due to a slow capacitive process.[4] Regardless, the calculated coverage compares favourably to the 5 x 10"10 mol/cm 2 surface coverage reported for the structurally similar 2, 2'-bipyridine on Au(l 11) at positive potentials.^ 4.4 Pyridine onAu(111) C V and capacitance measurements of pyridine were conducted in neutral and basic pH in order to compare the results to the corresponding results obtained for TP. Pyridine is readily soluble in H 2 0 and was added to the cell from a 10 m M stock solution. Shown in Figure 4-8, the cyclic voltammogram and capacitance plots of ImM pyridine at neutral pH are similar and show two sets of reversible peaks that are very similar to those in the literature.^] The peak at -0.55 V, labelled PY1, is associated with adsorption. Stolberg et al.H] found that initially, the pyridine molecules are adsorbed with the plane of the ring parallel to the surface of the electrode, via an interaction between the negatively-charged gold 65 electrode and the electron-deficient aromatic % system of the pyridine ring. On adsorption of pyridine, the capacitance becomes lower than in the absence of pyridine at these potentials, however the difference is relatively modest. A very sharp peak at +0.10 V, labelled PY2, is attributed to a two-dimensional phase transition in which the orientation of the pyridine molecules changes to an N-bonded Figure 4-8 CV and capacitance curves of a Au(l 11) electrode in the presence of 1 mM pyridine at pH 7 and pH 11, with the peaks labelled. Scans were performed in 0.05 M KCIO4. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blank curves under identical conditions in the absence of pyridine are shown with a dotted line. 66 configuration, in which the plane of the ring is normal to the surface of the electrode. This transition occurs roughly at the pzc (potential of zero charge), as bonding through the lone pair of the N is favoured on the positively-charged electrode. Positive of this peak, the capacitance drops significantly to a minimum of 9.6 up/cm 2 A low capacitance is associated with close-packed layers of dielectric organic molecules. Capacitance, C, is related to the dielectric constant 8 and the separation between the capacitor plates, d (where eo is the permittivity of free space), by d (4.2) Organic molecules typically have much lower dielectric constants than water (Table 4-1). In addition, by coating the electrode surface, the distance between the metal and the solution (in essence the distance between the two plates of a model capacitor) is increased. Both these effects serve to reduce the capacitance in the presence of a close-packed layer of adsorbed organic molecules. At potentials negative of the adsorption peak, the pyridine curves merge with that of the bare Au(l 11), indicating that pyridine is desorbed at these potentials. Table 4-1 Dielectric constants of selected pure substances'^! Compound e T (°C) H 2 O w 78.54 25 thiophene^ 2.76 16 pyridine^ 12.3 25 67 The adsorption peak is broadened here when compared to that seen by Stolberg et alJ 1] and there is some minor fine structure in P Y 2 . This indicates that the electrode surface is less uniform than that used by Stolberg et al., and may contain some defects at which the kinetics differ slightly from the majority of the surface. At lower concentrations, both peaks broaden and shift to more positive potentials. A l l peaks are pH independent, as shown in Figure 4-8. The only notable differences between the curves at pH 7 and pH 11 are the increase in current and capacitance at positive potentials and a decrease in the amount of negative current at negative potentials as the pH increases. Based on studies of OH" adsorption, on Au (111), the latter is likely the adsorption of OH" and the lifting of the (23 x V3) gold reconstruction near the positive potential limit.t3] 4.5 Thiophene on Au(111) The aqueous electrochemistry of thiophene within the double layer potential range of Au(l 11) was characterized in detail by C V and capacitance measurements for comparison to TP, as a thorough search of the literature revealed no previous studies. With a relatively low boiling point, thiophene is quite volatile at room temperature and tended to evaporate significantly overnight even when kept in a tightly capped flask. Therefore, experiments were kept as short as possible and fresh stock solutions were made immediately, before their injection into the cell. Typical cyclic voltammograms of 10 m M thiophene at pH 7 are shown in Figure 4-9 as the positive limit is gradually increased to show the relationships between the peaks, which have been labelled for clarity. 68 — ] — 1 — I — I — I — I — I — 1 — I — I — I — I — I — I — I — I — I — I — I -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E (V VS. SCE) Figure 4-9 CV and capacitance plots of Au(l 11) in the presence of 1 mM thiophene at pH 7 in 0.05 M KC10 4, showing the effect of increasing the positive potential limit on successive scans. Major peaks are labelled. CVs were at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blanks scans of the electrode under identical conditions in the absence of thiophene are shown with a dotted line. The C V in the presence of thiophene shows only very slight features negative of 0 V . However, the capacitance in the presence of thiophene already falls below that in the 69 absence of thiophene at potentials as negative as -0.6 V , suggesting that thiophene is adsorbed even at these potentials. As the positive limit increases, a peak labelled THI appears at +0.45 V in the C V (and at a slightly more negative potential in the capacitance) with a shoulder on the negative side. Scanning positive of the shoulder of THI is associated with a very slight peak near -0.1 V on the reverse scan. However, scanning positive of THI itself results in two significant changes. First, the capacitance drops very significantly and remains low until -0.3 V on the negative scan. Secondly, several peaks appear on the negative scan: TH2, a small peak at +0.3 V; two large peaks, TH3 at -0.2 V and -0.4 V that appear as a single, poorly-defined peak in the capacitance (and have been grouped together due to their close interrelationship); and TH4, a small sharp peak at -0.9 V . In addition, a couple of slight, poorly defined peaks appear between TH3 and TH4. Al l peaks appear to evolve with repeated scanning (Figure 4-10). On the first scan, THI shows a significantly larger peak current than on subsequent scans, and there is only a broad, slight shoulder on the negative side. On subsequent scans, the shoulder is greatly enlarged and the main peak grows smaller. The TH3 peaks are also shifted to more positive potentials on the first scan relative to their steady state potentials, and on repeated scanning, the more negative of the two peaks increases in size relative to the more negative peak. The current of TH4 (and of a small peak between TH3 and TH4) also increases with repeated scanning. As peak TH4 grows larger on the negative scan, TH5 becomes visible on the positive scan near +0.8 V . When the pH is increased by the addition of NaOH solution, the peaks are shifted to more negative potentials and THI and TH3 become narrower, without sub-peaks or 70 blank 10 mM thiophene, 1st scan 10 mM thiophene, steady state scan - i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E (V VS. SCE) Figure 4-10 A comparison of the first and steady state CVs for at Au(l 11) electrode in the presence of 10 mM thiophene at pH 7 in 0.05 M KC10 4, performed at a scan rate of 20 mV/s. Major peaks are labelled. A blank scan under identical conditions in the absence of thiophene is shown with a dotted line. shoulders. Figure 4-11 shows a comparison between 1 m M thiophene at pH 7 and at pH 11. At pH 11, TH1 and TH3 appear at 0.13 V on the positive scan and -0.62 V on the negative scan respectively. The separation between peak TH1 and TH3 remains the same (0.75 V) at both pHs if in the pH 7 case TH3 is considered to occur at average potential between TH3A and TH3B. Based on the C V and capacitance results, some observations can be made regarding the behaviour of thiophene on Au(l 11) within this limited potential range. The absence of an adsorption peak suggests that adsorption of thiophene on Au(l 11) is generally a 71 E o 60-4 ( H 3 O 20H blank, pH 7 1 mM thiophene, pH 7 blank, pH 11 1 mM thiophene, pH 11 TH3 T H 3 -I—i— i—i—|—i—|—.—|—i—|—i—|—i—|—i—|—i—| -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E (V vs. SCE) Figure 4-11 CV and capacitance curves of 10 mM thiophene on Au (111) at pH 7 and pH 11 in 0.05 M KC10 4 solution. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blank curves at each pH in the absence of thiophene are shown with dotted and dashed lines. kinetically slow process. In addition, the adsorbed layer does not lead to an immediate dramatic change in the capacitance at the electrode. Initially, the layer may be quite diffuse. Slow kinetics of thiophene adsorption were also reported by Su et al.l 7! in a study of thiophene on Au in 0.1 M HCIO4. They observed no peaks at all by C V although 72 adsorbed thiophene was imaged by S T M and a drop in the capacitance of the electrode was observed in the presence of adsorbed thiophene. Thiophene adsorption studies in the absence of potential control show that 15 h are required to reach an equilibrium between the thiophene in 10"2 to 10"4 M ethanolic solution and the thiophene adsorbed to Au(l 11), compared with a few seconds for thiols; and that initially thiophene forms diffuse layers in which the plane of the ring is parallel to the surface. C81 Peak T H I appears due to two oxidative processes (vide infra). TH3 is due to a related reduction process, as it only appears after scanning positive of t h i . Oxidation of thiophene typically results in oligomerization or polymerization and simultaneous charging or doping of the newly formed polymer. As described earlier, the accepted mechanism for thiophene polymerizationt9 - 1 1] involves the oxidation of the monomer at positive potentials to form a radical cation that can couple with other radical species nearby. Upon coupling, a dication is formed and removal of one proton per thiophene is required to regain neutrality, therefore this part of the process is pH-dependent. The longer conjugation length of the coupled product (a dimer, trimer or oligomer) results in these species having a lower oxidation potential than the monomer, such that it is immediately oxidized to a "doped" or charged species in which extra cationic charge is stored. It is expected that the charging process would be pH-independent. The C V and capacitance results are consistent with dimerization occurring. On the first scan, only monomer is adsorbed on the electrode. Therefore, T H I appears as a single peak, broadened near its maximum, at a potential similar to the potential of the main T H I peak in subsequent scans (Figure 4-10). This is consistent with almost simultaneous 73 oxidation and doping at the oxidation potential for the monomer (the broadening at the maximum being due to the close overlap of two peaks). TH3 on the negative scan is most likely due to reduction back to the neutral (undoped) species, which appears to be followed by desorption at negative potentials. The fact that TH3 shows more than one peak may indicate the formation of not just dimers but higher oligomers even on the first scan. This is certainly possible, given the fact that each thiophene has two reactive a carbons and the lower oxidation potential of newly-formed dimers would favour their immediate oxidation to the cation radical, allowing for further coupling. On subsequent positive scans, the readsorption of some of the dimer formed on the previous scan may be expected, as well adsorption of additional monomer from solution. At pH 7, the shoulder is seen on the negative side of the main TH1 peak, consistent with the doping of lower oxidation potential dimers produced on previous cycles. The main peak is smaller, as a smaller percentage of adsorbed molecules are monomer species. These would have a lower oxidation and reduction potential than the monomers and dimers, and therefore their reduction peaks would occur at more negative potentials. The increase in current for TH3, TH4 and the small peaks between them on subsequent scans are consistent with an increase in the proportion of longer chain species, which would be expected to desorb at more negative potentials relative to shorter chain species. However, a steady state in the C V of thiophene is reached quickly and after the second scan very little change in the size, shape or potential of the thiophene peaks was observed, giving no evidence of further oligomerization or polymerization. This is consistent with the work of Fujita et al.,11 2! which showed that more positive potentials are needed to produce polymer in aqueous solution. 74 The potentials at which the oxidative process for thiophene is observed are remarkably low. Thiophene polymerization is typically observed in organic solvents at a potential of +1.6 V to +2.0 V vs. S C E at a Pt electrode. t 9 l There are a few reports of thiophene polymerization in acidic aqueous media, t 1 3 - 1 5] In all cases, onset of polymerization was observed at considerably less positive potentials than required in organic solvents. Bazzaoui et alJ 1 3 ] and Jin and Xuef 1 4] attributed the low oxidation potential to the formation of 7C complexes between the monomer and the HCIO4, reducing the aromaticity of the rings; and to stabilization of the cation radical by the acid medium. Hu et a l J 1 6 ! observed that the oxidation potential of bithiophene was lowered in acetonitrile solution upon addition of perchloric acid, but attributed this to enhanced solution conductivity in the aqueous medium. Thiophene oligomerization on gold at a potential of 0 to +0.6 V vs. S C E in neutral aqueous electrolyte has also been reported^12] with polymerization occurring at +0.8 to +1.0 V . This is consistent with our results, and combined with the observation that the oxidation potential of thiophene is even lower in basic solution, it suggests that perhaps the low oxidation potential is due primarily to the use of an aqueous solvent rather than acidity per se. Indeed, it was found that in acetonitrile solution, a potential of +1.2 V vs. S C E was required to oxidize 15 m M TP (section 4.7). The oligomerization mechanism also explains many of the pH effects observed on thiophene electrochemistry. Based on the CVs, the shift in THI to more negative potentials is 79 mV/decade, supporting the argument that some thiophene molecules lose more than one proton to form chains longer than a dimer. If TH3 is due to reductive discharging as is believed, it should not be affected by pH. In that case, the shift in TH3 with pH would likely be due to a kinetic effect. The disappearance of the shoulder on the 75 positive side of TH1 may indicate the overlap of a pH dependent process (oxidative dimerization/oligomerization) with a pH independent process (charging) at high pH. That the reductive peaks also merge at high pH may indicate that some may involve a pH dependent process, or that fewer species are formed at high pH. The charge beneath both TH1 and TH3 peaks is the same at both neutral and basic pH. Integrating under TH1 was problematic due to other processes occurring near the positive side of the peak. The charge beneath TH3 was integrated using two different methods to account for the charging current. In the first, a two-point baseline was drawn between the inflection points at the positive and negative potential limits of the peak, and only the current above the baseline was integrated. This method may underestimate the charge, as contributions from other processes, such as hydrogen evolution on the negative side of the peak, may prevent the current from reaching the capacitive baseline. The second method involved integrating the current beneath the peak from j = 0 to the measured value of j and subtracting the capacitive current based on the average capacitance over the potential range of the peak. The capacitance is measured using a lower scan rate than the C V and is frequency dependent, therefore the charging current calculated this way is likely to be an underestimate. The two methods gave a charge of 50 uC/cm 2 and 43 uC/cm respectively. The actual faradaic charge should be somewhere between these two values, and will be assumed to be 47 ± 5uC/cm 2. Typically, the charging of polythiophene consumes 0.25 to 0.4 electrons per thiophene.!10! Assuming that TH3 is due to typical charging processes then gives between 1.9 ± 0.2 x 10"9mol/cm2 and 1.2 + 0.2 x 10"9 mol/cm2, corresponding to 8 A2/molecule to 14 A2/molecule. At most, this is half the 3.7 x 10"9 mol/cm 2 (or 4.5 A2) 76 coverage estimated by Matsuura et aU8i based on their surface IR spectroscopy study. However the area occupied by a single thiophene molecule adsorbed with the plane of the ring parallel to the surface was calculated to be 36 A2 and the area occupied by a thiophene adsorbed via the S with the plane of the ring normal to the surface was calculated to be 20 A2 (using U C S D Chimera and assuming a van der Waals radius of 1.7 A for C,n7] 1.85 A for Sl 6 l , and ignoring H). Based on these numbers, the thiophene layer on Au(l 11) observed is likely to be adsorbed in a close-packed, vertical S-bonded layer. The small discrepancy between the theoretical size of thiophene and the value calculated from the charge suggests a possible charging level of more than 0.4 electrons per thiophene. 4.6 Comparison of Electrochemistry of TP, Pyridine and Thiophene One of the challenges associated with qualitative electrochemical studies is that the results can be quite difficult to interpret. However, in the case of TP, some insight into the origin of the peaks can be gained by comparison with pyridine and thiophene. As with thiophene, TP shows no adsorption peak. However, a slight depression in the capacitance as well as evidence that will be presented in subsequent chapters suggest that adsorption occurs at ~ -0.8 V at pH 10.7, well negative of TP1. There is almost no change in the capacitance, suggesting that the layer is very diffuse, and that in all likelihood, TP adsorbs in a flat-lying configuration, with the plane of the ring and the long axis both parallel to the surface of the electrode. This would be consistent with the orientation of pyridine and bipyridine on Au(l 11) at negative potentials and the initial orientation of adsorbed thiophene.f1'5'8] 77 120 blank - < H 1 1 1 > | 1 1 1 1 1 1 1 1 . , , -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 E (V vs. SCE) Figure 4-12 CV and capacitance plots of a Au(l 11) electrode in the presence of 1 mM TP and 1 mM pyridine in 0.05 M KC104 at pH 7. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Peaks are labelled and the scan direction is indicated by the arrows. Blank curves in the presence of electrolyte only are shown with a dotted line. The peak TP1 has much in common with the pyridine peak PY2 corresponding to a phase transition from a flat-lying to tilted N-bonded state. Its pH independence is consistent with other evidence that TP1 is due to changes in the interfacial capacitance, 78 such as the strong concentration dependence of the peak shape and the relatively small amount of charge beneath the peakJ 1 7] Typical of peaks due to capacitive changes, both TP1 and PY2 are sharp and reversible, with minimal charge beneath the peak, and they broaden dramatically and shift to more negative potentials as the concentration of surfactant decreases. Scanning positive of both TP1 and PY2 leads to a similar drop in the capacitance positive of the peak. The similarities between PY2 and TP1 can be seen clearly in Figure 4-12. It is proposed that TP1 is due to the pyridine half of TP and that it is the result of a phase transition similar to that seen in pyridine from an initial flat-lying state to a perpendicular state bonded through the N and/or S atoms. The fact that the peak is shifted to more negative potentials in TP relative to pyridine is expected. Thiophene is 7i-electron rich relative to pyridine, and thus should act as a n donating group in TP. This would make the pyridine ring less electron deficient and favour a transition to the N-bonded state at more negative potentials. In addition, the interaction between the S and the gold surface may contribute to driving the transition of the pyridine ring from parallel to perpendicular. A large negative shift of this transition relative to pyridine is also observed for 2,2'-bipyridine.t5] While it is not possible to determine the conformation of the thiophene half of the molecule in this state, the surface coverage calculated from chronocoulometry experiments suggests that the thiophene may also be bonded via S at this potential, as will be discussed further in the next chapter. Unlike TP1, TP2 resembles no pyridine peaks, but instead shows similarities to thiophene peak THI in its irreversible behaviour and the low capacitance of the layer after scanning positive of it and reversing the scan direction (Figure 4-13). Both TP2 79 Figure 4-13 C V and capacitance plots in the presence of 1 mM TP and 1 mM thiophene at pH 7. Selected peaks are labelled, and the scan direction is indicated by the arrows. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. Blank scans of the electrode in the presence of electrolyte only are shown with a dotted line. and T H 1 show a larger maximum current and less current on the negative side of the peak on the first scan relative to subsequent scans, and show similar changes with pH. Thiophene peaks T H 1 and T H 3 and TP peaks TP2 and TP3 all shift negative and grow narrower, with shoulders and sub-peaks apparently merging into the main peak as the pH 80 is raised. Thiophene also shows a peak, TH4, similar to TP4 near -0.9 V . These similarities between TP peaks TP2, TP3 and TP4 and features in the C V and capacitance plots of thiophene suggest these peaks are due to the thiophene half of the molecule. The most reactive site on TP is expected to be the free oc-carbon of the thiophene, numbered 5' in Scheme 4-1. It is expected that TP could not polymerize within the limited potential range of the studies herein. However, given the evidence that thiophene undergoes a dimerization or oligomerization at these potentials, it is likely that peak TP2 is due to a similar dimerization and charging process, in which two TP molecules are linked via the 5' C. This would require the loss of 1 H (accompanied by 1 electron) per TP molecule, consistent with the pH dependence for the peaks; plus an amount of additional charge due to capacitive currents. TP3 comes from reductive discharging. As with thiophene, the differences between the first and subsequent scans are due to the fact that only monomer is present on the first scan, while both monomer and dimer are present on subsequent scans. This hypothesis is consistent with studies of 2-(2'-thienyl)pyridine-based polymers synthesized by chemical polymerization that show TP-based polymers can store charge and also show a large separation between oxidation and reduction peaks.[ 1 8> 1 91 Thiophene-based oxidation and reduction peaks were observed at +0.4 V and -1.7 V vs. Ag/Ag + , with a doping level of 0.15 electrons per TP. Pyridine reduction Scheme 4-1 Numbering scheme for TP carbon atoms 3 . 3 4 81 and oxidation peaks occurred near -2.1 V vs. Ag/Ag + , and would not have been accessible in the potential range of the TP experiments herein. The loss of features observed at high pH might be due to the overlap of pH dependent and pH independent processes. Additional peaks observed in the C V on holding at potential positive of TP2 are likely due to the formation of a new species (a dimer). The new TP1 peak, appearing 0.1 V negative of the original, is associated with a decrease in the original TP1 peak. In basic pH, the new TP1 peak is observed to be much larger relative to the original TP1 peak. This may be because the smaller negative potential limit at neutral pH (-0.8 V , compared to -1.0 V in basic pH) does not allow for as much of the new product to be desorbed, and hence more of this product can be observed on the subsequent scan. TP4 at -0.9 V is likely caused by the desorption of the new species, while the broad features growing in on the positive scan between -1.0 V and -0.5 V may be due to the re-adsorption of the new species. If this is the case, it appears that the new species does not diffuse easily from the electrode. 4.7 Electrochemistry of TP on Polycrystalline Au in Organic Electrolyte Given the significant differences between thiophene electrochemistry in aqueous vs. organic solvents, particularly with respect to the much lower oxidation potential in aqueous solvents, it was of interest to observe whether TP would show similar behavioural trends. Organic electrochemistry experiments were performed using a significantly different set up from the aqueous electrochemistry experiments. The concentration of TP was chosen to be 15 m M (much higher than in aqueous experiments). This concentration was 82 within the range of concentrations typically used in thiophene experiments, allowing for ease of comparison with the latter. TP is significantly more soluble in M e C N than in H 2 0 . Bis(pentamethylcyclcopentadienyl)iron or decamethylferrocene (DMFc) was used as an internal potential reference, since the Ag reference electrode is only a quasi-reference. The amount of DMFc added varied from run to run, but was consistently < 3 mM. As with Au(l 11) in aqueous solution, within the double layer potential region, the C V of polycrystalline Au in M e C N is featureless, showing approximately ideally polarizable behaviour (Figure 4-14). The surface area of the Au bead electrode was not known, so current rather than current density is plotted vs. E . On addition of DMFc, a reversible set of oxidation and reduction peaks can be seen. These were assumed to be at E1/2 = -0.12 V vs. S C E l 2 ° l for DMFc, and were used in order to reference all potential values vs. SCE. Whether DMFc was added before or after the addition of TP had a small effect on TP electrochemistry. However, blocking of the DMFc signal in the presence of TP made it impossible to reference the potentials vs. S C E in experiments where D M F c was added after TP. The first scan in the presence of 15 m M TP without the addition of DMFc shows a small peak at approximately +1.2 V (in the absence of DMFc, it was impossible to properly reference the potentials), followed by a large increase in the current. In the alternate case, where TP was added to an electrolyte solution already containing DMFc, the C V is featureless negative of +1.3 V except for the D M F c redox peaks. At +1.3 V , there is a large increase in the current. On subsequent scans, the size of D M F c redox peaks is reduced (Figure 4-14). After holding the electrode at positive potentials (2 to 3 83 min at +1.5 to +1.6 V vs. SCE), these peaks disappear altogether. Apparently, the oxidation of TP results in an insulating layer that blocks the transfer of electrons between the electrode and DMFc in solution. In contrast, holding the electrode at +1.7 V vs. S C E in the presence of D M F c but in the absence of TP results in a large reduction peak on the reverse scan (presumably due to the reduction of Au oxide). On subsequent scans, the DMFc redox peaks are unchanged from those observed before holding the electrode at positive potentials. 1 > 1 ' 1 1 1 1 1 1 1 ' r --1.0 -0.5 0.0 0.5 1.0 1.5 2.0 E (V vs. SCE) Figure 4-14 CVs of a polycrystalline Au bead in the presence of 0.1 M tetrabutylammonium hexafluorophosphate/MeCN electrolyte only (blank); DMFc and electrolyte; and 15 mM TP, DMFc, and electrolyte. The scan rate was 100 mV/s. 84 PTTP is almost insoluble in MeCN. Therefore, it was added dissolved in a small amount of CH2CI2 before addition to the cell, resulting in a mixed solvent (-5% C H 2 C 1 2 by volume); the concentration was unknown, as PTTP is only sparingly soluble in CH 2 C1 2 also, but it was <1 mM. Generally, PTTP has little effect on the DMFc oxidation process, even after scanning to potentials as positive as +2.8 V vs. SCE. Generally, in organic solvents, there is no evidence for any coupling reactions taking place in the presence of PTTP. This is consistent with the aqueous results. In the presence of TP, the onset of an irreversible oxidation occurs at +1.25 V vs. S C E , significantly above the positive potential limit of the aqueous experiments. This is consistent with the literature on the aqueous vs. organic electrochemistry of thiophene. 4.8 Summary C V and capacitance measurements show that TP adsorbs to Au(l 11) at negative potentials in State I, likely in a ^-bonded, flat-lying configuration. Scanning positive, a transition associated with pH-independent pseudocapacitive peak T P 1 occurs. This is similar to a peak seen in pyridine and bipyridine corresponding to a two-dimensional phase transition from a flat-lying state to a vertical, N-bonded state. In the case of TP, this likely corresponds to a state bonded through both the nitrogen and the sulphur. At positive potentials, a second transition occurs, characterized by a pH-dependent faradaic peak, T P 2 similar to that seen in thiophene. It is believed to be due to an oxidative dimerization/charging process. This is discharged on the negative scan at T P 3 . On subsequent scans, peaks due to the dimer are visible. In organic solvents, consistent with aqueous results for thiophene, oxidation occurs at much more positive potentials. 85 4.9 References (1) Stolberg, L . ; Morin, S.; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1991, 307, 241-262. (2) Lipkowski, J.; Stolberg, L . In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P., Ed.; V C H , 1992, pp 171-238. (3) Chen, A. , Lipkowski, J. J. Phys. Chem. B 1999,103, 682-691. (4) Visy, C ; Kankare, J. Electrochim. Acta 2000,45, 1811-1820. (5) Yang, D.; Bizzotto D.; Lipkowski J.; Pettinger B.; S., M . J. Phys. Chem. 1994, 98, 7083-7089. (6) Weast, R. C ; Astle, M . J.; Beyer, W. H. , Ed., CRC Handbook of Chemistry; Series; C R C Press Inc.: Boca Raton, 1983-1984, pp E50-51. (7) Su, G.-J.; Zhang, H . - M . ; Wan, L. -J . ; Bai, C . - L . Surf. Sci. 2003, 531, L363-L368. (8) Matsuura, T.; Nakajima, M . ; Shimoyama, Y. Jpn. J. Appl. Phys. 2001, 40, 6945-6950. (9) Tourillon, G. In Handbook of Conducting Polymers Vol. 1; Skotheim, T. A. , Ed.; Marcel Dekker: New York, 1986, pp 301-350. (10) Heinze, J. In Organic Electrochemistry; 4th ed.; Lund, H . and Hammerich, O., Ed.; Marcel Dekker: New York, 2001, pp 1309-1339. (11) Heinze, J. In Electrochemistry IV; Steckhan, E . , Ed.; Springer-Verlag: Berlin, 1990; Vol. 152, pp 1-47. (12) Fujita, W.; Teramae, N.; Haraguchi, H. Chem. Lett. 1994, 511-514. (13) Bazzaoui, E . A.; Aeiyach, S.; Lacaze, P. C. J. Electroanal. Chem. 1994, 364, 63-69. 86 (14) Jin, S.; Xue, G. Macromolecules 1997, 30, 5753-5757. (15) Mu, S.; Park, S.-M. Synth. Met. 1995, 69, 309-312. (16) Hu, X. ; Wang, G.; Wong, T. K. S. Synth. Met. 1999,106, 145-150. (17) Wandlowski, T.; Holzle, M . H . Langmuir 1996,12, 6604-6615. (18) Yamamoto, T.; Zhou, Z. H.; Maruyama, T.; Kanbara, T. Synth. Met. 1993, 55, 1209-1213. (19) Zhou, Z. H.; Maruyama, T.; Kanbara, T.; Dceda, T.; Ichimura, K.; Yamamoto, T.; Tokuda, K. J. Chem. Soc, Chem. Commun. 1991, 1210-1212. (20) Setyawati, I. A.; Rettig, S. J.; Orvig, C. Can. J. Chem. 1999, 77, 2033-2038. 87 5 Characterization of Capacitive Changes in Adsorbed TP by Chronocoulometry 5.1 Introduction Chronocoulometry was employed in order to further characterize the adsorbed states (State I and II) of TP on Au(l 11) in the potential range where changes in adsorption were reversible and where faradaic processes were believed to be absent (< -0.250 V vs. SCE). Information about relative charge density on the metal, C M , was recorded as function of potential, E . Because changes in current density and capacitance should produce corresponding changes in charge density, these data reinforce and complement the C V and capacitance results. In addition, because chronocoulometry is performed on a different timescale, the dependence of the electrode response to the kinetics of potential-dependent processes at the electrode-solution interface is expected to be different. Based on thermodynamic relationships, much can be learned about non-faradaic processes using this data, including adsorption energetics, coverage and orientation of different adsorbed states. 5.2 Charge Density Plots of GM, vs. E were derived from chronocoulometric measurements in basic solution performed at values of Ej -1.10 V to -0.25 V at 0.025 V intervals, using E d e s =-1.175 V to-1.10 V and Ebase —-0.950 V. The positive limit of E; was chosen to minimize complications that might arise from the formation of a new species positive of peak TP2; this potential was definitely below the onset of TP2 for the highest concentrations. However, as discussed in the previous chapter, for the lower concentrations, the oxidation 88 occurred at more negative potentials, and therefore some oxidation near -0.250 V was possible for these lower concentrations. The charge-potential curves for six different concentrations of TP, as well as two curves for Au( l 11) in 0.05 M K C 1 0 4 (pH = 7, and p H = 11) in the absence of TP are I — 1 — I — 1 — i — 1 — i — 1 — i — ' — I — 1 — i — • — r ~ -1.2 -1.0 -0.8 -0.6 -0 .4 -0.2 0.0 0.2 E(Vvs. SCE) Figure 5-1 Relative charge density as a function of potential, E, for a Au(l 11) in the presence of various bulk concentrations of TP at in 0.05 M KC104 at pH 11, from chronocoulometry measurements, a = 0 is marked with a dashed line. Dotted lines represent extrapolations of stable phases. Blank measurements made in the absence of TP at pH 7 and pH 11 are also shown. Chronocoulometry was performed using a desorption potential, E d e s , of between-1.175 V and-1.1 V and a resting potential, E b a s e , of-0.950 V. 89 shown in Figure 5-1. The two curves for A u ( l l l ) in the absence of TP diverge from one another near -0.20 V due to adsorption of OH" in the presence of basic solution.^] Because OH" is negatively charged, when it is adsorbed, there is a negative shift in the pzc. In the presence of TP, all curves merge with the bare Au curves near -0.95 V, indicating that TP is completely desorbed negative of this potential. There is a region of changing slope from -0.950 V to -0.800 V that does not correspond to any distinct features in the C V and capacitance curves. However, the increase of this feature with TP concentration and its absence in the absence of TP suggest that it is due to TP adsorption. 1 2 0 J 1 0 0 A 8 0 - I E D : 6 0 4 0 2 0 0 derivative of charge density ac voltammetry •1.2 - 1 . 0 -0 .8 - 0 . 6 E (V vs. SCE) - 0 . 4 - 0 . 2 Figure 5-2 Capacitance plots of 1 mM TP from differentiation of the charge density curve (in Figure 5-1) and from ac voltammetry measurements. Both measurements were performed in 0.05 M KC10 4 at pH 11. 90 This can be more clearly seen by taking the derivative of the curves. The derivative corresponds to the zero frequency (because it is not obtained by ac voltammetry) capacitance. A comparison of the zero frequency capacitance for lxlO" 3 M TP is compared to that obtained by ac voltammetry in Figure 5-2. In theory, the two plots should be very similar. The capacitance plots of 1 x 10"3 M TP generated using A C voltammetry and by differentiating the charge density curve show good correspondence with respect to minimum capacitances, and both show a sharp peak near -0.45 V . However, the curve derived from the charge density plot shows a peak at -0.80 V that is completely absent in the A C voltammetry plot. In addition, the peak at -0.45 V is significantly larger in the charge density derivative. Due to the fact that the capacitance in A C voltammetry is calculated from the response of the electrode to a voltage perturbation, features due to nonfaradaic changes in the coverage of adsorbed species will show frequency and/or scan-rate dependence if they are kinetically slower than the frequency of the perturbation and/or the scan rate. This is because the measured capacitance at constant bulk concentration is the sum of two terms, namely!2! [dE)^ {dE)r IdrJ.W* ,rue aM ( 5 - D where Ctme is the true capacitance and C a d d is the pseudocapacitance, dependent on kinetics. When measured by ac voltammetry, dE = d(E sin ox) where cois the ac perturbation frequency. Because C a d d depends on the adsorbate coverage, T, if CO is too fast relative to the change producing the peak, equilibrium will not be established, and the Cadd term will become smaller. Therefore, it can be inferred that the differences between 91 the two curves are due to relatively slow kinetics for the processes corresponding to the peaks. That the adsorption process is particularly slow is consistent with the long waiting times required to reach equilibrium between adsorbed and solution species. The region of changing slope in the charge density plot corresponding to adsorption is followed by a region of constant slope. This corresponds to a stable phase of constant capacitance. An extrapolation of this slope gives a hypothetical case where it is stable throughout the potential range in which TP is adsorbed. Following this extrapolation to O M = 0 for lxlO" 3 M TP (shown with a dotted line) can give an estimate of the orientation of the adsorbed molecules based on their dipole moments, using EN = {4np/e)rtm (5-2) where E N is the shift in the pzc, f is the dielectric constant of the inner layer, r m a x is the maximum surface concentration of the adsorbate, and //is the component of the effective dipole moment perpendicular to the electrode surface. Ji = juA- njuw and juA refers to the effective dipole moment of the adsorbate, n refers to the number of water molecules displaced by one adsorbate molecule, and fiw is the dipole moment of a water molecule at the electrode. A surface-enhanced infrared absorption spectroscopy study of water on Au(l 11) by Ataka et al.f4i found that water can take on multiple orientations on the surface, and that the relative populations of each orientation changes with potential. At potentials below the pzc, an orientation with hydrogen atoms closer to the electrode is preferred. At potentials slightly above the pzc, an orientation with the oxygen atom closer to the electrode is preferred, and at very positive potentials, the hydrogens face directly into solution (completely away from the electrode) due to coadsorption with perchlorate anions. However, near the pzc, a flat-lying orientation with the plane of the water 92 molecules parallel to the surface is preferred. For such an orientation, there is no net dipole moment due to cancellation by the image charge. Therefore, the polarity of juA determines the effective dipole moment perpendicular to the electrode surface. An extrapolation of the slope for the part of the curve positive of -0.80 V shows that at the maximum surface concentration, the pzc differs very little from that of the bare electrode (in the absence of high [OH], since OH" would be excluded from the surface by the adsorbed layer), indicating that the component of the dipole moment perpendicular to the surface of the electrode for the adsorbate is small. This evidence suggests that initially, TP is adsorbed with the plane of the rings parallel to the surface, on average. The absence of significant changes in the capacitance upon adsorption is similar to what is observed for pyridine adsorbed in its initial flat-lying configuration. However, pyridine shows an adsorption peak in the capacitance curve obtained at the same ac perturbation frequency, indicating that its kinetics of adsorption are faster than that of TP (Figure 4-12). Very slow kinetics have been previously observed for thiophene adsorption on Au(l 1 l).t5-6] Because thiophene is more 7i-electron rich than pyridine, the initial interaction between the 7t orbitals of thiophene or TP with the negatively polarized Au surface may be less favourable than that of pyridine with negatively polarized Au. In addition, the lower solubility of TP may cause it to form aggregates and slow its transport to the electrode. A second region of changing slope is seen in the charge density plots between -0.50 V and -0.20 V when [TP] > 1 x l O ^ M . This grows steeper and shifts to more negative potentials with an increase in concentration. This corresponds to and is consistent with peak TP1 in the C V and capacitance plots. 93 At high concentrations, the steep part of the plot culminates in another region of constant slope corresponding to another stable phase. (At lower concentrations, peak TP2 is shifted to more negative potentials and its onset masks or inhibits the occurrence of the plateau.) For this phase, the shift in pzc, E N , is approximately -0.43 V , implying an average orientation with a larger component of the dipole moment perpendicular to the electrode, and with the more electronegative end (the N and probably the S, since S-Au interactions are known to be favourable) bonded to the gold surface. This suggests that, like pyridine, TP undergoes a phase transition at TP1 from an aromatic rc-bonded orientation to a more vertical orientation, bonded through the N and/or S atoms. Unlike for pyridine!7] and bipyridinej8] but similar to 5,6-dimethyluracill9] the steep section of the plot does not pass through C»M = 0. For 5,6-dimethyluracil, there is evidence that the molecule remains with the aromatic ring parallel to the Au surface subsequent to the phase transition, which does not involve a reorientation, as there is no shift in the pzc and the capacitance remains quite high (16 uF/cm 2). In the case of TP, the evidence for a reorientation suggests that the charge on the metal and the dipole of the adsorbate are not as closely interdependent as they are with pyridine and bipyridine. The shift in pzc of -0.43 V for TP is also modest compared to that seen in pyridine on Au(l 11) of -0.86 V . This is likely due in part to the much smaller dipole moment of the thiophene moiety relative to pyridine. The gas phase electric dipole moments of these molecules are shown in Table 5-1. These only give a rough guideline for the observed electric dipole moments on the electrode, as the tilt angle and degree of organization of the molecules on the electrode can have a significant effect on the average effective dipole moment normal to the plane of the electrode surface. In addition, in the absence of 94 Table 5-1 Selected values of electric dipole moments for molecules in the gas phase Molecule Moment (debyes) W a t e r f 1.85 ± 1% Pyridinef10] 2.19 ± 2% Thiophenet10] 0.55 ± 5% TP(syn, 30° twist)!11! 4.815 TP (anti, 150° twist) Hi] 1.659 TP, the electrode is covered in water. Therefore, the dipole moment relative to water, rather than the absolute dipole moment, should be considered, and the change in the charge density on the electrode will be strongly dependent on the number of water molecules displaced by the adsorbate. Thiophene has a smaller gas phase electric dipole moment than even water, and thiophene ring will occupy half the space taken up by TP on the electrode, resulting in a much lower charge density than if the electrode were covered in pyridine rings only. The magnitude of EN, considering the dipole moment of gas phase TP in the syn conformation, may suggest that the average orientation of TP in this phase is tilted, with the component of the dipole moment perpendicular to the surface being relatively smaller, or that the two heteroatoms are not oriented syn to one another. While charge density was generally not measured positive of -0.250 V for the reasons explained earlier, a plot of charge density for 1 x 10~3 M TP to +0.200 V is shown in Figure 5-3 for illustrative purposes. Two sets of measurements were performed - the first to -0.250 V and the second from -0.250 V to +0.200 V (thus the -0.250 V point was 95 measured twice). C V and capacitance were run between the two chronocoulometry measurements. As can be seen, the two measurements at E = -0.250 V give two different values of C M -Typically, it was found that reproducibility was poor at E > -0.250 V. This may be due to the variably incomplete desorption and readsorption of the product formed here on successive steps. The slope increases steeply positive of -0.250 V , corresponding to peak bare Au, pH 7 bare Au, pH 10.7 1—•— i— •— i— •— i— 1 i 1 — i — < — i — 1 — i — > — i — > — i — ' — i — 1 — i -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E (V VS. SCE) Figure 5-3 Relative charge density for Au(l 11) in the presence of 1 mM TP to a positive potential limit of +200 mV in 0.05 M KC104 at pH 11, from chronocoulometry measurements. Blank measurements made in the absence of TP at pH 7 and pH 11 are also shown. Chronocoulometry was performed using Edes= -1.175 V and Eb a s e= -0.950 V. 96 TP2 in the C V and capacitance, and does not begin to flatten out until near its intersection with the bare Au curve. Interestingly, the charge consumed in this region is comparable to the amount consumed by the reorientation, despite the fact that the integrated charge beneath the C V peaks is significantly larger. A significant difference between the CVs and the chronocoulometry steps is that in the CVs, the adsorbed TP is always in State II prior to TP2, whereas in the step experiments, the potential is stepped directly to the potentials around TP2 from Eb a Se, which is -0.950 V. It is possible that this results in a lower coverage of the oxidation product, since it is not pre-organized as a close-packed layer. 5.3 Film Pressure and Surface Pressure Recall from section 2.1.2 that the influence of the potential, E , on the surface tension, y, at the metal-solution interface is described by the electrocapillary equation (2.3). The effect of an adsorbate on y is often expressed as film pressure, 7t, the lowering of the surface tension y due to the adsorption of the surfactant is given byt 3 - 1 2 ! where 0 is the surface coverage and the surface tension at a given potential, y(E), is calculated according tot 3' 1 2! Since the integration constant is the surface tension at the desorption potential, which is identical in the presence and absence of the adsorbate, it cancels on subtraction in Eq. (5.3). A plot of 7C vs. E is shown in Figure 5-4. (5.3) (5.4) 97 5 0 1 4 5 -4 0 -3 5 -3 0 -f 2 5 -l 2 0 1 * 1 5 -1 0 -5 -0 -1 X 1 0 " 6 M TP 1 X 1 0 " 6 M TP - ^ 3 x 1 0 ~ a M T P - r - 1 x 1 0 M TP 3 x 1 0 ^ M TP 1 x 1 0 M TP • 1 . 2 - 1 . 0 - 1 -- 0 . 8 - 0 . 6 E ( V vs. S C E ) - 0 . 4 - 0 . 2 Figure 5-4 Film pressure vs. potential on Au(l 11) in the presence of various bulk concentrations of TP, derived from charge density measurements in 0.05 M KC10 4 at pH 11. A n alternative to film pressure for measuring surface tension as a function of potential or concentration is surface pressure, <j>. In this case, charge, rather than potential, is defined as the independent variable through the use of Parson's function, where ^ = / + o-ME (5.5) Since y is calculated from O M V S . E , O M is known for a given value of y. Therefore, y was plotted as a function of O M and intermediate values were determined by linear interpolation between known values. In this way, % could be determined for a series of TP 98 concentrations, as well as for bare supporting electrolyte, at common values of Cm-Therefore, tf> at a given concentration, c, can be determined from the formula 0 = £[77>1=0 ~ %[TP}=c (5.6) A plot of ^ as a function of O"M is shown in Figure 5-5. As can be seen, 7t increases with both E and [TP]. Positive of the adsorption potential, for at least the three highest concentrations, there is a region with a relatively small positive slope, followed by a region with a steeper slope beginning near -0.40 V . These two regions correspond to State I and State II respectively in the C V and capacitance plots. The maximum fdm pressure of 43 mN/mis significantly lower than the 120 mN/m seen in pyridine.!7] This suggests that the TP layer is less closely packed, with weaker lateral interactions. The different sizes of the thiophene and pyridine halves of the molecules (due to the larger van der Waals radius of S and the larger quadrupole moment normal to the plane of the ring for thiophene)!14] might make packing more complicated, and there is no guarantee that the two rings are co-planar. The 7t vs. E and 0vs. rj electrocapillary curves are expected to be approximately bell-shaped. In the case of TP, the two regions corresponding to two different slopes in the 7t vs. E curves both appear to flatten out towards their positive limits. In the case with the smaller slope (E < -0.4V), if the positive half of the curve exists it is obscured by the curve producing the steeper slope (E > -0.4 V). In the case of the curve producing the steeper slope, overlap with the faradaic reaction begins at positive potentials, obscuring the positive half of that curve also. The shape of two overlapping bells is more clearly 99 seen for the 0 vs. a M curves due to the different scaling of the two regions (State I and II) of the plot. Whereas in the case of pyridine, the values of ^ and n are similar near the phase transition, for TP, 0is significantly larger than 7C. This is mostly because the phase transition does not occur near the pzc in TP, such that the o M E component of the Parson's coefficient (5.5), which is not a component of 7t, is relatively much larger compared to the y component, which is common to both n and (ft. 80-70-60-50-Is 40--z. -§ 30-1 20 10 0-I - • -1x10" b MTP • 1x10"5MTP 3x10"5MTP ^ I x l O ^ M T P -#-3x10" 4 MTP 1x10"3MTP • / •¥ mM •'• ** u • " -30 -25 -20 -15 -10 Charge density, a M (pC/cm ) 0 Figure 5-5 Surface pressure as a function of charge density for various concentrations of TP, derived from charge density measurements in 0.05 M KC10 4 at pH 11. 100 Surface excess, T, can be determined from either n or <p. T can be calculated from y as a function of the chemical potential of the adsorbate, u.A, based on the electrocapillary equation: r = -dy (5.7) Assuming a coverage of a monolayer only (no multilayers), Y will equal the surface coverage. Given that JU = JU°+RT In a ( 5 . 8 ) where a is the activity of TP. Since a = X, where X is the mole fraction of TP; and X is proportional to c, where c is the concentration of TP, d In a = d In c ^59^ Therefore, dju = RTd In c ^5 where R is the ideal gas constant and T is the temperature, assumed to be 295 K. This gives r = (s.ii) RTd In c dy Given the relation between jr. and y in (5.3) and the fact that 8 = 0 = 0, dine r dn F = P 7 V 1 (5.12) RTd In c A plot of T vs. In c is shown in Figure 5-6. 101 - -0.850 V - • -0.750 V -0.650 V - • - - 0 . 5 5 0 V In c Figure 5-6 Surface coverage vs. In [TP] from differentiation of film pressure curves. Similarly, T can also be determined by differentiating a plot of 0 v s . In c. Based on differentiation of % with respect to In c, the maximum T observed at a bulk concentration of 1 x 10"3 M TP at the negative plateau, near -0.70 V , corresponding to the flat-lying orientation, is 7 (± 5) x IO"11 mol/cm2. At -0.25 V, the maximum coverage is approximately 5 (±1) x 10"10 mol/cm2 by this method. Differentiating 0 v s . In c curves gives a similar value of 6 ( ± l ) x 10"10 mol/cm2. The theoretical area for TP 7t-bonded to the surface is 60 A2/molecule, giving a maximum coverage of 2.8 x 10"10 mol/cm 2 (calculated as for thiophene in section 4.5), while the theoretical maximum coverage for 102 7-6-5-CNT^ 4 E o | 3H o o 2-1 -0-A T • < -28 uC/cm' -25 pC/cm2 -22 uC/cm2 -19 pC/cm2 -16 pC/cm2 -13 pC/cm2 -10 pC/cm2 -7 uC/cm2 -13 -12 -11 -10 In c -9 -8 L , - * A i» • —*•— ^ o — i 1 | • —I -7 Figure 5-7 Surface coverage of TP from differentiation of surface pressure vs. In [TP] (smoothed). TP bonded with the rings normal to the surface is 5.2 x 10" mol/cm , based on an area of 32 A2/molecule. It can be seen that the errors are quite large in Figure 5-6 and Figure 5-7, and therefore the values are approximate. The errors in the chronocoulometry experiments for TP are larger than typically seen in the literature for two reasons. First, the limited solubility of TP restricted the range of available concentrations on which chronocoulometry could be performed. Secondly, the potential range was also restricted due to overlap with the oxidative process at TP2. This resulted in very few data points at 103 high values of o . However, the values from both methods agree with one another, and are close to the theoretical maximum coverage. The literature values of surface coverage for the similar-sized molecule 2,2-bipyridinet8] are 2 x 10"10mol/cm2 and 5 x 10"10mol/cm2 respectively for flat-lying, 7t-bonded and vertical TV-bonded orientations. While TP shows a similar coverage in its close-packed phase, its surface coverage for the aromatic 7t-bonded phase is comparatively low, suggesting that much of the electrode remains water-covered. The surface coverage of TP at -0.25 V is similar to the surface coverage of 4.6 (±0.4) x 10"10 mol/cm 2 calculated by assuming peak TP3 in the CV is due to a one electron process, integrating underneath it at equilibrium, and accounting for the capacitive charge.!15] This consistency suggests that the coverage likely remains the same after the transition at peak TP2 and during the reverse scan, where the capacitance remains constant (although the value assumes one electron per TP molecule, which may be an overestimate of charging and therefore an underestimate of surface coverage). As the slope is non-zero at the maximum value of T determined from the K vs. In c curves, it is possible that the actual maximum Gibbs excess or surface concentration is slightly higher. However, due to the poor solubility of TP in H2O, it was impossible to use a higher bulk concentration, and using a more positive potential limit resulted in overlap with the faradaic reaction. In the case of pyridine and bipyridine, almost all bulk concentrations reached r m a x , with lower bulk concentrations reaching r m a x at more positive potentials than higher bulk concentrations.!7<8] However, in the case of TP, only the maximum bulk concentration reached r m a x . This is because for lower bulk concentrations, the transition required to reach the state with this maximum concentration 104 occurs at more positive potentials, while the faradaic reaction occurs at more negative potentials (Figure 4-4). Therefore, State II, corresponding to this maximum concentration, is stable only at very high concentrations, close to the limit of TP solubility in aqueous solvents. 5.4 Adsorption Isotherm and Gibbs Energy of Adsorption The Gibbs energy of adsorption, AG ad S , was calculated by two different methods. The first involved extrapolating n vs. In c at r m a x to 7t = 0J81 This method assumes a Henry's isotherm, where there are no interactions between adsorbates such that Eq. (2.7) holds. When T = r m a x , P = 1/c. Substituting in Eq. (2.6) then gives A G = RT In c ( 5 1 3 ) However, it should be noted that the assumption of no interaction between adsorbates would not be expected to hold at r m a x . Therefore, the AG a d S calculated by this method includes the additional energy of intermolecular interactions between adsorbed molecules, and is often described as AGe= i. In the case of aromatic molecules, which are known to show attractive K-K interactions, this method should produce an overestimate of AG a d S (however, an underestimate AGads would result from a case where intermolecular interactions were repulsive). The second method for determining AGads was by fitting the ^data to a virial adsorption isotherm (Table 5-2) using Microcal Origin 6.O.1'31 The latter was achieved by superimposing the curves by shifting the <)> vs. In c curves parallel to the In c axis (using a correction factor, f(c); described in more detail in the Appendix) and fitting to the theoretical curve generated by substituting the surface equation of state for the virial 105 Table 5-2 Surface pressure equations for various adsorption isotherms!*-*] Isotherm F ( $ Virial ln^ [l + 4B0/RT]2-l 2 5 Henry \n(0/RT) Volmer ln(0/RT) + b<plRT + (\ + AB(j)IRT)2 -1 The above surface pressure equations give the relationship between surface pressure, 0, and the standard free energy of adsorption at zero coverage, AG, where, where F(0) = In c - AG/kT, and c is the bulk concentration or activity of the adsorbate. B and b are intermolecular interaction coefficients. adsorption model into the Gibbs adsorption isotherm (Figure 5-8) at a temperature of 295 K. The reduced chi-square value, %2/d.f., for this fit was 0.38, with the virial coefficient optimized at B = 1.8, corresponding to 92 A2/molecule. B is a measure of intermolecular interactions. A positive value indicates repulsive electrostatic forces between molecules, and B is predicted to be twice the molecular area for a rigid, circular molecule. The small value of B relative to the size of TP (for comparison, B = 120 A2/molecule for thiourea, a much smaller molecule, adsorbed on Hg)'"1 indicates that repulsive interactions are weak. It should be noted that this method favours data at lower concentration and charge values, and should give a value closer to the zero coverage AGads- Fitting to Henry and Volmer isotherms (Table 5-2) gave poorer fits, with %2/d.f. values of 7.9 and 0.99 respectively (Figure 5-8). Calculating AGads by extrapolating n vs. In c and fitting to a virial adsorption isotherm gave similar values of -26 (± 5) kJ/mol and -31 (±5) kJ/mol respectively. It is unexpected that the second method should give a higher value, as the former value describes mainly the close-packed orientation; any attractive intermolecular interactions would be included 106 c 4A 3A 2A 1 H 0 data virial fit Volmer fit Henry's fit T — i — • — i — • — i — > — i — ' — [ — •16 - 1 4 - 1 2 - 1 0 -8 -6 i 1 r - 2 0 In [TP] + f(o) Figure 5-8 Fit of the natural log of surface pressure vs. In [TP] + f(o) to various adsorption isotherms. Curves for various bulk concentrations have been shifted along the x-axis by f(o) for maximum overlap, as in Parsons.!13] in this value, and would make it more negative. The latter value describes mainly the more diffuse flat-lying state. Since significant repulsive interactions between the molecules are not expected (and are not indicated by the low value of B in the virial fit), the higher value for the second method may be due to error. Nevertheless, it should be emphasized that values from both methods are similar, and compare favourably to literature values for similar molecules. AGads has been calculated as -37 kJ mol"1 for pyridine on Au(l 11);[71 -30 and -50 kJ/mol for flat-lying and vertical orientations 107 respectively of 2,2'-bipyridine on Au(l 1 l)t 8l; and -22 kJ/mol for thiophene.!5! In the case of pyridine and 2,2'-bipyridine at low coverages (in a flat-lying orientation), AGads was calculated from the initial slopes of plots of K vs. In c, assuming a Henry's isotherm. 131 It was not possible to use this method for calculating AG a d S due to the limited range of concentrations and 7t values for this orientation due to the low solubility of TP. The AG a d S for the flat-lying orientation of 2,2'-bipyridine was also calculated from a fit to a Frumkin isotherm, while the AG ad S for the vertical, close-packed orientation was calculated as for TP, by extrapolating the linear segment of the % vs. In c plot to % = 0. The AGads for thiophene was calculated in terms of kT by fitting ER data to a Langmuir isotherm. 151 Therefore, it should be noted that because the method of calculating AG a d S differs between the study herein and the studies described of pyridine, 2,2'-bipyridine and thiophene, and among those studies themselves; they are not measures of exactly the same quantity in all cases and precise, direct comparisons cannot be made. Nevertheless, the various AG a d S values should be on a similar order of magnitude, and these numbers suggest that the interaction of TP with gold is comparable to that of similar species. However close-packed TP molecules bonded through the heteroatoms do not appear to exhibit the moderately attractive aromatic n-% interactions that are observed for close-packed, N-bonded 2,2'-bipyridine. 5.5 Summary Results from the chronocoulometry experiments on Au(l 11) in the presence of various concentrations of TP complemented those obtained by C V and A C voltammetry. From -0.950 to -0.800 V , a feature was visible in the charge density plots corresponding to the 108 adsorption of TP, which was too kinetically slow to be observed by C V or ac voltammetry. Features corresponding to State I, peak TP1, State II, and peak TP2 were also observed in the charge density, fdm pressure and surface pressure plots. A n analysis of chronocoulometry data confirmed that State I is a low-coverage phase adsorbed with the plane of the aromatic rings parallel to the gold surface, with a surface coverage of 7 x 10"" mol/cm2; and that State II is a more vertical close-packed and N and/or S-bonded state with a surface coverage of 5 x 10"10 mol/cm2. The Gibbs energy of adsorption of the close-packed state of TP was slightly lower than the full coverage A G a c i s observed for the close-packed state of 2,2'-bipyridine, suggesting that TP molecules do not exhibit the same moderately attractive intermolecular interactions. 5.6 References (1) Chen, A. , Lipkowski, J. J. Phys. Chem. B 1999,103, 682-691. (2) Bizzotto, D. Ph.D. Thesis, University of Guelph, 1996, p 20. (3) Stolberg, L . ; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1987,238, 333-353. (4) Ataka, K.- i . ; Yotsuyanagi, T.; Osawa, M . J. Phys. Chem. 1996,100, 10664-10672. (5) Matsuura, T.; Nakajima, M . ; Shimoyama, Y . Jpn. J. Appl. Phys. 2001, 40, 6945-6950. (6) Su, G.-J.; Zhang, H . -M. ; Wan, L. -J . ; Bai, C . - L . Surf. Sci. 2003, 531, L363-L368. (7) Stolberg, L . ; Morin, S.; Lipkowski, J.; Irish, D. E . J. Electroanal. Chem. 1991, 307, 241-262. 109 (8) Yang, D.; Bizzotto D.; Lipkowski J.; Pettinger B.; S., M . J. Phys. Chem. 1994, 98, 7083-7089. (9) Wandlowski, T.; Holzle, M . H. Langmuir 1996,12, 6604-6615. (10) Nelson, R. D. J.; Lide, D. R., Jr.,; Maryott, A . A . In CRC Handbook of Chemistry and Physics, 64th Edition; Weast, R. C , Ed.; C R C Press: Boca Raton, 1983, pp E-58. (11) Knop, J. V. ; Trinajstic, N.; Milun, M . ; Pejakovic, S. Rev. Roum. Chim. 1978, 23, 103-110. (12) Lipkowski, J.; Stolberg, L . In Adsorption of Molecules at Metal Electrodes; Lipkowski, J. and Ross, P., Ed.; V C H , 1992, pp 171-238. (13) Parsons, R. Proc. Roy. Soc. (London) 1961, A261, 79-90. (14) Flygare, W. H.; Benson, R. C. Molecular Physics 1971, 20, 225-250. (15) Chung, E . ; Bizzotto, D.; Wolf, M . O. Chem. Commun. 2002, 3026-3027. 110 6 Characterization of Faradaic Processes of Adsorbed TP by In Situ Fluorescence and Comparison to PTTP 6.1 Introduction As discussed in previous chapters, based on the electrochemical characterization of TP, peaks TP2 and TP3 in the C V (Figure 4-5) are due to oxidation and reduction processes. These were hypothesized to be dimerization and charging, based on the known electrochemistry of thiophene. In addition, electrochemical dimerization of thiophene derivatives adsorbed on Au has been previously reported.^ In the latter report, 6-(2,2':5',2"-terthien-5-yl)-hexanethiol on Au(l 11) was proposed to undergo intermolecular p\P'-coupling within the monolayer, as oc,cc'-coupling was not favoured by the proposed orientation of the adsorbed species. The occurrence of an electrochemical dimerization was supported by solution fluorescence measurements after the dissolution of the Au in N a C N / D M F / H 2 0 at 60°C. Since oc,a-coupling should be possible for adsorbed TP and is known to be chemically favoured, the oc,a-coupled TP dimer, 5,5'-Bis(2-pyridyl)-2,2'-bithienyl (PTTP) was independently chemically synthesized in order to compare its properties with those of TP in State HI and IV. The electrochemistry of PTTP was characterized by C V and capacitance. The photophysical properties of both TP and PTTP were compared in solution and in the solid state, and then used to interpret results of in situ spectroelectrochemical experiments under potential control. Changes in fluorescence spectra were measured in situ, complemented by data regarding the spatial distribution of fluorescence emission by fluorescence imaging. I l l 6.2 Electrochemistry of PTTP Cyclic voltammetry and capacitance measurements of adsorbed PTTP at pH 7 and pH 11 are shown in Figure 6-1. The plots show a slight decrease in the capacitance indicative of adsorption near -0.80 V vs. SCE. A single set of peaks is observed at -0.48 V and the anodic and cathodic peaks are separated by 0.08 V in the C V . The peaks appear near -0.46 V in the capacitance plots and show a smaller separation, implying slow kinetics. The relative currents of the C V Figure 6-1 CVs and capacitance plots of PTTP adsorbed on Au(l 11) from the air/solution interface at pH 7(a and c) and pH 11 (b and d) in 0.05 M KC10 4. Blank curves in the absence of PTTP are shown with a dotted line. CVs were performed at a scan rate of 20 mV/s. Capacitance measurements were performed at 5 mV/s, 25 Hz, 5 mV rms. 112 peaks, as well as the degree of hysteresis between the forward and reverse peaks of the capacitance were not very reproducible, especially in neutral solution, and appeared to be related to the amount of material on the electrode. The peaks observed could be due to adsorption, reorientation or charging. The absence of any pH dependence suggests that PTTP does not undergo further oligomerization within this potential range. At no potential was the minimum capacitance below 12 uF/cm , in contrast to a minimum capacitance of 7 uF/cm for a similar capacitance experiment in 1 m M TP solution, suggesting that the adsorbed PTTP layer formed by this method is relatively diffuse and/or disorganized. 6.3 Solution and Solid State Photophysical Characterization of TP and PTTP TP and PTTP are both fluorescent but have very different, characteristic emission spectra. Solution fluorescence and excitation spectra of TP and PTTP dissolved in dichloromethane are shown in Figure 6-2. (Spectra were obtained in CH2CI2 or CHCI3 because PTTP is poorly soluble in H 2 0.) TP shows a UV/Vis absorption maximum at 303 nm in solution, with a smaller peak at 263 nm. This agrees well with literature reports of absorption maxima at 302 nm and 263 nm.! 2! A single fluorescence emission maximum for TP was observed at 357 nm, close to the maximum of 362 nm reported in the literature.!3] Monitoring at 362 nm gave an excitation maximum of 311 nm. The excitation spectrum was almost identical to the absorption spectrum, but was slightly broadened and red-shifted. The absorption and emission spectra for TP are almost identical to those of thiophene!4! (km^ = 303 nm, X e m = 362 nm) and occur at longer wavelengths than those of 2,2-bipyridine!5! (A™ax = 282 nm, 235 nm, 7^m = 328 nm). 113 a) X (nm) Figure 6-2 Normalized fluorescence excitation and emission; and UV/Vis absorption spectra of a) 7.1 x 10"5 M TP and b) 2.2 x 10"5 M PTTP in CH 2C1 2. For PTTP, the fluorescence emission was measured with an excitation wavelength of 387 nm and the fluorescence excitation was monitored at 433 nm. For TP, the fluorescence emission was measured with an excitation wavelength of 311 nm and the excitation was monitored at 357 nm. PTTP excitation and emission occur at longer wavelengths than those of TP due to the longer conjugation length of PTTP. PTTP has an absorbance maximum at 389 nm, and it shows two emission maxima at 435 nm and 460 nm, with shoulders at 500 nm and 530 nm. Monitoring at any of these wavelengths gives an excitation maximum of 390 nm, 114 almost identical to the absorbance maximum. While the absorbance of PTTP is almost identical to that of terthiophene (kmax = 391 nm), its emission occurs at longer wavelengths than that of terthiophenef4] (Xem = 407 nm, 426 nm). This is consistent with studies of 2-(2'-thienyl)pyridine-based polymers, which show absorbance and emission spectra that are red-shifted relative to both polythiophene and polypyridineJ6-7] The red shift in absorption and emission for PTTP relative to TP is consistent with the longer conjugation length of PTTP. There is a much greater degree of overlap between T3 CD _N "56 E k_ o cu o c CO xi a_ o CO XI < solid state absorbance solution absorbance -o a> « E o CD O c CC X) o CO X) < solid state absorbance solution absorbance A. (nm) Figure 6-3 A comparison of solution and solid-state absorption spectra of TP and PTTP. 115 the absorbance and emission spectra for PTTP, and it is prone to self-absorption (the inner filter effect) at high concentrations. The vibrational fine structure seen in the emission spectrum of PTTP is typical of thiophene oligomers and is attributed to coupling of the electronic transition with a vibrational transition involving a C=C ring stretch.!8! Its presence suggests that the excited state is more planar and rigid than the ground state. I9! While solution absorbance spectra for PTTP were structureless, solid state UV/Vis spectra of TP and PTTP dropcast from dichloromethane onto quartz slides (Figure 6-3) were structured, with peaks at 382 nm, 399 nm and 430 nm. For TP, only a single broad peak was observed with a maximum near 280 nm. However, the absorbance bands for both TP and PTTP are broadened relative to the solution phase. As mentioned earlier, solution fluorescence spectra of TP and PTTP were originally measured in CHCI3 or CH2CI2. The addition of MeOH to increase solvent polarity had little effect on either the intensity or the spectrum of PTTP (Figure 6-4). No red shift was observed to indicate the occurrence of 7i-stacking or other forms of aggregation. In fact, there was a slight blue shift in both the excitation and the emission spectra (Table 6-1). Increasing the solvent polarity resulted in a small red shift in the emission maximum for TP (see Table 6-2). However, the fluorescence intensity for the same concentration of TP more than doubled from a solution that was mostly CH2CI2 to one that was mostly MeOH. This suggests that the excited state is stabilized in a polar environment. A similar decrease in £ and red shift in non-polar solvents is seen for pyrazine derivatives of bithiophenesJ10! 116 a) 1000-=3 m 800-w "c no 600-o 0 o 400-c _ o 200-o -0-200 b) 400 - i 3 W C 3 O O CD O CD O Crt CD 300-^  200 H 100-^  0 4 200 TP excitation in CH 2 CI 2 emission in CH 2 CI 2 - excitation in MeOH emission in MeOH excitation in H 2 0 emission in H 2 0 — i — 300 400 — I — 500 600 700 X (nm) m  p T T p i" \ excitation in CH CL \ 2 2 s emission in CH 2 CI 2 \ excitation in MeOH * emission in MeOH v i * \ s — I — 300 400 500 X (nm) 600 700 Figure 6-4 Solvent effects on the excitation (solid line) and emission (dotted line) of a) 1.5 x 10"5 M TP at excitation and emission maxima of 303 nm and 360 nm respectively and b) 7.6 x 10"6 PTTP at excitation and emission maxima of 390 nm and 460 nm respectively. Table 6-1 Solvent effects on the wavelengths of PTTP excitation and emission [PTTP] (solvent) excitation maximum (nm) emission maximum (nm) 7 x 10"5 M(CH 2 C1 2 ) 388 434, 459 7 x If/ 5 M (MeOH) 385 433, 457 117 Table 6-2 Solvent effects on the wavelength of the TP emission maximum [TP] (solvent) Emission maximum (nm) 1.5xlO"5M (CH 2C1 2) 357 1 .5x l0 5 M(MeOH) 360 1 . 5 x l 0 5 M ( H 2 O ) 365 l.OxlO"3 M (0.05 M KCIO4 /IXIO' 3 M NaOH/H 2 0) 369 6.4 In Situ Fluorescence Spectroscopy of TP and PTTP In order to compare State TV of TP with PTTP, fluorescence spectra of the adsorbed monolayers of both species were measured as a function of potential by in situ fluorescence. The electrochemical cell was positioned above the microscope objective, which focused the excitation light (340 nm to 380 nm) onto the working electrode from below. Light reflected and emitted from the electrode and the solution passed back through the objective, and the dichroic and emission filters allowed wavelengths > 430 nm to be detected by the spectrometer. Al l spectra shown in this section are background-subtracted spectra collected upon stepping to the desorption potential, E d e s , after holding at a given potential of interest, Ej, for a given length of time, typically 15 minutes. As described in Section 3.4.2, this background subtracted spectrum, AI, consists of the spectrum collected at E, (which is made up of background light only, as the fluorescence of adsorbed species is quenched when it is in contact with the metal electrode) subtracted from the spectrum at E d e s (made up of both background light and fluorescence from the species desorbed at E d e s that had 118 been previously adsorbed at Ej,). Therefore, AI is a measure of the fluorescence of the adorbate (adsorbed at a given value of Ej) upon its desorption at Ed e s-Control experiments in the absence of TP or PTTP gave AI = 0 at all potentials, indicating that light leakage and light scattering through the filters are constant. In situ fluorescence was performed on a layer of PTTP adsorbed on Au(l 11) from the air-solution interface as in section 6.2. The electrode was held at each value of Ej for 15 minutes before stepping to desorption. Plots of the spectra obtained at each potential are shown in Figure 6-5. 60 40 3 3, * 20 • raw data FFT filtered -0.800 V - 1 ' 1 ' 1 ' r --0.200 V -I ' 1 • 1 ' 1 ' 1 1 T-60 40 3 03, <i 20 0 J 0.600 V 0.000 V 400 450 500 550 600 650 700 I (nm) 450 500 550 600 650 700 X (nm) Figure 6-5 Background-subtracted fluorescence, AI, measured at E d e s , for a layer of PTTP previously adsorbed on Au(l 11) at various potentials, E i ( for 15 minutes. E d e s was -1.00 V and AI = I(E d e s) - I(Ej). A l l values of E are vs. SCE. 119 Upon stepping to E d e s , the background-subtracted fluorescence, AI, is significant after holding at the negative potential value of Ej = -0.800 V , suggesting that considerable adsorption takes place even at this potential, consistent with a noticeable depression in the capacitance at this potential relative to the bare electrode (Figure 6-1). As Ej becomes more positive, the background-subtracted fluorescence intensity upon desorption increases, up to a value of Ej = -0.600V. At Ej > -0.600 V , AI remains relatively constant. This is consistent with the single set of peaks in the C V and capacitance plots, the onset of which is near -0.600 V . Positive of this peak, the capacitance is relatively constant. The relatively long time scale used in the in situ fluorescence experiment (15 minutes) was designed to give enough time for this process to be complete by the time the potential is stepped to Ed e s , despite the fact that -0.600 V is negative of the peak maximum. In situ fluorescence was also performed in the presence of 1 m M TP (Figure 6-6). Under these conditions, after holding at Ej < -0.600 V , upon desorption there is negligible fluorescence after background subtraction. Beginning at Ei = -0.400 V, fluorescence characteristic of TP can be seen in the background-subtracted spectra upon desorption, and the fluorescence intensity increases for more positive values of Ei, reaching a maximum at -0.250 V . This is consistent with electrochemical data showing a maximum in the surface coverage and a minimum in the capacitance are reached at this potential. Generally, there is a rough correlation between AI and the surface coverage derived from chronocoulometry, T. However, AI is influenced by factors such as distance from the electrode, concentration of the local solution relative to the surface concentration, and shifts in the spectrum of the adsorbate, which may be affected by the 120 600 500 400 300 200 100 0 500 400 300-200 100 0 500-400 g - 300^ 3-< 2°<H 100 0 500H 400 g 300 ni, < 200-j 100 0 - i 1 r-• \ -0.800 V, 15 min. -0.600 V, 15 min. - I 1 j • r— -0.400 V, 15 min. -i ' 1 ' r 1 r— -0.250 V, 15 min. 400 450 500 550 600 650 700 \ I 1 J--0.200 V, 15 min. - l 1 1 1 i ' | 0 V, 15 min. 0.200 V, 15 min. ~I ' 1 ' 1 ' 1 ' 1 » 1 0.200 V, 30 min. 450 500 550 600 650 700 750 X (nm) Mnm) Figure 6-6 Background-subtracted fluorescence, AI, measured at the desorption potential, E d e s , in the presence of 1 mM TP. TP had been previously adsorbed on Au(l 11) at various potentials, Ej, for 15 minutes. E d e s was -1.00 V and AI = I(E d e s) - I(E|). Al l values of E are vs. SCE. 121 Figure 6-7 Changes in background-subtracted fluorescence intensity, AI, measured at the desorption potential, E d e s , as a function of time spent at E ( = +0.200 V vs. SCE. E d e s was -1.00 V and AI = I(E d e s) - I(Ei). 122 arrangement/packing of adsorbed molecules on the surface and the changes in their arrangement/packing after desorption. For Ej > -0.250 V , TP background-subtracted fluorescence upon desorption decreases, suggesting a loss of TP at the onset of dimerization. After holding for as little as 2 minutes at +0.200 V , a spectrum characteristic of PTTP can be seen upon desorption, and some of the vibronic structure can be distinguished as the time at Ej increases (Figure 6-7). It should be noted that this spectrum is not actually the spectrum of the species at +0.200 V , but of the desorbed species, which has undergone a reduction near -0.700 V (peak TP3 in the CV) prior to desorption at -1.000 V . This will be discussed further below. The spectrum of the species desorbed after a layer of TP was held at Ej = +0.200 V and the spectrum of chemically synthesized PTTP deposited on the electrode from the air/solution interface are extremely similar to that of chemically synthesized PTTP in chloroform solution (Figure 6-8). This strongly suggests that State IV is the product of electrochemical a,cx-coupling between the thiophene moieties of adsorbed TP. While (3,P-coupling, as proposed to occur within monolayers of 6-(2,2':5',2"-terthien-5yl)-hexanethiol, is theoretically possible,!1! it is not favoured sterically or chemically in the case of TP. And while Michalitsch et alJ 1 ! suggest that P,P-coupling is compatible with 7t-delocalization, they note that this is not well recognized in the literature. It seems unlikely that a P,P coupling product would show sufficient 7t-delocalization to resemble the a,a-dimer so closely. It should also be noted that the observations in the study of terthiophene derivatives referred to above may have been affected by the conditions used to dissolve the gold and solubilize the electrochemical product after electrosynthesis 123 (NaCN/DMF/H20 at 60°C) , hence an in situ characterization as used herein to study TP has significant advantages. One difference between the solution PTTP spectrum and spectrum of TP State IV shown in Figure 6-8 is that the latter can be interpreted as being either broadened or more 400 500 600 X (nm) Figure 6-8 A compar i son o f normal ized plots o f (a) a solut ion spectrum o f P T T P i n C H C 1 3 , (b) a background-subtracted in situ spectrum measured at E d e s o f P T T P that had been p rev ious ly adsorbed on A u ( l 11) at E; = -0.600 V and (c) a background-subtracted in situ spectrum at E d e s o f T P that had been previous ly adsorbed at Ej = -0.600 V and E; = +0.200 V . B o t h (b) and (c) were measured i n 0.05 M K C 1 0 4 at p H 11 and E d e s = -1.000 V . F o r (c) the T P concentration was 1 m M . F o r background-subtracted spectra, note that A I = I(E d e s)-I(Ei). 124 intense at longer wavelengths than the solution spectrum. There are no red shifts in the peak maxima that might suggest it-stacking of the molecules, despite the fact that small aromatic molecules such as pyridine,[11] bipyridine[12,13] and thiophene[14] appear to form organized, close-packed rows in S T M images. This may be because the desorption process (and prior events such as reduction and possibly re-orientation) destroys the long-range organization that the layer may have had while adsorbed; or it may be because TP and P T T P do not TC stack efficiently on the electrode even when adsorbed, due to the different sizes of the pyridine and thiophene moieties.The fact that TP fluorescence is observed at all is somewhat surprising (given that equivalent amounts of T P and P T T P show similar emission intensities) as TP generally shows little absorption between 340 and 380 nm and very little emission at wavelengths longer than 430 nm. It is possible that excitation and emission of TP may be red-shifted or broadened in our in situ experiments. Although little red-shifting of the emission spectrum due to aggregation was observed in solution experiments, adsorption to the surface is a special case, in which electrochemical data strongly suggest a close-packed arrangement involving a certain amount of intermolecular interaction during State II. This would explain the absence of fluorescence observed for State I (-0.600 V ) , in which the extent of intermolecular interactions is not expected to be significant. Therefore, a TP absorbance spectrum more similar to the solution spectrum, with little absorption between 340 to 380 nm, would be expected at -0.600 V . Red-shifting and broadening of spectra^, 16] have previously been observed for fluorophores attached to metal surfaces (Section 2.3.2), and it is possible that ordering similar to that on the surface and/or proximity to the electrode are sufficient to produce 125 these effects. However, with the available microscope optics, only a very small tail of the emission spectrum could be seen, thus it is impossible to tell for certain whether a red shift or broadening occurs, although broadening is extremely likely since it is observed for both the chemically synthesized PTTP and the product of TP oxidation. The background-subtracted fluorescence intensity of TP upon desorption, after holding the electrode at Ej = -0.250 V at 430 nm suggests that fluorescence at the emission maximum would likely be significantly higher than that observed for the species formed after holding at +0.200 V. Since no TP emission is observed upon stepping to E d e s for Ej = +0.200 V , it is unlikely that the low intensity of PTTP emission is due to incomplete dimerization. One possible explanation for the difference in emission intensities between TP and PTTP is their distance from the electrode on desorption. As previously mentioned, the effect of electrode-mediated quenching is inversely related to the cube of the distance between the surfactant and the electrode. With this type of relationship, even small differences in distance from the electrode can lead to large differences in fluorescence intensity. Being larger and less soluble, it might be expected that PTTP would not move as far away from the electrode as TP at the same desorption potential, and would therefore show less fluorescence intensity. Another possible explanation for the lower observed fluorescence of the dimer relative to the monomer is that either the monomer radical cation or some of the dimer has reacted with other species in solution and is no longer fluorescent. However, it can be seen in Figure 6-7 that when the electrode is held at +0.200 V for up to 60 minutes, the fluorescence intensity on desorption does not decrease with time spent at the oxidizing potential, suggesting that both the TP radical cation and the dimer are chemically very stable on the electrode, and 126 may be stabilized by inter-molecular interactions due to the organization of the monomer prior to dimerization. 6.5 In Situ Fluorescence Microscopic Imaging of TP and PTTP The in situ fluorescence of TP under potential control was also characterized by fluorescence microscopy. Before each experiment, a brightfield image was taken to ensure that the electrode was in focus and in order to see in detail features on the electrode surface that might influence the in situ images (Figure 6-9a). As can be seen, the electrode is not completely smooth, having many small features such as pits and scratches. Following the collection of the brightfield image, the U V filters (the same set used for in situ fluorescence spectrometry) were put in place, such that only emitted wavelengths should be detected. As in the spectrophotometer experiments, the electrode was held at a given potential, Ej, for a length of time, usually 10 min. Recall that for values of Ej at which the fluorophore is adsorbed, no fluorescence from adsorbed species will be observed due to electrode-mediated quenching. Therefore, any fluorescence at E; is background due to other sources of light such as non-adsorbed species. Ei, varied between -0.800 V and +0.200 V , and a background image was taken at each value of Ej. Following collection of the background image at Ej, the potential was stepped to the desorption potential, Ed e s , which was typically -1.000 V. Here, species that were adsorbed to the electrode at Ej become separated from the electrode due to desorption and displacement by water. Therefore, at Ed e s , fluorescence from these previously adsorbed 127 Figure 6-9 Images of the Au(l 11) electrode in the presence of 1 mM TP: (a) brightfield and with UV filters (b) at -0.200 V and (c) -1.000 V vs. SCE. species can be observed along with the background fluorescence. At Ed e s, images were taken at 2 s intervals over a period of up to 40 s after stepping to E d e s -Raw 8-bit images of the electrode in the presence of 1 m M TP at -0.200 V , where TP is adsorbed, and at -1.000 V after holding at -0.200 V are shown in Figure 6-9. These show that the background fluorescence from TP in solution has around 80% of the average intensity of the fluorescence at Ed e s- In addition, the optics result in an uneven distribution of light within the field of view, such that the middle of the image appears brighter. Because it was the fluorescence of only the adsorbed species that was of interest, images were interpreted with the background subtracted. Therefore, all images shown for given values of E, beyond this point are background-subtracted, greyscale adjusted images. The background was subtracted in a manner similar to that used to background correct the in situ fluorescence spectra. For each value of E;, the image at Ej (where only 128 -0.800 V -0.600 V -0.400 V Figure 6-10 The first in situ image after stepping to the desorption potential, E^s (with background subtracted i.e. AI = I(Edes) - I(E,)) taken of the electrode with UV filters in place after holding for 10 minutes at various potentials, Ej, in the presence of 1 mM TP solution. subtracted from the image at Edes (which includes both background fluorescence and fluorescence from the previously adsorbed species) to give a background-subtracted image, AI, according to Eqn. (3.4). Background-corrected, grey scale-adjusted images of the electrode after holding at different values of E; for 10 minutes each in the presence of 1 mM TP are shown in Figure 6-10. These images were the first taken immediately after desorption, before significant diffusion had taken place. It can be seen that for E , = -0.800 V , upon desorption there is negligible fluorescence in the background-subtracted image, suggesting that very little TP was adsorbed at this potential. For Ej = -0.600 V , where electrochemistry results show that TP is adsorbed in State I, significant fluorescence is observed over less than half the electrode in the background-subtracted image. A comparison to Figure 6-9a shows that initially, more fluorescence is visible at scratches on the electrode. This suggests that there are more adsorbed species at the scratches, or adsorbed species at the scratches are more fluorescent on desorption than species adsorbed on terraces, or both. In the first case, it is plausible that adsorption occurs preferentially at scratches on the electrode, which are regions that have greater surface roughness/surface area capable of interaction with the adsorbates and that may not be crystallographically <111>. In the second case, species may desorb further from the scratches than from the terraces; or it may be that intermolecular interactions are greater (i.e. there is some aggregation) for species adsorbed at the scratches, resulting in red shifting of the absorption and emission spectra that would favour the detection of their fluorescence under experimental conditions with the given optical filter set. The observation of fluorescence predominantly at scratches upon stepping to the desorption potential after holding the electrode at Ei = -0.600 V in the presence of 1 m M TP is in contrast to what has been observed for layers of octadecanol coadsorbed with 3 mol % fluorescent dye (l,r-dioctadecyl-3,3,3',3'-tetra methylindodicarbocyanine perchlorate), in which the morphology of the fluorescence image is unaffected by defects in the electrode.!17! 130 Very diffuse fluorescence can also be seen between the large scratches in the presence of TP in the background-subtracted image for Ej = -0.600 V , which may represent adsorption at smaller scratches or simply the amount of fluorescence observed for molecules that are evenly spaced, but not closely packed. It appears much of the electrode is still water-covered. This is consistent with the very low coverages observed by chronocoulometry. No fluorescence at all is observed by in situ spectrophotometry at this potential. However, this likely has to do with differences in the sensitivity of the detectors and differences in the degree of optimization and timescale of light collection for the camera vs. the spectrophotometer. For E; = -0.400 V , which is close to the phase transition to State II, the fluorescence intensity in the background-subtracted image upon desorption is still higher at scratches, but fluorescence is visible over the entire electrode. This suggests that the TP-Au interaction is stabilized by inter-molecular interactions. The unevenness of fluorescence over the electrode may be due to a spatial distribution of either concentrations or desorption rates. In either case, the image suggests that the phase transition to State II is not complete at this potential, and there may be multiple adsorbed states. In contrast, the background-subtracted images at desorption for -0.200 V (State II), 0 V (the middle of the oxidation peak TP2 in the CV) , and +0.200 V (positive of TP2) show very uniform distribution of fluorescence. Interestingly, in these images, the regions with large scratches that were illuminated at -0.600 V are dark. This may mean either that material remains adsorbed (and therefore quenched) in these scratches even at E d e s , or that State II and State III are not favoured in these areas of the electrode. 131 Although the camera C C D cannot distinguish between different wavelengths of light (and therefore cannot distinguish between TP and PTTP fluorescence), based on the in situ fluorescence spectra, the background-subtracted fluorescence upon desorption after holding at E ; = +0.200 V and part of that for Ej = 0 V comes from the TP dimer after reduction to the neutral species. From a macroscopic perspective, it appears that after dimerization, the layer retains the structure pre-organized in State II. Consistent with the in situ spectra, a major difference between the images of State II and that of the 30 25-20-? 15-< 10-5-• -0.800 V -0.600 V A -0.400 V r -0.200 V 0.000 V • 0.200 V • : , , ! » » » » „ i 1 i 1 i • i 1 r~ 0 10 20 30 40 time at desorption (s) Figure 6-11 A plot of average image intensity with time spent at the desorption potential, Edes, (with background subtraction i.e. AI = I(Edes) - I(Ej)) after holding for 10 minutes at different values of Ej. Edes was-1.00 V. 132 dimerized layer is the lower intensity observed with the dimer. While the initial background-subtracted desorption images for TP State II and the dimer are very similar, their behaviour after desorption is not. TP State II desorbs more quickly than the dimer, as demonstrated by a plot of average raw intensity at E d e s vs. time (with background subtraction) for different values of E; (Figure 6-11). A series of background-subtracted images taken at different times after desorption (Figure 6-12) show that the fate of the molecules after desorption depends very much on the state in which they were adsorbed. After holding at Ej = -0.600 V , at which electrochemistry data suggests the TP is adsorbed at low coverage with few intermolecular interactions, the features seen in the background-subtracted images upon desorption are blurred and expanded even 2 s after stepping to E d e s , suggesting significant diffusion and dispersion of the molecules away from the electrode. At 30 s after desorption, only a small amount of very diffuse fluorescence remains. Subsequent to holding at E; = -0.200 V , there is some blurring and expansion of features 2 s after desorption, but not to the same extent as for -0.600 V. In addition, starting at 2 s after desorption, fluorescence is seen in clumps on the right half of the image. These aggregates maintain their size and shape over 30 s, suggesting that they remain more or less stationary. It may be that the intermolecular interactions between the TP molecules in State II are strong enough that some of the structure is maintained even after desorption, forming aggregates that are not easily solubilized. For Ej = +0.200 V, the features remain even more distinct at 2 s after desorption. Interestingly, the fluorescence increases noticeably between 0 and 2 s, suggesting that the low intensity at 0 s is due, at least in part, to the proximity to the electrode. As the desorbed molecules diffuse further away 133 250 nm 250 \im 250 nm 250 nm Ei=-0.600 V Os 2 s 10s 30 s » • 250 nm 250 (im 250 jim 250 nm Ej =-0.200 V Os 2s 10s 30 s 250 fim 250 |ini 250 (im 250 nm Ei =+0.200 V Os 2 s 10 s 30 s Figure 6-12 Changes in background-subtracted fluorescence images ( A I = I ( E d e s ) - I(E;)) of a Au( 111) electrode in the presence of 1 m M TP after varying amounts of time following desorption. TP that had been adsorbed for 10 minutes at different values of E ; prior to stepping to the desorption potential, Edes. with time, the fluorescence increases. Some aggregates similar to those seen for Ej = -0.200 V are visible at 30s, but overall, it appears that, as might be expected, the desorbed dimer layer "holds together" better than the T P State II layer. Fluorescence microscopy results also lend support to the stability of the oxidized T P . After holding the electrode at various lengths of time, the average background-subtracted 134 fluorescence intensity difference on desorption does not decrease with time at Ej even up to 40 minutes. 6.6 Summary PTTP was synthesized chemically and characterized by cyclic voltammetry and capacitance on Au(l 11). Electrochemical characterization shows a set of broad, pH independent peaks at -0.48 V. The high capacitance of adsorbed PTTP suggests that PTTP forms a disorganized, low coverage layer when adsorbed from the air/solution interface. In situ fluorescence spectroscopy of adsorbed PTTP reveals that fluorescence is observed upon desorption after holding at potentials as negative as Ej = -0.800 V . PTTP fluorescence upon desorption increases as Ej becomes more positive, to a maximum at Ej = -0.600 V, and remains constant above -0.600 V. In situ fluorescence spectroscopy of TP shows no fluorescence corresponding to TP State I. However, fluorescence is observed upon desorption in the background-subtracted spectra after holding at Ej = -0.400 V (just positive of the transition to State II) and increases to a maximum at -0.250 V , corresponding to the potential of maximum surface coverage for State II. After holding at Ej = +0.200 V for as little of 2 minutes, a spectrum very similar to that of PTTP can be observed on desorption. In situ fluorescence imaging after holding at Ej = -0.600 V (State I), TP is adsorbed preferentially at features such as scratches and diffuses away quickly on desorption. After holding at -0.200 V (State II), TP forms a more uniform layer, but forms aggregates on 135 desorbing. After holding at +0.200 V , on desorption the uniform layer resembles that formed at -0.200 V , but desorbs more slowly. 6.7 References (1) Michalitsch, R.; E l Kassmi, A.; Yassar, A.; Lang, P.; Gamier, F. J. Electroanal. Chem. 1998, 457, 129-139. (2) Maestri, M . ; Sandrini, D.; Balzani, V . Chem. Phys. Lett. 1985,122, 375-378. (3) Bouwhuis, E . ; Janssen, M . J. Tetrahedron Lett. 1972, 3,1'il-l'Sb'. (4) Becker, R. S.; de Melo, S.; Macanita, A . L . ; Elisei, F. Pure Appl. Chem. 1995, 67, 9-16. (5) Henry, M . S.; Hoffman, M . Z. J. Phys. Chem. 1979, S3, 618. (6) Zhou, Z. H.; Maruyama, T.; Kanbara, T.; Ikeda, T.; Ichimura, K.; Yamamoto, T.; Tokuda, K. J. Chem. Soc, Chem. Commun. 1991, 1210-1212. (7) Yamamoto, T.; Zhou, Z. H. ; Maruyama, T.; Kanbara, T. Synth. Met. 1993, 55, 1209-1213. (8) Belletete, M . ; Mazerolle, L . ; Desrosiers, N.; Leclerc, M . ; Durocher, G. Macromolecules 1995, 28, 8587-8597. (9) Sato, T.; Hori, K.; Fujitsuka, M . ; Watanabe, A.; Ito, O.; Tanaka, K. J. Chem. Soc, Faraday Trans. 1998, 94, 2355-2360. (10) Hrdlovic, P.; Krajcovic, J.; Vegh, D. J. Photochem. Photobiol, A 2001,144, 73-82. (11) Andreasen, G.; Vela, M . E . ; Salvarezza, R. C.; Arvia, A . J. J. Electroanal. Chem. 1999, 467, 230-237. 136 (12) Cunha, F.; Tao, N. J.; Wang, X . W.; Jin, Q.; Duong, B.; D'Agnese, J. Langmuir 1996,72,6410-6418. (13) Dretschkow, T.; Lampner D.; T., W. J. Electroanal. Chem. 1998, 458, 121-138. (14) Noh, J.; Ito, E . ; Nakajima, K.; Kim, J.; Lee, H. ; Hara, M . J. Phys. Chem. B 2002, 706,7139-7141. (15) Kittredge, K. W.; Fox, M . A.; Whitesell, J. K. J. Phys. Chem. B 2001, 705, 10594-10599. (16) Pope, J. M . ; Buttry, D. A . J. Electroanal. Chem. 2001, 498, 75-86. (17) Shepherd, J.; Yang, Y. ; Bizzotto, D. J. Electroanal. Chem. 2002, 524-525, 54-61. 137 7 Conclusions 7.1 A Model for TP Adsorption on Au(111) Based on the electrochemical and spectroelectrochemical results described in this thesis, a model for the behaviour of a TP monolayer on Au(l 11) as a function of potential on the first potential scan to +0.2 V is shown in Scheme 7-1. According to this model, at potentials near -0.80 V vs. SCE, TP adsorbs as a monolayer with the plane of the rings parallel to the electrode surface (State I). Near -0.45 V , a two-Scheme 7-1 A model for behaviour of a TP monolayer on Au(l 11) as a function of potential on the first potential scan. - 1 . 0 - 0 . 8 - 0 . 6 - 0 . 4 - 0 . 2 CIO 0 2 0 . 4 E(Vvs. SCE) 138 dimensional phase transition occurs (TP1) to a close-packed layer of molecules oriented more normal to the surface, bonded through the heteroatoms (State II). At positive potentials, an oxidative dimerization similar to the oligomerization observed in thiophene occurs (TP2). It is proposed that the dimer is further oxidized (charged) (State III); when the electrode is scanned or pulsed to a negative potential, stored charge is discharged (TP3) and the neutral dimer, PTTP (State IV), is desorbed, as observed by fluorescence spectroscopy. On scanning the potential again in a positive direction, some dimer is re-adsorbed along with TP from solution, as shown by changes in the C V on subsequent scans. A more detailed characterization of each state follows. The slow kinetics of the adsorption process resulting in State I, the 71-bonded state, resemble those of thiophene more than pyridine. The capacitance of the TP layer adsorbed in this state is only slightly lower than that of the uncoated electrode. Fluorescence images suggest that adsorption occurs preferentially at scratches and defects on the electrode, and that the electrode remains largely water-covered, consistent with the low coverage for State I determined by chronocoulometry of 7 x 10" mol/cm . On desorption, State I is highly soluble and diffuses quickly from the electrode-solution interface (Figure 6-12). The fast kinetics of the phase transition to State II at high bulk concentrations resemble those seen in phase transitions of pyridine rather than thiophene. The negative shift in the pzc suggests that like pyridine, TP takes on a more vertical orientation with a dipole perpendicular to the surface, indicating that at least one of the heteroatoms is bonded to 139 the surface, and the known interactions of thiophene and pyridine with Au make it likely that both heteroatoms interact with Au via their nonbonding electrons. However, the low magnitude of this shift in the pzc relative to pyridine and 2,2'-bipyridine (Figure 5-1) suggests that perhaps TP in State II has a more tilted configuration than the latter molecules, or that the rings may not be co-planar. The surface coverage for State II was determined by chronocoulometry to be 5 x 10"10 mol/cm2, very similar to the coverage of the condensed phase for 2,2'-bipyridine. Fluorescence images show that State II adsorbs preferentially on flat areas of the electrode, having less preference for scratches and defects. On desorption, some of the intermolecular interactions/long-range packing seems to be retained, resulting in the formation of aggregates that diffuse only very slowly from the interface, indicating poor solubility (Figure 6-12). Between State II and State IV, two oxidative processes occur, one of which likely corresponds to a dimerization and another of which is proposed to be a further oxidation, resulting in State III. This species can undergo reduction and re-oxidation at potentials between -0.4 V and +0.4 V at neutral pH, indicating that this oxidation and this reduction comprise a chemically reversible process (unlike the irreversible dimerization) of the electrochemically generated dimer (Figure 4-7). It is interesting to note that the size and shape of the peaks corresponding to reduction is independent of bulk concentration and clearly involves only surface species. It is proposed that the oxidation and reduction are a. charging/discharging process involving partial oxidation and reduction. Subsequent to reduction, State IV, whose fluorescence spectrum is identical to that of PTTP, can be observed by fluorescence spectroscopy. Prior to desorption, State IV (and hence state III), like State II, are adsorbed on the same flat regions of the electrode, away 140 from defects and scratches. This is expected, as the molecules are pre-adsorbed in State II prior to the oxidation and reduction process generating State III and State IV. State TV desorbs more slowly than State II, as might be expected for a layer of larger, less soluble molecules. Like State II, it forms insoluble aggregates on desorption (Figure 6-12). Characterization of the electrochemically generated dimer shows that it produces sharp peaks in the C V similar to the TP phase transition peaks, but occurring at slightly more negative potentials (Figure 4-7). These likely correspond to a phase transition of the dimer similar to that observed for the monomer, from a flat-lying, low-coverage layer to a more condensed, vertical orientation. If the potential for the conversion (reduction) of State III to State IV is more positive than the potential for this PTTP phase transition, then a sharp phase transition peak is visible on the negative scan, just negative of peak TP3. Both PTTP and TP peaks are always visible on the subsequent positive scan, as the phase transition to a lower coverage state always results in desorption of PTTP, leaving room for the readsorption of TP on the subsequent scan. 7.2 Future Work A number of questions remain regarding the behaviour of TP on Au(l 11). First, very little is known about State III (proposed to be an oxidized dimer), which cannot be observed using the in situ fluorescence technique employed herein. Besides the dimerization, none of the redox processes that adsorbed TP undergoes have been definitively identified. It is not known how many electrons are involved in these processes, nor whether the oxidation and/or reduction are accompanied by changes in orientation or conformation. While the average orientation of adsorbed TP has been 141 inferred for State I and II, it would be interesting to see what the actual microscopic packing in each state might look like, considering the different sizes and electronic properties of the pyridine and thiophene rings. In the condensed phase, a head-to-tail alternating packing, with each thiophene ring sandwiched between two pyridine rings and vice versa might be expected to be favoured. Being able to observe the actual morphological differences between PTTP film adsorbed from the air-solution interface and electrochemically generated PTTP would confirm the reasons for their different electrochemical properties. Regarding the fluorescence results, it is puzzling that the apparent fluorescence intensity is lower for desorbed PTTP than it is for TP when the excitation and emission wavelengths should favour PTTP fluorescence. This may have to do with their relative distances from the electrode on desorption, or it may be because the PTTP shares the electrode with non-fluorescent species due to side reactions or incomplete reduction. In addition, it is not known whether the fluorescence observed for TP on the red edge of its spectrum is due to a 71-stacking-induced red shift or merely broadening. This could be determined using fluorescence spectroscopy with an appropriate optical filter set. Many of the questions outlined above could be answered by in situ vibrational spectroscopic studies, particularly in situ IR spectroscopy. Direct observation of State III, hypothesized to be an oxidized dimer, might be possible. Oxidation should result in significant charge redistribution, which would cause a shift in the vibrational frequencies. IR would also confirm that the linkages between TP molecules upon dimerization are actually a, a'. If side reactions occur on the electrode (such as reactions with water), their presence might be detected by IR also. 142 With IR spectroscopy, it is possible to distinguish not only between different chemical species but also between different surface orientations of adsorbed species, on account of surface selection rules. Information of this sort would complement the chronocoulometry data. For the phase transition of adsorbed TP from the 7t-bonded state to the more vertical orientation on the gold surface, it may be possible to determine whether both rings are equally tilted or whether there is a torsional angle between the rings, since separate signals are observed for pyridine and thiophene C - C , C - N and C-S stretches within the molecule. Furthermore, IR studies might be able to determine whether a conformational change occurs during the oxidative and reductive processes. In situ scanning probe techniques such as scanning tunnelling microscopy (STM) and atomic force microscopy (AFM) might be able to answer questions regarding the nanoscale structure of the adsorbed TP monolayers, including their stacking and their registry with respect to the Au surface. Using these techniques, it might be possible to determine why the 7t-bonded state seems to adsorb preferentially at scratches and defects, while other states adsorb preferentially on flat areas of the electrode. Because S T M is sensitive to electron density, it could potentially be used to image the oxidation and charging that produce State III. Both S T M and A F M might be employed to observe changes in conformation or orientation associated with oxidation and charging also. Using scanning probe techniques to image PTTP adsorbed from the air/solution interface would make it possible to identify the morphological differences between the latter PTTP monolayers and electrochemically generated PTTP monolayers that might be responsible for apparent differences in their behaviour. 143 Electrochemical quartz crystal microbalance (EQCM) studies might be used to give an independent estimate of the coverage of the dimer and/or the number of counterions required to balance any charge excess caused by oxidation/charging processes. This might make it possible to determine the number of electrons involved in the charging process. In addition to further characterization of TP, the adsorption of thiophene on Au(l 11) in aqueous solution produced results warranting further study. In situ fluorescence of thiophene might be able to determine the chain length of the products. As with TP, in situ IR might help determine the type of coupling, side products and orientation. Finally, it would be interesting to undertake studies of the binding of other multifunctional systems in which the strengths of the donor groups vary with respect to one another, such as 2-mercaptopyridine or 2-mercaptothiophene. 7.3 Concluding Remarks TP shows adsorption behaviour similar to that of pyridine and 2,2'-bipyridine. Interestingly, the pyridine half of the molecule appears to dominate the adsorption behaviour of TP (although the thiophene part of the molecule does have an influence as a n-electron rich substituent) despite the fact that thiophene might be expected to interact more strongly with gold. The thiophene half of the molecule confers electrochemical reactivity to the molecule that is influenced by the pyridine-like potentially-driven organization of the layer. Thus, the electrochemical generation of the dimer on the surface produces a highly organized layer of the dimer on Au(l 11) that would otherwise be impossible to obtain. 144 Comparative studies of thiophene and TP electrochemistry in aqueous and organic solvents demonstrate that the solvent and the solubility of the adsorbate can have a significant influence on the oxidation potential. This study also demonstrates the versatility of the in situ electrochemical fluorescence technique developed by Prof. Dan Bizzotto and Mr. Jeff Shepherd. Here, the technique has allowed characterization of the layer as it is formed at different potentials, as well as the identification of the product of a surface-confined electrochemical reaction on a gold electrode via its characteristic fluorescence spectrum. This is the first time that an electrochemical dimerization has been identified by in situ fluorescence. The ability to observe the spatial distribution of fluorescence from previously adsorbed species allows the inference of preferred adsorption sites, correlated to features on the electrode, for different states; and the properties of each state upon desorption. In fact, this study demonstrates clearly that use of multiple analytical techniques is useful and often necessary to characterize the behaviour of systems having some level of complexity. Overall, this system shows that multifunctional adsorbates can create chemically-tuned surfaces capable of switching between multiple states, and allow subtle comparisons of the influence of surface-adsorbate interactions between different functional groups. 145 Appendix Examples of Chronocoulometry Current Transients 0.0--0 .5--0 .650 V 0 .0 ] -0 .5 ] -0 .400 V 0 . 0 ] n -0.5-^ +0.100 V g -1-0H o -1 .5 ] e -1.0] c CD o -1.5 c -1.0H o -1.5 -2 .0] -2 .0] -2 .0 ] CD 4 9 2 50 100 150 -0.650 V 50 100 150 T i m e (ms) 50 100 T i m e (ms) 150 50 100 T i m e (ms) 150 Figure A - l A b o v e are the raw current transients measured upon stepping the A u ( l 11) w o r k i n g electrode to the desorpt ion potential , E d e s , f rom different potentials o f interest, Ej i n the presence o f 1 m M T P at p H 11. T h e y were used to calculate the values o f charge by integrating b e l o w each current transient f r o m the pre-point baseline to the measured current value over 150 ms. T h e integrated charge for the current transient measured at E = -0 .650 V is also shown above. Cor rec t ion for instrumental offsets and faradaic processes was performed by extrapolat ing to t ime = 0, as shown w i t h the dashed l ine , as descr ibed i n sect ion 3.3.4.1) 146 A note regarding the calculation of AG from the surface pressure curves Parsons (R. Parsons, Proc. R. Soc. London Ser. A , 261 (1961) 79-90) notes in his paper regarding the structure of the mercury-electrolyte interphase in the presence of thiourea that because the slope of the <|> vs. log c curves is independent of O", by shifting each curve along the x-axis by a correction term, f(a), the curves can be superimposed. F(o) was chosen at each value of a ot maximize the overlap between the curves. 147 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059339/manifest

Comment

Related Items