UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electro-fluorescence characterization of insoluble surfactants adsorbed on solid electrodes Shepherd, Jeffrey L. 2005

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

Item Metadata

Download

Media
831-ubc_2005-104406.pdf [ 19.34MB ]
Metadata
JSON: 831-1.0061078.json
JSON-LD: 831-1.0061078-ld.json
RDF/XML (Pretty): 831-1.0061078-rdf.xml
RDF/JSON: 831-1.0061078-rdf.json
Turtle: 831-1.0061078-turtle.txt
N-Triples: 831-1.0061078-rdf-ntriples.txt
Original Record: 831-1.0061078-source.json
Full Text
831-1.0061078-fulltext.txt
Citation
831-1.0061078.ris

Full Text

ELECTRO-FLUORESCENCE CHARACTERIZATION OF INSOLUBLE SURFACTANTS ADSORBED ON SOLID ELECTRODES  by  Jeffrey L. Shepherd  B.Sc, Laurentian University, Sudbury ON.  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF GRADUATE STUDIES CHEMISTRY  THE UNIVERSITY OF BRITISH COLUMBIA March 2005  © Jeffrey L. Shepherd, 2005  Abstract The development of an in situ technique for characterizing water-insoluble organic molecules adsorbed onto electrode surfaces was successfully accomplished. The developed technique is termed electro-fluorescence microscopy and was created through the union of epi-fluorescence microscopy and electrochemical methods. Initial investigations were conducted on a Au(l 11) electrode with adsorbed 1-octadecanol containing a small amount of a fluorescent dye molecule. Octadecanol adsorbed onto Au(l 11) was chosen as the initial starting point for the developed electro-fluorescence technique since a potential controlled adsorption/desorption process of these molecules to/from the electrode has been extensively studied. The system was then extended to monitor the behaviour of oleyl alcohol adsorbed on the A u ( l l l ) electrode.  Investigations of mixed monolayers containing various  concentrations of oleyl alcohol and octadecanol were also characterized with the developed technique. Because octadecanol and oleyl alcohol do not have a fluorescent moiety in their chemical make up, the surfactants were mixed with a small amounts of two fluorescent dye molecules which had the same alkyl chain length as the alcohols. These dye molecules efficiently mixed with the lipid-like alcohols and faithfully reported on the potential controlled behaviour of the surfactants. Fluorescence imaging revealed that the alcohol molecules desorbed from the electrode in a heterogeneous and aggregated morphology which did not diffuse from the metal with desorption time. The process was found to be repeatable for a given layer and reproducible on successive investigations. To extend the mechanism of the reductive desorption of self-assembled thiols from noble metals, the developed technique was also used to monitor the reductive desorption of a thiol containing a fluorescent moiety. In this investigation the electrode was a polycrystalline Au bead having a variety of surface features. The thiol was observed to selectively desorb from the various surface regions at certain values of potential.  ii  Dedication This thesis is dedicated to my loving wife Ruth Shepherd. Your patience and support during the past years have made the completion of this project possible. I can only offer you this dedication as a small token of my appreciation. Thank you so much for everything Ruth.  iii  Acknowledgments There are a number of people who's guidance, support and input have helped immensely during the course of this project. They all deserve a special thank you and I hope that nobody is overlooked in this section. Firstly, a special thank you to my supervisor Professor Dan Bizzotto who has guided me through this project. Your teaching and direction has been invaluable through the past years. I will always appreciate your outlook on science and your patience with your students. I would like to thank my family members who have made this experience richer. Thank you to my parents, brother and sister for support both spiritually and financially. Your thoughts and prayers have helped me through some tougher times during this project. Thank you to the members of the Bizzotto research group both past and present. These include Vivian Stauffer, Yanguo Yang, Robin Stoodley, John Agak, Dr. Eduard Guerra, Dr. Emily Chung, Dr. Pavel Freundlich, Amanda Musgrove and Aya Sode. It has been a pleasure sharing a laboratory with you.  You are all easy-going people which makes for an enjoyable work  environment. A special thank you to Eduard Guerra who's humour is simply amazing. You kept me laughing for a long time even after you left the lab. I will always remember the fun that was shared. Another special thank you is offered to my good friend Anand Thiurmalai who worked many late hours with me. Your friendship and company during the past few years will not be forgotten. Thanks to the mechanical shop for improving on the sketches that were given to you to create beautifully machined parts making up the microscope. A special thank you goes to Brian Ditchburn who created the spectro-electrochemical cell with a 0.17mm thick cover glass. Without this, our results would have been masked in poor quality images. Finally, I would like to thank Dr. Chad Sinclair for helping with the EBSD determination of the various surface facets on the polycrystalline Au bead. A thank you is also given to Dr. Mario Beaudoin for creating the stepped template of Si0 on the Au slide. I would also like to thank Alec 2  Kellas for building the light tight box that experiments were conducted in.  iv  Table of Contents Abstract  ii  Dedication  iii  Acknowledgments  iv  Table of Contents  v  List of Abbreviations  ix  List of Figures  xiv  List of Tables  xxiii  Chapter 1 1 Introduction 1.1 Objectives 1.2 Rationale 1.3 Scope of Thesis  1 2 3 5  Chapter 2 2 Theoretical Background 2.1 Qualitative Description of the Electric Double Laver  7 7  2.2  2.1.1  Electrochemical  2.1.2  Simple  2.1.3  Capacitance  2.5  2.6  7 8  of the Electrode/Solution  Interface  8  Quantitative Description of the Electric Double Layer 2.2.1  2.3 2.4  Concepts  Capacitor  The Gouy-Chapman-Stern  10  (GCS) Theory  10  The Metal Side of the Interface Thermodynamic Treatments of the Electrode/Solution Interface 2.4.1  The Gibbs  Adsorption  2.4.2  The Electrocapillary  13 16  Isotherm  16  Equation  19  The Adsorption of Neutral Molecules 2.5.1  Adsorption  Isotherms  2.5.2  Adsorption  of Organic  2.5.3  Chronocoulometry  21 21  Molecules  and the Back  onto Electrified Integration  Interfaces  24  Method  25  Insoluble Monolayer Characteristics 2.6.1  Structure  and Properties  of Monolayers  27 at the Air/Solution  Interface  28 2.6.2  Mixed  Monolayers  . . .  33  v  2.7  2.8  2.6.3 Monolayers Transferred to the Solid Support Factors Affecting Fluorescence from Molecules 2.7.1 Electronic Transitions and the Franck-Condon Principle 2.7.2 Fate of the Excited State Molecule 2.7.3 Dye Aggregates 2.7.4 Forster Energy Transfer 2.7.5 Fluorescence Near Metal Surfaces Fluorescence Microscopy 2.8.1 Carbocyanine Dyes 2.8.2 Microscope Resolution 2.8.3 The Epi-Fluorescence Microscope  34 34 34 37 40 42 46 48 49 49 52  Chapter 3 3 Literature Review 3.1 Adsorption of Lipids on a Mercury Electrode 3.2 Adsorption of Octadecanol Onto Au(lll) 3.2.1 Fluorescence Near a Metal Surface 3.3 Self-Assembled Monolayers of Thiols on Noble Metals 3.4 Luminescence at Electrode Surfaces 3.5 Summary  55 56 59 60 61 63 64  Chapter 4 4 Experimental Methodology 4.1 Systems Studied 4.2 Electrochemical Methods 4.2.1 Materials 4.2.2 Electrochemical Instrumentation 4.2.3 Electrochemical Procedures 4.2.3.1 Physically Adsorbed Surfactants on Au(lll) 4.2.3.2 Chemically Adsorbed Thiol on a Au Bead 4.2.4 Electrochemical Techniques 4.2.4.1 Cyclic Voltammetry 4.2.4.2 Differential Capacitance 4.2.4.3 Chronocoulometry 4.3 Spectroscopic Technique 4.3.1 Epi-fluorescence Microscopy 4.4 Description of the Programs Used 4.4.1 Electro-fluorescence Imaging Programs 4.4.1.1 image_scan.vi 4.4.1.2 image_adsorb_neg.vi and image_adsorb_pos.vi 4.4.1.3 potential_holding.vi 4.4.2 Data Analysis Programs 4.4.2.1 histogram_calc.m 4.4.2.2 count_features.scr  65 65 66 66 68 68 69 72 72 72 72 74 75 75 79 79 81 81 83 83 83 84  vi  Chapter 5 5 Electrochemical Investigations of C180H and OLA on Au(lll) 5.1 Electrochemical Characterization 5.1.1 CV and Differential Capacitance of OLA/C180H 5.1.2 Charge Density and Film Pressure of OLA/C180H 5.1.3 Electrochemical Characterization of DiIC18(5)/C180H 5.2 Summary and Conclusions  87 87 87 92 98 98  Chapter 6 6 Fluorescence Imaging of Physisorbed Alcohols on Au(l 11) 102 6.1 Imaging 3mol% DiIC18(5)/C18QH Adsorbed onto Autlll) 102 6.1.1 Collection and Treatment of Images 102 6.1.2 An Imaging/Potential Scan Investigation 103 6.1.3 Motivation for the 'Hotspot' Mask 106 6.1.4 Average Number and Mean Area of the Image Features 109 6.1.5 A Second Imaging/Potential Scan Investigation 111 6.1.6 An Imaging/Potential Step Investigation 113 6.1.7 Summary of the Results for 3 mol% DiIC18(5)/C180H on Au(lll) .116 6.2 Imaging a Two Dve/C18QH Laver on Autlll) 117 6.2.1 An Imaging/Potential Scan Investigation 117 6.2.2 Summary of the Results for the Two Dye/C180H Layer on Au(lll) 124 6.3 Imaging a Two Dye/OLA Laver on Audll) 125 6.3.1 An Imaging/Potential Scan Investigation 125 6.3.2 Summary of the Results for the Two Dye/OLA Layer on Au(lll) . . . 129 6.4 Imaging a Two Dve/25mol%OLA/C18QH Laver on Auflll) 130 6.4.1 An Imaging/Potential Scan Investigation 130 6.4.2 Summary of the Results for the Two Dye/25 mol% OLA/C180H on Layer Au(lll) 132 6.5 The Influence of Surface Irregularities on the Desorbed Structure 134 6.6 General Observations and Conclusions 136 Chapter 7 7 Fluorescence Recovery by Electrosorption 138 7.1 Fluorescence Decay and Recovery of 3 mol% DiIC18f5) in C18QH 138 7.1.1 The Collection and Treatment of Images 138 7.1.2 Fluorescence Decay and Recovery Under Constant Illumination ... 139 7.1.3 Implications and Possibilities for the Decay and Recovery 141 7.1.3.1 Fluorescence Recovery After Phobleaching (FRAP) 144 7.1.3.2 Triplet State or Exciplex Formation 144 7.1.3.3 Radiative Enhancement and Diffusion to the Metal 146 7.1.3.4 Light Independent Dye Aggregation 149 7.2 Proposed Kinetic Model 150 7.2.1 Model Assumptions 150 7.2.2 Kinetic Model Derivation 152 vn  7.2.2.1 Modification to the Model . 7.3  7.3.1  7.4  156  Fluorescence Decay and Recovery of 3 mol% DiIC18(5VOLA Fluorescence  Decay and Recovery  Under Constant  Illumination  General Observations and Conclusions  158 . . . 158  160  Chapter 8 8 Selective Desorption of BODIPY-C10-SH from a Au Bead 162 8.1 Electro-fluorescence Characterization of the Selective Desorption Process  162  8.2  8.3 8.4  8.1.1  Modification  8.1.2  The Collection  to the Electrochemical and Treatment  Setup  162  of Images  163  Electrochemical and Epi-fluorescence Investigations 8.2.1  Electrochemical  8.2.2  Electro-fluorescence  163  Characterization  163  Characterization  165  Selective Reductive Desorption General Observations and Conclusions  169 173  Chapter 9 9 Summary and Conclusions 9.1 Summary  9.2 9.3  175 175  9.1.1  Electrochemical  9.1.2  Electro-fluorescence  Characterization  9.1.3  Electrochemical  9.1.4  Selective Reductive  of Physisorbed  Characterization  Control over Radiative Desorption  Alcohols  of Physisorbed Decay  of a SAM  Alcohols  175 . . . . 176 179 180  Proposed Mechanism for the Desorption and Re-adsorption Events 180 Suggestions for Future Study and Modification to the Developed Technique  183 References  187  Appendix A l  199  Appendix A2  223  viii  List of Abbreviations Chapter 1  LB SAMs C180H OLA GCS  Langmuir-Blodgett self-assembled monolayers 1-octadecanol cis-9-octadecen-l-ol, or Oleyl Alcohol Gouy-Chapman-Stern  Chapter 2  IPE C d e e E q / q q C IHP OHP Xj 4> o' x 4> a a a k T n° z e (J) K N 4> C C pzc 0  M s  dl  ;  2  2  d s  M  A  0  H  D  ideally polarized electrode capacitance distance between the plates in a capacitor dielectric constant permivity of free space voltage charge current charge on the metal surface charge in solution double layer capacitance inner Helmholtz plane outer Helmholtz plane distance from the metal surface where the IHP resides potential at the IHP charge density at the IHP distance from the metal surface where the OHP resides potential at the OHP charge density at the OHP charge density in solution charge at the metal surface Boltzmann constant temperature number concentration of ions in a reference lamina signed charge of an ion charge of an electron electrostatic potential inverse of the Debye length Avagadro's number potential of the metal surface Helmholtz or inner layer capacitance diffuse layer capacitance potential of zero charge ix  * 0 X  fee  i4 a  P  a G Y  r  MX L WE REF E  0 Cbulk K *i *-i  ^  ads  E* (0  ^, ^/  0 0 =0 0  EMB PP  L W  P  P  T H, P  Pi  outer of Volta potential chemical potential inner or Galvani potenital Fermi level work funciton surface potential face centered cubic area alpha phase beta phase thermodynamic property of the dividing surface Gibbs free energy interfacial tension surface excess concentration relative surface excess M and X" electrolyte neutral molecule working electrode reference electrode potential of the working electrode with respect to the reference fractional monolayer coverage concentration of the bulk ratio of the absorption rate constant to the desorption rate constant adsorption rate constant desorption rate constant gas constant Gibbs free energy of adsorption lateral interaction parameter film pressure potential at which the surface tension (y*) is known frequency adsorption potential desorption potential presence of adsorbed surfactant absence of adsorbed surfactant contact angle electromicrobalance density of Wilhelmy plate length of Wilhelmy plate width of Wilhelmy plate thickness of Wilhelmy plate immersion depth of Wilhelmy plate into solution density of liquid +  x  gravitational force constant liquid expanded state of a compressed monolayer liquid condensed state of a compressed monolayer collapse pressure equilibrium spreading pressure equilibrium spreading pressure mole fraction interfacial tension between solid and vapour interfacial tension between solid and liquid interfacial tension between liquid and vapour excited state wavefunction ground state wavefunction transition dipole moment electronic state vibrational state electronic coordinate nuclear coordinate fluorescence quantum yield rate constant for relaxation via fluorescence rate constant for competing deactivation pathways natural excited state lifetime of a fluorophore measurable mean fluorescence lifetime fluorescence intensity at some time after exposure to light sensitizer molecule acceptor molecule distance separating the center of dipoles in molecules natural excited state lifetime of a fluorophore intensity at a distance from a metal intensity at infinite separation from a metal Forster critical distance wavelength of sensitizer fluorescence sensitizer quantum yeild absorption of an absorber layer under vacuum wavelengths refractive index Forster energy transfer energy transfer constant numerical aperature point spread function diffraction limited resolution charged coupled device  mean molecular area xi  PC PS HMDE DOPC AmB STM NR FTIR PM-IRRAS SFG EQCM ECL  phosphatidylcholine phosphatidylserine hanging mercury drop electrode dioleoyl phosphatidylcholine Amphotericin B scanning tunneling microscopy neutron reflectrometry fourier transform infrared spectroscopy polarization modulation infrared reflection absorption spectroscopy sum frequency generation electrochemical quartz crystal microbalance electrogenerated chemiluminescence  Chapter 4 DiIC18(5) BODIPY-C10-SH WE CE RE SCE HPLC CV MS GS C hm  Kc  rms CO ^base  ^des  WD ''ads image I Io ^ads^o  AI/I AI  0  1,1 '-dioctadecyl-3,3,3',3'-tetramethylindodicarbocyanine perchlorate 4,4-difluoro-1,3,5,7-tetramethyl-8-[(10-mercapto)]-4-bora-3a,4adiaza-s-indacene working electrode counter electrode reference electrode saturated calomel electrode high performance liquid chromatography cyclic voltammetry metal solution interface gas solution interface capacity imaginary current real current A C voltage root mean squared frequency base potential variable potential desorption potential working distance adsorption potential variable image potentail image intensity image intensity without fluorophore flat field corrected intensity mean histogram intensity or fluorescence intensity mean histogram intensity or fluorescence intensity  xii  Chapter 5  r  min  Yo  Yi Y2  71  minimum capacitance electrocapillary equation for an uncoated electrode electrocapillary equation for the potential defined state 1 electrocapillary equation for the potential defined state 2 film pressure  Chapter 6 PECVD  plasma enhanced chemical vapour deposition  Chapter 7 FRAP SERS [A] [Ap] [X] kp k  a  [A(t)j [Ap(t)l [A ] q  fluorescence recovery after photobleaching surface enhanced raman spectroscopy concentration of dye monomer concentration of photobleached dye concentration of non-absorbing dye aggregate photobleaching rate constant aggregation rate constant time evolution of the concentration of monomer time evolution of the concentration of photobleached dye concentration of quenched dye monomer  Chapter 8 EBSD SEM  electron back scattered diffraction scanning electron microscope  Chapter 9 SPR  surface plasmon resonance spectroscopy  xiii  List of Figures Figure  2-1 Schematic depiction of a simple capacitor. When the capacitor is connected to a power source, charge collects on the plates 9  Figure  2-2 Proposed model for the double layer region including specifically adsorbed anions. The Inner Helmholtz Plane (IHP) and Outer Helmholtz Plane (OHP) are shown in the figure with the corresponding positions from the metal surface (x), potential ((b), and charge density (o). Taken from [22] 9  Figure  2-3 (a) Double layer capacitance described by the (GCS) model. C is a series capacitor made up of the Helmholtz and diffuse layer capacitances, (b) The calculated potential profile through the solution side of the double layer. (1:1 non-specifically adsorbing electrolyte in water at 25°C). Taken from [22] 12  Figure  2-4 The variation of electrostatic potential near a charged surface, (a) Short range image potentials and (b) long range Coulombic potentials, (c) Total potential resulting from both Coulombic and image potentials showing the independence of distance from the surface over the range 10" to 10" cm approximately. Taken from [27] 15  d  5  3  Figure  2-5 The principle low index planes for fee crystals showing the (100), (110) and (111) faces. The right column represents the cutting plane of the unit cell and the left column shows the resulting crystal face. For clarity, atoms in the center of the cube faces have been omitted unless they are contained within the cutting plane. Taken from [29] 17  Figure  2-6 Measured work functions for various Cu surfaces. measured for the smoothest surface. Taken from [30]  Figure  2-7 The variation in the concentration (c,) of component / within the z coordinate between two bulk phases a and p. Taken from [31]. . 18  Figure  2-8 (a) Actual system showing the two pure phases a and P and the interfacial region, (b) Hypothetical model system representing the Gibbs dividing surface 18  Figure  2-9 Simulation of the dependence of film pressure on the bulk concentration of the adsorbate for adsorption isotherms described by Henry, Langmuir and Frumkin. The lateral interaction parameter (a) is indicated on the curves. Taken from [32] 23  Figure  2-10 Differential capacitance of a Hg electrode in I M K N 0 at 25°C (dotted line). The solid lines represent the capacitance of the same electrode in contact with octyl alcohol. The arrows represent the frequency (kHz) used in the capacitance measurement. A depression in capacitance is noticed near potentials around the pzc. Pseudo-capacitance peaks are noted at potentials positive and negative of the pzc. Taken from [23] 26  The largest work function is 17  3  xiv  Figure 2-11 A table showing the effectiveness of functional groups in providing attraction to water and the strength of the monolayer film as a result of the hydrocarbon chain length. Taken from [46] 29 Figure 2-12 Schematic depiction of a monolayer in a Langmuir trough film balance. The movable barrier allows for compression and decompression of the organic layer. The Wilhelmy plate is used for the measurement of film pressure 29 Figure 2-13 Schematic representation of a Wilhelmy plate of known height, width and length immersed into solution. The plate is connected to an electromicrobalance (EMB) for the measurement of force. The contact angle between the water and plate is measured by 0. 30 Figure 2-14 A graph showing the surface pressure vs. area isotherm curve. The various states of the monolayer at different compressions are shown in each region. Taken from [47] 32 Figure 2-15 A representation of the interfacial tension between liquid-vapour (LV), solid-vapour (SV) and solid-liquid (SL) for a liquid drop resting on a smooth solid surface. A contact angle, 0 is formed between the liquid and solid 32 Figure 2-16 Possible molecular distributions that result from various mixing of two surfactants. The films can be (a) miscible, (b) complete separtated, or (c) slightly immiscible 35 Figure 2-17 Types of monolayer deposition on a solid support, (a) X-Deposition, (b) Y-Deposition and (c) Z-Deposition. Taken from [53] 35 Figure 2-18 The horizontal touching method for transferring insoluble molecules to a hydrophobic surface 36 Figure 2-19 A Jablonski diagram showing the emission processes for electronic transitions. Electronic states A and B have the same multiplicity but C does not 39 Figure 2-20 Monomer and dimer formation in dye monolayers. The mixed monolayers of the dye represented at the top of the figure, and cadmium arachidate were deposited on a glass substrate. E is the extinction coefficient of the dye, and a is the number of dye molecules per unit area. The monomer band is 490 nm, and the dimer band is at 460 nm. Taken from [53] 41 Figure 2-21 Exciton band energy diagram for a molecular dimer with parallel transition dipoles. The ovals correspond to the molecular profile, and the double arrow indicate the polarization axis. Taken from [60] 43 Figure 2-22 Exciton band energy diagram for a molecular dimer with in-line transition dipoles. xv  Taken from [60]  43  Figure 2-23 A multilayer system for measuring the energy transfer between sensitizer (S) and acceptor (A) at various separation distances. S absorbs U V radiation and fluoresces in blue. A absorbs blue radiation and fluoresces in yellow. Taken from [53] 45 Figure 2-24 Calculated (line) and experimental luminescence (symbols) of S with separation distance from A. Taken from [53] 45 Figure 2-25 Calculated (lines) and experimental (symbols) excited state lifetimes for a Cu/Eu+3/air system at various separation distances from Cu. The excited state lifetime is small at small separations between the dye and metal. Taken from [63] 47 Figure 2-26 Typical substituents in cyanine dyes with the general structure X(-CH=CH) -CH=Y. The dominant factor in determining A is n. Taken from [74] 50 n  max  Figure 2-27 Schematic representation of Airy discs. Light is passed through a small slit (1), the objective (2) creating an Airy disc (3) due to diffraction rings from the objective. The top of the figure represents the intensity distribution of the Airy disc. Taken from [77] 51 Figure 2-28 A schematic representation of overlapping Airy discs. When the two disc are far apart they are easily resolved. When they begin to overlap (—) and ( ) a limit d is reached which is known as the maximum diffraction limited resolution. Taken from [77] 51 a  Figure 2-29 A schematic representation of the light path in epi-illumination. The light is passed through an exciter filter selecting a band of light. This light is directed through the objective to the sample. Fluorescent (and scattered) light is collected through the objective. Long wavelength light (fluorescence) is allowed to pass through the dichroic filter. A final filter to remove any excitation light is the barrier filter. This light is effectively coupled onto the whole array of detectors of the CCD 53 Figure 2-30 Transmission spectra for a filter cube consisting of an exciter, dichroic and emission filter assembly. Data taken from [78] 54 Figure 4-1 Ball and stick representation of the surfactants that physisorb onto Au(l 11) in the ideal all trans configuration, (a) Octadecanol or C180H and (b) cis-9-Octadecen-l-ol or OLA 67 Figure 4-2 Ball and stick representation of the fluorescent probes that are mixed with the surfactants (a, b) and that chemisorb onto the metal surface (c) 67  xvi  Figure 4-3 Schematic representation of the three electrode cell used in the electrochemical characterization. The WE is either Au(l 11) or a polycrystalline Au bead, the CE is a Pt or Au coil, and the RE is a saturated calomel electrode (SCE). The RE is connected to the cell through a salt bridge 70 Figure 4-4 (a) Full scale CV (b) double layer CV and (c) capacitance for Au(l 11) in contact with 0.05M KC10 . Measurement of capacitance used a 5mV rms, 25Hz A C perturbation and the capacitance was calculated assuming the interface can be modeled as a series RC circuit. 71 4  Figure 4-5 Schematic depiction for the procedure of spreading the insoluble alcohols onto the electrolyte surface and the transfer of the alcohols to the Au(l 11) electrode surface. The hanging meniscus arrangement is also shown 73 Figure 4-6 Schematic depiction for the procedure of forming the SAM externally from the cell and the subsequent rinsing and transfer to the electrochemical cell. The bead is submerged through the electrolyte surface rather than forming a hanging meniscus. The fluorescent thiol is not present at the gas/solution interface 73 Figure 4-7 A depiction of the potential steps used in chronocoulometry. E is chosen at a potential where the system can equilibrate. E is a variable potential that increments in the positive direction after each measurement cycle. E is the desorption potential value where the surfactant is removed from the metal surface. The charging current is measured between the step from E to E 76 base  var  dcs  var  des  Figure 4-8 Absorption and emission spectra for (a) DiIC18(5), (b) Fluorescein and (c) BODIPYC10-SH with the excitation and emission filters of the Chroma Set 41008 and UM-WIBA overlayed. The spectra for DilC 18(5) was taken from [ 196]. The spectra for fluorescein was taken from the Molecular Probes reference standard [197] and the spectra for BODIPY was taken from a similar molecule to BODIPY-C10-SH [198] 77 Figure 4-9 Quantum efficiency of the Kodak KAI-2092 CCD. Taken from [200]  78  Figure 4-10 Spectroelectrochemical cell used in the electro-fluorescence experiments. The coverglass above the objective is 0.17mm thick. The objective is interchangeable with a 1 OX objective and the SPOT CCD can be switched with a spectrophotometer. The filter is on a rotation device to allow for interchange of filters during one experiment 80 Figure 4-11 Schematic depiction of the potential control during the acquisition of images in the double layer potential region. The values of E are not an exact representation of the experimental values and may vary for a particular experiment, (a) represents the potential variation for image_scan.vi. (b) represents the potential variation for image_adsorb_neg.vi and (c) represents the potential variation for image_adsorb_pos.vi. The symbol ^ indicates xvii  the potential where the images were acquired.  82  Figure 4-12 Schematic depiction of the procedure for rendering masked AI/I images from which the average histogram is calculated. The average histogram is calculated excluding zeros created from the mask. Fluorescence intensity is assumed proportional to the average grey scale 85 0  Figure 4-13 (a) Calculated mean histogram intensity for the images of the Au(l 11) electrode in air using the UM-WIBA (o) or Chroma (0) filter cube. The decay observed for the UM-WIBA filter set was corrected assuming a linear decay over 5 minutes (300 sec) intervals. The correction (•) is shown in (b) 86 Figure 5-1 Cyclic voltammetry (left column) and differential capacitance (right column) for Au( 111) in the absence (•••) and presence (—) of the surfactant. Thin lines represent the negative going potential scan and bold lines represent the positive going scan. The capacitance was calculated using a 5 mV rms AC potential perturbation consisting of 25Hz 88 Figure 5-2 Measured capacitance at 0.150 V/SCE for the various mixtures of OLA in C180H. The error bars reveal that only 0, 25, and 100 mol% layers are reproducible 91 Figure 5-3 Charge density curves measured for the various mixtures of OLA/C180H adsorbed onto Au(l 11). The pzc for the Au(l 11) electrode in contact with KC10 occurs at a potential 0.255 V/SCE. The maximum film pressure is obtained where the charge density for the coated and uncoated electrode cross. . 93 4  Figure 5-4 Calculated film pressure for the various mixtures of O L A and C180H adsorbed on Au(l 11). The film pressure was calculated using the back integration method. The inset shows the variation in the maximum film pressure with O L A content 95 Figure 5-5 Schematic representation of three electrocapillary curves for three states of the adsorbed layer. y represents the potential region for the adsorbed layer. Yi represents the disrupted layer at potentials negative of the pseudo-capacitance peaks and y represents the electrocapillary curve for the uncoated electrode. The film pressure is calculated by difference between the electrocapillary curves. The lowest energy will be maintained as represented by the bold line 96 2  0  Figure 5-6 Cyclic voltammetry and differential capacitance for the various concentrations of DilC 18(5) in C180H. The inset shows the variation in the minimum capacitance (measured at 0.150 V/SCE) with dye content 99 Figure 5-7 The variation in maximum film pressure measured at the pzc for the various mixtures ofDiIC18(5)inC180H 100  xviii  Figure 6-1 (a) Fluorescence images, (b) capacitance and (c) calculated fluorescence intensity obtained from an image/potential scan investigation of 3 mol% DilC 18(5)/C 180H adsorbed on Au(l 11). In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a) 104 Figure 6-2 The masking procedure: Thefirstadsorption I/I image (a) shows intense features on the electrode. The 'hotspot' mask purposely overestimates these features to effectively mask these intense regions. These 'hotspots' are eliminated from the calculation of the mean histogram intensity by application of the mask to the AI/I image. For completeness, the same example is given for the normalized desorption image (b) 108 0  G  Figure 6-3 (a) Outlined fluorescence features, (b) number of image features and (c) the mean area calculated for the data presented in Figure 6-1. The open and closed symbols represent the negative and positive potential scan directions respectively 110 Figure 6-4 Results obtained for a second image/potential scan investigation for the same adsorbed layer of 3 mol% DiIC18(5)/C180Ff showing (a) the fluorescence images, (b) capacitance, (c) number of image features, and (d) their mean area. In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a) 112 Figure 6-5 Results obtained for a image/potential step investigation for the same adsorbed layer of 3 mol% DilC 18(5)/C 180H showing (a) the fluorescence images, (b) capacitance, (c) number of image features and (e) their mean area. In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a) 114 Figure 6-6 An image/potential scan investigation for an adsorbed C180H layer containing 2 mol% 5-octadecanoylaminofluorescein and 2 mol% DilC 18(5). The images superimposed on the capacitance plot were modified as AI/I„ and then pseudo-coloured to green (fluorescein) and red (DilC 18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot 118 Figure 6-7 Image statistics obtained for the two dye/C180H layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively 120 xix  Figure 6-8 (a) A schematic representation of the stepped template used to measure the fluorescence intensity of an adsorbed layer of 2 mol% DiIC18(5)/2 mol% 5octadecanoylaminofluorescein/C180H separated from a 300 nm thick Au surface by controlled thicknesses of Si0 . (b) The measured fluorescence intensity for both DilC 18(5) and 5-octadecanoylaminofluorescein. The fit from Equation 2-56 for the DilC 18(5) is shown in (c) where (o) represents the data and the line represents the fit. 122 2  Figure 6-9 An image/potential scan investigation for an adsorbed O L A layer containing 2 mol% 5octadecanoylaminofluorescein and 2 mol% DilC 18(5). The images superimposed on the capacitance plot were modified as AI/I and then pseudo-coloured to green (fluorescein) and red (DilC 18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot 126 0  Figure 6-10 Image statistics obtained for the two dye/OLA layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively 128 Figure 6-11 A n image/potential scan investigation for an adsorbed 25 mol% OLA/C180H layer containing 2 mol% 5-octadecanoylaminofluorescein and 2 mol% DilC 18(5). The images superimposed on the capacitance plot were modified as AI/I and then pseudo-coloured to green (fluorescein) and red (DilC 18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot 131 0  Figure 6-12 Image statistics obtained for the two dye/OLA layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively 133 Figure 6-13 The influence of surface irregularities on the desorbed structure of C180H. Image (a) is a brightfield illuminated electrode with adsorbed surfactant before potential cycling. In (b) the pits from electrochemical and mechanical polishing are captured. Image (c) shows the defect mask applied to the desorption Al/I image of Figure 6-1. 135 0  Figure 7-1 (a) Measured capacitance and (b) the calculated fluorescence intensity of (c) selected images in a fluorescence decay and recovery investigation of 3 mol% DiIC18(5)/C180H. xx  The small intense feature in the images is due to a small shift in the electrode during measurements allowing some intense features to be observed after subtraction of an I /I image. The mask used in the calculation of fluorescence intensity was exaggerated to cover this feature. The labelled images correspond to the labels in (b) 140 ads  0  Figure 7-2 (a) Capacitance and (b) calculated fluorescence intensity of (c) selected images in a fluorescence decay and recovery investigation of a separate 3 mol% DilC 18(5)/C 180H. The data is similar to Figure 7-1. The images in (c) are labelled corresponding to the labels in (b) 142 Figure 7-3 A fluorescence decay and recovery investigation of a 3 mol% DiIC18(5)/C180H layer during variable exposure times to the excitation source. The first, second and third decays shown in (b) have light on:off for lmimlmin, lmin:2min and lmin:5min respectively without changes in the duty cycle of the potential perturbation shown in (a). The images in (a) correspond to the labels in (b) 145 Figure 7-4 The fluorescence decays of Figure 7-3 normalized to the first desorption image in each decay 147 Figure 7-5 The measurement of four fluorescence decays from a 2 mol% DiIC18(5)/2 mol% 5octadecanoylaminofluorescein/C180H layer. The first two decays were measured for fluorescence of DilC 18(5) (top panel) and the last two decays were measured on the same adsorbed layer and on the same region of the electrode surface for 5octadecanoylaminofluorescein fluorescence (bottom panel) 147 Figure 7-6 Normalized fluorescence spectra of DilC 18(5) measured between 660 and 740 nm during the 20 minutes spent at desorption for a typical fluorescence decay. The solid line represents the spectrum measured when the layer was first desorbed, and the dotted line after twenty minutes of desorption 151 Figure 7-7 (a) The first and last fluorescence decays from Figure 7-1 with the kinetic model fits overlaid. The bottom panel of (a) represents the same decays/fits on log scale. The symbols defined in the legend are the experimental results and the solid line represents the calculated fits. The model parameters are shown in (b) 155 Figure 7-8 A schematic depiction of the fluorescence intensity at different separation distances between the metal and fluorophore. The model assumes a distance of separation of d* where fluorescence quenching is minimal. At a position d , some of the fluorophores will be quenched and not accounted for in the model 157 0  Figure 7-9 (a) Measured fluorescence decays of DiIC18(5)/OLA at -0.600 (o), -0.650 (0), -0.700 (A), and -0.750 V/SCE (>). The line represents the kinetic model fit for each decay. The aggregation rate constants are shown in (b) 159 xxi  Figure 8-1 Measured capacitance for a S A M coated Au bead electrode. The capacitance for the bare electrode (•••) was measured before the S A M formation. The capacitance curves for the SAM coated electrode (symbols) were measured during the imaging experiment. One S A M coated electrode was exposed to the desorption limits of E = -0.900 (a), -1.250 (b), -1.300 (c), 1.350 (d) and -1.400 V/SCE (e) 164 Figure 8-2 Measured fluorescence images and calculated intensity of the reductive desorption of the S A M from the Au bead. The calculated fluorescence intensity plots (a - e) correspond to the capacitance measurements in Figure 8-1 (a - e). Selected images are represented at the bottom at the potentials labeled on the graph 167 Figure 8-3 Delineation of features in the desorption images at (a) -.1350, (b) -1.300, and (c) -1.250 V/SCE with the outlines displayed in the middle column. The application of the outlines to the desorption image at -1.300 V/SCE is displayed in the third column 170 Figure 8-4 (a) SEM image of the etched bead revealing surface heterogeneity, facets and grain boundaries, (b) Optical image of the etched bead positioned as in the experiment with the surface features labeled, (c) Optical image with the grain boundaries highlighted. Image (c) with the fluorescent features of E = -1.250 (d), -1.300 (e) and -1.350 V/SCE (f) applied. 171 des  Figure 9-1 Proposed mechanism for the desorption/re-adsorption of organic molecules from/to the electrode surface. Fluorescence images taken from experiment are placed at unique potential regions around a typical C180H/Au(l 11) capacitance curve. The fluorescence images are cropped and enlarged according to the reference image. An enlarged schematic depiction of the interface is shown beneath the fluorescent images. In the schematic representation, the shading or gray value represent the fluorescence intensity. Black represents non-fluorescent and white represents fluorescent 181  xxu  List of Tables Table 4-1 The measured extinction coefficients for DilC 18(5), the Fluorescein reference standard, and a similar compount to BODIPY-C10-SH. The extinction coefficient was measured at the indicated wavelength for the dye dissolved in the indicated solvent 68  xxiii  Chapter 1 1  Introduction Surface active molecules that form organized compounds in the aqueous phase such as  micelles have been used in numerous applications. Not only are these molecules a critical component of biological cell membranes, but they have found both industrial and general application as soaps and detergents because of their unique interfacial properties. When these molecules adhere to an interface, the surface properties are modified. This modification has been accomplished on solid substrates for applications including corrosion, lubrication and catalysis [1] optical devices [2] electronic devices, and sensors including both chemical and biochemical [3,4]. Furthermore, the adsorption of these lipid-like organic molecules onto the solid support, offer an ideal environment for the immobilization of biologically important proteins for characterization with various surface probe microscopies. The lipid-like layer in this scenario acts as a cushion for the deposited protein to overcome denaturation during the characterization. One important variable that is capable of further modifying the character and function of the adsorbed surfactant is the electrical potential of the interface. Thus, when adsorbed onto electrode surfaces the properties of the interface can be changed for a desired purpose. Organic modification of noble metal electrodes is most commonly achieved through the selfassembly of thiol molecules [5] although Langmuir-Blodgett (LB) deposition is also a common surface modification method [2,4]. Self-assembled monolayers (SAMs) of thiols on Au is of particular interest due to the ease of forming the interface and because of the strong covalent bond formed between the sulfur group and the metal composing a robust interface. The lack of oxidation on the A u surface makes this a stable material in a variety of environments and chemical and biosensors are now becoming increasingly studied [6-8]. By changing the charge on the metal surface with electrochemical methods, the character and therefore function of the adsorbed monolayer can be modified. Electric potential can induce conformational changes in the adsorbed layer and in some cases result in the desorption of the organic from the metal surface. This allows for a wider scope of chemical and biochemical sensing because the orientation or placement of the adsorbed layer can be chemically tuned to respond with reactants in solution. For biosensors, the  1  immobilization of thiols having a biological receptor for selectively binding ligand molecules from the surrounding phase, is desirable. One example includes immobilizing oligonucleotides (receptor) onto an electrode surface for detection of DNA (ligand) hybridization [9]. A variety of functional groups on the surface would result in a material with multiple selectivity and thus a larger range of application. The ability to arrange the adsorbed molecules containing the different functional groups in a desired spatial distribution would increase the specificity and increase the ease of characterization. Thus, methods to induce selective adsorption and/or desorption of the molecules may be required. A common method to create mixed SAMs is through a thiol exchange reaction where the thiol coated metal is introduced to a solution containing a different thiol. A thiol exchange reaction occurs when a different thiol replaces a previously adsorbed thiol. Electrochemistry can also be employed to form a patterned surface with a more specific spatial distribution of the different functional groups. It is understood that different crystal faces have different binding energies for the Au-S bond [10]. This property can be exploited for the selective desorption of a thiol from a desired face of a polycrystalline metal surface and the inclusion of a new thiol from a separate solution resulting in the desired mixed monolayer. The electrical potential of the interface is then a powerful variable in modifying the chemical properties of these unique interfaces. However, the application of such interfaces should not go without proper characterization of the potential induced behaviour. While electrochemical methods can detail the character of the adsorbed layer, they are essentially average techniques and as such are limited to reporting the average properties of the system. The merging of electrochemistry with in situ spectroscopic techniques will help develop a more overall picture of the response of these unique interfaces to the electric variable. The development of such an in situ technique was the primary goal of this thesis project.  1.1  Objectives The spectro-electrochemical characterization of water-insoluble surfactants deposited onto  well-defined electrode surfaces is of principle investigation in this lab. The lipids or lipid-like materials in these investigations are typically spread as a monolayer at the gas/electrolyte interface in the electrochemical cell. The transfer of these molecules onto the electrode surface is achieved 2  by touching the uncharged metal to the floating monolayer. This method of coating the electrode surface renders a unique interface where all of the surfactant material is confined to the surface of the electrolyte or the electrode surface. As such, these surfactant materials do not exist in a reservoir in the bulk electrolyte as would be in the case for water-soluble organic compounds. These systems are then metastable and unique to the adsorption arena. The long term goal of the Bizzotto lab is to understand the influence of transmembrane potentials on the biological cell. The transmembrane potential arises due to a difference in ion concentration between the inside and outside of the cell wall. From this perspective, lipid-like molecules deposited onto the electrode surface resemble V2 of the biological cell. The organic layer will eventually incorporate membrane proteins and the electric potential of the electrode will be changed. This however, cannot be done until the full characterization of the lipid-like cushion layer is first accomplished which requires more than just electrochemical investigations. The goal of this thesis project was to develop a non-destructive in situ technique to monitor the potential induced changes of the organic modified surface. This was accomplished through the union of fluorescence microscopy with electrochemical methods and is termed electro-fluorescence microscopy. A n inverted epi-fluorescent microscope was used to monitor the potential induced behaviour of lipid-like surfactants adsorbed onto the electrode surface. If the molecules in question do not have a fluorescent moiety in their chemical make up small amounts of a fluorescent probe were mixed with the insoluble surfactant.  Initial studies were performed on a 1-octadecanol  (C180H) coated Au(l 11) electrode surface but other systems were also studied during the course of this project. These systems include the physisorption of fluid oleyl alcohol (OLA) or mixed monolayers of C180H and OLA on Au(l 11) and the chemisorption of a fluorescent thiol onto a polycrystalline Au bead. Studies such as these are involved in the current motif for biochemical or chemical sensor array development. Provided that the spatial resolution of the electro-fluorescent technique is on the order of the sensor array, our method may contribute to the development and characterization of such devices.  1.2  Rationale The charge of the metal support, controlled by electric potential, plays a significant role in 3  modifying the physical state of the adsorbed surfactants. Previous studies of the octadecanol coated A u ( l l l ) electrode have shown that extreme negative potentials can induce desorption of the surfactant which is replaced by adsorbed water.  This was confirmed with electrochemical  measurements and in situ optical techniques including electroreflectance, elastically scattered light and preliminary fluorescence investigations [11-17]. In these investigations it was proposed that the desorbed molecules remain near, but not on the electrode surface and will re-adsorb onto a less negatively charged metal. Since the re-adsorption of these molecules cannot happen from a reservoir in the bulk electrolyte, the desorbed molecules were proposed to exist in an aggregated morphology, however, this was not directly observed. Furthermore, the fate of the desorbed molecules near the electrode surface was not fully described. The C 1 8 0 H / A u ( l l l ) interface was chosen as a starting point for the developed electrofluorescence technique since the adsorption/desorption process was extensively studied and some questions remained open. The system was then extended to monitor the behaviour of oleyl alcohol (OLA) adsorbed on the Au(l 11) electrode. Investigations of mixed monolayers containing various concentrations of OLA/C180H were also characterized with the developed technique. Because C180H and OLA do not have a fluorescent moiety in their chemical make up, the surfactants were mixed with a small amount of a fluorescent carbocyanine dye molecule having the same alkyl chain length as the surfactants. Carbocyanine dye molecules are commonly used as biomembrane probes because they partition into cell walls [18-21]. Thus the carbocyanine probe is believed to exists in the lipid-like C180H or O L A matrix and faithfully report on the behaviour of the surfactant. Initially, the electrochemical characterization of C180H, OLA and mixed monolayers of OLA/C180H were conducted because a full and exhaustive understanding of the electrochemical characteristics of the system are required before results are meaningful from the in situ technique. Furthermore, C180H was studied electrochemically with the incorporation of small amounts of the carbocyanine dye molecule to understand any disruptive effects that the dye may cause in the electrochemical character of the surfactant in question. A 3 mol% mixture of dye/C180H was imaged under full potential control using the developed technique. To verify that the carbocyanine dye was effectively reporting on the potential induced behaviour of C180H, a second system was studied where C180H was mixed with two dye molecules and the desorbed structures compared.  4  Since the incorporation of the two dye molecules had little disruption effects to the electrochemical character of the adsorbed surfactant, the adsorbed OLA and OLA/C180H mixed monolayers were studied with both dye molecules. As a final investigation, a fluorescent molecule bearing a thiol moiety was self-assembled onto a polycrystalline Au bead and the reductive desorption monitored.  1.3  Scope of Thesis The work presented in this thesis encompass a large body of knowledge from various  scientific disciplines. Chapter 2 concisely outlines the fundamental theories necessary for a complete understanding of the electrochemical and spectroscopic results.  The fundamental method of  characterizing the adsorbed surfactants is based on electrochemical techniques thus, a background of the theories for the electrode/solution interface is initially covered in Chapter 2. Some simple electrochemical concepts are dealt with first, which lead into an in-depth description of the electric double layer as described by the Gouy-Chapman-Stern (GCS) theory.  This is followed by  thermodynamic treatments of the interface including the Gibbs adsorption isotherm, the electrocapillary equation and the Langmuir adsorption isotherm. Characteristics of the insoluble monolayer present at the gas/solution and metal/solution interface are described next and will facilitate the understanding of the method by which the surfactant is transferred to the metal surface. A detailed description of the fluorescence theory including the factors that affect fluorescence intensity are also outlined in the chapter. This section is highly concentrated on the quenching of luminescence of fluorescent dye molecules near a metal surface.  A brief description of the  microscopic technique is given at the end of the chapter and details the maximum or diffraction limited spatial resolution that can be obtained with the fluorescent microscope. Chapter 3 is a review of previous studies relevant to the topic of this thesis. The first concept covered is the initial studies of the adsorption of phospholipids onto a Hg electrode. The techniques and results of this work greatly motivated the study of the adsorption onto planar solid electrodes. The electrochemical and spectroscopic characterization of the C180H/Au( 111) interface is reviewed followed by the pioneering studies measuring the excited state lifetimes of fluorescent dye molecules separated from a metal surface. The chapter ends with a brief review of SAMs on noble metals as this is the most common organic modification of electrode surfaces. 5  The experimental methodology is presented in Chapter 4. This section gives a summary of the systems studied in this thesis as well as the materials, methods, and instrumentation. The in situ technique is described, and the procedure by which the monolayers are deposited onto the electrode is summarized. Experimental results for the electrochemical characterization of C180H, OLA and mixed monolayers of OLA/C180H adsorbed onto Au( 111) is given in Chapter 5. This section also presents the amount of disruption that the carbocyanine fluorescent dye molecule introduces to the adsorbed layers of C180H. Fluorescence imaging of these layers adsorbed on Au(l 11) is detailed in Chapter 6, and the observation of unique photophysical behaviour of the carbocyanine dye separated from the metal surface is given in Chapter 7. The final section of experimental results is given in Chapter 8 describing the observation of a selective reductive desorption process of a SAM coated polycrystalline Au bead electrode. The results are summarized in Chapter 9 drawing conclusions from the various studies. This chapter will aim to bring together the observed results for the development an overall picture of the potential induced adsorption/desorption process of these organic modified interfaces. suggestions for future study are also presented.  6  Some  Chapter 2 2  Theoretical Background This chapter briefly outlines the fundamental theories of the electrode/solution interface, the  properties of insoluble monolayers, and fluorescence from monolayer assemblies near metal surfaces. A description of the electric double layer as proposed by the Gouy-Chapman-Stern theory is detailed first. The electrocapillary equation is described next, focused on molecular adsorption onto electrode surfaces. Finally, the properties of insoluble monolayers at solid or liquid interfaces and a description of the factors that affect fluorescence intensity are presented.  2.1  Qualitative Description of the Electric Double Layer Ions in the bulk of solution exist in a spherically symmetric environment that is electrically  neutral within a given region. This cannot be said for ions that exist near a boundary such as an electrode surface. For a charged electrode, ions near the interface experience electrostatic forces that are not present in the bulk. These forces can result in a large concentration of ions near the electrode to counterbalance the surface charge. Furthermore, water molecules (in aqueous solutions) may take on a preferred orientation at the electrode surface inducing a net dipole. It is the charges and oriented dipoles extending from the metal to the bulk solution that make up the electric double layer.  2.1.1 Electrochemical Concepts Under certain conditions, an electrode/solution interface may have a range of potentials where charge transfer  reactions (oxidation/reduction) do not occur because they are  thermodynamically or kinetically unfavourable. In this range of potentials the electrode is termed an ideally polarized electrode (IPE). Although no charge is transferred within this range of potentials, adsorption and desorption of molecules to and from the electrode may still occur and nonfaradaic currents may arise due to changes at the electrode/solution interface. These may include changes in the electrode area, or composition/orientation of the material at the electrode surface. Within the IPE potential region, the electrode/solution interface is analogous to a capacitor for which simple relations exist.  7  2.1.2  Simple  Capacitor  A simple capacitor is schematically represented in Figure 2-1. The capacitance (C, measured in farads (F)) is a function of the distance separating the two plates (d) and the dielectric material (e) between the plates. Applying a voltage (E, measured in volts (V)) results in the accumulation of charge on the plates (q, measured in Coulombs (C)) according to the equation ££  q  n  C  =  —  d  -  =  —  E  •  2-1  The term e is the permitivity of free space. A variation in the potential will induce a charging 0  current (/, measured in amps (A)) dq  I=—=  C  dt  dE  2-2  dt  This is referred to as a capacitive current, and is different from current due to electron transfer. For a circuit consisting of n capacitors in series, the total capacitance is given by 1 Cseries  1  1 1 + —+•••• + — Q  2-3  ^2  These simple relations are useful when describing the electrochemical characteristics of a real electrochemical system within the IPE potential region.  2.1.3  Capacitance  of the Electrode/Solution  Interface  At a given potential, a charge may reside at the metal surface (q ) which is counterbalanced M  by a build up of opposite charge in solution (q ). This separation of charge results in the double layer s  capacitance (C ) that is a function of potential. Three theories in describing the electric double layer dl  have been given by Helmholtz, Gouy and Chapman, and Gouy, Chapman and Stern [22]. This has led to the most complete picture of the electric double layer as shown in Figure 2-2. Helmholtz, proposed that the interface was analogous to a simple capacitor. It was believed that the charge on the metal surface was counterbalanced by a thin section of charge in solution. Gouy and Chapman proposed that the charge in solution would exist more diffusely packed since the conductivities of metals and solution are distinctly different. They introduced a diffuse layer into the solution side of the Helmholtz model. A later refinement by Stern accounted for the finite size of ions suggesting 8  Battery  88 o  +•  Capacitor  + + + +  Figure 2-1 Schematic depiction of a simple capacitor. When the capacitor is connected to a power source, charge collects on the plates.  Figure 2-2 Proposed model for the double layer region including specifically adsorbed anions. The Inner Helmholtz Plane (IHP) and Outer Helmholtz Plane (OHP) are shown in the figure with the corresponding positions from the metal surface (x), potential ((J)), and charge density (a). Taken from [22]. 9  that closest approach of an ion to the metal surface would be its radius after loss of the solvation shell. This treatment introduces an inner layer between the metal and the diffuse layer of solution. As shown in Figure 2-2, the solution side is broken up into several layers. The inner layer exists closest to the electrode surface and contains solvent molecules and sometimes specifically adsorbed ions or molecules. The inner layer is made up of the inner Helmholtz plane (IHP) and the outer Helmholtz plane (OHP). The IHP resides at position x, from the metal surface and is defined as the closest approach of the electrical center of a specifically adsorbed anion after loss of solvation shell. The potential and charge density at the IHP is given by (J) (V) and a ' (pC/cm ) respectively. 2  ;  Solvated ions can approach the metal only to a distance x defined as the closest approach of the 2  electrical center of the nearest solvated ion. This defines the OHP and the potential at point x is 2  given by cj>. The diffuse layer of solution extends from the OHP to the bulk of solution. This is a 2  three dimensional layer containing the nonspecifically adsorbed ions in solution and has a charge density a . The total excess charge density on the solution side of the interface (a ) is given by s  _.S'  where a  M  _ /  .  —d  ~.M  f  A  is the charge density on the metal. The double layer capacitance is made up of two  capacitors in series (inner and diffuse layer) as shown by the Gouy-Chapman-Stern ( G C S ) theory.  2.2  Quantitative Description of the Electric Double Layer  2.2.1  The Gouy-Chapman-Stern (GCS) Theory The G C S theory is completely derived in [22]. The model treats the electric double layer in  the absence of specific adsorption. The solution is divided into small sections (laminae) of equal thickness starting at the OHP and extending into the bulk electrolyte. The laminae represent the population of ionic species with varying energies (with respect to a reference laminae in the bulk). The total charge per volume of any lamina is calculated using the Poisson-Boltzmann equation describing the gradient of potential from the OHP into the bulk of solution. This is represented as (  dx  '  MTn°  X=X2  10  A  2  ( ze<p~ ^ sinh 2kT  2-5  where, k is the Boltzmann constant (J K" ), Tis temperature (K), n° is the number concentration of 1  the ions in the reference lamina (cm ), z is the signed charge of the ion, e is the charge on an electron 3  ( Q , and c\> is the electrostatic potential (V). A profile of the potential in the diffuse layer can be accomplished by integrating Equation 2-5 between the limits of (J) and (p to obtain 2  tanh  ^ zecp^ 4kT  2-6  zecp ^  (  2  tanh  4kT  where K is represented as  2n°z e ^ K eekT f  2  2  2-7  The inverse of K is the Debye length representing the diffuse layer thickness. Taking c = n°/N  A  (where N is Avagadro's number) and water (e = 78.49) at 298 K, the Debye length is 3.04 x 10"  8  A  (z" c' cm). For dilute solutions, the Debye length, and hence the diffuse layer thickness increase. 1  m  In the absence of specific adsorption, only water is present in the inner layer and the charge density at any point from the electrode surface to the O H P is zero. Therefore, the total potential drop from the metal surface (cb ) across the double layer in the absence of specific adsorption is 0  given by 2-8  <t> = 02 " 0  The distribution of potential across the double layer can be visualized in Figure 2-3(b). Potential drops linearly from the metal to the OHP, then exponentially into the bulk of solution. A relation between the charge density on the metal and (\) can be obtained using a Gaussian enclosure and the a  Gauss law to obtain .M  ( ££,  d(p_ dx  = (MT££ n )  0 2  0  11  sinh  ze(p  2  2kT  2-9  > E  Figure 2-3 (a) Double layer capacitance described by the (GCS) model. C is a series capacitor made up of the Helmholtz and diffuse layer capacitances, (b) The calculated potential profile through the solution side of the double layer. (1:1 non-specifically adsorbing electrolyte in water at 25°C). Taken from [22]. d  12  Substituting in cb from Equation 2-9 results in 2  G  m  <7 v X M  ze  sinh  = (SkTeey)  2  2kT  eeo  Y  2-10 )  The double layer capacitance can be found by differentiating a with respect to (b . Inverting and M  0  simplifying yields  x,  1  C  dl  1  ££  + 0  2-11 (2ee z e n°/kT) 2  cosh(ze</> /'2JCT)  2  2  o  2  which resembles two capacitors in series in simplified form as J  c  1  1_  ~c  +  2-12  c  ^dl  The double layer capacitance (C ) is a function of the inner layer or Helmholtz capacitance (C ) and dl  H  the diffuse layer capacitance (C ) (see Figure 2-3(a)). The inner layer capacitance is independent D  of potential and resembles a simple capacitor. The double layer capacitance is dominated by the smallest value of the capacitors in series. For low electrolyte concentrations, C has a minimum dl  at the potential of zero charge (pzc) because the capacitance of the diffuse part of the double layer is small. The minimum occurs at the pzc (where o = 0 and therefore cb = 0), since no excess of M  2  counter ions is required in the diffuse double layer to balance the zero charge on the metal. Measuring the capacitance for low electrolyte concentrations of a non-specifically adsorbing electrolyte is a method for determining the pzc of a particular metal/electrolyte interface [23,24].  2.3  The Metal Side of the Interface The G C S theory treats the metal as a hard wall that carries charge. A more realistic  description of the metal is given by the self-consistent jellium model [25,26]. The metal atoms are relatively fixed in position however, the electrons are free to move in response to an electric field. Electrons can then 'spill out' onto the surface creating dipoles and surface potentials. Consider bringing a unit test charge from zero energy to a point x in front of a charged metal surface. Zero 13  energy is defined at infinite separation from the charged metal surface where no external influences of charge exist. Since the charged metal surface will exert a force on the surrounding phase, work will be required to move the test charge from zero energy across the electric force exerted by the metal. The forces acting on the test charge include Coulombic and image forces. Coulombic interactions are long range (Figure 2-4b) and image forces arise from an equal and opposite charge induced in the metal analogous to an electrostatic mirror [27] when the test particle and metal surface are in close proximity (Figure 2-4a). The point x is defined as a region near enough to the metal surface so that the test charge experiences neither Coulombic nor image interactions. The work per unit charge (potential), required to bring the test particle from zero energy to this special distance is defined as the outer or volta potential (i|/). The work required to bring the test charge from zero energy to the interior of the charged phase is the electrochemical potential (ji , measured in J mol" ) 1  and the work required to cross an uncharged phase is the chemical potential (u., measured in J mol" '). Combining the two, the work done per unit charge when the test particle is moved from zero energy across a charged surface phase and into an empty interior is known as the inner or Galvani potential ((b). Electrochemical potential can be written as Ji - /LL + ecp  2-13  with e being the charge of an electron. Some intrinsic properties of metals such as the Fermi level (E , units of eV) or work function F  (O, units of eV) differ from one metal to another and can affect the energetics of the electrons. The Fermi level is a definition in describing the distribution of electrons among the available energy levels in a metal. In a metal it is defined as the maximum energy that an electron will have at absolute zero temperature. Thus at 0 K, all energy states with energy less than the Fermi level are populated while those energy states greater than E are empty. The work function is defined as the F  minimum energy required to remove an electron from the interior of a metal to a position just outside the surface. Those electrons that 'spill out' onto the surface create a dipole. A surface potential (%) is created from the dipole and affects the work function of the metal according to the following equation. - O = jl + e%  14  2-14  Distance from metal s u r f a c e (in c m )  Distance f r o m metal surfaceJin cm)  (a)  (b)  Distance from metal s u r f a c e (in cm) (C)  Figure 2-4 The variation of electrostatic potential near a charged surface, (a) Short range image potentials and (b) long range Coulombic potentials, (c) Total potential resulting from both Coulombic and image potentials showing the independence of distance from the surface over the range 10" to 10" cm approximately. Taken from [27]. 5  3  15  The surface potential and work function can be further influenced by the arrangement of metal atoms at the surface.  For face centered cubic (fee) metals such as Au, Ag and Cu, the surface  crystallography can be described by Miller Indices. Cutting the metal along an index plane to expose a certain face allows for many different surfaces. For example, the low index planes described by the Miller indices 100, 110, 111 are shown in Figure 2-5. The unit cell for the (100) surface has four-fold symmetry and a square lattice. The (111) face is more densely packed having three-fold symmetry and a triangular or hexagonal lattice, and the (100) face has two fold symmetry or a rectangular lattice. From the packing density, the (111) face is the most dense surface followed by (100) and (110). These are just the principal or low index planes, but a large number of surfaces can be constructed, and one of the more important is the (210) surface. This is the least dense surface of the fee class with the lowest work function as shown in Figure 2-6. Generally, more densely packed surfaces have larger work functions. This can be rationalized by the effect that surface morphology has on the surface potentials. The higher the density of atoms at the surface, the smoother the electron density is at the surface because any roughness or less dense packing will induce additional dipole moments thereby changing x, and lowering the work function [28].  2.4  Thermodynamic Treatments of the Electrode/Solution Interface A well used thermodynamic treatment of an interface is the Gibbs Adsorption Isotherm. This  is a general description of an interface but the outcome can be extended to produce the electrocapillary equation which is important for electrochemical systems.  2.4.1 The Gibbs Adsorption Isotherm An interface of surface area^4 is created when two phases, a and P, come in contact. Within the interfacial region there exists a mixed concentration of the species in the two phases. This can be visualized in Figure 2-7 showing the variation in concentration of a species along a distance coordinate. In schematic form, the interfacial region consists of some finite area as seen in Figure 2-8a. A hypothetical dividing plane (Gibbs dividing plane) with zero thickness is placed in the interfacial region as shown in Figure 2-8b and the concentration of the species is discussed in terms of excesses or deficiencies within the interface compared to the pure phase. Assigning the symbol 16  Figure 2-5 The principle low index planes for fee crystals showing the (100), (110) and (111) faces. The right column represents the cutting plane of the unit cell and the left column shows the resulting crystal face. For clarity, atoms in the center of the cube faces have been omitted unless they are contained within the cutting plane. Taken from [29].  f A l / eV  0.6  0.7  0,8 U  50  100  1 50 ° < / °  Figure 2-6 Measured work functions for various Cu surfaces. The largest work function is measured for the smoothest surface. Taken from [30]. 17  Figure 2-7 The variation in the concentration (c,) of component i within the z coordinate between two bulk phases a and p. Taken from [31].  a) Actual  b) Model  Hypothetical Gibbs dividing surface  Interfacial region  Figure 2-8 (a) Actual system showing the two pure phases a and P and the interfacial region, (b) Hypothetical model system representing the Gibbs dividing surface. 18  a to denote the thermodynamic properties of the dividing surface, one can write a  o"  8  2-15  n - n - n - nf t  t  i  which states that the total amount of component i in the dividing surface (also known as the surface excess) is given by the total amount of component i in the system, minus the amount of component / in the two pure phases. The same can be said for any extensive thermodynamic property such as Gibbs free energy, volume, entropy and internal energy. Therefore, at constant temperature and pressure, the following relation holds  dG° =  dG  dA+Ys  dA  ji^n*  2-16  The first term is known as interfacial tension represented by the symbol y (N m" ), and the second 1  term describes the species in the phases. Invoking the Gibbs-Duhem equation, the final form of the Gibbs adsorption isotherm at constant temperature and pressure is  -J =£iY4u,  2-17  7  This equation describes the interfacial tension between two phases (in a two component system). The term F, (known as the surface excess concentration (mol cm" )) describes the amount of species 2  i in the interfacial region and is mathematically represented as rf  r  CT  =  2-18  —  A Values for n° and therefore T° are dependent on the choice of dividing surface and are not experimentally measurable without a reference system. The dividing surface is often chosen such that n° and therefore I" / are zero. The relative surface excess (T ) denotes the value of r,° for the 1  j(1)  dividing surface that makes n° equal to zero. The final form of the isotherm can be extended to include charged species resulting in the electrocapillary equation. 2.4.2 The Electrocapillary Equation The Gibbs adsorption isotherm was applied to a real electrochemical system of an IPE 19  immersed in an MYelectrolyte containing a neutral molecule L. A full account of the derivation can be found in [22] and just some key equations are given here.  The general form for the  electrocapillary equation is (  -dy  = o-  dE_ +  \  (  ^  X MX  M  ^MX  d  M  +  +  1  XH.O  L-  1  A  HjO  H0 2  djl  1  2-19  The term a dE_ arises from the expressions  dp.™ - dp" " = -Fd(4, E  -<p" )=-FdE_ EF  WE  2-20  and 2-21 where F is the Faraday constant (C mol" ) and E_ is the potential of the working electrode with 1  respect to the reference.  The negative subscript is a convention describing that the reference  electrode responds to an anionic component of the system. Equation 2-19 is referred to as the Gibbs-Lippman equation, which under constant temperature, pressure and composition simplifies to the Lippman equation dE ' T,P,n=  M  o  2-22  This relation has proven very important in the study of adsorption onto Hg electrodes since the surface tension can be measured directly. The electrocapillary equation has been used to measure the adsorption of neutral molecules onto the electrified interface. Through cross differentiation, some useful relations can be obtained. For example, the Essin-Markov coefficient is related as <2E\ MX J  doM  A  2-23  and describes the degree of specific adsorption. This value should change with electrolyte concentration, just as there should be a change in the pzc due to specific adsorption changes [23]. The derivative is zero in the absence of specific adsorption. For specifically adsorbed anions, and 20  constant charge density, the pzc moves in the negative direction to compensate for the adsorption of anions. For cations, E moves in the positive direction at constant charge density. ±  The electrosorption valency is another term that arises out of the cross differentiation of the electrocapillary equation da  M \  2-24  revealing the dependence of the free energy of adsorption on the electrical variable. These equations are all applicable to Hg electrodes where the surface tension can be directly measured. For solid electrodes, a modified approach based on the electrocapillary equation is used.  2.5  The Adsorption of Neutral Molecules Many isotherms that thermodynamically describe adsorption processes exist and some of the  more common ones are described here.  2.5.1 Adsorption Isotherms The most common isotherm describing monolayer adsorption onto a surface is given by Langmuir. The fractional monolayer coverage (6) is related to the bulk concentration of adsorbing molecules (C , (mol cm" )) by 3  bulk  ~7Z~Q  =  K  C  bulk  2  "  2  5  where K is the ratio of the adsorption rate constant to the desorption rate constant. This value can be described by an Arrhenius equation in the following form AG  (  K  K=- -= k 1  exp  f l  ^  RT  2-26  Relating the isotherm in terms of the surface excess concentration results in  r max  KC„  21  2-27  where 0 is replaced by I7rmax. Thus, a knowledge of the surface excess concentrations can result in a value for the free energy of adsorption. Furthermore, using the Gibbs adsorption isotherm the surface excess concentration can be related to surface tension resulting in a method for determining the film pressure. In this case 2-28  1-  max / where n is the film pressure (N m" ), and R is the gas constant (J mol" K" ). The choice of isotherm 1  1  1  is dependent on the monolayer characteristics. For example, Figure 2-9 reveals the simulated dependence of film pressure with concentration described by the isotherms of Langmuir, Henry and Frumkin. The Henry isotherm describes the adsorption of molecules in the limit of low coverages, or low bulk concentrations. Under these conditions the Langmuir isotherm simplifies to  r  KG bulk  2-29  and the film pressure becomes  n = RTT  2-30  The Frumkin isotherm takes into account possible interactions between adsorbing molecules which is assumed not to occur in the Langmuir isotherm. A lateral interaction coefficient (a) describes the positive or negative interaction between neighbouring molecules on the surface. The Langmuir isotherm then becomes  -r  2-31  exp  V  max /  and the film pressure is  K=  -RTT,  In 1- r V  r ] + a( r ^ r  2-32  max /  V  max /  In Figure 2-10, all isotherms follow Henry's behaviour at the limit of low surface coverages. The Frumkin model reveals that a positive lateral interaction increases the adsorption and therefore the 22  0  0.2  0.4 0c/r  0.6  0.8  m a x  Figure 2-9 Simulation of the dependence of film pressure on the bulk concentration of the adsorbate for adsorption isotherms described by Henry, Langmuir and Frumkin. The lateral interaction parameter (a) is indicated on the curves. Taken from [32].  23  film pressure. The opposite trend is observed for a negative interaction coefficient. These relationships allow for the determination of the Gibbs free energy of adsorption. Other methods are available for the determination of molecular adsorption onto electrode surfaces for which the film pressure can be calculated. This can be achieved through capacitance, and charge density measurements [32]. To extract information on molecular adsorption from strictly capacitive measurements requires the determination of the zero frequency capacity. This is possible with liquid electrodes such as Hg, however, this treatment usually fails for solid electrodes due to frequency dispersion. Frequency dispersion occurs with solid electrodes [33-36], owing to the surface roughness. Therefore, the zero frequency capacity is unknown for solid electrodes. To circumvent this problem, chronocoulometric methods (a zero frequency method) for determining the charge density of the solid electrode are used from which the film pressure can be calculated.  2.5.2 Adsorption of Organic Molecules onto Electrified Interfaces Amphiphilic molecules can be physisorbed onto the electrode surface lowering the interfacial tension. The interfacial tension can be calculated using the electrocapillary equation for both the organic coated and uncoated electrode and the difference is film pressure. With Hg electrodes, interfacial tension can be directly measured at various electrode potentials with different concentrations of organic species. Using the Lippman equation (Equation 2-23), electrocapillary curves for the adsorption of various organic compounds onto Hg from the bulk solution have been measured [23,37,38]. Capacitance can be extracted from the electrocapillary curves using C =  dE  2-33 - / T,P,n  V  M  - J ,P,Hj T  From these equations, the charge density and interfacial tension could also be measured from the integration and double integration of the capacitance curve respectively. This however, requires the knowledge of interfacial tension at a given value of potential. More specifically, E  <J = \ CdE M  24  2-34  and E  y=f-jjCdE  2-35  2  E*  can be evaluated if E* is a potential for which the surface tension (y*) is known. Because of frequency dispersion, this value is difficult to achieve with solid electrodes. Furthermore, the charge in a real capacitor is a function only of potential. In real electrochemical systems, the charge is a function of both the electrode potential and the surface coverage, o (E,Q). Thus, the capacitance can be written as  _ (do) dE)  C=l  9  +  (do_ [dO  E\  dd_ dE  2-36  where the first term is the true capacitance, and the second term is called the pseudo-capacitance, that results from potential dependent changes in surface coverage on the electrode. Figure 2-10 shows the differential capacitance of a Hg electrode coated with octyl alcohol. The adsorption of organic molecules results in a low capacitance near the pzc.  On either side of this potential, the  adsorption/desorption of the surfactant molecules is revealed by the pseudo capacitance peaks. These peaks can also be taken to represent dramatic changes in the orientation of the surfactant at the metal surface. Moreover, these peaks are frequency dependent and decrease from an equilibrium value at co = 0, to zero at infinite frequency. To eliminate the frequency dependence on the measurement of charge density, a technique of back integration can be used [32]. This method involves the use of chronocoulometry and has been applied in the calculation of film pressure for many solid electrode/adsorbate systems [39-43].  2.5.3  Chronocoulometry and the Back Integration Method Chronocoulometry is a zero frequency method for determining the charge density on the  metal. The method involves quickly pulsing potential and measuring the capacitive charging current for coated and uncoated electrode surfaces. The current transients are integrated to obtain the charge density on the metal and the interfacial tension is found through integration of the charge. More 25  0.24  0.4  0  -0.4  -0.8  S vola t  Figure 2-10 Differential capacitance of a Hg electrode in I M K N 0 at 25°C (dotted line). The solid lines represent the capacitance of the same electrode in contact with octyl alcohol. The arrows represent the frequency (kHz) used in the capacitance measurement. A depression in capacitance is noticed near potentials around the pzc. Pseudo-capacitance peaks are noted at potentials positive and negative of the pzc. Taken from [23]. 3  26  specifically  y(E)^-\a dE+y(E ) M  2-37  f  f  E  The integration limits are potentials where the organic molecules are adsorbed (is,) and desorbed (E ). f  Under these conditions, the integration constant, y(Ef) does not depend on the presence or absence of the organic molecules because there is no adsorption at E . Thus, the film pressure can be f  calculated through the back integration technique through the following equation E,  71 = 7  0 = o  -7g=  E,  G dE  \  M  E  -  I <J  dE  M(M)  2-38  E,  f  because the constants of integration cancel. The terms 6 and 6 = 0 represent the presence or absence of the organic molecules on the electrode respectively. The relative Gibbs excesses can be obtained through the differentiation of film pressure with respect to the In of the bulk concentration of the organic species as  r =  RTdlnc  2-39 ^ T,P,E  The free energy of adsorption can be calculated using Henry's adsorption isotherm with extrapolation to zero coverage. In this region, the film pressure should be linearly related to the bulk concentration of the organic molecules in solution. This is one example of many possible methods to calculate the free energy of adsorption.  2.6  Insoluble Monolayer Characteristics The characteristics of monolayers consisting of long chain surfactants that form Langmuir-  Blodgett (LB) films are of importance to this thesis. Much of the work presented is on the potential induced behaviour of these lipid-like surfactants adsorbed on a metal surface in the presence of an electric field. The epi-fluorescence characterization of these monolayers require the incorporation of fluorescent molecules. 27  2.6.1 Structure and Properties of Monolayers at the Air/Solution Interface The ability to obtain monolayers by depositing and organizing molecules onto the surface of liquids and the handling of these monolayers was pioneered by Langmuir and Blodgett [44,45]. Monolayer assemblies at the air/water interface are achieved by depositing a small amount of an amphiphile solution in a volatile solvent onto a clean water surface. The monolayer is formed by the spontaneous spreading of the amphiphile as the solvent evaporates. Amphiphiles have been extensively used because they are composed of both hydrophilic and hydrophobic parts from which the monolayer formation is driven by the interaction of the head group with water and repulsion of the hydrocarbon tail into the gas phase. Common surfactants include fatty acids such as stearic acid or long chain alcohols. The quality of the monolayer is dependent on the length of the hydrocarbon chain and head group as shown in Figure 2-11.  Using a Langmuir trough, the packing or  compression of the monolayer can be altered by confining the film to a defined area. A typical Langmuir trough is depicted in Figure 2-12 where the moving barrier will compress or relax the film. Incorporation of a Wilhelmy plate allows measurement of surface tension. As shown in Figure 2-13 the Wilhelmy plate (usuallyfilterpaper) is connected to a microbalance and suspended in the liquid between the barriers of the trough. The forces acting on the plate are gravitational force, surface tension, and buoyancy. The plate of known density (p ), length (L ), width (W ), and p  p  p  thickness (T ) is immersed to a depth H, in a liquid of density p,. The net downward force is given p  by  r( p  (pp)s{ P P p)  F=  L  W  + 2  T  T  +  w )(co*e)p  P x g  T w H, p  p  2-40  where y is the surface tension of the liquid, 0 is the contact angle between the liquid and the plate, and g is the gravitational force constant. This equation is simplified by maintaining a completely wetted plate with a contact angle of 0 = 0, and a constant H,. The film pressure (or interfacial tension lowering as a result of the surfactant) is found by  n  AF  - y water  J film  2W  2-41  p  provided that W » T . Measurements of surface pressure versus compression area render a surface p  28  Very weak  Weak  (no film)  (unstable films)  Hydrocarbon —CH.J —GH Br —CH C1 —NO  —CHjOCH, —GiR 4OCH3 —COOGH,  2  Strong (stable film with G i chain) 6  —CIIoOH  2  Very strong ( d e chain compounds dissolve) — SO,-  —COOH  —OSO3-  —GN  — C,H S'0 -  —CONH2  — NH,+  4  4  — C H = NOH —CH4OH —CH COCH  3  2  3  —NHGONH2 —NHCOGH3  Figure 2-11 A table showing the effectiveness of functional groups in providing attraction to water and the strength of the monolayer film as a result of the hydrocarbon chain length. Taken from [46].  Wilhelmy plate  Monolayer Movable barrier Q C X X C X X ) C ) C ) O Q Q ( X X X p >//////*  • • ./• y /  .'-y y y y y .r J s y y ,  r  ,/  •//////,  Water VV t i l , V I <• //////////////y/y%/^ Trough  ss ? s s S / s s r s * ,  Figure 2-12 Schematic depiction of a monolayer in a Langmuir trough film balance. The movable barrier allows for compression and decompression of the organic layer. The Wilhelmy plate is used for the measurement of film pressure.  29  r—  w  ^  Figure 2-13 Schematic representation of a Wilhelmy plate of known height, width and length immersed into solution. The plate is connected to an electromicrobalance (EMB) for the measurement of force. The contact angle between the water and plate is measured by 0.  30  pressure/area isotherm  (TI  vs. A) useful in characterizing the physical states of the floating  monolayer. A schematic representation of such an isotherm for a simple amphiphilic molecule with the corresponding monolayer states is depicted in Figure 2-14. At large areas, where the molecules are sufficiently far apart, lateral interaction such as van der Waals forces are negligible and the molecules will have an average energy equal to the average translational kinetic energy of VikT for each degree of freedom. The plane of the surface is two dimensional leading to a total kinetic energy of kT. This region is known as the gaseous region described by 2-42  KA=kT  where A is the area occupied by the molecule measured in A /molecule and TX. is the surface pressure. 2  This relation holds when the distance between the molecules is large and u is very small. If any association between the molecules takes place, then the measured values of TL4 would be less than kT. Upon compressing the monolayer, the liquid region is attained. In general, there are two types of liquid films known as the liquid expanded (L ) and liquid condensed (L ) films. If the film ex  co  pressure is extrapolated to zero, the value of A obtained is much larger than that obtained for closely packed films. The L film is established at higher compressions. The liquid to solid phase transition c o  is obtained when the molecules adhere very strongly to each other through van der Waals forces in the chain-chain interactions. At these high compressions, the TC VS. A isotherm generally shows no change in the film pressure at high A and represent a condensed film. Further compression can result in a collapsed film for which the characteristic collapse pressure iz , is the highest surface pressure col  to which a monolayer can be compressed without detectable movement of the molecules in the film to form a new phase. One important film pressure is the so-called equilibrium spreading pressure iz  eq  (ESP) where the bulk surfactant on the liquid surface is in equilibrium with its monolayer. This  can be achieved by depositing an excess of surfactant onto the water surface resulting in a reservoir or lens of the excess amphiphile. The value of the ESP depends on the forces that hold the molecules in the monolayer, as well as the forces that favour the molecules retention in the reservoir. This is similar to the wetting of solids as described by Young's equation Jsv-Ysl  =  7/ COS0 v  2-43  A schematic representation for the wetting of solids is shown in Figure 2-15 describing the 31  1  W  20  1  A  40  1  1  50  60  -LI—X  1  100  L_  200  Area (A /molecule) compression of. barrier 2  —  Figure 2-14 A graph showing the surface pressure vs. area isotherm curve. The various states of the monolayer at different compressions are shown in each region. Taken from [47].  Vapour  Ysi  Solid  Figure 2-15 A representation of the interfacial tension between liquid-vapour (LV), solid-vapour (SV) and solid-liquid (SL) for a liquid drop resting on a smooth solid surface. A contact angle, 0 is formed between the liquid and solid.  32  relationship between solid-vapour (^v), solid-liquid (si), liquid-vapour (Iv) interfacial tensions, and the contact angle 6. For wetting to occur, the contact angle must be less than 90°. For a reservoir of surfactant on the liquid surface, the spreading is usually discussed in terms of the spreading coefficient defined by S  =7 o,aw  I  - (7 aw  + 7 ow  \l  I  2-44  ) oa )  where ow, oa, and aw correspond to the organic/water, organic/air, and air/water interfaces respectively. For spreading to occur, the spreading coefficient must be greater than zero, again requiring contact angles less than 90°. This equation assumes that y  ow  is at its equilibrium value and  non-equilibrium conditions require different spreading coefficients. The ESP is an important characteristic allowing for reproducible monolayers to be formed without the use of compression.  2.6.2  Mixed Mon olayers In any two-component monolayer the following description for the average area per molecule  under 'ideal mixing' can be written as  A = XiA + X2A  2-45  2  where A, and A are the area per molecule in pure monolayers of components 1 and 2 respectively, 2  and Xi " X2 ana  a r e t n e  mole fractions of the components. Similarly, the expression for iz is !2  ^12 =  +  Xl^2  "  2  46  These relations can be written in terms of the excess function  A=  "  +  x  2  47  with a similar relation for n . For ideal mixing, the excess terms are zero. According to Gains [46] ex  the expression for the excess free energy of mixing can be calculated according to K  ^G  ex  2A)h  = \[A -( A l2  X{  + Z  l  "  2  48  0  The extent of mixing can be found by plotting A or n against l2  !2  Linear relationships represent  ideal mixing and any deviation from linearity points to non-ideal behaviour. Monolayers of two  33  components can mix completely, partially, or can be immiscible. Examples of these mixed films are shown in Figure 2-16 for simple amphiphilic molecules.  2.6.3  Monolayers  Transferred  to the Solid  Support  Many of the refined methods for obtaining monolayers on a solid substrate transferred from the Langmuir film were accomplished by Blodgett [48-51]. Depending on the type of substrate, and the mode of deposition, three structurally different multilayers are recognized and shown in Figure 2-17. X-type multilayers are organized in a substrate-tail-head-tail-head etc, geometry formed by the sequential hydrophobic attachment of monolayers onto the plate by immersion only. The sequential hydrophilic attachment of monolayers onto the plate upon withdrawal results in the substrate-head-tail-head-tail etc, or Z-type geometry. The most common arrangement is the Y-type formed by substrate-tail-head-head-tail etc, built up by both dipping and withdrawing the plate through the floating monolayer. Of importance to monolayers transferred to planar electrode surfaces is the horizontal lifting method developed by Langmuir and Schaefer [52]. This method involves touching a hydrophobic material to the floating monolayer. The monolayer is transferred to the hydrophobic surface upon contact as shown in Figure 2-18.  2.7  Factors Affecting Fluorescence from Molecules A brief description of the fluorescence theory will be given in this section. General  fluorescence from molecules will be described first, followed by a description of fluorescence from molecules in multilayer assemblies. This will lead to concepts of fluorescence energy transfer and will be linked to fluorescence near a metal surface. The descriptions in this section will aim to provide the basis for understanding the results obtained with the in situ electro-fluorescence technique.  2.7.1  Electronic  Transitions  and the Franck-Condon  Principle  Luminescence arises from the radiative deactivation of molecules existing in an excited electronic state. Electronic transitions occur when the molecule is exposed to visible or ultraviolet wavelengths. The electrons in a molecule interact with the incoming radiation creating an induced 34  a)  b)  c)  Figure 2-16 Possible molecular distributions that result from various mixing of two surfactants. The films can be (a) miscible, (b) complete separtated, or (c) slightly immiscible.  r f .1* t*  f '  /jT  f  J" J' j  J"X  J. X y  V  J  .1' .-F .ii" .J" .it .f J • J' jf .1' ,f-*r ,.f /r" -|'  .JP  rnnhn'hi r \X 1  .i'  fsifft  >• x V  •  /, ' / v / <  Figure 2-18 The horizontal touching method for transferring insoluble molecules to a hydrophobic surface.  36  dipole moment that is necessary for the electronic transition. The probability of an electronic transition is given by 2-49 where V/- and V,° represent the wavefunctions for the excited and ground electronic states and d is the dipole moment operator. The integral fl  2-50  is known as the transition dipole moment and the intensity of the spectroscopic transition is dependent on this value. When  is zero, the transition is said to be forbidden. When this value  is non-zero transitions, are allowed. Furthermore, as a result of the selection rule AS = 0, the transition from the ground state to the excited state must occur between states having the same multiplicity. Transitions where the selection rule is broken (triplet excited state to a singlet ground state) can occur, but with a low probability. The Franck-Condon principle has been used to estimate the strength of electronic transitions. It is based on the view that the electronic transition occurs much more rapidly than does the redistribution of the nuclei in the molecule, because the nuclei are more massive than the electrons. Quantitatively, the transition dipole moment may be stated as  2-51 where e and v represent the electronic and vibrational states respectively. With r being the electronic coordinates and R the nuclear coordinates, the second factor in the equation describes the overlap between the vibrational wavefunctions for the initial and final states. Since the intensity of the transition depends on the square of the transition dipole, the greatest intensity will occur for the transitions that have the greatest overlap in the vibrational wavefunctions. This occurs as a vertical transition from a low vibrational ground electronic state to a high vibrational excited electronic state.  2.7.2  Fate of the Excited State Molecule A molecule in an excited electronic state has numerous pathways to relax to the ground  electronic state. Fluorescence occurs by a radiative deactivation process through the emission of a 37  photon with rate constant k . The fluorescence quantum yield (tpy) describes the efficiency of this f  process amongst alternate deactivation pathways and is generally given by k  f  2-52  A I k ,  The denominator represents the sum of ./V deactivation pathways, each with a rate constant  The  natural lifetime of a fluorophore (T ), is the inverse of k , or lifetime that would be observed in the 0  f  absence of competing deactivation paths. The quantum yield can also be expressed as T  (D  f  f  =  —  2-53  *o  where x is the measurable mean fluorescence lifetime. Fluorescence intensity is related to the measurable mean fluorescence lifetime of a single fluorophore through the equation  = e  x  2-54  where (Ij) is the intensity of fluorescence at some time after the fluorophore is exposed to a short t  pulse of light. Thus, competing events can decrease fluorescence intensity by decreasing the mean lifetime.  Figure 2-19 is a Jablonski diagram revealing the possible radiative pathways for  deactivation. The ground state (A) and electronic states (B) have the same multiplicity while C has a different multiplicity. The excitation pathway results in the promotion of an electron to one of the vibrational levels in the excited electronic state. Collisions of the excited molecule with neighbours or solvent will result in the relaxation of vibrational energy to the ground vibrational level of the excited state. If a photon is emitted, the process is known as fluorescence, and the emitted photon is of lower energy than the excitation energy. This difference is known as the Stokes shift. If the photon is not emitted after the vibrational relaxation, intersystem crossing to a triplet state can occur. The relaxation from this triplet state to the ground state by emission of a photon is phosphorescence. Fluorescence and phosphorescence are the typical radiative decay pathways for the excited state molecules, however, non-radiative pathways also exist. 38  Intersystem crossing  Relaxation B  -**  C Excitation  Fluorescence  Phosphorescence  A Figure 2-19 A Jablonski diagram showing the emission processes for electronic transitions. Electronic states A and B have the same multiplicity but C does not.  39  Molecules in an excited state are highly reactive compared to their corresponding ground state due to the additional energy. For long lived triplet states, a chemical reaction may occur faster than the decay rate. Excimers are formed through the reaction of an excited state molecule with a ground state molecule. The excimer will have an emission spectrum that is different from the monomer [53]. Photobleaching or photodecomposition is another common fate for excited state molecules which results in the progressive loss of fluorescence intensity with irradiation time. Rost [54] believes that photobleaching is due mainly to a photochemical reaction of the excited state molecule rendering a non-fluorescent product. Quenching is another phenomenon that can result in complete deactivation of the excited state. This process has been linked to factors such as temperature, oxygen concentration, impurities, and concentration quenching [54]. Furthermore, the energy of the excited state molecule may be transferred to acceptor molecules through fluorescence resonance energy transfer. This transfer of energy can also occur to metal surfaces [53].  2.7.3 Dye Aggregates Much of the fluorescence measured in this thesis was from desorbed monolayer assemblies near a metal surface. The dense packing of the fluorophores can significantly affect the absorption and emission spectra and the metal can quench fluorescence through non-radiative energy transfer. For example, the spectral properties of a carbocyanine dye/arachidate (C H COOH) mixed 19  39  monolayer deposited on a glass substrate is dependent on the mixing ratio [53]. In molar ratios ranging from 1:1 to 1:100 of dye:arachidate, the absorption spectra reveal an increasing band at shorter wavelengths with decreasing arachidate content as shown in Figure 2-20. Red-shifted absorption bands are also possible and have been observed when mixing octadecane (C H ) with 18  38  a cyanine dye at a molar ratio of 1:1 [55]. A blue-shifted absorption band is typical of H-aggregates, where the chromophores are packed in a sandwich arrangement [56]. A brick stonework arrangement of chromophores results in the red-shifted absorption band typical of Scheibe or J-aggregates [55,57,58]. This packing was spectroscopically confirmed with scanning force microscopy [59]. The packing of the chromophores results in dipole-dipole interactions that are responsible for the magnitude of the spectral shift. The 40  "-"-CO  CiC?»-CH=CH. I C  18 37  i  H  C  18 37 H  5  400  450  500  550  600  Wavelength [hmj  Figure 2-20 Monomer and dimer formation in dye monolayers. The mixed monolayers of the dye represented at the top of the figure, and cadmium arachidate were deposited on a glass substrate. E is the extinction coefficient of the dye, and a is the number of dye molecules per unit area. The monomer band is 490 nm, and the dimer band is at 460 nm. Taken from [53].  41  molecular exciton model [60] describes this mechanism for dimers, trimers and higher order aggregates. The excited state resonance interaction between the dipoles is treated electrostatically. Figure 2-21 represents the exciton band energy diagram for molecular dimers. The transition dipoles can be oriented parallel, in-line or oblique to each other resulting in different band splitting. In the parallel arrangement, the out-of-phase dipole corresponds to a lowering of energy (£') and the in-phase dipole interaction raises the energy (£"'). The transition dipole moment is given by the vector sum of the individual transition dipole moments. Therefore, the transition from a ground state to excited state E' is forbidden since the transition dipole is zero, while a transition to E" is allowed. This results in the blue shifted absorption for a singlet-singlet electronic transition in the dimer. This arrangement corresponds to the sandwich packing of aggregates. The opposite band splitting is obtained for in-line orientated transition dipoles shown in Figure 2-22.  Here, the in-phase  arrangement of the transition dipoles leads to electronic attraction producing an excited state E', while the out-of-phase arrangement causes repulsion. Thus, transitions from a ground state to E' are allowed while transitions to E" are forbidden resulting in the red-shifted absorption band.  2.7.4 Forster Energy Transfer In 1946 [61] Forster described the radiationless transfer of electronic excitation energy from a donor molecule (sensitizer) S, to an acceptor molecule A through long-range resonance dipoledipole interactions for an S-A pair. There are three parameters required for this transfer of energy to occur. Firstly, there must be significant overlap between the fluorescence spectrum of the sensitizer with the absorption band of the acceptor. This is calculated through the Forster overlap integral. Secondly, the energy transfer between the molecules depends on the distance separating them. The transfer occurs through an electric dipole-dipole interaction for which the energy is proportional to i?" ,where R is the distance separating the center of the dipoles. From a Coulombic 3  standpoint, this requires that R is large compared to the distance between the two charges in each dipole. Lastly, the orientation of the dipoles must be accounted for and is done so by a so-called orientation factor. The radiationless transfer of electronic excitation energy from sensitizer to acceptor results the quenching of S by A. The process is considered "long-range" because the two molecules do 42  Parallel transition dipoles  E  Dipole phase relation  Dimer levels  Monomer levels  Figure 2-21 Exciton band energy diagram for a molecular dimer with parallel transition dipoles. The ovals correspond to the molecular profile, and the double arrow indicate the polarization axis. Taken from [60]. I n - l i n e transition  dipoles  E £"•£'  Monomer levels  Dimer levels  If  Dipole phase relation  Figure 2-22 Exciton band energy diagram for a molecular dimer with in-line transition dipoles. Taken from [60].  43  not in come in contact. Forster was able to describe the energy transfer rate through the following relation that generally holds for separation distances between 10-100 A.  2-55  Here, x is the measured fluorescence lifetime of the sensitizer in the absence of the acceptor, and s  R is the distance between the center of the chromophores. At the Forster critical distance R , 50% 0  of the excitation energy is transferred to the acceptor. At this specific distance, the probability of relaxation by energy transfer equals the probability of relaxation in the absence of the acceptor. In Forster's derivation, the energy transfer is believed to occur between a single donor acceptor pair, and therefore, the dipole-dipole interaction is used. Kuhn and coworkers measured fluorescence intensity with separation distance between S and A [53] revealing a inverse quartic dependence to the quenching of S. They achieved separation between S and A with non-fluorescent spacer molecules. In this investigation, S absorbs U V radiation and fluoresces in blue. The acceptor absorbs blue light and has yellow fluorescence. The sandwich arrangement used in these experiments is shown in Figure 2-23. When S and A are in close proximity and irradiated with ultraviolet light, the excitation energy of S is transferred to A and yellow fluorescence is observed. At large separations, the energy is not transferred to A and blue fluorescence is measured. Finally, if the sensitizer is absent, no fluorescence is observed because A does not absorb U V radiation. From dimensionality considerations, the energy from a sheet-like arrangement of dipoles is given by a R' dependence. 1  Therefore the resonance condition requires an R' proportionality The quenching of S was found 6  to follow (  2-56  V where Forster critical distance is given by 2-57  44  3 dark  2  yellow  blue  A  S  t  t  t  ultraviolet  t  t  t  f  radiation  Figure 2-23 A multilayer system for measuring the energy transfer between sensitizer (S) and acceptor (A) at various separation distances. S absorbs U V radiation and fluoresces in blue. A absorbs blue radiation and fluoresces in yellow. Taken from [53].  hv  r = 1/20 I  I  CuH37  CieH37  r =1/10 I  I  C»Hj7  C H]7 U  Figure 2-24 Calculated (line) and experimental luminescence (symbols) of S with separation distance from A. Taken from [53].  45  Here, q is the sensitizer fluorescence quantum yield in the absence of the acceptor. A s  AS  is the  absorption of the absorber layer under vacuum wavelength and X is the wavelength of sensitizer s  fluorescence. The refractive index of the medium is given by n, and the term a depends on the direction of the transition dipole moments of S and A. Figure 2-24 illustrates the experimental and calculated values for this scenario. The experimental values match well with the expected values.  2.7.5 Fluorescence Near Metal Surfaces Fluorescence near metal surfaces is further complicated because mechanisms affecting both radiative and non-radiative pathways exist. Non-radiative decay into the metal substrate occurs at close proximity while oscillations in fluorescence intensity are observed at larger separation distances. This variation in fluorescence intensity with separation distance from a metal has been studied by various authors [62-64] and a variety of reviews exist in the literature [65-67]. The pioneering work in this field was conducted by Drexhage [68-71]. R. Chance and coworkers have calculated the fluorescence emission and energy transfer for a variety Eu /mirror systems based on +3  this experimental work. Again, the chromophore is treated as an oscillating dipole placed near the metal surface. Figure 2-25 shows the experimental and calculated results. The calculations show that the Forster transfer (b ) predicts an inverse cubic dependence on the decay rate constant with ET  distance through the following relation. b  = (3d-  2-58  3  ET  For thick metals, this holds due to simple dimensionality considerations, but fails for thin metals if P is a constant. Chance and coworkers separated the radiative and non-radiative effects from the metal and calculated a value for p for two values of length (d, and d ). The distance separating S and 2  the metal is given by d„ and the thickness of the metal is given by d . The calculations reveal that 2  for large values of d (thick metal film) the energy transfer rate constant b , varies as d, ~ , so that 3  2  ET  P is constant. For thin metals (small d ), the energy-transfer rate constant varies as d, ~ , requiring 4  2  that P is proportional to d,' . 1  According to Figure 2-25, variations in the intensity at various distances suggests larger complexity of the quenching due to the mirror. As previously mentioned, non-radiative energy transfer to the metal occurs at close proximity between S and the metal. The oscillating dipole of 46  d  I  2  3  4  T  1  1  i  1000  2000  5  6  7  8  1  r  1  1  <  3000  4000  9 r  5000  d (A)  Figure 2-25 Calculated (lines) and experimental (symbols) excited state lifetimes for a Cu/Eu+3/air system at various separation distances from Cu. The excited state lifetime is small at small separations between the dye and metal. Taken from [63].  47  the fluorophore can drive surface plasmon modes at the metal surface effectively quenching fluorescence [72,73]. At large separation distances, coupling with non-radiative surface plasmon modes is negligible. When the oscillator S is in front of a metal surface, a mirror image (of opposite sign) S' is induced. Therefore, the field of S interacts with its mirror S' with some time delay and oscillations are observed as a result of constructive and destructive interference. In summary, a variety of interactions exist for which the absorption/emission of fluorescent dye molecules can be altered. Aggregation of dye molecules forming dimers, trimers and higher order aggregates can result in blue or red shifted absorption spectra. Furthermore, distant dependent non-radiative energy transfer of an excited state molecule can occur to neighbouring species. For absorbing dye molecules, this transfer is governed by the Forster equation with a R' distance 6  dependence between 5* and A for a single S-A pair. For acceptors in a sheet-like arrangement, the dimensionality of the system decreased and the energy transfer is proportional to an R' separation 4  distance. The fluorescence intensity is further complicated when the fluorophore is present near a metal surface. Close proximity between fluorophore and a thin metal results in fluorescence quenching with a R' separation dependence. For thick metals, the energy transfer varies as R' . In 3  4  these cases of energy transfer, fluorescence is quenched at close proximity between S and A. An increase in fluorescence intensity results at larger separation distances.  2.8  Fluorescence Microscopy The relations given in the preceding sections detailed a number of possible fates for the  excited state molecule. Given the complexities of the possible radiative and non-radiative pathways associated with fluorescence, it can become a tedious task to understand and quantify fluorescence data.  Fluorescence microscopy has historically been used in determining the distribution of  fluorescent dyes within a biological matrix. From this perspective, many complications can be eliminated by choosing the appropriate environment for the fluorophore. Eliminating fluorescence quenchers, oxygen, and metal substrates can lead to well-defined images of fluorophore distribution within a given matrix. This section of the thesis will describe some principles of fluorescent microscopy.  48  2.8.1  Carbocyanine  Dyes  One family of dye molecules that are widely used for hydrophobic labelling in biological applications are cyanine dyes. In addition to their large quantum yields, high extinction coefficients, and long wavelength absorption, they have a reasonably sized head group and have become desirable for labels that do not disturb the system being probed. The basic structure of the cyanine family of dyes is depicted in Figure 2-26. The general structure of the dyes is given by X(-CH=CH) -CH=Y n  where the substituents X and Y typically contain nitrogen groups. The factor n determining the chain length between the chromophores is the largest factor influencing the wavelength of maximum absorption. This group of dyes readily self-associate in the form of J and H-aggregates. The resulting aggregates can have altered or shifted absorption and emission spectra when compared to the monomer. In some cases, the aggregates have been rendered as non-absorbing or non-fluorescing species [75]. These aggregates are commonly referred to as J and H aggregates since their discovery in 1936 [57]. Association has been observed in aqueous solution and on solid substrates [76].  2.8.2  Microscope  Resolution  Historically, microscopy has been used to magnify a specimen of interest. The magnification is achieved through the objective that produces a magnified image of the object in what is known as the intermediate image plane. The eyepiece then magnifies the intermediate image in the same manner as a magnifying glass. The total magnification is given by the objective magnification multiplied by the eyepiece magnification but does not alone determine the clarity of an image. The clarity of an object is dependent on the ability of the optics to efficiently collect the light that is reflected, diffracted or emitted from the sample. This is typically achieved through objectives that have a large collection angle, or a large numerical aperture (NA). Objectives with a large NA allow for higher resolution images by reducing diffraction. Figure 2-27 shows light passing through a small slit (1) that propagates through the objective (2) which behaves as a circular aperture creating a circular diffraction pattern known as an Airy disk (3). Therefore the light passing through the slit is not imaged as a bright spot, but rather as a slightly blurred spot with diffraction rings around it. An objective with a large N A collects the diffraction orders more efficiently resulting in a decrease 49  X  Y  Figure 2-26 Typical substituents in cyanine dyes with the general structure X(-CH=CH) -CH=Y. The dominant factor in determining A is n. Taken from [74]. n  max  50  >  Figure 2-27 Schematic representation of Airy discs. Light is passed through a small slit (1), the objective (2) creating an Airy disc (3) due to diffraction rings from the objective. The top of the figure represents the intensity distribution of the Airy disc. Taken from [77].  X  Figure 2-28 A schematic representation of overlapping Airy discs. When the two disc are far apart they are easily resolved. When they begin to overlap (—) and ( ) a limit d is reached which is known as the maximum diffraction limited resolution. Taken from [77]. 0  51  in the size of the Airy disc. Another way of describing spatial resolution is through the point spread function (PSF) which measures the intensity of the Airy disc as shown in the top of Figure 2-27. Obviously, the more concentrated the spot is, the better the resolution. The maximum resolution is defined as the ability to resolve two Airy discs as schematically shown in Figure 2-28. When the Airy discs are separated by a large distance, they are easily resolved, however, when they overlap, a limit is reached where the profiles display two bright spots with a valley between them. The Raleigh criterion describes the diffraction limited resolution as  d  =  ^  IN A where X is the wavelength of light propagating through the objective. This is the maximum resolution that the microscope can achieve.  2.8.3  The Epi-Fluorescence Microscope The basic principle of fluorescence microscopy is to illuminate a fluorescent dye molecule  with visible or ultraviolet light and collect fluorescence with a detector (typically a charged coupled device, CCD). The principles are similar to the transmission microscope with some additional features. For the ability to view relatively weak fluorescence, the excitation light must be separated from fluorescence. This is achieved with the three filters shown in the inverted epi-arrangement in Figure 2-29. Visible or U V light is directed toward an exciter filter which transmits a limited band of wavelengths (hash marked arrows). The dichroic beam splitter reflects short wavelength excitation light to the objective. The fluorescent light from the fluorophore is directed back to the dichroic beam splitter that allows only longer wavelengths (fluorescence) by reflecting shorter (excitation) wavelengths. The fluorescence and any excitation light that may have leaked through the dichroic beam splitter are directed toward the barrier filter. This filter transmits only the fluorescence. Typical transmission spectra for such a filter arrangement is shown in Figure 2-30. As a final note, the fluorescence is collected in a coupler that connects the detector to the microscope. The coupler is used to focus or expand the light onto the detector and fill the CCD array. 52  Fluorophore  if Objective Exciter Filter J l 4- t s Dichroic Filter Light  Barrier Filter  CCD Array Figure 2-29 A schematic representation of the light path in epi-illumination. The light is passed through an exciter filter selecting a band of light. This light is directed through the objective to the sample. Fluorescent (and scattered) light is collected through the objective. Long wavelength light (fluorescence) is allowed to pass through the dichroic filter. A final filter to remove any excitation light is the barrier filter. This light is effectively coupled onto the whole array of detectors of the CCD.  53  -0.2  -1 300  1  350  1  400  1  450  1  1  500  550  1  600  1  650  1  700  1  750  1 800  Wavelength (nm)  Figure 2-30 Transmission spectra for a filter cube consisting of an exciter, dichroic and emission filter assembly. Data taken from [78].  54  Chapter 3 3  Literature Review The work presented in this thesis covers a wide range of topics and methods for  characterizing water-insoluble amphiphilic molecules adsorbed on solid electrode surfaces. Adsorption studies onto electrode surfaces were initiated with soluble organic molecules. The adsorption of these compounds from the bulk of solution onto the electrified interface was studied using Hg electrodes because the interfacial tension can be directly measured. Furthermore, Hg has an atomically smooth surface that is easily reproduced by extruding a fresh drop. Therefore, electrochemical results are not clouded by frequency dispersion, and a reproducible electrode surface is easily obtained. While the study of soluble adsorbates is important, water-insoluble organic surfactants are also of interest, especially for membrane investigations. Only a few groups have studied the electrochemical behaviour of these molecules at the metal surface. When the organic material exists at the air/solution interface, adsorption to the electrode surface must occur through deposition techniques (typically LB methods). The formation, structure, properties and potentials uses of LB films have been widely covered in a variety of reviews [2,4,79-81]. In 1966, Gains published a book describing Langmuir films with some mention to LB films as well [46]. This book also includes a brief description of the Langmuir-Schaeffer technique of depositing monolayers to a solid substrate through the horizontal lifting method. Although LB techniques have been widely used to form monolayer and multilayer assemblies onto a solid substrate, the self assembly of alkane thiol molecules is a popular method for coating noble metal electrodes.  Preceding this, however, were the electrochemical investigations of  insoluble organic adsorption on Hg electrodes in the late 1930s and early 1940s. These studies have led to a variety of electrochemical methods for characterizing the metal/surfactant interface. Since then, there have been technological advances in creating and maintaining single crystal surfaces and the adsorption of water-insoluble organic molecules onto the solid electrode is receiving more attention. Planar solid surfaces represent a more "real world" situation, and a variety of in situ methods of characterizing the potential induced behaviour of these organic modified surfaces are available.  55  This section of the thesis provides some background to the electrochemical and in situ characterization of electrodes coated with water-insoluble organic materials. The systems described include both physisorbed and chemisorbed surfactants on Au electrodes. The chapter begins with a description of the early electrochemical studies of lipid-coated Hg. This will be followed by the electrochemical investigation of octadecanol-coated Au surfaces, focusing on Au(l 11). A description of the photophysics of dye molecules separated from a metal surface will ensue and a detailed review of self assembled monolayers of alkane thiols onto Au surfaces will also be presented. The chapter ends with a description of some investigations that have measured luminescence at electrode surfaces.  3.1  Adsorption of Lipids on a Mercury Electrode  In the late thirties and early forties, Frumkin and Gorodetskaya studied the adsorption of various water-insoluble surfactants such as cetyl alcohol (1-hexadecanol), oleic acid (cis-9octadecenoic acid) and myristic acid (tetradecanoic acid) on a Hg electrode [82,83]. Adsorption onto the Hg surface resulted in a depression of capacitance and the monolayer thickness was estimated. More recently, the study of lipid adsorption onto a dropping Hg electrode evolved from the work of Miller [84]. Lipid monolayers were spread at the solution surface in a Langmuir trough, and a polarized Hg drop was passed through the interface at various values of mean molecular area (MMA) for the monolayer film. Using a variety of electroactive ions (Cu , A g and Tl ), the permeability 2+  +  +  of these adsorbed monolayers was characterized using polarography. At high film compressions, the monolayer showed a resistance to permeability. This effect was most notable when the Hg surface was uncharged (pzc). This suggests that when Hg is passed through well-ordered floating lipid films, a compact lipid layer is formed on the electrode. Changing the potential negative or positive of the pzc induced orientational changes in the adsorbed monolayer resulting in an increase in ion permeability. In a similar fashion, the lipids egg phosphatidylcholine (PC) and bovine spinal cord phosphatidylserine (PS) were investigated on the Hg surface.  When adsorbed, the lipid  monolayers produced a flat low-capacitance region around the pzc. At potentials either positive or negative of the pzc, peaks in the capacitance were observed and assigned to adsorption and desorption of the monolayer [85]. A review by Miller [86] describes similar results using 56  radioactively labelled oleyl alcohol. Under potential control, the Hg drop was passed through the compressed floating monolayer within seconds of being freshly extruded. The drop was allowed to expand in the monolayer until dropping into a cup existing in the subphase. The area of Hg and the radioactivity measured in the cup were used to calculate the surface concentration of lipid transferred to the electrode. Using various potentials and film pressures of the gas/solution interface, the maximum adsorption was found for potentials near the pzc for Hg. Desorption of the surfactant was achieved at large negative potentials. These complimentary results suggest that the polarization of the metal/solution interface can dictate the organizational quality of the adsorbed lipid layer. The biomimetic research of Nelson, Guidelli and Moncelli extended Miller's approach by spreading lipid monolayers onto the hanging mercury drop electrode (HMDE). Nelson has studied the adsorption of phospholipids (namely dioleoyl phosphatidylcholine, (DOPC)) onto the Hg electrode [87-90]. The method used to coat the Hg electrode was achieved by pushing the electrode through a monolayer spread at the equilibrium spreading pressure (ESP). This method of depositing the surfactant resulted in reproducible monolayer-coated surfaces characterized by low capacitance values. The monolayers were characterized with differential capacitance measurements at various potentials. The capacitance achieved a minimum value of 2 pF cm" at potentials near the pzc. This 2  minimum was taken to indicate a highly ordered monolayer transferred to the metal from the floating monolayer at a film pressure of the ESP. Applying a negative potential resulted in pseudocapacitance peaks which were assigned to dramatic changes in the adsorbed monolayer structure. At extreme negative potentials, the monolayer was desorbed from the electrode and replaced by adsorbed water. The quality of the monolayer at potentials corresponding to the pseudo-capacitance were studied by the reduction of metal ions [91,92]. Results showed an increased permeability of the monolayer through these pseudo-capacitive transition peaks. The effect of toxins on the adsorbed layer has also been studied [93,94] as well as ion transport through lipid layers with and without incorporated Gramicidin ion channels [95-100]. The investigation of thallium transport through gramicidin ion channels in DOPC monolayers has also been conduced by Moncelli and Guidelli [101-105]. Recently, Bizzotto et al. quantitatively characterized one of the pesudo-capacitance peaks using chronocoulometry and film pressure measurements [106]. It was proposed that the second 57  phase transition represents the growth and coalescence of defects. At the adsorption and desorption potentials, A C impedance spectroscopy characterized the state of adsorbed surfactant (at adsorption potentials) or adsorbed water molecules (at desorption potentials) [ 107]. At the adsorption potential, an increase in high frequency resistance was observed indicating a decrease in the mobility of electrolyte ions near the interface. This was assigned to the formation of a condensed film of the lipid on the metal surface. At the desorption limit, the same decrease in electrolyte mobility was observed. This was explained by a decrease in the exchange rate of the water molecules that exist between the desorbed surfactant and metal. The water layer was proposed to be 'ice-like' in relaxation time allowing for the lipid molecule to exist near the electrode by adhesive forces. The high quality of adsorbed DOPC lends itself to studying the disruptive effects of bioactive drugs with lipid layers.  Recently, the interaction between the lipid monolayer and various  formulations of Amphotericin B (AmB), were studied with capacitance and metal ion (Tl ) reduction +  [108]. Changes in the lipid order were found to occur through interaction with the drug. The mean size and number of pores formed in the monolayer were estimated by fitting the reduction current transients to a model of a random array of microelectrodes. The deposition of DOPC onto the Hg surface can also occur through the spreading of vesicles from the subphase [109]. Deposition occurred from three systems: DOPC from a monolayer at the air/solution, from 100 nm DOPC liposomes in the electrolyte, and a mixed system consisting of the co-adsorption of the monolayer and liposomes. Adsorption from the liposomal suspension was distinctly different compared to that adsorbed from the floating monolayer. This was manifested by the larger decrease in minimum capacitance for the liposome system.  The adsorption of vescicles from the subphase was  characterized with our recently developed method of in situ electro-fluorescence microscopy [110]. The fusion of DOPC liposomes with a previously adsorbed dye containing layer was observed. This was revealed by large changes in fluorescence after the fusion event. The fusion is suggested to occur by the interaction of the liposomes in the potential-created defects in the adsorbed DOPC monolayer resulting in a hybrid layer where the dye molecules diffuse into the freshly formed dye free vesicles. Some lipid exists far enough from the electrode surface, resulting in the measured fluorescence.  58  3.2  Adsorption of Octadecanol Onto A u d l D Among other insoluble surfactants, monolayers of octadecanol adsorbed onto low index  planes of Au was studied by Bizzotto [11]. Electrochemically, the characterization included cyclic voltammetry, differential capacitance, charge density and film pressure measurements [12,13]. Similar to the early work of adsorption onto Hg, octadecanol was spread at the air/solution interface in a Langmuir trough. For different film pressures of the floating monoalyer, the surfactant was deposited to the metal surface under potential control through the horizontal touching method. The adsorption was found to be dependent on electrode potential, the electrode surface and the film pressure of the floating monolayer. The organic monolayer transfers from the air/solution interface to the metal with a ratio equal to unity at the pzc. When the metal surface is charged, the transfer ratio drops below unity and at sufficiently negative potentials the transfer to the metal does not occur. Thus, the structure and stability of the organic film is strongly potential dependent. The electrochemical behaviour under conditions of sweeping potential for adsorbed octadecanol on Au( 111) was described in a review [14]. The differential capacitance for an uncoated electrode was measured first. The surfactant was then spread onto the air/solution interface, and the monolayer deposited onto the electrode at the pzc. The capacitance for the organic coated electrode was measured in a restricted potential range and then compared to the uncoated electrode. At positive potentials near the pzc, the adsorption is characterized by a low value of capacitance. At the negative limit, the capacitance merges with that for the clean Au surface indicating desorption of the surfactant. This behaviour is similar to the adsorption of insoluble surfactants onto the Hg surface. On the return (positive) scan, the low value of capacitance is re-established. Furthermore, on continuous cycling, the shape of the capacitance plot was unchanged indicating that the adsorption/desorption process is reproducible. Further investigations, were conducted where the monolayer was spread in a Langmuir trough and compressed. Once deposition occurred, the film was decompressed at the air/solution interface. Even at these low compressions, the electrochemical data gave the same results as for monolayers in the compressed state opening a question as to the fate of the desorbed molecules. The possibility of trapped aggregates below the electrode was then tested with elastically scattered light investigations [15,16]. Monochromatic light was directed at the electrode surface at a 45° angle of 59  incidence and the elastically scattered light was detected at a 45° angle with respect to the electrode surface and 90° to the plane of incidence. The scattered light was obtained at various potentials for coated and uncoated electrode surfaces. At the desorption limit, the amount of scattered radiation increases above that for the bare electrode. The intensity of scattered radiation decreases upon readsorbing the molecules. The scattering signal was reproducible on numerous scans, indicating that the adsorption/desorption is driven by the formation of surfactant aggregates that are trapped beneath the electrode. On re-adsorption, the aggregates spread onto the metal reforming the well-ordered adsorbed layer. The experiments conducted on the octadecanol/Au(l 11) system spurred  further  characterization by a number of authors. By incorporating a small amount of a fluorescent probe into the octadecanol matrix, Bizzotto and Pettinger conducted preliminary investigations of the desorption of octadecanol using confocal fluorescence imaging [17]. Images at adsorption and desorption were measured indicating that the aggregates were separated from the electrode surface upon desorption. Zawisza et al. [111,112] have studied this system using scanning tunneling microscopy (STM), neutron reflectometry (NR), fourier transform infrared spectroscopy (FTIR) and the in situ method of polarization modulation infrared reflection absorption spectroscopy (PMIRRAS). The layers studied include monolayer and multilayers of octadecanol on the metal surface. For monolayers, pressure measurements indicate that beyond 12 mN m" the monolayer is in a 1  compressed state and at lower film pressures in a decompressed state. PM-IRRAS measurements have indicated a 10° tilt angle to the surface normal. Therefore, the molecules are nearly vertical at positive adsorption potentials. At film pressures lower than 12 mN m" , the tilt angle increases 1  progressively with decreasing film pressure. However, these measurements probe average properties of the system and the layers are assumed to be uniform.  3.2.1  Fluorescence  Near a Metal  Surface  Investigations of the radiative pathways for fluorophores near the metal surface are sometimes referred to as radiative decay engineering [113]. Proximity of fluorophores to metal surfaces can impact fluorescence intensity, quantum yield, and excited state lifetimes [53,65,68,69,71,114], Furthermore, the magnitude of these effects is influenced by the distance 60  separating the fluorophore from the metal. Kuhn and co-workers[53,l 14] were among the first to investigate this process by measuring the absorption and emission of light from fluorescent dye molecules at various separation distances from an evaporated silver mirror on a glass substrate. At small separations (<50 nm) the excited-state lifetime approaches zero due to non-radiative energy transfer into the metal. This quenching efficiency was found to decrease as the third power of increasing separation [65]. At close proximity, quenching is dominant, however, at specific separations where standing waves can interact with the oscillating dipole, enhancement of fluorescence can occur. Such events are discussed in recent reviews [113,115,116]. It is shown that a nearby rough thin metal substrate can increase the radiative decay resulting in an enhancement of fluorescence intensity by factors up to 1000. These effects however, are restricted to the interaction of low quantum yield chromophores near rough metal surfaces. The electro-fluorescence set-up described in this thesis is capable of monitoring fluorescence from monolayers near the metal surface. One advantage of this technique is that electric potential can change the distance of separation (adsorption/desorption) avoiding the use non-fluorescent spacer molecules that result in a fixed separation distance.  3.3  Self-Assembled Monolayers of Thiols on Noble Metals The term self-assembly is widely and loosely used, which causes confusion. Self-assembled  monolayers (SAMs) have been defined as ordered molecular assemblies that are formed through the spontaneous adsorption of a surfactant headgroup that has specific affinity for a substrate. Through this definition, the field of SAMs cover a wide range of systems, though, the most common case is thiol adsorption on noble metals followed by silane based systems on Si0 . With thiol SAMs on 2  noble metals, the self-assembly is driven by a covalent bond between the metal and the sulfur group of the thiol as well as lateral interactions between neighbouring molecules [117]. For example, these interactions may include van der Waals forces, hydrogen bonding, or donor-acceptor interactions that can increase the packing density resulting in more stable adsorbed layers. SAMs on noble metals have received attention for their possible application in molecular electronics [118], chemical sensors [6,7] biochemical sensors [8], corrosion inhibition [119,120], and chemically tunable surfaces [121123] to name a few.  Usually, the S A M is formed on planar surfaces although SAMs on 61  nanoparticles are also receiving attention [ 124]. With the many application prospects of these unique interfaces, the formation, stability and reproducibility of the S A M has been the subject of much discussion [125-129]. A significant amount of literature has described the growth mechanism of SAMs on Au(l 11) surfaces because of the relative ease of preparation and cleaning of the metal. Furthermore, n-alkanethiols are the most common surfactant used because they are fully saturated and fairly simple compounds from a chemical perspective. Under conditions of 'full coverage', alkanethiols consisting of chain lengths between CIO and C20 result in a hexagonal (V3 x V3 )R30° arrangement on Au( 111) with the hydrocarbon chains tilted about 30° from the surface normal [130132]. The growth of alkane thiols from solution onto Au(l 11) surfaces has been widely studied [133-137] and a simple mechanism proposed using sum frequency generation (SFG) [138]. The initial stage of adsorption is fast and characterized by chemisorption of the head group to the metal. This process is followed by a slower step of the straightening of the alkyl chain. The final and slowest step involves the reorientation of the terminal groups. While the proposed mechanism is simplistic, the quality of the adsorbed thiol is dependent on the alkyl chain length, solution cleanliness and terminal head group. The adsorbed S A M is robust but the quality of the layer can be altered with changes in temperature. An increase in temperature can result in thermal desorption [139,140] or conformational defects [141] which defects are not annealed out by the lowering of temperature. Forming a S A M in a controlled fashion on electrodes has been completed by holding the potential at a value that favours adsorption. For thiol SAMs, very negative potentials can induce reductive desorption in strongly alkaline solutions which are required to avoid hydrogen evolution [ 10,142-161]. The desorption potentials differ depending on the crystallographic faces of the metal upon which the S A M is formed [10,162]. Reforming the adsorbed layer via an oxidative process is possible although it is dependent on the length of the alkyl chain. For small chain lengths (n = 28) irreversible desorption occurs whereas longer chain lengths (n > 16) can undergo the oxidative re-adsorption [150,152,153,163]. Furthermore, experiments using single-crystal electrodes have shown that this reductive desorption process is dependent on crystallographic orientation and on the length of the alkyl chain [10,147,150]. This has also been observed in physically adsorbed systems [164].  This process was investigated by Morin through electrochemical and FTIR studies 62  [148,151,154,158]. In-situ methods such as electrochemical quartz crystal microbalance (EQCM) [156,165] and S T M [161,166-171] have been employed to monitor the reductive desorption but the fate of the desorbed thiolate is not well understood. In the interest of forming a surface with more than one type of chemical functionality, thiol exchange reactions are commonly used [166,172,173]. These reactions involve replacing a weakly adsorbed thiol by one that is more strongly adsorbed. This difference in adsorption strength may be a result of the relative solubility between the competing molecules in solution or perhaps the sticking probability [117]. Furtheremore, the relative lateral interaction between the neighbouring molecules may provide a larger probability for one molecule to displace the existing thiol. Even though the thiol exchange reaction is commonly used for the creation of a surface with multiple functionality, creating a desired surface pattern with this method is difficult. Surface patterning can be achieved using lithography [117] but the search continues for more flexible types of chemical patterning. The ability to remove a portion of the original S A M and allow a different thiol to adsorb in the void has been accomplished [ 149,166,170,174-177]. Again, achieving this with a desired pattern is difficult. Exploiting the separation in reductive desorption potentials may assist in the creation of multifunctional surfaces. For example, an ideally shaped nanocrystal of Au exists in a truncated octahedron geometry and contains both the (111) and the (100) surfaces [178]. Thus, the selective removal of thiol from one face may be achieved through the use of electric potential.  3.4  Luminescence at Electrode Surfaces The investigation of luminescence at a metal electrode surface is brief in the literature,  presumably due to metal mediated quenching of the excited species. However, electrogenerated chemiluminescence (ECL) of molecules on or near the electrode surface has been reported [179]. In these investigations an illumination source is not required because an excited state molecule is generated through an electron transfer reaction. The electrogenerated species reacts with a molecule in the electrolyte to form the excited state molecule which can relax through the emission of a photon. The molecule Ru(bpy)  2+ 3  is commonly studied and E C L from layers of this molecule  adsorbed on the electrode surface through L B or S A M films have been reported [180-182]. It has been suggested that luminescence is observed if the rate of the radiative process is comparable to the 63  metal mediated quenching rate [183], but the decay in luminescent intensity with numerous measurements also suggests loss of the molecule from the interface. Since luminescence is not generated in the bulk solution, this possibility is difficult to determine. Aside from these reports, there is little literature on measuring luminescence from the electrode surface, especially in using a fluorescent probe for means of characterization. Fluorescent thiol molecules bound to electrode surfaces have only been briefly touched on in the literature [184-188], but the fate of a desorbed species has not been directly observed until recently [189]. The experiments described in this thesis use fluorescent molecules as a probe for imaging the potential induced response of physisorbed and chemisorbed molecules on the electrode surface [110,189-193]. The fluorescence images acquired in these investigations were used to characterize the structure of molecules near but not on the electrode surface.  3.5  Summary The adsorption of water-insoluble organic materials onto the metal surface has been studied  for a variety of electrochemical systems. The work presented in this thesis utilizes the techniques and methods for characterizing the interface electrochemically. By incorporating an in situ technique of fluorescence microscopy, the adsorption/desorption of C180H to/from the A u ( l l l ) electrode surface will be further characterized. Furthermore, mixed monolayers of OLA/C180H will be investigated using the developed technique to understand how the fluid OLA molecules effect the solid C180H matrix. As a final investigation, the selective reductive desorption of a fluorescent thiol molecule will be monitored, and the fate of the desorbed thiolate revealed.  64  Chapter 4 4  Experimental Methodology The instrumentation, materials, methods and programs used will be described in this chapter.  Electrochemical instrumentation and methods are outlined first, followed by details of the materials used in the investigations. A description of the in situ electro-fluorescence microscopy set-up will ensue after establishing a foundation for the electrochemical techniques. This will be followed by a brief description of the data acquisition, and image analysis programs.  4.1  Systems Studied The spectro-electrochemical investigations of organic modified gold electrodes was studied  in this thesis. Au was the working electrode in all investigations and was either a single crystal with the 111 face exposed, or a bead electrode with a polycrystalline surface. Au(l 11) was used in the investigations of physisorbed alcohols, and the bead electrode was used in the study of a chemisorbed thiol molecule. Three surfactants were used to modify the electrode surface. The alcohols that were physisorbed on the Au( 111) electrode include 1 -octadecanol (C180H) and cis-9octadecen-1-ol (oleyl alcohol, or OLA). These molecules were studied in both single and mixed proportions on the A u ( l l l ) surface. Octadecanol (CH -(CH ) -CH OH) is a saturated water3  2  16  2  insoluble alcohol that is solid at room temperature. Oleyl alcohol (CH -(CH ) -CH=CH-(CH ) 3  2  7  2  7  CH OH) is an unsaturated water-insoluble alcohol (cis C-9) and is a liquid at room temperature. 2  When using the electro-fluorescence technique, these surfactants were mixed with a small amount of the fluorescent carbocyanine dye, l,r-dioctadecyl-3,3,3',3'-tetramethylindodicarbocyanine perchlorate (DilC 18(5)). In some experiments, a mixture of DilC 18(5) and the fluorescent dye, 5octadecanoylaminofluorecein was used.  These fluorescent dye molecules contain saturated  hydrocarbon tails (18 carbons) extending from the fluorescent moiety, allowing for efficient mixing with the alcohols. Chemisorption onto a polycrystalline A u bead was accomplished using a fluorescent molecule (BODIPY) bearing a thiol moiety. The thiol molecule 4,4-difluoro-1,3,5,7-tetramethyl-8[(10-mercapto)]-4-bora-3a,4a-diaza-s-indacene (BODIPY-(CH ) -SH) is solid at room temperature. 2  65  10  4.2  Electrochemical Methods Preceding any fluorescent characterization of the interface, the system was first probed with  electrochemical techniques. A description of the materials, equipment and experimental methods will be given in this section.  4.2.1 Materials All electrochemical investigations were conducted in a standard three electrode cell with either a Au(l 11) single crystal or a polycrystalline Au bead serving as the working electrode (WE). The preparation of the single crystal electrode including cutting and polishing is detailed in a previous thesis [194]. The gold bead electrode was prepared by flaming a Au wire to the point of melting and slowly removing the metal from the flame. This is a similar method to the creation of Pt bead electrodes with a polycrystalline surface [195]. Both the single crystal and polycrystalline electrodes were electropolished in perchloric acid according to a literature procedure [194]. The counter electrode (CE) was a coil of either Pt or Au wire and a saturated calomel electrode (SCE) was used as a reference (RE), connected to the working solution through a salt bridge. For the physisorbed alcohol/Au(l 11) systems, the supporting electrolyte was KC10 triply recrystallized 4  from Millipore water (> 18 MQ cm). The electrolyte ranged in concentration from 0.025M to 0.05M. Previous analysis of the Au(l 11)/KC10 system revealed little difference in electrochemical response 4  within this concentration range [16]. The electrolyte in the chemisorbed thiol/Au bead system was 0.1M NaOH (Fluka suprapure, used as received). C180H and OLA (shown in Figure 4-1) were acquired from SIGMA with HPLC grade purity of 99% or better. The carbocyanine dye, DilC 18(5) shown in Figure 4-2a was purchased from Molecular Probes and was used as received.  Also shown in the figure is 5-  octadecanoylaminofluorescein (b) and BODIPY-C10-SH (c). The fluorescein dye was purchased from Molecular Probes (used as received) and the thiol molecule was synthesized by A. Kell and M . Workentin at the University of Western Ontario [189]. Separately, the surfactants were dissolved in HPLC grade chloroform (Fluka) to approximately 3mg/ml. All electrochemical experiments were conducted in an argon atmosphere (Praxair, passed through a Supelco hydrocarbon filter). The measured extinction coefficients for the fluorescent dyes is presented in Table 4-1. Literature for 66  a) Octadecanol (C180H)  b) cis-9-Octadecen-l-ol (OLA)  Figure 4-1 Ball and stick representation of the surfactants that physisorb onto Au(l 11) in the ideal all trans configuration, (a) Octadecanol or C180H and (b) cis-9-Octadecen-l-ol or OLA  a) Carbocyanine Dye DilC 18(5)  b) 5-octadecanoylaminofluorescein  c) HSC10BODIPY  Figure 4-2 Ball and stick representation of the fluorescent probes that are mixed with the surfactants (a, b) and that chemisorb onto the metal surface (c). 67  DilC 18(5) was available from Molecular Probes [196], but for 5-octadecanoylaminofluorescein and BODIPY-C10-SH, only data for similar compounds were available. Data from the reference standard for fluorescein (Fluorescein/pH 9.0) [197] was used and a compound similar to BODIPYC10-SH (BODIPY FL CI2) [198] are shown below.  Fluorescent Molecule (Molecular Probes Product Name)  Extinction Coefficient (cm" M" )  Solvent  DilC 18(5)  260 000  Methanol  644  Fluorescein/pH 9.0  48 000  Methanol  497  BODIPY FL C  87 000  DMSO  505  1  12  (nm)  1  Table 4-1 The measured extinction coefficients for DilC 18(5), the Fluorescein reference standard, and a similar compound to BODIPY-C10-SH. The extinction coefficient was measured at the indicated wavelength for the dye dissolved in the indicated solvent.  4.2.2 Electrochemical Instrumentation Electrochemical investigations were conducted on a Fritz Haber Institute potentiostat (FHIELAB #0599 or FHI-ELAB C050-0298) with an EG&G Princeton Applied Research (PAR) model 5204 lock-in amplifier for impedance measurements. A National Instruments data acquisition card (model PCI-5062E, 16-bit, 333 kHz) was used for the collection of data and in some cases used to control potential.  4.2.3 Electrochemical Procedures All glassware making up the electrochemical cells were cleaned in a hot bath of 1 part sulphuric acid and 1 part nitric acid (v:v) for a minimum of two hours.  The glassware was  thoroughly rinsed with Millipore water then filled and allowed to soak for over 12 hours. The electrolyte was introduced into the cell and purged with Ar thereby removing any dissolved oxygen. A constant blanket of Ar was always present above the solution surface to prevent oxygen from reentering the cell. After de-oxygenation of the electrolyte, the salt bridge was filled first on the side with the supporting electrolyte and then the reference side with a saturated KC1 solution. The CE 68  and W E were flamed annealed a number of times before introduction into the system and the W E was contacted to the electrolyte in the hanging meniscus arrangement. The solutions were separated by a stopcock allowing for contact between the SCE and the working solution avoiding possible chloride contamination of the electrolyte.  Figure 4-3 shows a schematic diagram of the  experimental setup for standard electrochemical measurements with the W E in the hanging meniscus arrangement. To confirm the cleanliness of the system, cyclic voltammetery (CV) and differential capacitance measurements were conducted and compared to literature. Figure 4-4 shows the full scale C V (1.250 V/SCE to -0.800 V/SCE) (a), C V of the double layer potential region (b) and differential capacitance of the double layer potential region (0.650 V/SCE to -0.800 V/SCE) (c) for Au(l 11) in contact with the 0.05M KC10 . This figure demonstrates the quality of the single crystal 4  electrode, recorded over a wide range of electrode potentials in the absence of the insoluble monolayer. The characteristic oxide and reduction peaks can be used as fingerprints for the identification of the crystallographic orientation of the electrode surface. The full scale C V agrees well with literature [199] indicating that the electrode was well oriented and properly prepared. The double layer potential window displays the capacitive currents in the double layer region and a relatively flat capacitance for Au(l 11) in contact with KC10 . It is within the potential region 0.150 4  to -0.800 V/SCE that C V and capacitive measurements for the organic coated electrode were conducted thereby avoiding oxidation or reduction reactions. Similar characterization for the Au bead immersed in NaOH was conducted preceding the S A M formation. After these initial checks, further  electrochemical characterization  such as  charge  density  measurements using  chronocoulometry (referred to as step experiments) was accomplished.  4.2.3.1 Physically Adsorbed Surfactants on Au(lll) The procedures described here are those of Bizzotto [11].  The surfactant/chloroform  mixture was deposited onto the electrolyte surface, achieved by depositing an aliquot of the mixture onto the gas/solution interface and allowing the monolayer to form at the equilibrium spreading pressure (ESP). Previous electrochemical studies of octadecanol/Au(l 11) were performed in a Langmuir trough and the results indicate that a layer spread at the ESP reproducibly transfers to the electrode surface [14]. After total evaporation of the solvent, the W E was flame annealed, dried, 69  Potentiostat & Lock-in Amplifier Computer andDAQ  Figure 4-3 Schematic representation of the three electrode cell used in the electrochemical characterization. The W E is either Au(l 11) or a polycrystalline Au bead, the C E is a Pt or Au coil, and the R E is a saturated calomel electrode (SCE). The R E is connected to the cell through a salt bridge.  70  - I — i — ' — ' — • — i — • — ' — ' — i — • — ' — • — i — ' — > — ' — i — • — • — ' — i — •  -80  -0.8  -0.4  0.4 E/Vvs SCE  0.0  0.8  1.2  10 H o  H — i — i — i — i — i — i — i — ' — i — i — i — i — i — i — i — <  -0.9  -0.6  0.0  -0.3  0.3  0.6  E/VvsSCE Figure 4-4 (a) Full scale C V (b) double layer C V and (c) capacitance for Au(l 11) in contact with 0.05M KC10 . Measurement of capacitance used a 5mV rms, 25Hz A C perturbation and the capacitance was calculated assuming the interface can be modeled as a series RC circuit. 4  71  and allowed to cool in Ar for at least 5 minutes before the electrode was touched to the surfactant/electrolyte surface and upon contact, the potential of the W E is 0 V/SCE. The electrode was raised to form a hanging meniscus thereby ensuring that electrochemistry was only conducted at the polished 111 face of the single crystal. This procedure is depicted schematically in Figure 4-5.  4.2.3.2 Chemically Adsorbed Thiol on a Au Bead A S A M of BODIPY-C10-SH was formed on the Au bead by submerging the metal for 10 minutes into a 3 mg/ml solution of the thiol molecule dissolved in chloroform. The thiol coated bead was removed and rinsed in water followed by sonication in methanol. After alcohol evaporation, the bead was introduced to the electrochemical cell and submerged into solution rather than forming a hanging meniscus. This procedure is schematically represented in Figure 4-6.  4.2.4 Electrochemical Techniques 4.2.4.1 Cyclic Voltammetry A C V involves linearly sweeping the potential and measuring the resulting current. Each C V consisted of a potential sweep (20mV/sec) to a positive potential limit at which point the linear ramp was reversed and swept to a negative potential limit. Each experiment began with the measurement of a C V in the full scale potential window (1.250V/SCE to -0.800V/SCE for Au(l 11)/KC10 and 4  0.150V/SCE to -1.300V/SCE for Au bead/NaOH ). To investigate the double layer region, the potential window was narrowed to 0.650V/SCE and -0.800V/SCE for the Au/KC10 system and 4  0.00V/SCE to -1.300V/SCE for the bead/NaOH system. The electrode was always characterized by this method before the surfactant was introduced to the system. The Labview program used to acquire the C V data was based on a similar program described in a previous thesis [11].  4.2.4.2 Differential Capacitance Within the double layer potential region, the M S interface can be simply modeled as a capacitor. The capacitance of the interface was measured in the double layer region immediately following the double layer C V measurements. The capacitance was measured using a sinusoidal voltage perturbation (25 Hz, 5 mV rms). The resulting current was analyzed by a lock-in amplifier 72  C 1 8 0 H , O L A or a Mixture dissolved in chloroform  Electrochemical Cell  0 V/SCE  Electrochemical Cell  Hanging Meniscus  Electrochemical Cell  Figure 4-5 Schematic depiction for the procedure of spreading the insoluble alcohols onto the electrolyte surface and the transfer of the alcohols to the Au(l 11) electrode surface. The hanging meniscus arrangement is also shown.  BODIPY-C10-SH dissolved in chloroform  Methanol and water rinse  Submerge Bead  NaOH  Beaker  Beaker  HfHHnft.  i  Electrochemical Cell  Figure 4-6 Schematic depiction for the procedure of forming the S A M externally from the cell and the subsequent rinsing and transfer to the electrochemical cell. The bead is submerged through the electrolyte surface rather than forming a hanging meniscus. The fluorescent thiol is not present at the gas/solution interface.  73  and the in-phase and quadrature components measured. The in-phase and out-of-phase components of the resulting current (real and imaginary currents) were collected and used in the calculation of differential capacitance. Assuming the interface can be modeled as a capacitor and resistor in series, calculation of the capacitance was accomplished using equation 4-1.  *  i  (  hm  C =  1+  Vm  4-1  re  \ im l  J  Here, C is the calculated interfacial capacitance (u.F/cm ), i 2  im  and i are the imaginary and real K  current densities (u-A/cm ), V is the amplitude of the A C voltage (mV rms) and co is the frequency 2  ac  (Hz). The program used to measure the differential capacitance is described in a previous thesis [11].  Briefly, the program was modeled after the C V program with two main differences, the  collection of a second data channel and the calculation of capacitance from the acquired data.  4.2.4.3 Chronocoulometry The chronocoulometric method for measuring the charge density of the metal surface was accomplished following the procedures of Lipkowski et al. [32]. These experiments involve measurement of the capacitive charging current due to a sequence of potential steps. These current transients are used in the determination of charge characteristic of the MS interface. Three potentials are invoked in this technique, E  basc  ,E  var  and E . Generally, E dcs  basc  is a potential chosen such that the  interface can reach equilibrium. This can be at positive (adsorption) potentials or at negative (desorption) potentials. For this work it is crucial to choose a value of E layer is not disrupted. In this thesis, E seconds and E  dcs  base  basc  such that the surfactant  was chosen to be 0 V/SCE with a waiting time of 10  was chosen to be -0.825 V/SCE or -0.850 V/SCE where the surfactant was desorbed  from the metal surface. The variable potential, E  var  began 25 mV/SCE more positive than E  dcs  and  increased in +25mV increments with each cycle. A n experiment would be as follows: the W E potential is set at E  base  for 10 seconds before stepping the potential to E  waiting period ensuring equilibrium. The potential is then pulsed to E E  basc  is again established. The current transients for the pulse from E  potential step will be exactly the same but with E  var  74  dcs  var  var  for another 10 second  for 100 milliseconds before  to E  des  is measured. The next  becoming more positive. A schematic depiction  of the sequence of potential steps is shown in Figure 4-7. From the current transients, the charge density was calculated from integration by a program similar to that described elsewhere [11]. The program calculated the charge density for each transient collected.  4.3  Spectroscopic Technique  4.3.1 Epi-fluorescence Microscopy An inverted epi-fluorescence microscope (Olympus 1X70) was employed for 'looking up' onto electrode surface while in the electrochemical cell in the hanging meniscus (Au(lll)) or submerged (Au bead) arrangement. The microscope is equipped with a 50X objective (numerical aperture (NA) = 0.5, working distance (WD) = 10 mm) or a 1 OX objective (NA = 0.3, WD = 10mm). White light from a DC-powered 75W xenon short-arc bulb (UXL-S75XE) was directed towards a filter cube that transmits a band of light specific to the fluorophores absorption spectrum. Two filter cubes were used. A Chroma filter cube (Set 41008) was used for the measurement of fluorescence from DilC 18(5) and consists of three filters; excitation (590-650nm), dichroic ( 640nm) and emission (660-740nm). A n Olympus filter cube (UM-WIBA) was used to measure fluorescence from 5-octadecanoylaminofluorescein and BODIPY-C10-SH and is made up of three filters; excitation (455-480nm), dichroic (500nm) and emission (515-550nm). The excitation and emission filters are overlaid onto the spectra of all three dyes in Figure 4-8. The fluorescent light was collected through the microscope objective and directed to the filter cube then to the digital camera or spectrometer. The monochromatic digital camera (SPOT Real Time (RT)) was purchased from Diagnostic Instruments. The camera contains a Kodak (model KAI - 2092) interline transfer charged coupled device (CCD). The CCD is Peltier cooled to 37°C below ambient to reduce dark noise and contains 1520 x 1080 pixels. Each pixel is 7.4 pm by 7.4 pm in size. The fluorescence image was focused onto the CCD with a 0.76X coupler (Diagnostic Instruments) allowing full fill of the CCD. The quantum efficiency of this CCD is shown in Figure 4-9. All images were acquired using SPOT software with 2x2 binning resulting in an image that is 760 x 540 pixels with a resolution of 8 bits (256 grey levels). With the 50X objective the images are roughly 0.2mm wide by 0.3mm high with a resolution of 0.385 pm/pixel. The diffraction limited resolution describing the limit at which two points can be separately distinguished is 0.854 pm for a wavelength of 700 nm (UM-WIBA 75  0.4 0.2  J-'base A  0.0  T  -0.2 -0.4 -0.6 -0.8 -1.0  E des  "X"  0  —r~ —! 25 50  1  1  1  1—  1  I  1  1  1  1  1  1  1—  75 100 125 150 175 800 825 850 875 900 925 950 975 Time / sec  Figure 4-7 A depiction of the potential steps used in chronocoulometry. E is chosen at a potential where the system can equilibrate. E is a variable potential that increments in the positive direction after each measurement cycle. E is the desorption potential value where the surfactant is removed from the metal surface. The charging current is measured between the step from E to basc  var  des  var  76  -Absorption — Emission]  a)  A;. .  Chroma Set 41008 Excitation Filter  '^\  / /  Chroma Set 41008 Emission Filter  \  ;  \  :,  '''''  \  t '• \  0 4480  530  580  630  ^  680  730  Wavelength (nm)  b)  UM-WIBA j Excitation Filter i  A / '  / |\  \ / ;\ \ / j \  1:  \ i  / \ i  380  430  480  UM-WIBA  I Emission 1 Filter  i "\  530  630  Wavelength (nm)  c)  n  UM-WIBA''  r  :  i UM-WIBA  Excitation j Filter i  / 1 ^  ; Emission ; Filter  / -1 • \ I A '•• \  /  / ;:\ '1 ' ^  20  20  480  530  Wavelength (nm)  Figure 4-8 Absorption and emission spectra for (a) DilC 18(5), ( b) Fluorescein and (c) BODIPYC10-SH with the excitation and emission filters of the Chroma Set 41008 and UM-WIBA overlayed. The spectra for DilC 18(5) was taken from [196]. The spectra for fluorescein was taken from the Molecular Probes reference standard [197] and the spectra for BODIPY was taken from a similar molecule to BODIPY-C10-SH [ 198].  77  < 0.00 J  300  600 Wavelength (nm)  Figure 4-9 Quantum efficiency of the Kodak KAI-2092 CCD. Taken from [200].  I 9 0 0  emission) and 0.650 pm for wavelength of 532 nm (Chorma filter emission). A l l fluorescence imaging experiments were performed in a light tight box effectively eliminating ambient light. The spectro-electrochemical cell used for the in situ electro-fluorescence investigations was a modified from the standard electrochemical cell.  To allow for movement of the entire  electrochemical cell above the objective, the salt bridge was connected directly to the cell. Filling the salt bridge was accomplished by flowing the working solution through the stopcock. Concentrated KC1 was not used in the reference side of the salt bridge. However, to avoid chloride contamination, the SCE was still separated from the working solution. The spectro-electrochemical cell was specially designed with a viewing port for the objective at the bottom of the cell. The glass window in the cell was a 0.17mm thick coverglass slide. This thin optical window resulted in images of excellent quality. The experimental setup for in situ electro-fluorescence investigations is shown schematically in Figure 4-10.  4.4  Description of the Programs Used  4.4.1 Electro-fluorescence Imaging Programs The programs described here were written in Labview and were designed to control the potential of the W E while acquiring fluorescence images from the surfactant coated electrode surface. As such, three potentials were used in the programs: E , E ads  dcs  and E  i m a g c  corresponding to  potentials where the surfactant is adsorbed, desorbed or intermediate between these states. These programs were written in combination with a sequence of images acquired and then saved by the SPOT program. The number of images to be acquired within a specified time interval were defined with this software package. Image acquisition parameters such as binning, bit-depth, exposure time and gain were also defined using SPOT software. At the specified time, the images were acquired and then automatically saved to a desired folder. Labview sets the potential to a predefined value and waits until the image is acquired. After image acquisition, Labview steps to another potential and waits for the next image to be acquired and saved. After the collection of each image the potential, capacitance and image parameters were written to a spreadsheet which is saved at the termination of the image/potentail sequence.  79  CE  RE  WE  Potentiostat & Lock-in Amplifier Computer andDAQ  Xe Arc Lamp Spot RT CCD Figure 4-10 Spectroelectrochemical cell used in the electro-fluorescence experiments. The coverglass above the objective is 0.17mm thick. The objective is interchangeable with a 10X objective and the SPOT CCD can be switched with a spectrophotometer. The filter is on a rotation device to allow for interchange of filters during one experiment.  80  4.4.1.1 image_scan.vi This Labview program was written to control the potential of the W E and measure capacitance during the acquisition of images. A n array of potentials between E  ads  and E  des  were  defined as well as the image acquisition parameters. The program was initiated by holding at E  ads  until the first image in the sequence was acquired by SPOT. Labview then stepped the potential negatively to the next value in the array (E  imagc  ) awaiting acquisition of the next image in the  sequence. The program continued in this manner until E  dcs  where E  i m a g e  positive (return) direction. The program was terminated when the E  ads  was then incremented in the potential was reached on the  return scan. The potential variation for this program is depicted in Figure 4-11(a). For the remainder of the thesis, these experiments are referred to as an imaging/potential scan sequence.  4.4.1.2 image_adsorb_neg.vi and image_adsorb_pos.vi This Labview program is similar to imgae_scan.vi except the potential perturbation is different. Specifically, the program was written to understand the effect of the potential perturbation on the structure of the desorbed surfactant. It works in a similar fashion to image_scan.vi but imposes a different potential stepping procedure. The array of potentials for image_adsorb_neg.vi is not cyclic and varies negatively from E E  ads  ads  to E  dcs  only. The program was initiated by holding at  until the first image in the sequence was acquired by SPOT. Labview then stepped the potential  negatively to the next value in the array (E  imagc  ) awaiting acquisition of the next image in the  sequence. Immediately after the image was acquired, the potential was stepped to E  ads  for 10  seconds. After the 10 second waiting period, the potential was stepped negatively to the next value in the array (E  image  ) awaiting acquisition of the next image in the sequence. This potential stepping  procedure was repeated until the negative desorption limit and the program terminated. The potential variation for this program is depicted in Figure 4-ll(b). Performing the same potential stepping/imaging procedure for the positive potential scan requires consideration of the hysteresis observed between the positive and negative going potential scans. The potential was set to E  ads  for  5 seconds and then pulsed to E , desorbing the layer for 5 seconds. The potential was then stepped des  to E  imagc  where a fluorescent image was collected. In this manner, the layer is desorbed and then the  re-adsorption process can be investigated. This was achieved using image_adsorb_pos.vi. The 81  0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 Time ( s e c o n d s )  b)  E  0  30  60  90  120  150  180  210 240  180  210  Time (seconds)  c)  90  120 150  240  Time (seconds)  Figure 4-11 Schematic depiction of the potential control during the acquisition of images in the double layer potential region. The values of E are not an exact representation of the experimental values and may vary for a particular experiment, (a) represents the potential variation for image_scan.vi. (b) represents the potential variation for image_adsorb_neg.vi and (c) represents the potential variation for image_adsorb_pos. vi. The symbol ^ indicates the potential where the images were acquired. 82  combination of image_adsorb_neg.vi and image_adsorb_pos.vi constitutes what will be referred to as a image/potential step sequence.  4.4.1.3 potential_h.olding.vi This program was designed to measure the changes in fluorescence intensity with time. Labview set the potential to E  dcs  and maintained this value for a predetermined amount of time.  During this holding time, a sequence images of the interface were acquired under the control of SPOT software.  4.4.2  Data  Analysis  Programs  All images were analyzed using IMAGE PRO PLUS 4.5 and dipimage [201] for M A T L A B . The average grey scale of the image will be considered proportional to fluorescence intensity since a typical CCD response is linear with intensity [202].  4.4.2.1 histogram_calc. m The raw images collected in an imaging experiment were analyzed by this Matlab script that is completely described in Appendix A1. A l l images in the sequence were arranged in an image array and flat field corrected by division of an I image (I/I ). Generally, I is an image of the 0  0  0  electrode surface with the filter cube in place, before the introduction of surfactant. The intensity of this image results from a small amount of light that leaks through the filters making up the cube. This serves to correct for any nonuniform illumination. All I/I images in the array had an adsorption 0  I/I image (I /I ) subtracted from them yielding AI/I images. A mask was applied to the array of 0  ads  0  0  AI/I images to eliminate any intense features that may bias the calculation of the average image 0  intensity. This mask is referred to as a 'hotspot' mask, and was created from the I /I image. An ads  0  edge enhancing filter (gradient magnitudefilter)was applied to the I /I image to outline the intense ads  0  features. The outlines were then closed to fill the features, and the image was made binary (black and white) through a threshold function. The features of the mask were assigned a pixel value of 0 and the remainder of the image were assigned a pixel value of 1. The array of AI/I images were then 0  multiplied by the hotspot mask. The average histogram value of the image was calculated excluding 83  the zeros created from imposing the mask. This process is schematically depicted in Figure 4-12. If an I image was not available, the raw I 0  ads  image in the sequence was subtracted from all others  images in the array rendering A l images. In this situation, the hotspot mask was again applied and the average histogram was calculated excluding the zeros created from the mask. The program is also able to delineate features of the desorbed surfactant by application of the gradient magnitude edge enhancing filter to the AI/I or AI images. These images are then thresholded and statistics such 0  as average area, and number of particles was obtained using an IMAGE PRO PLUS 4.5 routine. The consistency of the lamp intensity was quantified by measuring images of the electrode surface in air with the Chroma and UM-WIBA filter cubes engaged during a typical image/potential scanning routine. This was performed with the same image acquisition parameters used in an electro-fluorescent investigation. The UM-WIBA filter had a non-linear decay in the measured intensity over the time taken for the imaging sequence. The decay was presumably due to a temperature effect causing a shift in the transmission range of one of the filters in the cube. As a result, all of the average histogram values of the fluorescence images obtained with the UM-WIBA filter cube were corrected for this decrease in the incident intensity. The process by which the intensity is corrected assumes a linear decay in the intensity over 5 minute intervals. The slope and intercept of the decay in 5 minute intervals was used to correct for the intensity decay with time. Figure 4-13 shows the lamp intensity with time (a), calculated from the histogram_calc.m routine for the Chroma and UM-WIBA filters. The corrected lamp intensity from the UM-WIBA filter is shown in (b).  4.4.2.2 count_features.scr This is an IMAGE PRO PLUS 4.5 macro that calculates the number and average area of the particles within an image. The analysis was run on the binary images output from histogram_calc. m. A watershed filter was applied to the images which segregates touching particles and the particle statistics were determined. Features that were at the edge of the image or had a minimum diameter smaller than the spatial resolution of the microscope were not counted in the routine. As a safeguard, the minimum diameter was 1 pm which is slightly larger than the diffraction limited resolution.  84  AI/Io  Intense features from Iads/Io 'hotspot' mask  Masked AI/I images  Figure 4-12 Schematic depiction of the procedure for rendering masked Al/I images from which the average histogram is calculated. The average histogram is calculated excluding zeros created from the mask. Fluorescence intensity is assumed proportional to the average grey scale. D  85  -0.025 -0.050 o  Chroma  o  UM-WIBA  -0.075 -0.100  '  — i  300  600  900  1500  1200  1  1—  •  1800  2100  Time / s e c  b)  0.025 0.000 -0.025 -  < -0.050 o  UM-WIBA  •  U M - W I B A Corrected  -0.075 -0.100  —•  0  1  300  •  1  600  •  1  •  900  1  1200  •  1  1500  1  1  1800  •  1  1  2100  Time / s e c  Figure 4-13 (a) Calculated mean histogram intensity for the images of the Au(l 11) electrode in air using the UM-WIBA (o) or Chroma ( 0 ) filter cube. The decay observed for the UM-WIBA filter set was corrected assuming a linear decay over 5 minutes (300 sec) intervals. The correction (•) is shown in (b).  86  Chapter 5 5  Electrochemical Investigations of C180H and OLA on Au(lll) The electrochemical characterization of lipid-like layers physisorbed onto Au(l 11) is detailed  in this chapter. The adsorption/desorption of octadecanol and oleyl alcohol both in single and mixed proportions to/from a Au(l 11) surface is demonstrated and characterized using cyclic voltammetry, differential capacitance, charge density and film pressure measurements. To determine the viability of the electro-fluorescence technique, the influence of one of the fluorescent molecules (DilC 18(5)) on the electrochemical character of the adsorbed layer of C180H is determined.  5.1  Electrochemical Characterization  5.1.1  CV and Differential  Capacitance  of  OLA/C180H  Experiments were conducted on an interface modified by various adsorbed layers which consist of 0, 25, 50, 70 and 100 mol% OLA in C180H. A stack plot of the C V and differential capacitance measured for the mixed layers is shown in Figure 5-1. The left and right columns represent the C V and capacitance measurements respectively. For both columns, pure C180H is situated at the top with increasing OLA content going down the column. The dotted line represents the C V and differential capacitance for Au(l 11) in the absence of adsorbed surfactant and the bold and thin lines represent the positive and negative going potential scans for the surfactant coated electrode respectively. The C V and capacitance for (a) C180H are very similar to those reported previously [12,13,16,203]. At positive potentials (near the pzc, 0.255 V/SCE), a decrease in the capacitive current occurs in the C V for the C180H coated electrode. This corresponds to a differential capacitance value near 6 U.F cm" in comparison to values greater than 30 |iF cm" in the 2  2  absence of the organic molecules. The large decrease in capacitance results from the adsorption of a low dielectric material onto the metal surface replacing adsorbed water. The minimum capacitance near the pzc is characteristic of organic adsorption onto the electrode surface. When scanning potential in the negative direction, the C V and capacitance display a series of peaks between -0.150 and - 0.325 V/SCE. These pseudo-capacitance peaks are due to potential dependent orientational changes in the adsorbed layer. The increase in capacitance at these potentials indicate that the  87  3 2 1 1 a ) C180H 0 -1 3 2 1 0 -1  20 10 _ :  v.-  0  b ) 25mol% O L A  :  20 10  -7  0 3 CN i 2 E 1 o < 0 -1 —  c) 5 0 m o l % O L A  eg  E o  .  UL  —  3 2 1 0 -1  :  CA7V  —  •  o  d) 70mol% O L A f  -  10 '0  J  10  j  ----••-""v 0  3 ] 2 1 0 -1  X  -  20 '-  ^-rr:  "  20 '-_  e) O L A  :  20 10 '-_  1  i  1  1  ' i  1  1  1  i '  1  1  i  1  1  1  i  1  1  1  -0.8 -0.6 -0.4 -0.2 0.0 E / V vs S C E  i  0  1  0.2  :  -i—|—i—i—i—|—i—r—i—j—i—i—i—|—i—i—i—j—i—i—i—|—i  -0.8 -0.6 -0.4 -0.2 0.0 E / V v s SCE  0.2  Figure 5-1 Cyclic voltammetry (left column) and differential capacitance (right column) for Au( 111) in the absence (•••) and presence (—) of the surfactant. Thin lines represent the negative going potential scan and bold lines represent the positive going scan. The capacitance was calculated using a 5 mV rms A C potential perturbation consisting of 25Hz.  88  adsorbed layer is disorganized allowing water to penetrate. Continuing the potential scan in the negative direction increases the capacitance to values near 13 p;F cm' remaining fairly constant over 2  a large range of potentials. Nearing -0.600 V/SCE, an increase in the capacitance reveals the onset of desorption followed by coincidence of capacitance with the water coated surface near 17 [J.F cm' . 2  This merging is not as clear in the C V because the initial stages of hydrogen evolution cloud this observation. Assuming the interface can be modeled as a series RC circuit, the C180H molecules are displaced from the metal surface and replaced by water at these negative potentials. On the return scan, capacitance remains coincident with that for the water covered electrode until values just positive of -0.400 V/SCE where a slight decrease in capacitance occurs. This decrease in capacitance is proposed to arise from contact adsorption of C180H during the initial stages of readsorption [11,204]. The C V reveals a small peak at this potential which is followed by a sharper re-adsorption peak appearing in the capacitance plot as a sharp depression reestablishing a capacitance value near 6 \x¥ cm" . The proposed method for re-adsorption involves the contact 2  adsorption of aggregated C180H molecules followed by spreading [11,204]. This process of adsorption/re-adsorption is repeatable over many hours as long as slow potential scans are used. The C V and differential capacitance curves presented in (e) for pure OLA are similar to C180H in that depressed capacitance is observed at positive potentials, and desorption is revealed at negative potentials. However, the peaks present in the C V and the capacitance for OLA occur at slightly shifted potentials as compared to C180H. At positive potentials, the minimum capacitance measured for OLA is 3.5 |J,F cm" in comparison to 6 U.F cm" for CI 80H. This suggests that OLA 2  2  forms a more compact or defect free layer on the electrode surface. Since the dielectric constant for C180H and OLA are similar, the lower value of capacitance measured for OLA indicates that the Au surface is more effectively blocked from the aqueous electrolyte. This may result from a more well-ordered or highly packed monolayer of OLA in comparison to C180H. However, when considering the cis configuration of OLA in comparison to the linear C180H molecules, an wellordered OLA layer may not be realizable. Therefore it is possible that the OLA layer is defective in terms of molecular packing, but is arranged on the electrode surface in such a manner to effectively block water from penetrating the interface. In contrast, the adsorbed solid C180H layer may contain defective regions which allow water to adsorb on the electrode. The relative magnitude 89  of the minimum capacitance suggest this possibility. When continuing the potential scan in the negative direction, the low capacitance value is maintained until -0.150 V/SCE and the first pseudo-capacitance peak is observed at -0.200 V/SCE. This is followed by another pseudo-capacitance peak at -0.375 V/SCE. The capacitance increases through these peaks, achieving a value just above 12 [iF cm" at -0.400 V/SCE. Therefore, the 2  adsorbed OLA layer is slightly disorganized through defect formation similar to C180H. A general increase in capacitance is observed at potentials negative of the pseudo-capacitance peaks in contrast to the relatively constant capacitance for C180H. OLA may not maintain a stable state during these potentials due to its fluidity. At -0.500 V/SCE, the capacitance for OLA, is significantly high (16.5 p,F cm" ) compared to that measured for C180H (13 \i¥ cm" ). This demonstrates that the onset of 2  2  desorption for OLA occurs at less negative potentials than C180H. Desorption of OLA is observed at -0.700 V/SCE to -0.800 V/SCE. The general trend on the positive scan of potentials shows a gradual decrease in capacitance for OLA. Deviation in capacitance from the water coated electrode, occurs with a decline in capacitance beginning at -0.500 V/SCE. This was not observed in C180H until more positive values of -0.400 V/SCE just before the contact adsorption peak. For OLA, this peak is observed at -0.375 V/SCE again at more negative potentials than for C180H. The minimum capacitance is again established at the positive potential limit demonstrating the re-adsorption of OLA onto the electrode. The introduction of various amounts of OLA in C180H (Figure 5-1 b, c, and d) cause significant changes in the characteristics of the adsorbed layer. With increasing OLA content, the peaks in the capacitance curves shift slightly in the negative potential direction. This negative shift is observed at potentials where capacitance diverges from that for the water covered surface and at potentials where the depression in capacitance follows contact adsorption. The pseudo-capacitance peaks also show a negative shift with increasing OLA content. These changes also coincide with the evolution of the peaks in the C V . The differences are indicative of a more organized or less defective layer that is stable on the electrode surface as compared to pure C180H. A deviation in the minimum capacitance is observed with increasing OLA content. Figure 5-2 shows the minimum capacitance for the various mixtures of OLA in CI 80H. The capacitance was measured at the positive potential limit (0.150 V/SCE) averaged over several data sets with the 90  mol%OLA Figure 5-2 Measured capacitance at 0.150 V/SCE for the various mixtures of OLA in C180H. The error bars reveal that only 0, 25, and 100 mol% layers are reproducible  91  error indicated on the graph. The general decrease in C  mjn  with OLA content suggests a more defect  free layer of OLA adsorbed onto the Au surface at potentials close to the pzc. A slight deviation from ideal mixing was observed indicating that CI 80H and OLA are somewhat immiscible. These observations are slight compared to completely immiscible mixed layers adsorbed on Au( 111) [203 ]. The minimum capacitance for C180H, OLA, and 25 mol% OLA/C180H were reproducibly measured as indicated by the error bars in the plot. Layers containing 50 and 70 mol% OLA/C180H were not created as reliably and required close inspection during step experiments (chronocoulometry) because the stepping procedure may disturb the adsorbed layer.  5.1.2 Charge Density and Film Pressure of OLA/C180H A more quantitative measure of the quality of the adsorbed layers was obtained with charge density measurements (chronocoulometry or step experiments). In these investigations the charge density was calculated from the current transients resulting from a potential step from a positive potential to the desorption potential. Therefore, since capacitance is related to the charge density by the integration of charge with respect to potential (Chapter 2), the charge density curves can be compared to the capacitance measured during the negative scan of potentials. Because the potential stepping procedure may reduce the quality of the adsorbed layer, only those layers which 'survived' the stepping procedure were used in the analysis (as determined by the similarity in the capacitance measurements before and after steps). The charge density curves determined for mixtures of OLA/C180H are shown in Figure 5-3. The charge density for an adsorbed layer of C180H (•) is similar to previous reports [12,13,16,203]. For potentials between 0.150 and -0.100 V/SCE a region of constant slope arises and signifies a potential range of constant capacitance, calculated to be 8 |iF cm" , similar to that measured by A C voltammetry. At more negative potentials, the charge density 2  decreases sharply between -0.150 and -0.250 V/SCE consistent with the pseudo-capacitance peaks measured at these potentials during the negative potential scan. More negative of -0.250 V/SCE, the charge density decreases less rapidly and is relatively constant until -0.500 V/SCE much like the fairly constant capacitance measured over this potential range. Negative of -0.500 V/SCE the charge density becomes consistently more similar to the values for the uncoated electrode surface signifying the desorption of the organic molecules.  92  -0.8  -0.6  -0.4 -0.2 0.0 E / V vs SCE  0.2  0.4  Figure 5-3 Charge density curves measured for the various mixtures of OLA/C180H adsorbed onto Au(l 11). The pzc for the Au(l 11) electrode in contact with KC10 occurs at a potential 0.255 V/SCE. The maximum film pressure is obtained where the charge density for the coated and uncoated electrode cross. 4  93  The inclusion of OLA changes the region of constant slope at positive potentials. This yields a decreasing value for the capacitance with increasing OLA content, similar to that measured in the A C voltammeric measurement. The increase in OLA content also shifts the pzc to values similar to the uncoated electrode surface. These changes reveal a changing orientation of the adsorbed layer, with the small dipole in the head group of C180H oriented normal to the electrode surface and moreperpendicular to the electrode for OLA. It is also possible that the more defective CI 80H layer contained water that was strongly organized resulting in a shift in the pzc. The steep decrease in charge density at potentials negative of -0.150 V/SCE is observed with the OLA mixtures, but occurs at negatively shifted potentials compared to C180H consistent with the shift in the pseudocapacitance peaks. Furthermore, a change in the slope is observed between -0.300 V/SCE to -0.400 V/SCE for pure OLA resulting from the observed peak in the C V in this potential range. From the charge density plot, the film pressure can be calculated from the area between the curves for the coated and uncoated electrode. The was accomplished using the back integration technique and the results are shown in Figure 5-4. The curve for the C180H coated metal (•) reveals three partial parabolas (three separate electrocapillary curves) indicated by two weak inflection points. In keeping with the negative potential scan direction, the first inflection point is at -0.250 V/SCE and the second at -0.600 V/SCE. The three electrocapillary curves are shown schematically in Figure 5-5 and correspond to three different adsorbed states of the surfactant. The three adsorbed states are defined by the potentials where the surfactant is adsorbed (y ), slightly 2  disrupted after the pseudo-capacitive potential region (y0 and desorbed (y ). This figure is only a 0  schematic representation of the electrocapillary curves for the three adsorbed states and is similar, but not an exact representation to the obtained film pressure curve depicted in Figure 5-4. From a thermodynamic standpoint, the lowest energy will be maintained as shown by the bold line. Therefore, as the interfacial tension between the metal and surfactant is changed by either the electric variable or charge density on the metal, the minimum energy will be maintained by moving from one electrocapillary curve to the next.  The crossing point of the two electrocapillary curves is  discontinuous and results in changes in the capacitance since the capacitance is the second derivative of interfacial tension with respect to potential (Chapter 2). At positive potentials, the electrocapillary curve for the adsorbed state (y ) will be followed. This results in the large film pressures (y - Y2) 2  0  94  60.0  -0.8  -0.6 -0.4 -0.2  0.0  0.2  0.4  E / V v s SCE Figure 5-4 Calculated film pressure for the various mixtures of O L A and C180H adsorbed on Au(l 11). The film pressure was calculated using the back integration method. The inset shows the variation in the maximum film pressure with OLA content.  95  Y  Y (E) 0  7U(E) = y (E) - Y (E) 0  E / V v s SCE  2  •  Figure 5-5 Schematic representation of three electrocapillary curves for three states of the adsorbed layer. Y 2 represents the potential region for the adsorbed layer. Yi represents the disrupted layer at potentials negative of the pseudo-capacitance peaks and y represents the electrocapillary curve for the uncoated electrode. The film pressure is calculated by difference between the electrocapillary curves. The lowest energy will be maintained as represented by the bold line. 0  96  measured at positive potentials in Figure 5-4. The y curve will be maintained until -0.250 V/SCE 2  where the system moves to the Yi electrocapillary curve manifested by the inflection point at this potential in Figure 5-4. The crossing point of the two electrocapillary curves at this potential results in the observed pseudo-capacitance peaks. The difference between Y and Yi results in the decreased 0  film pressures measured negative of -0.250 V/SCE seen in Figure 5-4. The last crossing point between the electrocapillary curves is at -0.600 V/SCE where the increase in capacitance signifies desorption of the organic molecules. This is consistent with an interfacial tension equivalent to a water covered electrode (Y ) at these negative potentials resulting in the minimal film pressure. It 0  is stressed that Figure 5-5 is a simplified schematic depiction of the electrocapillary curves. Furthermore, the electrocapillary curves for the negative and positive going potential scan directions would be slightly different due to the hysteresis observed in the desorption and re-adsorption process. From Figure 5-4 C180H has a maximum film pressure that is slightly higher than the equilibrium spreading pressure of C180H on the water surface as measured previously [13]. With the incorporation of OLA, a shift in the potential of maximum adsorption occurs consistent with the shift in the pzc measured in the charge density plot. The inset, shows the variation of the maximum film pressure with increasing OLA content. The maximum film pressure was largest for the 70 mol% OLA/C180H mixture consistent with the largest depression in minimum capacitance for this mixture. Again, this data indicates a slight deviation from ideal mixing. The electrochemical characterization indicates that the two components do not form an ideally mixed layer, but display more ideal character than a previously studied mixed layer [203]. The average properties in the C V and capacitance measurements show some compatibility due to the similar nature of the molecules. However, the difference in layer phase (solid vs. liquid) does suggest possible phase segregation that may give rise to some non-ideal mixing.  The  electrochemical characterization described above show an average behaviour of the layers making it difficult to conclude if mixing is purely ideal or non-ideal. The in situ fluorescence measurements may clarify this issue by probing the nature of the desorbed layer. Since C180H adsorbed onto Au( 111) has been well characterized in the literature, DilC 18(5) was mixed in small concentrations with CI 80H to determine the disruption that the dye introduces in the adsorbed layer. 97  5.1.3  Electrochemical  Characterization  of  DiIC18(5)/C180H  Mixed layers composed of 0, 1, 3, and 5 mol% DilC 18(5) in C180H adsorbed on Au(l 11) were studied to estimate the degree of disruption that the fluorescent dye introduced into the adsorbed layers. A stack plot of the C V and capacitance measurements for the mixed layers is shown in Figure 5-6. From the C V and capacitance data, the effects of the dye appear slight, and most noticeable in the C V peak heights. A broadening of the peak at -0.250 V/SCE, and a decrease in the peak at -0.125 V/SCE occurs with increasing dye content (plots a to d). The minimum capacitance (shown in the top of the figure) does not vary much with dye content, but the pseudocapacitance peaks for both the adsorption and desorption processes undergo some subtle changes. Namely, the height of the adsorption peak, increases with dye content indicating a slight decrease in organizational quality of the adsorbed layer. These peaks are kinetic in nature, and larger peaks generally point to a more disrupted layer. Charge density and film pressure were also measured for these dye containing layers and only subtle differences are noticed, as observed in the maximum film pressure measured at the pzc shown in Figure 5-7 revealing a slight decrease with the incorporation of the dye. The variation is fairly linear and decreases by 6 mN m" at the 5 mol% ratio. When comparing the electrochemical data, 1  it appears that only a small amount of disruption occurs as a result of the inclusion of dye molecules. With such a small change in the film pressure with increasing dye content, fluorescence images of the surfactant were acquired with 3 mol% concentrations of the dye. This small disruption to the adsorbed layers is acceptable so that enough dye can be included to obtain measurable fluorescence.  5.2  Summary and Conclusions Mixed layers of OLA and C180H adsorbed onto A u ( l l l ) were characterized using  electrochemical methods. The C V and capacitance measurements for pure OLA and C180H indicate that OLA forms a more well-ordered and defect free layer on the electrode surface. Both molecules adsorb on the electrode at positive potentials and undergo orientational changes when potential is swept in the negative direction. The slight disorder of the adsorbed layer is manifested by the shift in the pseudo-capacitance peaks. At more negative potentials, the molecules were displaced from the electrode by water, revealed through the merging of the capacitance curves in 98  3 2 1 0 -1  \  a)  C180H  H  20  :  V \  10 0  3  b)  ?  1mol%dye  jj  CM  :  E 20 o UL 10 ':  •  o 1 < 0 :r -1 3 2 1 0 -1  O c)  3mol%dye , ,~„  Ji  i'.  20  i  111  i  1 11  i  111  1  11  :  10 -j 0  :  20 10 ; 0  i ' i • i 1  J  ' \  .v.:_,_  3 1 d) 5mol%dye 2 1 0 -1 1  0  1  -0.8 -0.6 -0.4 -0.2 0.0 E / V vs SCE .  : 1  I  1  1  1  I  1  1  1  I  1  1  ' I  1  1  1  I  1  1  1  -0.8 -0.6 -0.4 -0.2 0.0 E / V v s SCE  0.2  I '  0.2  Figure 5-6 Cyclic voltammetry and differential capacitance for the various concentrations of DilC 18(5) in C180H. The inset shows the variation in the minimum capacitance (measured at 0.150 V/SCE) with dye content. 99  100  the absence and presence of the surfactant. OLA re-adsorbs onto the electrode surface at slightly negative shifted potentials compared to C180H, and result in a lower value for the minimum capacitance. Mixtures of the two alcohols reveal features characteristic of both pure OLA and C180H. The increase in OLA character can be measured indirectly through the formation of a more defect free, lower capacitance layer. Using step experiments and the back integration technique, the film pressure for these adsorbed layers was obtained. Results indicate that the film pressure increases with the inclusion of OLA. This supports the conclusion of a better ordered layer formed by adsorbed OLA. Layers of C180H, 25 mol% OLA/C180H and O L A were reproducible and reliably adsorbed onto the electrode surface.  Fluorescence imaging of the potential induced  behaviour of OLA, C180H as well as the 25 mol% OLA/C180H will be studied as these are the most reproducible adsorbed layers. The inclusion of DilC 18(5) in CI 80H was studied to determine the degree of disruption to the adsorbed layer caused by introducing the dye. For small concentrations of the dye, only slight changes, most notably in the decrease of maximum film pressure was observed. The measurements of charge density and film pressure are sensitive methods for quantifying the disruption. Since only a small change in the film pressure (decrease of < 4 mN m"') occurred for the 3 mol% DiIC18(5)/C180H layer, this concentration was determined suitable for the electro-fluorescence investigations. The change in the minimum capacitance for this 3 mol% layer is slight and during in situ experiments, capacitance will be used to confirm the continued quality of the adsorbed layer. From the results obtained for the DilC 18(5)/C 180H layers, it is expected that small amounts of the fluorescent dye will introduce only small deviations in the electrochemical characteristics of the adsorbed layers of OLA and 25 mol% OLA/C180H. With the small deviations observed in the C180H layers, it is believed that the dye mixes well with the organic layer and can be used to faithfully report on the potential induced changes within the adsorbed layers. During the acquisition of fluorescence images, the capacitance will be measured as an indication of the quality of the adsorbed layer during the in situ investigations. As a minimum standard, the capacitance curves during fluorescence imaging must not deviate significantly from the capacitance measured for the dye-free layers. This will serve as a check to ensure that the imaging technique is characterizing the potential induced behaviour of these lipid-like surfactants adsorbed onto the electrode surface.  101  Chapter 6 6  Fluorescence Imaging of Physisorbed Alcohols on Au(lll) This chapter describes the electro-fluorescence characterization of physisorbed alcohols  adsorbed onto a A u ( l l l ) electrode surface.  Initial fluorescence imaging investigations were  performed on adsorbed C180H containing 3 mol% DilC 18(5). Slow potential sweep and potential pulse experiments were conducted on this adsorbed layer, and the desorbed surfactant morphology revealed. To verify that DilC 18(5) was reporting on the potential induced behaviour of C180H, a second investigation was conducted where C180H was mixed with two fluorescent dyes; DilC 18(5) and 5-octadecanoylaminofluorescein and the image features compared. The nature of the desorbed molecules were significantly similar demonstrating that the dye molecules accurately reported on the behaviour of the adsorbed/desorbed alcohol. The desorbed C180H layer was heterogeneous in nature and made up of small aggregated features which did not change in morphology as long as slow potential sweeps were used. To probe the potential dependent nature of a mixed monolayer, a 25 mol% OLA/C180H layer was imaged and the mixing of the two surfactants investigated.  6.1  Imaging 3moI% DiIC18f5VC18QH Adsorbed onto A u f l l l ) The intensity of fluorescence measured from the interface is affected by several factors,  including distance-dependent quenching by the metal, light induced photobleaching and other processes (e.g. aggregation, exciplex formation), that may change the photophysical response of the fluorescent dye. The image analysis (calculated mean histogram intensity, delineation of features in the images) is dependent on the intensity of the image, therefore a complete description of the results must consider any known effects that can modify the fluorescence intensity.  6.1.1 Collection and Treatment of Images Images of the organic coated metal surface were collected using the Labview programs image_scan.vi, image_adsorb_neg.vi and image_adsorb_pos_cap.vi described in Chapter 4. Before fluorescence imaging, a C V and capacitance measurement of the interface were acquired in the double layer region to determine the quality of the adsorbed layer. A l l fluorescence images for a  102  given adsorbed layer were acquired from the same region of the electrode surface with the same image acquisition parameters to determine the reproducibility of the desorption/re-adsorption process. In each investigation, the metal/surfactant interface was exposed to constant illumination from the light source. The raw images collected in each procedure were treated by the Matlab histogram_calc.m script (fully described in Appendix A l ) producing flat-field corrected (AI/IJ images of the interface.  6.1.2 An Imaging/Potential Scan Investigation The adsorbed 3 mol% DilC 18(5)/C 180H layer was first imaged using an imaging/potential scan routine carried out by the image_scan.vi Labview program detailed in Chapter 4. Briefly, the potential was held at an adsorption value (0.150V/SCE) for 30 seconds during which time an image of the interface was captured using a 10 second acquisition. Immediately after the 30 second wait period, the potential was stepped by -25 mV/SCE for another 30 second wait period and 10 second image acquisition time. This process was repeated until -0.800V/SCE and the potential then incremented positively by +25 mV/SCE steps. The cycle was complete when the layer re-adsorbed at 0.150 V/SCE.  This procedure produced fluorescence images of the interface during slow  'potential scan' of effectively less than 1 mV/sec. In total, 78 images of the interface were acquired, each separated by 30 seconds. The images of this imaging/potential scan analysis are presented in Figure 6-1.  This figure displays (a) selected Al/I images of the interface at the potentials indicated 0  by the letter legend on (b) the capacitance plot. The 'hotspot' mask was not applied to the images presented in Figure 6-la and the dark features present in the fluorescent images are due to the subtraction of the I /I image. While the 'hotspot' mask was not applied to the images presented, ads  0  the use of this mask was necessary during the calculation of the mean histogram intensity (fluorescence intensity) shown in Figure 6-lc.  The negative and positive potential scans are  represented by the open and closed symbols respectively. The values of capacitance measured during the imaging cycle (Figure 6-lb) are characteristic of an adsorbed layer of 3 mol% DiIC18(5)/C180H as previously detailed in Chapter 5, and verifies that the interface optically investigated was a typical C180H coated Au(l 11) surface. The fluorescence image taken at the adsorption potential (image A) is dark and featureless due to metal-mediated quenching of the 103  200 \im  a)  I  -0.8  J  -0.6  -0.4 -0.2 E/VvsSCE  K  0.0  L  0.2  M  -0.8  N  -0.6  O  -0.4 -0.2 E/VvsSCE  Hotspot  0.0  0.2  Figure 6-1 (a) Fluorescence images, (b) capacitance and (c) calculated fluorescence intensity obtained from an image/potential scan investigation of 3 mol% DiIC18(5)/C180H adsorbed on Au(l 11). In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a).  adsorbed surfactant containing dye molecules. This is also observed in the calculated fluorescence intensity at positive potentials (Figure 6-lc). When imaging the interface at progressively more negative potentials, the fluorescence images remain quenched even through the pseudo-capacitive peak potential region (image B). This demonstrates that even though dramatic changes are occurring in the layer (revealed through the increase in capacitance), the organic molecules still remain close enough to the electrode so that fluorescence quenching by the metal dominates. This observation supports the existence of defects in the adsorbed layer within this range of potentials rather than partial desorption. A similar observation was made previously when investigating the interface with elastic light scattering measurements [16]. At roughly -0.550 V/SCE (image C) on the negative potential scan, the fluorescence image remains strongly quenched and only a slight increase in the calculated fluorescence intensity is noted. Only negative of -0.600 V/SCE, where the capacitance begins to show a general increase toward values close to that of the water covered surface, is there a dramatic increase in the fluorescence intensity (Figure 6-lc) and features become apparent in images D to F. This increase in both the fluorescence intensity and capacitance indicate the onset of desorption induced by electric potential. The capacitance suggests that desorption is complete between potentials of -0.700 and -0.800 V/SCE. The fluorescence intensity plateaus within this negative potential region. The images F to H reveal that the molecules are displaced from the electrode surface in a heterogeneous and aggregated morphology that do not diffuse from the electrode surface. This lack of diffusion is most clearly observed when tracking the cluster of desorbed molecules outlined in images H to L, indicating that there is a strong interaction causing the desorbed surfactant to remain in place. Previous studies using impedance spectroscopy on a similar modified interface (lipid coated Hg) have shown that the water molecules that displace the surfactant from the metal surface exist in a rigid 'ice-like' structure at potentials where the surfactant is desorbed [107]. Due to the rigid structure of the ordered water molecules at these negative potentials, it is possible that surfactant remains in place due to an adhesive interaction between the desorbed surfactant and structured water. Continuing the potential increments in the positive direction clearly reveals hysteresis between the desorption and re-adsorption processes. This is shown in both the mean histogram intensity (Figure 6-lc) and in the number of image features. The fluorescence intensity, although slowly decreasing with increasing potential, is significant over the 105  potential range -0.800 to -0.500 V/SCE. Just positive of-0.500 V/SCE, a slight inflection point is observed preceding a more significant decrease in both the calculated fluorescence intensity and in the number of features in the images (image N). This decrease occurs at potentials just negative of potential region where the capacitance becomes slightly lower due to an adsorbed state prior to spreading. The ability to measure fluorescence at this potential supports the proposed existence of a contact adsorbed state preceding re-adsorption of the layer. In the contact adsorbed state, the molecules are far enough away from the metal surface to avoid total fluorescence quenching. Approaching more positive potentials around the pseudo-capacitance peaks (-0.250 to -0.100 V/SCE), the fluorescence intensity is minimal. The adsorbed layer is re-established at the positive potential limit resulting in quenched fluorescence. These results demonstrate that potential controls the process of desorption/re-adsorption for the DiIC18(5)/C180H surfactant layer. Electric potential manipulates the distance separating the surfactant and the metal through variation of the interfacial tension. Desorption is indicated by an increase in capacitance followed by an increase in the measured fluorescence intensity. The desorbed molecules appear organized in clusters and do not diffuse from the vicinity of metal surface within the spatial resolution of this technique (roughly 1 p,m). The desorption/re-adsorption process proceeds via different mechanisms as shown by the hysteresis in the capacitance, fluorescence intensity, and features in the images. On the positive potential scan, incomplete quenching of the dye molecules near the pre-adsorbed potential region supports the existence of a contact adsorbed state of the molecules preceding full re-adsorption.  6.1.3 Motivation for the 'Hotspot' Mask As shown in Figure 6-lc, an unexpected result was noted in the calculated fluorescence intensity during the re-adsorption process. The values of fluorescence intensity during the positive potential scan decrease below zero between -0.200 and 0.150 V/SCE. Fluorescence intensity was calculated by subtracting an I /I image. The negative values of intensity require a decrease in the ads  0  fluorescence intensity of the re-adsorbed layer which is not expected since fluorescence should be quenched by the metal. Since the minimum capacitance is re-established at positive potentials, the layer is believed reform a well-ordered layer after the imaging cycle. The procedure of normalizing 106  the fluorescence intensity corrects for variations in the lamp intensity over the duration of one imaging cycle. Therefore, a decrease in fluorescence intensity suggests that some parts of the adsorbed layer exist far enough away from the electrode surface to be photobleached or photodecomposed. This may be a direct result of the manner in which the layer is deposited onto the electrode surface.  Rather than using a Langmuir trough to deposit a well-ordered and  compressed monolayer from the gas/solution interface to the metal surface, the organic layer was formed at the equilibrium spreading pressure (ESP) in the electrochemical cell and deposited onto the electrode surface. While this method allows for reproducible transfer onto the electrode surface, some small excess reservoirs or crystallites present on the gas/solution interface may also be transferred to the metal. This is supported by the existence of the so-called fluorescence 'hotspots' on the electrode surface. Figure 6-2a shows that these features are highly intense even when adsorbed on the electrode. This figure shows the first adsorption I/I image in the experiment G  depicted in Figure 6-1 (before subtraction of an I /I image). The mask is shown in the center ads  0  column and the masked regions were exaggerated to be effective. The mask is applied to the adsorption I/I image (shown at the right side of the figure) covering the intense features. For D  completeness, the same procedure is shown in Figure 6-2b applying the mask on the desorption Al/I  0  image. The exaggerated features in the mask occupy about 10 % of the image; however, any fluorescing regions not caught by the mask would continually photobleach resulting in the decrease in fluorescence at the positive potential region. This could explain the small negative intensity seen in the calculated fluorescence intensity. The mask is applied to the images during the calculation of fluorescence intensity because these features saturate the CCD even at the adsorption potential. The intensity of these features did not change with a variation in potential so were not considered in the characterization of the potential dependent behaviour of the adsorbed surfactant.  These adsorbed layers transferred from the  gas/solution interface at the ESP have always been considered to be uniform. The measurements of capacitance and PM-FTIRRAS [205], have described monolayer adsorption onto the electrode surface. However, these techniques can only probe the average properties of the organic coated surface. As a result, these small crystallites on the metal surface are not readily observed or recognized as being present in these investigations because they only make up a small portion of the 107  Hotspot mask  Iads/Io  Masked Iads/Io  b)  ••  Hotspot mask  Ides'I-o  • •  Masked Ides^o  Figure 6-2 The masking procedure: The first adsorption I/I image (a) shows intense features on the electrode. The 'hotspot' mask purposely overestimates these features to effectively mask these intense regions. These 'hotspots' are eliminated from the calculation ofthe mean histogram intensity by application ofthe mask to the AI/I image. For completeness, the same example is given for the normalized desorption image (b). 0  0  108  adsorbed surfactant. Therefore, while a small region of the images can be cropped to eliminate the intense features, the masking routine allows for a larger representation of the overall behaviour of the adsorption/desorption process. Thus, the 'hotspot' mask can be considered as an elegant cropping method that removes features which do not appreciably change with electric potential and that are a product of the non- ideal deposition procedure.  6.1.4 Average Number and Mean Area of the Image Features The results from the imaging/potential scan procedure shown in Figure 6-1 are believed to accurately represent the desorption/re-adsorption events because the electrochemical character of the re-adsorbed layer is not disturbed. The number and average area of the features within each image were calculated using the programs histogram_calc.m and count_features.scr described in Chapter 4 and the results of this analysis are shown in Figure 6-3. The features are shown in (a) and the calculated number of aggregates with applied potential are shown in (b) with the average area of the particles in (c). The potential dependent variation in the number of aggregates (Figure 6-3b) is similar to the calculated fluorescence intensity in Figure 6-lc. No features were observed in the images at positive potentials (image A of Figure 6-3a) due to fluorescence quenching from the metal. The number of features increase at potentials negative of -0.600 V/SCE similar to the calculated histogram plot, and reaches a maximum at the desorption limit of -0.800 V/SCE. On the positive potential scan, the number of features remains significant between -0.800 to -0.700 V/SCE and decreases at more positive potentials. Between -0.500 and -0.400 V/SCE, a slight inflection in the number of features is observed consistent with the calculated histogram intensity. The images within this region (L to N) show a general decrease in the number of aggregates indicating that not all of the desorbed molecules re-adsorb at the same potential, again supporting the concept of a potential regime for contact adsorption preceding a fully re-adsorbed layer. Furthermore, the change in the mean area of the aggregates (Figure 6-3c) begins to show a slight increase following contact adsorption at a potential of -0.275 V/SCE. This slight increase in the average area reveals the process by which the contact adsorbed aggregates spread on the electrode to reform the low capacitance layer. As the surfactant spreads, the relative area or diameter of the cluster would increase resulting in the observed increase in the calculated area (Figure 6-3c). However, this result 109  Figure 6-3 (a) Outlined fluorescence features, (b) number of image features and (c) the mean area calculated for the data presented in Figure 6-1. The open and closed symbols represent the negative and positive potential scan directions respectively.  is not conclusive because it is based on the few features that remain visible at this potential (image N). Therefore, any conclusions based on such a small sample size should be considered carefully. A second imaging/potential scan investigation was performed on the same adsorbed layer to determine if the manner in which the layer is desorbed remains the even after a number of scans.  6.1.5  A Second Imaging/Potential  Scan  Investigation  On the same adsorbed layer as shown in Figure 6-1, another image/potential scan analysis was performed on the same region of the electrode with the same acquisition parameters. The results of this analysis are shown in Figure 6-4 and can be directly compared to Figure 6-1 and 6-3. The images presented in Figure 6-4a are almost identical in structure to those in Figure 6-la. This is most easily observed when tracking the group of features outlined in the images H to L between the two data sets. The largest difference between the two investigations is the decrease in fluorescence intensity by approximately 50% for this set of data, seen by comparing the image pallets or the average histogram intensities (Figure 6-4c). Since there is no obvious change in the desorbed structure, this decrease in fluorescence intensity is attributed to photobleaching of the fluorescent molecules on repeated exposure to light. The fluorescence intensity reveals a slight difference at potentials negative of -0.700 V/SCE between the two data sets. On the first imaging cycle (Figure 6-1 c), the intensity plateaus in this potential region; however, for the second imaging cycle (Figure 6-4c) a significant decrease in intensity is observed. Since the first imaging cycle resulted in more intense images of the interface, the decay in fluorescence may not have been as readily observed because small changes would be hidden by the larger overall intensity. The large decrease in fluorescence intensity at the desorption potential for the second imaging scan was not expected. Since the molecules are desorbed at the negative potential limit, a decrease in fluorescence intensity suggests that the desorbed molecules are being progressively photobleached as a result of the length of illumination time. This is complimented by the continual decrease in fluorescence intensity on the return scan. Alternatively, the desorbed molecules may be slowly drifting back to the electrode surface with the time spent at the desorption potential. Although the origin of this effect cannot yet be stated definitively, this concept is dealt with in greater detail in Chapter 7. The decrease in fluorescence intensity for this second set of data makes the image analysis somewhat more difficult; 111  200 pm 1.6  H  G  F  E  E/VvsSCE  D  C  B  A  E/VvsSCE  Results obtained for a second image/potential scan investigation for the same adsorbed layer of 3 mol% DiIC18(5)/C180H showing (a) the fluorescence images, (b) capacitance, (c) number of image features, and (d) their mean area. In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a). Figure 6-4  however, the number and average area of the aggregates is shown in Figure 6-4 (d and e) are fairly consistent with the first imaging cycle shown in Figure 6-3. While some slight deviation in the calculated mean histogram and image analysis can be expected due to irreversible photobleaching, the data sets are similar. Therefore, the adsorption/desorption process is repeatable for a given layer, consistent with the reproducible capacitance values obtained during numerous potential cycles as detailed in Chapter 5. To understand the effect that the type of potential perturbation has on the adsorbed layer, a potential stepping procedure was conducted on the same adsorbed layer, described in the next section.  6.1.6  An Imaging/Potential  Step  Investigation  An image/potential step cycle was conducted on the same layer immediately following the second image/potential scan routine using the programs image_adsorb_neg_cap.vi and image_adsorb_pos_cap.vi described in Chapter 4. In this routine, a potential step to 0 V/SCE is added for 10 seconds between each image. Therefore, the desorbed molecules were only illuminated for 20 seconds compared to 30 seconds for the image/potential scan routines. On the return (positive) potential scan, the organic molecules were desorbed prior to the next image potential. This desorption step is required because of the hysteresis between desorption and adsorption processes. Characterizing the positive steps of potential requires a potential step to -0.800 V/SCE for 5 seconds before the next imaging potential is established. Although this stepping procedure is not the same as in the chronocoulometry measurements, this investigation may help understand why the layers are sometimes disrupted during the potential stepping procedure. However, it should be noted that this imaging procedure involves about twice as many potential steps than does a typical chronocoulometric investigation. The potential pulse perturbation to 0 V/SCE has a dramatic influence on the desorption structure and the re-adsorption process as shown in Figure 6-5. From the increase in capacitance on the positive potential steps (b), the adsorbed layer is clearly perturbed from the potential stepping routine.  This increase in capacitance has also been observed during chronocoulometric  investigations. The fluorescence images (a) for the negative set of potential steps are weak. The 113  H  G  F  E  D  C  B  A  4^  -0.8  -0.6  -0.4 -0.2 E / V vs S C E  0.0  0.2  -0.8  -0.6  -0.4 -0.2 E / V vs S C E  0.0  0.2  Figure 6-5 Results obtained for a image/potential step investigation for the same adsorbed layer of 3 mol% DilC 18(5)/C 180H showing (a) the fluorescence images, (b) capacitance, (c) number of image features and (e) their mean area. In all curves the open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labels in (b) correspond to the labelled images in (a).  potential step perturbation is not creating the same type of aggregates as seen for the sequential scanning experiments. The formation may require a growth process to create the aggregates and allow them to separate from the electrode.  However, at desorption potentials the fluorescence  features in the images can be seen. The structure of the image at -0.800 V/SCE (image H) is significantly different from that obtained in the scanning procedure (image H from Figure 6-1 and 6-4). While the features present in the outlined region are maintained, new aggregates are now observed within the image. This demonstrates that the manner in which the electric variable perturbs the organic layer impacts the formation of desorbed features. This is further manifested in Figure 6-5 in the curves for the fluorescence intensity, number of desorbed aggregates and their mean area. In all the curves, a maximum is obtained at potentials just negative of -0.400 V/SCE on the positive direction of potential steps. Therefore, the layer does not suffer from the same degradation in measured intensity during the positive potential steps. Presumably, the re-adsorption step protects the layer from photobleaching because the molecules in the excited state are quenched when adsorbed onto the metal. Therefore, fluorescence remains more intense on the positive set of potential steps between -0.800 to -0.400 V/SCE because the layer has been photobleached less. Alternatively, the molecules may be further from the electrode surface due to the stepping procedure consistent with the disrupted capacitance measurement. While the absolute values for the number of aggregates and mean area remain similar to those observed in the slow potential scan investigations, the potential dependence are distinctly different. Rather than decreasing at more positive potentials, the number of aggregates remain close to the maximum over a large potential range (-0.800 to -0.400 V/SCE). The average area of the desorbed aggregates (Figure 6-5e) goes through a maximum at potentials just negative of contact adsorption. A rapid decrease in the average area due to contact adsorption is observed at more positive potentials. Interestingly, at the most positive potentials where the molecules are re-adsorbed, the fluorescence intensity is greater than zero. Again this result coupled with the increase in capacitance at the same potentials suggests that the re-adsorbed layer is disrupted as a result of the stepping procedure. Therefore some of the molecules are not re-adsorbed which results in the measured fluorescence at these potentials. As previously mentioned, the increase in fluorescence intensity during the positive potential steps may be due to the decreased exposure of the desorbed species to the excitation light. However, 115  the increase in fluorescence intensity by over a factor of 2 in the calculated mean histogram (compared to the second imaging/potential scan in Figure 6-4c) and the increase in the calculated mean histogram at positive potentials, implies that this explanation must be coupled with other effects. It is more likely that the potential stepping procedure is more intrusive to the adsorbed layer than the slow potential sweep. The increase in capacitance and the calculated mean histogram intensity at positive potentials both point to a more disrupted adsorbed layer as a result of the potential stepping procedure. However, the potential stepping procedure has a different time allowed to create the desorbed aggregates. Therefore, this difference in time may affect the desorbed structure of the molecules and their distance from the electrode.  6.1.7 Summary of the Results for 3 mol% DiIC18(5)/C180H on Au(lll) The desorption/re-adsorption of a 3 mol% DilCT8(5)/CT80H layer from/to a A u ( l l l ) electrode surface was imaged using fluorescence microscopy. At positive potentials, fluorescence intensity was reduced due to metal-mediated quenching. The quenching efficiency became less as the potential was swept in the negative direction. An increase in capacitance at potentials negative of -0.600 V/SCE was followed by an increase in the fluorescence intensity indicating the onset of desorption. The surfactant molecules desorbed from the electrode in an aggregated state and were not observed to diffuse away from the electrode surface. For a given adsorbed layer, the fluorescence results were repeatable provided that slow potential scans were used. The creation of desorbed aggregates were found to be dependent on the method by which they were displaced from the electrode surface. Slow potential sweeps maintained a well-ordered layer upon re-adsorption to the electrode and potential steps introduced irreversible disorganization to the re-adsorbed surfactant. The disorganization may have been due to the more aggressive potential stepping procedure compared to the slow potential sweep investigations. During the potential steps, the time spent at desorption was less than during the slow potential sweeps. This difference in time may result in the formation of less stable desorbed aggregates that did not reform as an ordered layer when readsorbed.  These results demonstrated the capability fluorescence imaging provides for the  investigation of the adsorption/desorption process of C180H provided that DilC 18(5) accurately reported on the behaviour of the adsorbed alcohol. Carbocyanine dye molecules are commonly used  116  as biomembrane probes because they partition into cell walls [18-21]. Thus, it is likely that DilC 18(5) will not segregate to form a domain separate from C180H, but rather exists in the lipidlike C180H matrix faithfully reporting on the behaviour of the surfactant. To verify that this is indeed the case, C180H was mixed with both 2 mol% 5-octadecanoylaminofluorescein and 2 mol% DilC 18(5) and investigated with the imaging/potential scan procedure.  6.2  Imaging a Two Dve/C18QH Laver on A u d l D The investigations of a 2 mol% 5-octadecanoylaminofluorescein and 2 mol%  DiIC18(5)/C180H adsorbed layer involved imaging the under identical conditions for both dyes sequentially. For example, the layer was first examined with an imaging/potential scan routine measuring the fluorescence from DilC 18(5), immediately followed by an image/potential scan routine measuring the fluorescence from 5-octadecanoylaminofluorescein.  The images were  acquired from the same adsorbed layer and the same region of the electrode/surfactant and treated according to the regular AI/I routine. In the interest of clarity and for the ease of comparison, the 0  treated  images  were  pseudo-coloured  red  for  D i l C 18(5)  and  green  for  5-  octadecanoylaminofluorescein and then combined creating images containing green, red or mixed fluorescence.  6.2.1 An Imaging/Potential Scan Investigation The adsorbed layer of C180H containing 2 mol% 5-octadecanoylaminofluorescein and 2 mol% DilC 18(5) was first characterized using an imaging/potential scan routine measuring DilC 18(5) fluorescence followed by the same imaging routine while measuring the fluorescence from 5-octadecanoylaminofluorescein on the same region of the surfactant/electrode region. The resulting combined pseudo-coloured images are shown in Figure 6-6 superimposed onto the capacitance plot. While the capacitance measurement is similar to the single dye investigation, a slight increase in the measured minimum capacitance is observed at the positive potential region for this set of data. This deviation can be expected since the monolayer now contains a total of 4 mol% dye. The electrochemical characterization of C180H containing various amounts of DilC 18(5) in Chapter 5 revealed that above a 3 mol% concentration, the minimum capacitance increases with a 117  N  -1.0  -0.8  -0.6  -0.4  -0.2  0.0  0.2  0.4  E / V vs S C E Figure 6-6 An image/potential scan investigation for an adsorbed C 1 8 0 H layer containing 2 mol% 5octadecanoylaminofluorescein and 2 mol% DilC 18(5). The images superimposed on the capacitance plot were modified as Al/I and then pseudo-coloured to green (fluorescein) and red (DilC 18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot. 0  corresponding decrease in the maximum film pressure. However, in concentrations lower than 3 mol%, the fluorescence intensity of the images would suffer. Since the change in the capacitance plot is slight it is believed that the adsorbed layer containing two dyes is representative of a typical adsorbed layer. The overall character of the images presented in Figure 6-6 are similar to those presented earlier. At positive potentials the fluorescence of both dye molecules are quenched due to the proximity of the adsorbed layer to the metal. Fluorescence remains quenched through the pseudocapacitance peaks similar to the images presented in Figure 6-4a. Slightly positive of -0.600 V/SCE on the negative potential scan, an increase in fluorescence intensity of fluorescein (green) is observed without significant contribution from DilC 18(5) (red) (Figure 6-6 image C). When potential is made progressively more negative, the overall fluorescence intensity of the images increase due to the desorption of the surfactant containing dye, and the images appear heterogeneous and aggregated in structure (images D and E). Nearing the negative potential limit, a significant contribution from DilC 18(5) is observed in images F and G and the structure is similar to that presented in Figure 64a. The overall colour of images F and G in Figure 6-6 result from a mixing of the pseudo-colours indicating that the two dye molecules mix efficiently in the C180H layer and does not segregate into separate regions of the C180H layer. At the negative potential limit (-0.800 V/SCE) the aggregated nature of the desorbed surfactant is clear as reported by both dye molecules. Again, the desorbed molecules do not diffuse away from the electrode surface. As might be expected from the results presented in Figure 6-4a a decrease in fluorescence intensity of DilC 18(5) is observed at the negative potential limit shown in Figure 6-6 (image H). Interestingly, the fluorescence from 5octadecanoylaminofluorescein remains constant over this negative potential range. On the positive potential scan, the molecules are re-adsorbed in a similar fashion as previously observed. Image analysis was performed on the images and the data in Figure 6-7 represents (a) the calculated fluorescence intensity, (b) number of image features and (c) their mean area for each dye used in this investigation. The data obtained for DilC 18(5) are nearly identical to those measured for the single dye/C180H layer presented in Figure 6-4 (c, d and e). This demonstrates that the fluorescence response  of DilC 18(5)  is relatively unaffected  by the presence  of 5-  octadecanoyalaminofluorescein. However, there are some slight differences between the trends for 119  -0.8  -0.6  -0.4 -0.2 E/VvsSCE  0.0  0.2  -0.8  -0.6  -0.4 -0.2 E/VvsSCE  0.0  0.2  Figure 6-7 Image statistics obtained for the two dye/C180H layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively.  120  the two dye molecules in the same layer. Namely, the calculated fluorescence intensity of the fluorescein dye (Figure 6-7a) begins to increase at less negative potentials (-0.400 V/SCE) compared to DilC 18(5) (-0.600 V/SCE). Furthermore, the fluorescence intensity of DilC 18(5) decreases at the negative limit while the intensity for fluorescein reaches a maximum. The increase in fluorescence intensity of fluorescein at less negative potentials compared to DilC 18(5) is most likely due to a difference in the distance-dependent metal-mediated quenching profiles for the two dye  molecules.  This  may  explain  why the  image  aggregates  reported  by 5-  octadecanoylaminofluorescein are less defined than for DilC 18(5). If the fluorescence from the fluorescein dye is less sensitive to the separation distance from the metal, a more general increase in the background fluorescence intensity may obscure the contrast needed to see the surfactant aggregates. Should this be the case, the fluorescein molecules would report the onset of desorption more clearly because the intensity of fluorescence would be greater than DilC 18(5) at the same separation distance from the metal. To verify this possibility, the fluorescence intensity at known separation distances from a Au surface was measured by an ex situ method. A unique template consisting of Au coated with a steps of Si0 was created by Dr. Mario Beaudoin in the Advanced 2  Materials and Process Engineering Laboratories (AMPEL) at the University of British Columbia. On a clean glass slide, an adhesive layer of Cr (100 nm) was deposited using an electron beam evaporator. A 300 nm thick layer of Au was deposited onto the Cr layer with sputter evaporation onto which S i 0 in steps of known thicknesses were deposited using Plasma Enhanced Chemical 2  Vapor Deposition (PECVD).  In total, 4 S i 0 steps were deposited onto the Au layer with 2  thicknesses of 10, 20, 30 and 40 nm as shown schematically in Figure 6-8a. A layer of 2 mol% DiIC18(5)/2 mol% 5-octadecanoylaminofluorescein/C180H was transferred to the stepped surface from a monolayer present at the ESP similar to the electrochemical experiments. A minimum of 5 fluorescence images were acquired at each stepped surface and were treated by divided by an I  0  image of the bare Au surface. These I/I images were then averaged and the mean histogram 0  intensity calculated. The relative fluorescence intensity was found after correcting for the different image exposure times required for each Si0 step containing adsorbed surfactant. The fluorescence 2  intensity as a function of separation distance from the metal is shown in Figure 6-8b. When themolecules are adsorbed onto the Au surface, fluorescence is effectively quenched for both dyes. 121  a)  2dye/CI8()ll 4 x 1 0 n m Glass 3 0 0 n m Au 1 0 0 n m Cr  Glass Substrate  d / nm  Figure 6-8 (a) A schematic representation of the stepped template used to measure the fluorescence intensity of an adsorbed layer of 2 m o l % D i I C 1 8 ( 5 ) / 2 m o l % 5octadecanoylaminofluorescein/C 180H separated from a 300 nm thick Au surface by controlled thicknesses of Si0 . (b) The measured fluorescence intensity for both DilC 18(5) and 5octadecanoylaminofluorescein. The fit from Equation 2-56 for the DilC 18(5) is shown in (c) where ( o ) represents the data and the line represents the fit. 2  122  Upon increasing the separation between the fluorophore and metal, an increase in the fluorescence intensity is noticed for both dyes. As proposed, the fluorescence intensity increases more rapidly with an increase in separation for the fluorescein dye. Therefore, use of the fluorescein dye molecule allows a more accurate assessment of on the onset of organic desorption since the fluorescence intensity increases at smaller separation from the metal in comparison to DilC 18(5). However, DilC 18(5) can report on the morphological changes in the desorbed layer more accurately because of the contrast needed to observe the surfactant aggregates. Further insight into the separation distance between the metal and surfactant at the desorption potentials was possible by fitting the DilC 18(5) data in Figure 6-8b to Equation 2-56 (described in Chapter 2). The data was fit to this equation using the standard deviation of the 5 images per step as a weighting factor and the fit is shown in Figure 6-8c. The experimental data was fit to the equation generating d = 37 nm and ljl = 228. Ideally an experimental measure of 1J1 would be 0  0  0  preferred over a fitted value but this was not possible due to limitations in obtaining a thick enough step of Si0 . This fitting routine was also applied to the data for the 5-octadecanoylaminofluorescein 2  molecule but the results were not reliable. However, the fitting parameters for the DilC 18(5) data were reasonable and may give insight into the distance that these molecules desorb from the metal. This estimation must assume that the desorbed molecules in the electrochemical characterization are similar to those on the stepped S i 0 surface. In the experimental investigations, the desorbed 2  molecules are separated from the electrode surface by water molecules. Although the Si0 /surfactant 2  interface is not the same as the Au(l 1 l)/surfactant interface, separations using S i 0 are reasonable 2  since the refractive index values for water and Si0 are similar. An 1„ image in the electrochemical 2  setup can be obtained by measuring an image of the floating monolayer. However, flat field correcting this image with an I image was not possible due to the inability to focus on the 0  gas/solution interface in the absence of fluorophore molecules. In general, the floating monolayer was an order of magnitude greater than the most intense desorption image obtained in an electrofluorescence investigation. Since the value of d describes the Forster critical distance where 50% 0  of the molecules are quenched, the larger intensity of the floating monolayer compared to the desorbed layer suggests that the desorbed molecules are closer to the electrode than the d value 0  (37nm) obtained from the fitting parameter. However, this is only an estimation and further  123  investigations are required before this can be confirmed. Although the fluorescence from the fluorescein molecules could not be accurately fit to the general equation, the raw data (Figure 6-8b) does confirm that the fluorescence intensity for this molecule increases more rapidly with an increase in separation compared to DilC 18(5). This explains the observed increase in fluorescence intensity of the fluorescein compound at less negative potentials. The other significant difference between the potential dependence of fluorescence intensity for fluorescein in comparison to DilC 18(5) is the lack of a fluorescence decay at the negative potential limit for the fluorescein compound (Figure 6-7a). Negative of -0.700 V/SCE, the fluorescein fluorescence intensity continues to increase in contrast to the decrease in fluorescence intensity for DilC 18(5) at the same potentials. This is a clear indication that the reduction in fluorescence intensity is due to a fluorescence decay process affecting only DilC 18(5) molecules and not because of a drift of the desorbed surfactant molecules back to the electrode surface with time since this would affect both molecules similarly. On the positive potential scan, hysteresis is observed in the fluorescence intensity plot for fluorescein/C180H (Figure 6-7a) and an inflection in the curve is noted just negative of -0.400 V/SCE preceding contact adsorption consistent with DilC 18(5). At these potentials, the number of aggregates show a rapid decrease due to the contact adsorption and spreading of the surfactant.  6.2.2  Summary  of the Results for  the Two Dye/C180H  Layer on  Au(lll)  The two fluorescent dye molecules reported on the desorption/re-adsorption process of the same C180H layer similarly. Observations consistent between the two sets of data included quenched fluorescence and featureless images at adsorption potentials.  Both dye molecules  remained quenched through the pseudo-capacitance peak potential region supporting the concept of defects in the adsorbed layer. At more negative potentials, the fluorescence increased due to the onset of desorption and images in both data sets revealed a heterogeneous and aggregated structure of the desorbed molecules. Some deviations were noticed due the difference in metal-mediated fluorescence quenching between the two dye molecules. This was revealed as an increase in fluorescence intensity at less negative potentials as compared to DilC 18(5). A fluorescence increase 124  was observed at -0.400 V/SCE from the fluorescein dye in comparison to -0.600 V/SCE for the DilC 18(5). On the return scan hysteresis was observed as reported by both dye molecules showing the different mechanisms for the desorption/re-adsorption processes. Some slight but important differences were noted when comparing the fluorescence response of the two dyes. Namely, the fluorescence intensity for the fluorescein dye remained large at the most extreme negative potentials in comparison to the decrease of fluorescence intensity for the DilC 18(5) dye at the same potential limit. Since the two molecules mixed well into the C180H layer, this deviation points to a fluorescence decay of the DilC 18(5) molecules, and not from a slow drift of the desorbed layer back to the electrode surface with time. These results demonstrated that two dye molecules can be mixed with the surfactant without significant disruption to the character of the adsorbed layer. Furthermore, the use of these two dyes allow extraction of more details on the desorption/re-adsorption process since the metal-mediated fluorescence quenching efficiency is different for the two dyes. These two dyes were mixed with OLA for the in situ characterization of the more fluid monolayer described next.  6.3  Imaging a Two Dye/OLA Layer on A u d l l ) In Chapter 5, the electrochemical characteristics of adsorbed OLA were shown to be slightly  different than those of CI 80H, presumably due to the fluid nature of the OLA adsorbed layer. A smaller minimum capacitance was observed as well as a negative shift in potentials for the onset of desorption. These results indicated that the OLA formed a more well-ordered adsorbed layer on the electrode.  Imaging the effect of potential on this interface will show the differences in the  adsorption/desorption process for the two alcohols. Electro-fluorescence investigations were conducted on a layer of O L A containing both 2 mol% DilC 18(5)  and 2 mol% 5-  octadecanoylaminofluorescein.  6.3.1  An Imaging/Potential  Scan  Investigation  The 2 mol% DiIC18(5)/2 mol% 5-octadecanoylaminofluorescein/OLA surfactant mixture was analyzed with an imaging/potential scan routine using the same parameters as for the imaging of C180H. The acquired data is shown in Figure 6-9. The measured capacitance is almost identical 125  to  -1.0  -0.8  -0.6  -0.4 -0.2 E / V v s SCE  0.0  0.2  0.4  Figure 6-9 An image/potential scan investigation for an adsorbed O L A layer containing 2 mol% 5-octadecanoylaminofluorescein and 2 mol% D i l C 18(5). The images superimposed on the capacitance plot were modified as AI/I and then pseudo-coloured to green (fluorescein) and red (DiIC18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot. 0  to that obtained in the absence of the fluorescent dye molecules (Chapter 5), except for the small increase in the minimum capacitance measured at positive potentials. In agreement with the results for the two dye/C180H investigation (Figure 6-6), the fluorescence from both dyes are quenched at positive (adsorption) potentials. As potential is scanned negatively, features begin to appear in the images which contain fluorescence from both dye molecules (images F and G). As determined from the electrochemical characterization in Chapter 5, the potentials for desorption are slightly shifted in the negative direction. This is also observed in the imaging investigation by a featureless image at -0.650 V/SCE (image E) in comparison to the C180H investigation (image E in Figure 6-6). Continuing the potential scan in the negative direction, the capacitance for OLA (Figure 6-9) continues to increase to values near that for the water covered interface as the molecules desorb. A corresponding increase in the fluorescence intensity is observed in image H with a noticeable decrease in the fluorescence intensity from DilC 18(5). The desorbed images contain features that are again aggregated in structure and do not diffuse from the metal surface, however, at the negative potential limit distinct intense features unique to OLA are observed. These intense features were not noticed for fluorescein/CI80H and suggest a slightly different manner in which the desorbed molecules are created from the fluid OLA matrix. Initiating the positive potential scan results in a slight increase in the measured fluorescein intensity. This may be due to a slight reorganization of the desorbed OLA molecules. The re-adsorption of the molecules during the positive potential scan proceeds in a similar fashion to the C180H studies revealing the typical hysteresis (images I to M) and the decrease in fluorescence intensity of both dye molecules when re-adsorbed at the positive potential limit. The calculated fluorescence intensity, number of image features and their mean area are presented in Figure 6-10. The general characteristics of these plots are consistent with the two dye/C180H investigation shown in Figure 6-7. While the measured fluorescence intensity of fluorescein/OLA is consistent with fluorescein/C180H, a noticeable difference is observed when comparing the fluorescence intensity of DiIC18(5)/OLA and DiIC18(5)/C180H. Of particular interest is the more drastic decrease in the fluorescence intensity of DilC 18(5) in OLA at negative potentials (Figure 6-10a). This suggests that the process by which the DilC 18(5) fluorescence decreases occurs more efficiently in OLA compared to C180H. This is also manifested by the 127  a)  1.5 n  1.5  o DilC18(5) 1.0  * Fuorescein  A  1.0 x  X  s  °0.5  °0.5 <  s  <  0.0 0.0  b)  600  $ 650000  g 500  S 400  g> 4 0 0 CD CT300 cn  ra300 O)  <  200  <  200  *  100 0  «  100  c)  0  25  2 5 -,  E  " E 20  20  ro 15  ro 15  < 10 ro  < 10  H  -i  c  ro a  5  5  —i—i—I—i—i—i—I—i—i—r~  0  -0.8  -0.6 -0.4 - 0 . 2 0.0 0.2  E / V v s SCE  0  -0.8  -0.6 -0.4 -0.2  0 . 0 0.2  E / V v s SCE  Figure 6-10 Image statistics obtained for the two dye/OLA layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively.  128  decrease in the number of features at negative potentials presented in (b). Interestingly, while the curve for the fluorescence intensity of DiIC18(5)/C180H continually decreases on the return potential scan (Figure 6-7a), a significant increase or recovery of fluorescence is observed from -0.800 to -0.750 V/SCE for DilC 18(5) in OLA (Figure 6-10a). This fluorescence recovery occurs without changes in the structure of the desorbed OLA aggregates (images G to J of Figure 6-9). Positive of -0.700 V/SCE, the calculated fluorescence intensity of DiIC18(5)/OLA decreases resulting in featureless images of the interface. At more positive potentials, fluorescence is fully quenched due to the re-adsorption of DiIC18(5)/OLA. The calculated number of features in the images (Figure 6-10b), is consistent with the negative going curve for DiIC18(5)/C180H (Figure 6-7b) until -0.750 V/SCE where the rapid decrease in fluorescence intensity occurs. Since the fluorescence is minimal at the negative potential limit, the features could not readily be detected by the image analysis procedure. The calculated mean area of the features (Figure 6-10c) at the negative limit do show a slight decrease; however, this is based on very few particles and as such, any conclusions must be tempered. For the positive potential scan, the number of features increases and only a small region of hysteresis is noted (-0.775 to -0.575 V/SCE). Overall, the general characteristics for OLA and C180H are fairly consistent until the negative potential limit. When desorbed, the DilC 18(5) dye molecules in OLA undergo a rapid decrease in fluorescence intensity. The process by which this occurs will be detailed in Chapter 7. Although the fluorescence is recovered at the initiation of the re-adsorption scan, the decrease in fluorescence intensity clouds analysis of the re-adsorption process.  6.3.2 Summary of the Results for the Two Dye/OLA Layer on  Au(lll)  The results obtained for the analysis of the two fluorescent dyes mixed with OLA differed from those obtained for the same two dyes in CI 80H presumably due to the fluidity of the adsorbed OLA layer. While the adsorption/desorption process was again observed to occur through the creation of desorbed aggregates, some interesting differences between the two surfactants were noted. Namely, the negative shift in potentials for desorption of OLA was confirmed with fluorescence images and intensity, similar to that described in Chapter 5. Other differences between C180H and OLA measured with the fluorescence method included a more rapid decrease in 129  fluorescence intensity for DilC 18(5) in OLA as well as a more clear observation of surfactant aggregates at the desorption potentials reported by fluorescein. Both results were a consequence of the fluid nature of OLA which must influence how the desorbed aggregates were formed. The rate by which DilC 18(5) undergoes a decrease in fluorescence intensity is studied in greater detail in Chapter 7. The electro-fluorescence technique has been successfully used to monitor the desorption/readsorption events of C180H and OLA from/to a A u ( l l l ) electrode using two fluorescent dye molecules. The electro-fluorescence investigation of mixed alcohol adsorbed layers is described next.  6.4  Imaging a Two Dve/25mol%OLA/C18QH Laver on A u ( l l l ) In Chapter 5, the electrochemical characterization of mixed OLA/C180H layers revealed  almost ideal mixing. Of these investigations the 25 mol% OLA/C180H was the most reproducible. While there is little difference in the desorbed structure between C180H and OLA, the electrofluorescence technique may be capable of monitoring the mixed response of the two alcohols since the measured fluorescence intensity of DilC 18(5) seems to be dependent on the fluidity of the surfactant matrix. An image/potential scan investigation of adsorbed 25 mol% OLA/C 180H mixed with the two fluorescent dye molecules is described in the next section.  6.4.1 An Imaging/Potential Scan Investigation The results of an imaging/potential scan investigation for an adsorbed layer of 2 mol% DilC 18(5)/2 mol% 5-octadecanoylaminofluorescein/25mol% OLA/C180H is shown in Figure 6-11. The capacitance has similar character to the 25 mol% OLA/C 180H mixtures measured in Chapter 5. As seen from the images, no clear distinction can be made between the C180H and OLA components of the mixed layer. This is a reasonable result since the alcohols did show nearly complete miscible character from the electrochemical characterization in Chapter 5. Furthermore, since the desorbed features of pure OLA and pure C180H in the previous fluorescence investigations were similar, the ability to measure the response of one component preferentially may be dependent on the dye used and its ability to segregate into one phase or the other. 130  I  -1.0  J  -0.8  K  -0.6  L  M  -0.4 -0.2 E / V v s SCE  N  O  0.0  0.2  0.4  Figure 6-11 A n image/potential scan investigation for an adsorbed 25 mol% O L A / C 180H layer containing 2 mol% 5octadecanoylaminofluorescein and 2 mol% D i l C 18(5). The images superimposed on the capacitance plot were modified as A l / I and then pseudo-coloured to green (fluorescein) and red (DilC 18(5). The images were then combined to create a combination of colours containing red, green or mixed pseudo-fluorescence. The open and closed symbols represent the negative and positive potential scan directions respectively. Capacitance was measured as detailed in the experimental section. The labelled images correspond to the potentials indicated on the capacitance plot. 0  The calculated fluorescence intensity of DilC 18(5) in the mixed layer is shown in Figure 612a. The more rapid decay and slight recovery at the negative potential limit characteristic of DilC 18(5) in OLA is observed. Presumably, the incorporation of OLA in C180H is creating a more fluid phase of the entire mixed monolayer. Because the potentials where the rapid decay occurs are slightly shifted toward the positive direction (more typical of C180H), the data suggests that OLA is mixing with the C180H component of the layer. The curve for fluorescence intensity of fluorescein for the negative scan in potential (Figure 6-12a open symbols) is almost completely dominated by C180H. Interestingly, the re-adsorption scan (closed symbols) shows a rapid decrease in fluorescence intensity which was not observed in the previous investigations. It was noted in Figure 6-10a that the intensity of fluorescein in OLA slightly increased in the measured intensity at the initiation of the return potential scan. This was attributed to a slight reorganization of the desorbed OLA molecules. Since the fluorescence intensity measured in the positive potential scan (Figure 6-12a) constantly decreases, the OLA reorganization may be impeded due to interference by the C180H matrix. This is consistent with the lack of intense fluorescence features in the desorbed images that were observed in the previous fluorescein/OLA investigation (Figure 6-9). The number of features and their mean area reported by both dyes (Figure 6-12b and c) do not reveal any dominance of one component in comparison to the previous investigations of pure OLA and C180H.  6.4.2 Summary of the Results for the Two Dye/25 mol% OLA/C180H on Layer Au(lll) Electrochemical characterization of mixed monolayers of OLA/C180H was conducted in Chapter 5 and revealed mostly miscible character. Since the desorbed structures of OLA and C180H are very similar, the electro-fluorescence method could not unambiguously separate the response of one alcohol from the other. The only significant result showing a mix of both C180H and OLA was in the measured fluorescence intensity of the DilC 18(5) fluorescent dye. This curve revealed the more efficient decrease in fluorescence intensity at the negative limit (characteristic of OLA) that was shifted slightly in the positive direction (characteristic of C180H). While surfactants that show complete immiscible character or have significant differences in the structure of desorbed images would be less complicated to characterize with the developed technique, the results of the analysis 132  E/VvsSCE E/VvsSCE Figure 6-12 Image statistics obtained for the two dye/OLA layer showing (a) the calculated fluorescence intensity, (b) number of image features, and (c) their mean area. The left and right columns represent the data for DilC 18(5) and fluorescein. The open and closed symbols correspond to the negative and positive scan directions respectively.  133  do not show any appreciable separate response from the two alcohols indicating that the molecules are well mixed.  Although some combined response of the two alcohols was observed in  fluorescence intensity plot of DilC 18(5) in the mixed monolayer, this is attributed to the difference in the fluorescence response between the two dye molecules, demonstrating the importance of choosing an appropriate fluorophore molecule in these investigations.  6.5  The Influence of Surface Irregularities on the Desorbed Structure The developed technique of electro-fluorescence microscopy has described the desorption/re-  adsorption process of single component surfactants. A l l of the data reported so far has shown the fluorescence response from the dye molecules at the interface. Using brightfield illumination, the desorbed structure could be compared to defects on the electrode surface to verify if surface irregularities are related to the desorbed structure. Before the layers were exposed to potential cycling or the imaging procedure, a brightfield image of the electrode surface was acquired immediately after the electrode was touched to the electrolyte surface, allowing for the correlation of defects on the electrode surface. As an example, the experimental data from Figure 6-1 will be used. Figure 6-13a is the brightfield illuminated image of the electrode surface with an adsorbed octadecanol layer before the fluorescence image measurement. Some defects such as pits due to mechanical and electrochemical polishing of the electrode surface can be seen. These features are enhanced and a mask is created (b) with the defects shown in black. After creation of defect outlines, the application of this mask to the fluorescence desorption image demonstrates no correlation of these defects with the resulting desorbed layer. If the defect sites were acting as regions where desorption would be initiated, then the largest intensity in the images should be clustered somewhat near these regions which is not observed. The alternative scenario of the defect sites being the last regions of the electrode where desorption may occur would tend to result in the lowest intensity regions clustered around these defects, an effect again not supported by the images. Therefore defects on the electrode are not playing a significant role in the morphological evolution of the desorbed layer. This observation was consistent with the characterization of the dyes containing two dye molecules for both C180H and OLA.  134  Figure 6-13 The influence of surface irregularities on the desorbed structure of C180H. Image (a) is a brightfield illuminated electrode with adsorbed surfactant before potential cycling. In (b) the pits from electrochemical and mechanical polishing are captured. Image (c) shows the defect mask applied to the desorption Al/I image of Figure 6-1. 0  135  6.6  General Observations and Conclusions The  developed  technique  of  electro-fluorescence  microscopy  monitored  the  adsorption/desorption process of surfactants adsorbed onto a A u ( l l l ) electrode surface. This chapter demonstrated the potential controlled fluorescence response of various dye containing adsorbed layers. Initially, a 3 mol% DiIC18(5)/C180H layer was investigated. Results from this analysis demonstrated that the desorbed C180H molecules existed in a heterogeneous structure composed of many small features which did not diffuse away from the electrode surface. An increase in the measured capacitance was followed by an increase in fluorescence intensity reaching a plateau at the desorption limit. On the positive scan of potentials, the fluorescence intensity and number of features within the image remained large until a more dramatic decrease was observed at potentials just negative of the potential region for contact adsorption. This result strengthened the hypothesis that a contact adsorbed state of the layer is established preceding full re-adsorption. A second potential sweep investigation was conducted on this same adsorbed layer and on the same region of the electrode surface. Results proved that the manner by which the desorbed aggregates form and re-spread onto the electrode surface was repeatable as long as slow potential scans were used. Large potential step perturbations demonstrated that the desorbed aggregates were formed via a different mechanism than slow potential sweeps. This was manifested by the change in structure and morphology of the desorbed species resulting in a disrupted re-adsorbed layer. The ability of the DilC 18(5) fluorescent dye to report on the potential induced behaviour of C180H was tested by mixing the surfactant with DilC 18(5) and 5-octadecanoylaminofluorescein and separately imaging the fluorescence of each dye molecule from the same region of the electrode/surfactant layer. Results showed that the desorbed structure of the C180H molecules were similar whether they were reported by DilC 18(5) or 5-octadecanoylaminofluorescein. By using a mixture of dye molecules, different fluorescent properties between the dye molecules was observed. The fluorescein molecule were found to have less efficient metal-mediated quenching and therefore reported on the surfactant molecules at less negative potentials than DilC 18(5). Furthermore, the DiIC18(5) molecule were observed to undergo a decrease in fluorescence intensity at the negative potential limit, whereas the fluorescein compound did not. This was attributed to a fluorescence property unique to the DilC 18(5) molecules. This fluorescence decay occurred more efficiently in  136  OLA presumably due to the fluid nature of the matrix. This fluorescence decay will be put to a kinetic model described in Chapter 7. The desorbed structure for pure C180H and OLA layers were found to be very similar. The main difference was in the calculated mean histogram intensity at the negative potential limit. In mixed proportions, the electro-fluorescence technique could not unambiguously separate one alcohol component from the other because of the miscible character of the layer. Some mixing character was observed for the fluorescence intensity versus potential plot for DilC 18(5) indicating that OLA forms a more fluid layer to the overall mixed surfactant.. In general, this method for monitoring surfactants containing dye molecules is powerful and has extended the description of these unique systems. The method of characterization is not restricted to planar electrode surfaces as will be shown in later chapters. The unique response of DilC 18(5) at the negative limit of potentials is intriguing, and is detailed in the next chapter.  137  Chapter 7 7  Fluorescence Recovery by Electrosorption This chapter describes the potential dependent photophysics of the fluorescent dye molecules  DilC 18(5) and 5-octadecanoylaminofluoresce when adsorbed onto a metal surface. A 3 mol% mixture of DilC 18(5) in C180H was used for the initial characterization of a unique fluorescence decay and recovery event. At desorption potentials, intense fluorescence from DilC 18(5) was observed to undergo a strong decay with the time spent at desorption. The fluorescence decay was independent of the time exposed to the excitation light and a recovery of nearly 85% was observed after a subsequent re-adsorption/desorption cycle. A kinetic expression describing two parallel routes for this fluorescence decay was developed and fit to the experimental data. Thefirstpathway is described as the irreversible photodestruction of the dye molecules and the second pathway as the reversible fluorophore aggregation resulting in a decrease in monomer fluorescence. The mixture of 3 mol% of DilC 18(5) in OLA was also characterized by the same method and the decrease in fluorescence intensity was more rapid as a result of the fluid nature of the OLA matrix. This potential controlled recovery of fluorescence was not observed for 5-octadecanoylaminofluorescein supporting the aggregation hypothesis since the DilC 18(5) carbocyanine dye is known to aggregate in solution [76].  7.1  Fluorescence Decay and Recovery of 3 mol% DiIC18(5) in C18QH  7.1.1  The Collection  and Treatment  of  Images  Imaging was conducted at adsorption (E  ads  =  0 V/SCE) and desorption (E  des  =  -0.800 V/SCE)  potentials where the fluorescent dye molecules are either in contact with (adsorbed) or separated from the metal surface. A fluorescence decay experiment was accomplished using the potential_holding.vi  Labview program described in Chapter 4. The procedure involved holding the  potential for a total of 5 minutes at E  ads  at which time fluorescence images of the electrode surface  with adsorbed surfactant were acquired at 1 minute intervals. The potential was then pulsed to  E  des  for 20 minutes where a series of fluorescence images were captured at 1 minute intervals. This routine of 5 images at E  ads  followed by the 20 images at E  des  138  will be referred to as 'one decay'. If the  capacitance maintained values characteristic of an undisturbed adsorbed layer during the imaging, 5 to 6 decays were attempted sequentially on one adsorbed layer from the same region of the electrode with the same acquisition parameters. In some cases, contact between the electrode and solution was lost due to electrolyte evaporation and the number of decays recorded was hindered. The average histogram (fluoresecnce intensity) was calculated from the Al/I images similar 0  to the histogram_calc. m program described in Chapter 4, except that thefind_particles.m subroutine was replaced with a fitting routine calc_decays.ni described in Appendix A2. The fitting routine, calc_decays.m used the calculated average histogram values from the AI/I images which are then D  fit to a model described at the end of this chapter. The 'hotspot' mask was applied to the AI/I during 0  the calculation of the mean histogram intensity ensuring that the analysis only involved regions that show a variation in fluorescence intensity with applied potential.  7.1.2  Fluorescence Decay and Recovery Under Constant Illumination A series of fluorescence images were collected for six consecutive decays (i.e. six repetitions  ofthe decay collection routine) from a 3 mol% DiIC18(5)/C180H layer adsorbed on Au(l 11). The results of this analysis are shown in Figure 7-1. The measured capacitance versus time (a) display the values typical for the adsorption and desorption of 3 mol% DiIC18(5)/C180H in comparison to Chapters 5 and 6. The calculated fluorescence intensity (AI/I ) versus time is shown in (b). The first 0  five fluorescence intensity data points were zero due to the reduction in fluorescence lifetime as a result of the adsorption of surfactant (see the capacitance values). When holding the potential at E , des  a strong decay in fluorescence intensity occurred over 20 minutes. The surfactant was re-adsorbed after 20 minutes thereby quenching the fluorescence and decreasing capacitance. In the subsequent decay cycle, the first E  des  image resulted in a surprising 92% recovery of fluorescence intensity.  Subsequent decays also revealed this potential controlled process of decay and recovery demonstrating the repeatable nature of the event. In all cases, the recovery was not 100%, however a large portion of the intensity was always recovered after the re-adsorption/desorption cycle. The images for this process in the first decay are shown in (c). Image A (first adsorption potential image) showed no fluorescent features due to metal mediated quenching. The first E  des  image in the first  decay (B) was aggregated and highly fluorescent consistent with the desorption structure shown 139  a)  b)  25.0 g 20.0 ° 15.0 ^10.0 5.0 o 0.0  „  0 i 8 6 4 2H 0 8 6 4 2 0 2  ~\— —i—'—i— —i— —i— —i— —i— —i—'—i— —i— —i— —i—•—i—<—i— —i— —i— —i  -10  1  0  1  1  1  1  1  1  1  1  1  1  1  10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 Time (min)  200 urn  c)  B4  B5  C  Mask  D  Figure 7-1 (a) Measured capacitance and (b) the calculated fluorescence intensity of (c) selected images in a fluorescence decay and recovery investigation of 3 mol% DilC 18(5)/C 180H. The small intense feature in the images is due to a small shift in the electrode during measurements allowing some intense features to be observed after subtraction of an I /I image. The mask used in the calculation of fluorescence intensity was exaggerated to cover this feature. The labelled images correspond to the labels in (b) ads  140  0  in Chapter 6. Within one minute at desorption, the fluorescence intensity of the image had decreased to almost 50% from the original value yet did not change in structure or morphology (Bl). The decay in the image intensity continued with time resulting in an almost featureless image after 20 minutes (B5). When re-adsorbed at the beginning of the second decay (C), fluorescence was again quenched except for the intense features that arise due to a small shift in the electrode. These features were caught by the exaggerated feature mask used in the analysis (also shown in the figure). The recovery of fluorescence upon desorbing the surfactant in the second decay is shown in image D. When comparing the desorption images B and D, the recovery of fluorescence clearly occurs without changes in the structure or morphology of the desorbed surfactant. This observation is consistent with all the desorption images throughout the 6 decays (images not shown). This series of decay and recovery clearly illustrate the link between fluorescence recovery and the changes in potential which effectively manipulates the separation between the electrode and the desorbed surfactant.  It is also clear that the process is repeatable for one adsorbed layer.  Separate investigations on different adsorbed layers of DilC 18(5) were conducted to understand the overall reproducibility of this unique recovery event. The same analysis on a separate layer is shown in Figure 7-2 where only five decays were measured. Immediately apparent is the similarity between the image features, decay and recoveries in comparison to Figure 7-1 although the fluorescence intensity of the desorption images are larger in Figure 7-2c. From the consistent results for these two separate layers, the fluorescence decay and recovery event is confirmed to be repeatable process controlled by electric potential which is capable of manipulating the separation distance between the surfactant and metal. This fluorescence recovery was reproducibly achieved, and although the magnitude of the fluorescence was variable between data sets, the degree of the fluorescence recovery was always greater than 80%. Since the recovery due to electrosorption is not 100% fluorescence will continually decrease to an undetectable level with repeated investigations on the same layer.  7.1.3  Implications  and Possibilities  for the Decay and  Recovery  The observed recovery of fluorescence intensity with changes in potential, or more strictly changes in separation of the dye from the metal, was unexpected. 141  A number of possible  -0.2  -I—'—i—•—•—>—i—'—•—>—i—'— — —i—<—'—•—i—'—•— —i— — — —i—'—<— 1  0  20  1  40  1  60  80  1  1  100  1  1  120  Time (min)  A  B  Bl  B2  B3  Figure 7-2 (a) Capacitance and (b) calculated fluorescence intensity of (c) selected images in a fluorescence decay and recovery investigation of a separate 3 mol% DiIC18(5)/C180H. The data is similar to Figure 7-1. The images in (c) are labelled corresponding to the labels in (b)  142  explanations for this phenomena are explored and discussed in the following sections. The observation of fluorescence recovery is not new, however it is usually associated with Fluorescence Recovery After Photobleaching (FRAP) investigations [206]. In FRAP, a small region containing fluorophore is exposed to an intense pulse of radiation effectively photobleaching the dye molecules. This creates a small non-fluorescent section of the layer (micrometer sized). The recovery of fluorescence in this region is governed by the lateral diffusion of surrounding unbleached dye molecules into the bleached region. Therefore, it is not the photobleached dye molecules that recover the ability to fluoresce but rather a process of incoming 'new' or un-photobleached molecules into the bleached region. The other possibilities that may explain the potential dependent fluorescence recovery include triplet state trapping or exiplex formation due to extended illumination times. Lengthy exposure to the excitation light will continually promote ground-state fluorophores to an excited state where a triplet state, by way of intersystem crossing may be populated. The long lived triplet state may result in phosphorescence resulting in the observed fluorescence decay under constant illumination because these wavelengths are beyond range defined by the filter set used of the microscope. The observed recovery upon re-adsorption may be due to these triplet states relaxing to the ground state by way of energy transfer into the metal. Also, with extended illumination times, the high-energy excitedstate molecule may complex with a neighboring ground-state molecule to form an exciplex. The exciplex may be non-fluorescent, or fluoresce at strongly red-shifted frequencies that are beyond the filter set of the microscope. Re-adsorption of the surfactant may disrupt the exciplex, yielding monomers able to fluorescence when re-desorbed from the metal. The presence of a metal surface may provide other alternative explanations for the observed recovery. According to recent literature, for certain separations between the surfactant and the metal, a significant enhancement of fluorescence maybe observed [113,115,116,207]. This enhancement region may occur for separations on the order of nanometers. It is possible that upon desorption, the surfactant present in this region of enhancement drifts out of this special region with time reducing the measured fluorescence. Even if this enhancement region is not present, it is still possible that the desorbed molecules are slowly drifting toward the metal surface with desorption time, becoming progressively quenched. Unfortunately, capacitance cannot detect a slow drift of these molecules  143  toward the electrode because it only senses what is on the metal surface and not nearby. In the following sections, these possibilities will be dealt with in turn through various experiments and known facts about the system.  7.1.3.1 Fluorescence Recovery After Phobleaching (FRAP) Although the electro-fluorescence set-up does not involve an intense pulse of radiation, photobleaching may still occur due to the lengthy illumination time of 20 minutes. The images collected in Figure 7-1 and Figure 7-2 however, clearly illustrate the lack of diffusion or change in structure/morphology of the desorbed surfactant layer with desorption time. This lack of diffusion is consistent even after a number of adsorption/desorption cycles. Therefore, the diffusion of fluorophore molecules from the surrounding region into the bleached region is unlikely since the diffusion lengths needed would be quite large (> 100 Lim). Another explanation for the fluorescence recovery for this system is required.  7.1.3.2 Triplet State or Exciplex Formation In the case where triplet state or exciplex formation could account for the observed decay and recovery, a variation in the illumination time would effect the decay curves. From this perspective, if a triplet state were to be populated, a decrease in the illumination time experienced by the desorbed surfactant would allow the triplet state to relax via a phosphorescent photon within milliseconds [208-210]. Therefore, a decrease in the illumination time should drastically change the decay rates. Furthermore, the concentration of an exciplex would also decrease because the fluorescent molecule would not exist in the excited state for long periods of time. This would again change the decay rates for light on versus light off. These possibilities were tested by blocking the excitation light for various lengths of time during the monitoring of the fluorescence decay. These investigations of variable light exposure were conducted without changes in the electrical variable duty cycle. Figure 7-3 shows the results five decays for which the excitation radiation was manually blocked at various intervals over the 20 minutes spent at the desorption potential. All images collected at E  ads  were acquired during  uninterrupted.illumination. The first decay was performed with the exposure to excitation light 144  A  B  Bl  B2  B3  Figure 7-3 A fluorescence decay and recovery investigation of a 3 mol% DiIC18(5)/C180H layer during variable exposure times to the excitation source. The first, second and third decays shown in (b) have light on:off for lmin: lmin, lmin:2min and lmin:5min respectively without changes in the duty cycle of the potential perturbation shown in (a). The images in (a) correspond to the labels in(b). 145  alternating between on and off at 1 minute intervals (1 minute on/1 minute off), resulting in 10 fluorescence images over 20 minutes at E . From the capacitance values in (a), it is clear that the des  monolayer was desorbed for the typical 20 minutes over this period and only the duty cycle of the light source was changed. The second decay was acquired with the light on for the full 20 minutes revealing the typical fluorescence decay. The third decay was acquired with the light alternating between 1 minute on and 2 minutes off, yielding seven fluorescence images during the 20 minutes spent at E , Another typical decay with uninterrupted illumination was acquired for the fifth decay des  cycle, and the last decay was acquired with the light alternating between 1 minute on and 5 minutes off yielding 4 fluorescence images during the 20 minutes spent at E . Clearly, the decays for des  uninterrupted and intermittent light levels do not deviate significantly in shape. This is more clearly observed when the decay trends are normalized to the intensity of the first desorption image within the decay cycle as shown in Figure 7-4. The length of exposure to the excitation light is not an important factor in both the decay and recovery process. From this analysis, it can be stated that triplet state or exiplex formation is not causing the observed decay and recovery. This indicates that there is a fluorescence decay pathway that exists which is independent of the exposure to light. However, even though the exposure to light was intermittently changed, a lack of 100% recovery was still observed. It becomes evident that there also exists an irreversible loss of fluorescence. This indicates that there are at least two possible processes that describe the fluorescence decay and recovery events.  7.1.3.3 Radiative Enhancement and Diffusion to the Metal Although it is possible that the dye molecules are desorbing from the electrode into a special window where radiative enhancement would occur, this possibility is unlikely in this particular system because of the nature of the metal substrate. Much of the literature on the subject of fluorescence enhancement due to the metal surface is closely related to a surface enhancement effect similar to that in Surface Enhanced Ramman Spectroscopy (SERS). Here, the enhancement is most efficient due to the coupling of light with surface plasmons on very rough surfaces [211] of metals such as Ag, Au, and Cu. The description of fluorescence enhancement at a metal is commonly described for silver island thin films where the surfaces are rough [212,213] and the enhancement 146  o II  -1_  1.0  I  L  o o A > +  0.8  <  0.6  :  >  0.4  § 0.2  <  L  _i  l_  9  o o > & o > o  1 st decay 2nd decay 3rd decay 4th decay 5th decay  9  0.0  -|  1  1  r  ~i  1  1  5  0  r  n  1  1  "T  -  10  1  T  1  15  20  Time / min Figure 7-4 The fluorescence decays of Figure 7-3 normalized to the first desorption image in each decay.  0.20 -|  o DilC18(5)  S 0.10 0.00  ooooooooooooo  oooooeoooooo  o 5-Octadecanoylaminofluorescein  0.30 °0.20 < 0.10  s  :  o.oo  :  ooo»  T"  5  10  15  20 25 30 Time (min)  35  40  —I  45  1  —  50  Figure 7-5 The measurement of four fluorescence decays from a 2 mol% DiIC18(5)/2 mol% 5octadecanoylaminofluorescein/C 180H layer. The first two decays were measured for fluorescence of DilC 18(5) (top panel) and the last two decays were measured on the same adsorbed layer and on the same region of the electrode surface for 5-octadecanoylaminofluorescein fluorescence (bottom panel).  147  is particularly evident for low quantum yield fluorophores [207]. Since the Au(l 11) surface is very smooth compared to Ag island films, the possibility of a radiative enhancement region near the electrode is unlikely. Although this is not definitive, the possibility of this enhancement region will be described in more detail later in the chapter. At present, the developed technique cannot detect a slow drift of the molecules back to the metal surface with time. Since the electro-fluorescence characterization of the mixed monolayers of 2 mol% 5-octadecanoylaminofluorescein, 2 mol% DilC 18(5) in C180H conducted in Chapter 6 show efficient mixing of the two dyes within the CI 80H matrix, this system can be used to test this hypothesis. If the desorbed layer is drifting back to the metal surface with time, the decay and recovery should be observed for both DilC 18(5) and 5-octadecanoylaminofluorescein within one layer. To test this possibility, a mixture of 2 mol% 5-octadecanoylaminofluorescein and 2 mol% DilC 18(5) in C180H was studied in the decay experiments. In total, 4 decays were measured for this mixture adsorbed on the Au(l 11) electrode. Thefirst2 decay cycles measured the fluorescence from DilC 18(5) immedietely followed by 2 decay cycles measuring the fluorescence from 5octadecanoylaminofluorescein on the same region of the electrode/surfactant. If the dye molecules are  slowly  drifting  back  to  the  electrode  surface,  both  D i l C 18(5)  and  5-  octadecanoylaminofluorescein should undergo the observed decay and recovery process. The results of the decay investigation for this mixed monolayer are shown in Figure 7-5. Immediately apparent, is the decay and recovery for the DilC 18(5) dye in C180H. These decay and recovery curves are similar to those previously shown even with a lower concentration. The fluorescein dye clearly does not undergo a similar decay and recovery process and only shows a constant decay in fluorescence consistent with irreversible photobleaching or photodestruction. These results indicate that a slow drift to the electrode is not responsible for the observed decay and recovery. Therefore, the decay and recovery is specific to the carbocyanine dye DilC 18(5). This accounts for the deviation in the measured fluorescence intensity curves during the adsorption desorption cycles described in Chapter 6. The fluorescence from the fluorescein dye molecules were consistently at a maximum at the desorption limit regardless of the supporting matrix (C180H or OLA). Conversely, at negative (desorption) potentials, the DilC 18(5) fluorescent molecules were observed to rapidly decrease in fluorescence intensity. This was observed when DilC 18(5) was present in C180H or OLA.  148  Furthermore, the decrease in fluorescence was more efficient when DilC 18(5) was present in the more fluid OLA matrix. This suggests that the process by which the decay occurs depends on the fluidity of the supporting matrix.  7.1.3.4 Light Independent Dye Aggregation The clear indication of a dominant light independent pathway for the fluorescence decay and recovery of DilC 18(5) has been shown. Cyanine dyes have been observed to undergo selfassociation to form aggregates. Since their discovery in 1936 [57], the aggregated complexes have been commonly referred to as J or H aggregates. Cyanine dye aggregation has been observed in solution, on the water surface, and on solid surfaces, and the aggregation has been shown to be reversible with temperature and specific to the aggregate environment [76,214-216]. These aggregated dye complexes have a significant effect on the spectral properties of the dye. Compared to the monomer, an aggregated complex may have shifted absorption/emission spectra [214] as described in Chapter 4. There have been many investigations of aggregates in both solution and on solid phases mainly for their application in photography [216]. Red-shifted fluorescence is typical of J-aggregates and blue-shifted absorption is typical of H-aggregates. In some cases, the H aggregated complex has been rendered non-fluorescent [75,217,218]. Some reports have described the structure or framework of the aggregate and its effect on the photophysics of the dye, [58,219] yet the number of monomers that make up an aggregate has not been clearly defined. The aggregate formation process has been shown to be reversible with changes in temperature [215]. Experiments conducted on layers of DilC 18(5) at the water and glass surface [76] have been described in the literature. It was found that the aggregation of DilC 18(5) occurred more readily on the water surface because of the less restricted microenvironment. Furthermore, at low concentrations, the dye molecules were observed not to aggregate on the solid glass surface. These results suggest that the light-independent pathway dominant in the observed decays may be the result of formation of an aggregated complex of carbocyanine dye. This aggregated complex will significantly change the nature of the photophysics of the dye, possibly altering the fluorescence emission spectrum to wavelengths beyond the filter cube transmission range. Measurements of the fluorescence spectrum of 3 mol% DilC 18(5) in C180H during one 149  decay cycle is shown in Figure 7-6. The spectra were acquired with the fluorescence filters in place, therefore only a restricted wavelength range could be examined. In the figure, a fluorescence spectrum measured after 1 minute at E  des  minutes at E  des  followed by an emission spectrum measured after 20  are compared. After normalization at 675nm, no change or shift in the emission  profile over 20 minutes was observed and suggests that any shift must be considerably larger than 60 nm so that the emission lies outside the wavelength range of measurement. A more likely possibility is that the absorption spectrum for the aggregated species lies outside of the excitation filter wavelength range. Aggregation then 'protects' the dye from photobleaching but reduces fluorescence since the complex is no longer absorbing photons. Alternatively, the new aggregate may not fluoresce at all. In both scenarios, the decrease in fluorescence will occur because fluorescence can only be generated from the monomer form of the dye, which is decreasing with time at the desorption potential due to aggregation. Re- adsorption onto the metal surface may have the effect of reverting the aggregate to its monomer form which is able to fluoresce once re-desorbed. This phenomenon is not expected for 5-octadecanoylaminofluorescein since fluorescein dye molecules do not readily undergo aggregation, although in multilayer assemblies it has been observed [220]. This aggregation hypothesis will be tested through fitting to a simple model developed next.  7.2  Proposed Kinetic Model  7.2.1  Model Assumptions There are at least two pathways responsible for the fluorescence decay of DilC 18(5) with  time spent at desorption. One path must occur irreversibly since the recovery is not 100%. This pathway is believed to be dependent on exposure to excitation light and it is most likely due to the photobleaching or photodestruction of the fluorescent molecules. The second pathway is light independent and causes changes that are reversible once the surfactant layer is re-adsorbed onto the solid metal surface. This is believed to be a light independent aggregation of DilC 18(5) which revert to the monomer form upon re-adsorption.  The kinetic model reflects these two pathways.  Experimentally, only the fluorescence intensity (AI/I ) can be measured. In the model it is 0  assumed that this is proportional to the concentration of monomer dye molecules. Also, since the separation between the fluorophore and metal strongly influences the radiative lifetime, the model 150  650  675 700 725 Wavelength / nm  750  Figure 7-6 Normalized fluorescence spectra of DilC 18(5) measured between 660 and 740 nm during the 20 minutes spent at desorption for a typical fluorescence decay. The solid line represents the spectrum measured when the layer was first desorbed, and the dotted line after twenty minutes of desorption.  151  must assume a minimum separation from the metal where quenching effects are minimal. Therefore, the model assumes that the desorbed layer exists far enough away from the metal such that even partial quenching does not occur. Furthermore, since the aggregates revert to the monomer form by re-adsorption, it is assumed that each new decay begins at time zero. Therefore, each of the consecutive decays is treated such that time is reset to zero for the first desorption potential image in each decay cycle.  7.2.2  Kinetic Model  Derivation  A ground-state molecule of the carbocyanine dye (A) can undergo three principal reactions. In order to fluoresce, A must absorb a photon, become excited and then relax to the ground-state via the emission of a fluorescent photon. Secondly, A can absorb a photon and irreversibly photobleach with a rate constant k , resulting in a new compound or compounds (A ) that are no longer p  P  fluorescent and therefore unobservable by the experiment. Lastly, two ground-state molecules of the dye could form an aggregated complex (X) with a rate constant k , creating a non-fluorescent a  complex. The formation of X is reversed upon re-adsorption onto the metal surface, creating monomers that can again participate in fluorescence. Within the model, the first reaction is ignored since the excited-state lifetime of a fluorescence event is short when compared to the other two processes. The remaining two parallel processes are described by A A +A  )A  kp  7-1  p  K  )X  7-2  Experimentally, the fluorescence intensity is used as a measure of the concentration of free unassociated monomer dye molecules (A) and only the emitted light within the frequencies of the microscope band-pass filter are observed. Assuming that the normalization correction to the images results in a constant flux of photons over the image region analyzed, the change in A with time can be written as  152  Integration of this equation, using the intensity of the first desorption image for each decay (AA as an integration limit at (t ), yields the evolution of [A] with time as shown below a  [A(t)] =  7-4 k +  k A -k A e- r< k  p  a  o  a  o  The experimentally measured changes in fluorescence intensity with time are fit to this general equation along with one additional piece of data.  The difference in intensity between two  consecutively desorbed layers, just after the potential step from E  represents the experimentally  ads  measured amount of photobleached dye. This value (A), can be found by calculating the difference in the image intensity of the images B and D from Figure 7-1. From the model, this value can be found through the use of Equation 7-1 in the following form.  =  dt  7-5  k [A] n  Inserting Equation 7-4 into the value of [A], results in the following equation.  (k A (t) p  A k -k A e- '') k  p  +  0  a  a  0  7-6  log  Using this equation, the amount of photobleached dye (A ) can be calculated and compared to the p  experimental loss in fluorescence intensity. This value is also used within the fitting routine as an extra parameter. Previous studies on this system have shown that the desorbed molecules will scatter light more efficiently than adsorbed molecules [16]. This would appear as a constant offset to the measured histogram intensity which is not accounted for in the model. Although the scattered light is attenuated by the filter set used, a small amount of leakage can be expected. To account for this extra contribution, Equation 7-4 was modified with the addition of a constant term d. Fitting of the model to each individual experimental decay cycle was performed using the data shown in Figures 7-1, 7-2 and 7-3. The fitting routine (described in Appendix A2) utilized a least-squares fitting method based upon the Matlab optimization routine using a Nelder-Mead 153  simplex method. The resulting fit to the data is shown in Figure 6-7a. Overall, the fit of the modified Equation 7-4, is in agreement with the experimental data with regression coefficients of >0.997. Both rate constants, k and k as well as the constant, d are output from the model. The a  p  model also calculates A from Equation 7-6 and compares this to the experimentally calculated value. The experimental data (symbols) and the resulting model fit (lines) for the first and last decays of Figure 7-1 are shown in Figure 7-7a. The bottom panel of (a) shows the same decays on log scale. Clearly the decays are not a simple first order process. The fitting parameters for all the decays of Figure 7-1 are shown in Figure 7-7b. With the inclusion of the scattered light constant parameter, the model fits the experimental data very well. The resulting fit parameters are shown (b) as a function of the decay cycle number. The rate constants (k and k), scattered light offset (d), a  and the amount of photobleached A (A) both experimental and calculated from the model, are compared in the next discussion. Generally, k is an order of magnitude larger than k and increases with decay number. This a  is a surprising result as k was expected to be relatively constant. The scattered light contribution a  is small and slowly decreases with decay number. This decrease in the scattered light is expected as previous investigations have revealed that the efficiency of scattering decreases with cycle life as a result of the slow decay in quality of the adsorbed surfactant [16]. The calculated and experimental values of yl overlap well and show a very slow decrease in magnitude with increasing decay number. p  The relative magnitudes of the rate constants indicate that the reversible aggregation process dominates the observed trends and only a small amount of irreversible photobleaching occurs. This result fits with what is expected since the aggregation process effectively protects the monomer from photobleaching and since the recoveries are so large, the aggregation must dominate. The slight decrease in A is expected since exposure of the desorbed layer to light increases with decay number. From this perspective, the layer is repeatedly exposed to light and photobleaching removes some of the fluorescing A molecules before efficient aggregation dominates. It would be expected that the molecules most affected by photobleaching would be those furthest from the electrode first since their lifetimes are much longer due to less efficient metal mediated quenching of the fluorescence. This results in a decrease in the rate of photodecomposition and therefore a decrease inA with decay p  number. 154  a)  2.0  s  -i  1.5 -  A First Decay  to H  o Sixth Decay  < 0.5 0.0 -  ^ O O O O O O O O O O O O O O  10 1 i 10" i  <  1  ° O Q  10" 5  b)  3.00 • | 2.00 ^ 1.00 ^"0.00  1 — 1 10 15 Time (min)  20  "™1  "7  J  - 0.20 E 0.15 E 0.10 0.05 0.00 Experimental Calculated  0.40 n 0.30 1 af-0.20 0.10 0.00 0.15 0.10 •o 0.05 0.00  n  4 i  1  2  1  1  3 4 5 Decay Number  1  6  Figure 7-7 (a) The first and last fluorescence decays from Figure 7-1 with the kinetic model fits overlaid. The bottom panel of (a) represents the same decays/fits on log scale. The symbols defined in the legend are the experimental results and the solid line represents the calculated fits. The model parameters are shown in (b). 155  7.2.2.1 Modification to the Model The increase in k by over a factor of 2.5 with decay number reflects that the system is more a  complex than the simple model proposed. This discrepancy results from the assumption that the fluorescence intensity is a measure of all the fluorescing monomers of A in the interfacial region. The model assumes that all the dye molecules exist far enough from the metal such that quenching is not important. This must be reconsidered. Figure 7-8 schematically illustrates the simplified depiction of the interface with the organic film represented as a slab separated from the metal. The relative fluorescence intensity (based on a typical quenching curve) is also shown. At present, the model assumes that at desorption the molecules are at a distance d* beyond the effects of quenching. At this distance, all the molecules within the layer are able to fluoresce and can be measured. As a result, the layer at d* in this slab depiction is fully fluorescent and represented as a uniform grey value. At a different separation distance d , where 50% of the fluorescence is quenched, the slab is a  represented as a nonuniform grey level. If the surfactant layer exists close to the electrode surface, a certain population of A, called A will not fluoresce and therefore will not be considered in the model. Although these molecules are not able to fluoresce as efficiently, they are able to participate in the aggregation process.  Thus, these quenched molecules can still form aggregates with  themselves, A -A , or with a fluorescing partner, A -A. This possibility explains the increase in k q  q  q  a  with decay number since the rate of aggregation is no longer simply measured by Equation 7-2 because the fluorescence intensity is not an accurate measure of the total concentration of the monomer dye molecules A and A ^but only of the fluorescing monomer species [A]. The real rate of aggregation k ' and its relationship to the determined k is more accurately defined in the following a  a  relation  k [A]  =  2  a  k' [A+A [ a  q  7-7  which can be simplified to (  7-8 V  156  0  20  40  60  80  100  distance from electrode surface Figure 7-8 A schematic depiction of the fluorescence intensity at different separation distances between the metal and fluorophore. The model assumes a distance of separation of d* where fluorescence quenching is minimal. At a position d , some of the fluorophores will be quenched and not accounted for in the model. 0  157  With increasing exposure to light, the amount of fluorescing^, decreases while the amount of nonfluorescing A remains constant and is manifested as an increase in k with decay number since A a  q  participates in the aggregation process as an invisible partner. The kinetic model fits the experimental data very well with some slight modification to the assumptions. This rate of decay at desorption explains the observed differences in histogram intensity during the adsorption/desorption slow potential scan cycles measured in Chapter 6. It was observed that a rapid decrease in fluorescence intensity occurred for the desorbed DiIC18(5)/OLA layers. Since OLA is a liquid, the rapid decrease in fluorescence intensity at potentials past -0.700 V/SCE can be attributed to a more rapid aggregation process within the fluid matrix of OLA. This process was investigated by measuring the fluorescence decay and recovery of DilC 18(5) in OLA at various negative potential limits, revealing the effect of the supporting matrix on the rate of aggregation.  7.3  Fluorescence Decay and Recovery of 3 mol% DiIC18(5")/QLA  7.3.1 Fluorescence Decay and Recovery Under Constant Illumination The fluorescence decay and recovery of DilC 18(5) in OLA was studied at four negative potential limits; -0.600, -0.650, -0.700 and -0.750 V/SCE. Preliminary investigations revealed that the fluorescence was undetectable after 10 to 15 minutes spent at E  des  due to a more rapid decay than  DilC 18(5) in C180H. The decay cycles were then modified: 5 images a.tE at 30 second intervals. ads  Imaging at E  des  was modified to 20 images over 10 minutes. Therefore, the number of images  remains the same as the previous decay cycles and only the time spent at the different potentials has changed. The decays and recoveries (a) along with the aggregation rate constants (b) are shown in Figure 7-9. For the decay and recoveries, the symbols represent the calculated histogram values from the experiment and the lines represent the kinetic modelfitto the curves. In (a), more negative potential values results in a more rapid decay in fluorescence. This suggests that the aggregation of D i l C l 8(5) occurs more rapidly as the layer is subjected to more negative potentials which is related to the distance away from the electrode surface. This also suggests that aggregation occurs more rapidly as the layer becomes progressively more desorbed. This is also manifested by the lower 158  a)  0.30 _o0.20  o -0.600 V/SCE  < 0.10 0.00  Qccco  ^cocoo  0.30 _o0.20 ^  o -0.650 V/SCE  0.10 0.00  ocxx>o  0.30 _o0.20 ^  A -0.700 V/SCE  0.10 0.00  ,  A  A A A A A A A A^A . .  A A A A  A A A A A A A A A  f  0.30 _o0.20  > -0.750 V/SCE  < 0.10 **f*[^D t> D p t>t»  0.00  »M)t)pt>l>l>PI>>h.  10  15  20  25  frfrfr t>tJMi'fDt>DI>t!'t>l>>  30  35  T i m e (min)  b)  30.0 n 25.0 1— ' c 20.0 E 15.0 10.0 5.0 0.0 - i  > -0.750 V/SCE  : :  A -0.700 V/SCE o -0.650 V/SCE  :  :  :  o  o -0.600 V/SCE  Decay Number  Figure 7-9 (a) Measured fluorescence decays of DiIC18(5)/OLA at -0.600 (o), -0.650 (0), -0.700 ( A ) , and -0.750 V/SCE (>). The line represents the kinetic model fit for each decay. The aggregation rate constants are shown in (b).  159  amount of photobleached (A ) product with increasing negative potentials. As the aggregation rate p  increases, the molecules are protected from photobleaching. As such the recoveries are larger for the curves that have less photobleaching due to a more rapid aggregation. Also apparent in the analysis is the slight decrease in the intensity of the first value in the desorbed cycle (A ). It would 0  be expected that as the potentials approach more negative (desorption) values that the intensity of A would increase. This however, is again a result of the more rapid aggregation. It becomes more a  difficult to accurately obtain a value for A because the image decreases in intensity so rapidly over a  the 10 seconds spent during image acquisition. In fact, at the most negative potential (-0.800 V/SCE), the decay was so rapid that an accurate decay trend could not be reliably obtained. When studying the aggregation rate constant output from the kinetic model, the values increase rapidly with increasing negative potentials. At -0.600 V/SCE, the rate constant is comparable with that obtained at -0.800 V/SCE for DilC 18(5) in CI 80H. As the negative potential limit increases, the decay rate constants increase by an order of magnitude. This demonstrates that the rapid loss of the DiIC18(5) fluorescence intensity during the slow potential scans presented in Chapter 6 (Figure 6-9), are a result of the increased aggregation rates of DilC 18(5) in the fluid OLA. As the layers become increasingly more desorbed, the aggregation rate is increased and a dramatic loss of fluorescence intensity is observed. On the positive potential scan when the molecules become contact adsorbed, the fluorescence intensity again increases rapidly because the aggregated molecules are destabilized, and the aggregation rate hindered as a result of the re-adsorption process.  7.4  General Observations and Conclusions The carbocyanine dye DilC 18(5), was observed to undergo a repeatable fluorescence decay  and recovery process when separated from the metal surface in a potential controlled fashion. The desorption of the dye molecules into the aqueous phase resulted in a strong decrease in the calculated fluorescence intensity with time. Unexpectedly, a recovery of lost fluorescence was repeatably observed upon re-adsorption to values greater than 80%. The decay of fluorescence was attributed to two parallel pathways describing irreversible photobleaching and reversible aggregation of the dye molecules when separated from the metal. It is believed that the aggregated species is a nonabsorbing complex that reverts to the monomer form when re-adsorbed onto the metal surface. The 160  aggregation process was found to be the dominant factor in the decay pathway. Although the non-absorbing complex was not spectroscopically observed in the present experimental set-up, the decay and recovery were not observed for 5-octadecanoylaminofluorescein, a dye molecule that does not readily aggregate. This demonstrates that the desorbed molecules cannot be slowly drifting back to the electrode surface with time. This was a key factor in confidently ascribing the decay of the carbocyanine dye to an aggregation process. A kinetic model was applied to the decay pathways to describe the rate of aggregation of the carbocyanine dye. The aggregation rate constant is roughly 10-40 times greater than the rate constant for the photodecomposition process. When the carbocyanine dye is present in a more fluid monolayer (OLA), the decay and recovery was observed to occur by an order of magnitude more rapidly in comparison to C180H. This was attributed to the fluidity of the O L A phase allowing the DilC 18(5) molecules to diffuse and aggregate more efficiently. This aggregation process was used to describe the large decrease in fluorescence intensity during the slow potential scans of DiIC18(5)/OLA observed in Chapter 6. The process of desorption allows the molecules to aggregate more rapidly as the desorption process continues resulting in the observed decrease in fluorescence at the most negative potentials. Upon re-adsorbing the molecules in the positive potential scan, the aggregates become destabilized and the resulting monomers may again fluoresce. This accounts for the observed increase in intensity on the positive return cycle of the slow potential scan. Overall, these results demonstrate the possibility for electrochemical control over the radiative characteristics of a fluorophore. This control stems from the ability of electric potential or charge to manipulate the interfacial tension of the metal/solution interface, which controls the distance/separation between the metal surface and the desorbed surfactant layer. The observation of aggregation to form a non-absorbing complex was shown only indirectly and direct proof of the existence of these aggregates will require the measurement of the changes in the absorption spectrum of the desorbed molecules over time.  161  Chapter 8 8  Selective Desorption of BODIPY-C10-SH from a Au Bead The reductive desorption of a fluorescent molecule bearing a thiol moiety from selective  regions of a polycrystalline gold bead electrode is described in this chapter. The collection of capacitance during an image/potential scanning investigation revealed irreversible character when scanning to large negative potentials. An increase in the measured capacitance was followed by an increase in the fluorescence intensity due to reductive desorption. The slight solubility of the thiolate molecule in solution resulted in slow diffusion of the thiolates away from the electrode surface when desorbed. The diffusion process prevented oxidative re-adsorption of the thiol during the positive return scan in potential enabling the observed selective desorption.  8.1  Electro-fluorescence Characterization of the Selective Desorption Process  8.1.1  Modification  to the Electrochemical  Setup  An alkaline electrolyte (0. I M NaOH) was used to reduce hydrogen evolution resulting from the extreme negative potentials required to reductively desorb the surfactant. Typically, stronger basic solutions are used, however, the thin window in the spectro-electrochemical may be etched in this situations which would reduce image quality. As a result, the reductive desorption peak in the CV was masked by hydrogen evolution and only capacitance measurements will be described. The self assembled monolayer (SAM) was formed on the electrode externally from the cell by immersing the bead in a 3mg/ml solution of the thiol dissolved in chloroform for 10 minutes. Longer deposition times were studied (up to 20 hours) with no significant change to the minimum capacitance. The bead was then rinsed with water, followed by sonication in a methanol solution to remove any physically adsorbed surfactant. After evaporation of methanol, the S A M coated electrode was pushed through the electrolyte surface to ensure that the entire bead was in contact with the working solution. The method of forming the S A M externally from the cell results in the absence of fluorophore at the gas/solution interface in contrast to the physically adsorbed systems. In the S A M investigations on the polycrystalline Au bead, a monolayer of fluorescent thiol is present only at the metal surface until the molecules are desorbed.  162  8.1.2  The Collection  and Treatment  of  Images  Fluorescence images were acquired with the lOx objective to capture as much of the bead surface as possible. A l l images were acquired with an exposure time of 50 ms and a gain of 1 through five imaging/potential scanning sets on one adsorbed layer. The program was modified to increment the potential in -50 mV steps beginning at 0 V/SCE. Five characteristic desorption values were predetermined for each imaging cycle (E = -0.900, -1.250, -1.300, -1.350 and -1.400V/SCE) des  where the potential was held for the acquisition of 3 desorption images to monitor the diffusion of desorbed thiol. The potential was then stepped back to 0 V/SCE in +50 mV increments. The next image/potential scanning routine was conducted on the same S A M coated electrode with the more negative E  dcs  value. A l l potential steps were separated by 2 seconds, resulting in a scan rate of  roughly 25 mV/s. All images were analysed using dip_image [201] for M A T L A B and Image Pro Plus 4.5. The average gray scale was calculated using a M A T L A B procedure similar to histogram_calc.m outlined in Appendix A l . In this analysis, the subroutine find_features.m was not used. An I was not 0  available due to the inability to accurately reposition the bead before and after the S A M formation. Therefore, the fluorescence intensity was calculated from AI images after the subtraction of an I  ads  image.  8.2  Electrochemical and Epi-fluorescence Investigations  8.2.1  Electrochemical  Characterization  The thiol coated electrode was introduced to the electrochemical cell at 0 V/SCE and the stability of the S A M was first probed in a limited potential range with a CV prior to fluorescence imaging and capacitance measurements. Figure 8-1 shows the capacitance measured for one S A M coated electrode during the five image/potential scanning routines having different E  des  limits.  Therefore, one S A M coated electrode was exposed to five different desorption limits each more negative than the last. Beginning with (a), the freshly modified electrode surface has a capacitance of 1.2 u,F cm" . This is significantly lower than the capacitance for an un-coated electrode at the 2  same potential (> 80 U.F cm" ). During the slow potential increments to the negative desorption limit 2  ofE  des  = -0.900V/SCE the capacitance remains low and relatively constant. This low capacitance 163  -1.4  -1.2 -1.0  -0.8 -0.6 -0.4 E / V v s SCE  -0.2  0.0  Figure 8-1 Measured capacitance for a S A M coated Au bead electrode. The capacitance for the bare electrode (•••) was measured before the S A M formation. The capacitance curves for the S A M coated electrode (symbols) were measured during the imaging experiment. One S A M coated electrode was exposed to the desorption limits of E = -0.900 (a), -1.250 (b), -1.300 (c), -1.350 (d) and -1.400 V/SCE (e). 164  demonstrates that the S A M is a well-ordered monolayer layer despite the bulky head group and short assembly time. The lack of an increase in capacitance at the negative limit demonstrates that these potentials do not induce desorption and only a small increase in the capacitance at the positive limit was observed on the return scan (solid symbols). This suggests very slight disruption or disorder to the SAM. The subsequent imaging/potential scan (b) was made more negative to E  des  = -1.250  V/SCE where a significant increase in capacitance is observed. This demonstrates the instability of the monolayer toward larger negative potentials through penetration of the S A M by the electrolyte. This is believed to happen through a similar mechanism to the physically adsorbed surfactants by potential created defects in the layer caused by the initial stages of desorption. On the return scan, the capacitance has noticeably increased to 3 p,F cm" . Since the capacitance has not merged with 2  that for the bare electrode, it is expected that some surfactant remains on the metal surface. Exposing the same S A M coated electrode to further negative potentials (E = -1.300 V/SCE) shown des  in (c), introduces the drastic changes in the adsorbed layer as a result of desorption process. At the start of the negative potential scan, the capacitance begins at much higher values than the initially coated S A M electrode in (a), and increases to values just above 20 U.F cm" at the negative limit. The 2  sharp increase in capacitance at the negative potential region suggests significant desorption of the thiol molecules. This is further demonstrated in (d) for a negative limit of E  des  = -1.350 V/SCE  where the capacitance merges with that for the uncoated electrode surface. In both (c) and (d), a large increase in capacitance remains for the positive potential scan indicating that the thiol molecules are not re-adsorbing onto the metal surface. This suggests that the electrode is becoming continually stripped of more thiol with each new desorption limit. As a final excursion, the same layer was exposed to a negative limit of E  des  = -1.400 V/SCE. Even at positive potentials, the  capacitance is high enough to be consistent with the bare electrode capacitance indicating that all the thiol has been removed from the metal surface. The slight offset could be the result of different immersion depths of the bead between the bare run and the imaging procedures. The selective removal of the S A M from regions of the metal surface will be shown with fluorescence imaging  8.2.2  Electro-fluorescence  Characterization  The calculated fluorescence and selected images for the desorption process are shown in 165  Figure 8-2. These images and fluorescence values were measured at the same time as the capacitance measurements in Figure 8-1 and will be discussed together. When the negative potential limit is restricted to E  des  = -0.900 V/SCE, only a slight disruption to capacitance is  observed which corresponds to the slight increase in fluorescence measured (a). This does not require full desorption of the molecules but rather a change in the adsorbed monolayer such that the fluorescent head group of the S A M is slightly further away from the metal surface. This change in fluorescence intensity without desorption of the molecules was also observed by Gaigalas in the study of the dynamics of immobilized fluorophores [185,186] and fluorescein-labelled single-strand oligonucleotides immobilized onto gold [184]. The images from A to D reveal the structural changes in the monolayer as a result of a change in the electrical variable. In order to observe any fluorescence within this region of potentials, the exposure time was changed from 50 milliseconds and a gain of 1 to 5 seconds with a gain of 2. Even at this longer exposure time, only slight changes in the desorption potential images were noticed. Image A is quenched as expected, and a relatively uniform increase in intensity was observed in image B preceding the negative limit. The three images at E  des  = -0.900 V/SCE (images C I to C3) reveal some heterogeneity in fluorescence, with  the most intense regions at the bottom right, and the top center of the image. The dark spot at the bottom right is a result of the subtraction process eliminating the only 'hotspot' in the image similar to the dark regions previously observed with the C180H and O L A investigations in Chapter 6. Fluorescence intensity measured in the scan to the second desorption limit is shown in (b). The dramatic increase in capacitance is followed by an increase in the fluorescence intensity at the negative limit. Features in the images become apparent at -1.150 V/SCE (image F) which are more clearly observed at the negative limit (images G l to G3). The fluorescence at this desorption potential is much more intense even when measured with an image exposure times of 50 milliseconds and gain of 1. Holding the potential at the negative limit for three images allows enough time for the slow diffusion of desorbed molecules. This is most clearly manifested when comparing the difference between image G l and G3. The coupled mechanism of diffusion, and the excited state relaxation of the molecules not fully desorbed on the positive scan result in the decrease in fluorescence intensity of image H. The structure of the desorbed layer is considerably different than those obtained for the physically adsorbed systems in Chapter 6. In those investigations, only 166  e)  03  60  positive s c a n  X 40 • o 0  20  < negative s c a n  1  1 mm  < d)  60 n =! 4 0 H  Ol  02  M  <  A A A A  A A A A A  A A A A A A A  N' 60 1  O  40  KI  ^ 2 0 <  b)  -j  K2 -*  A\* ,  n  02  03  P  KI  K2  K3  L  Gl  G2  G3  H  CI  C2  C3  D  K3  4  L  060  Ol  G3  ™ 40 H  f  OH 60 i  a)  a: i I  X40 • o  e 20 • 9  o  i ' ' '  -1.4  i ' ' '  -1.2  i '  -1.0  • • i • ' '  -0.8  i ' ' ' i  -0.6  -0.4  -0.2  0.0  E / V vs S C E  Figure 8-2 Measured fluorescence images and calculated intensity of the reductive desorption of the S A M from the Au bead. The calculated fluorescence intensity plots (a - e) correspond to the capacitance measurements in Figure 8-1 (a - e). Selected images are represented at the bottom at the potentials labeled on the graph.  a small amount of the adsorbed layer contained the fluorescent molecule. Because of this there was contrast in the images and the aggregates revealed. The features of the desorbed S A M , while heterogeneous, are significantly larger in shape and structure. The S A M appears to be displaced from the metal surface in selective sections. More specifically, in image G l , the elliptic region of fluorescence at the bottom right, and the rectangular region at the top centre will be important regions of the electrode surface on the next desorption limit. For graph (c), the increase in capacitance is again followed by an increase in fluorescence. Since the capacitance on the return scan was significantly altered as a result of irreversible desorption, the features in the desorption images should reflect a significant amount of desorbed thiol. This is observed when potential is held at the negative limit ofE  des  = -1.300 V/SCE. The desorption images (images KI to K3) again reveal  a heterogeneous structure and slow diffusion of the thiolates. Interestingly, the features that were present in the preceding scan (i.e. the oval and rectangular shaped features in image Gl) are not present in the images KI to K3. This is a clear indication that the diffusion of the desorbed thiolates on the previous scan render these molecules unable to re-adsorb on the metal surface. Thus, fluorescence is not observed from these regions on subsequent scans. Conversely, there are two distinct regions in image KI that have yet to fluoresce. These regions exist as the curved shape in the left of the image, and the darkened region in the upper right corner. The next scan (negative potential limit of E  des  = -1.350 V/SCE) shows a similar trend in the fluorescence intensity to those  measured previously. A slight increase in intensity at the negative limit is observed however, the magnitude of fluorescence is lower as a result of small amount of thiol available on the electrode to desorb. The features in the images O l to 03 are exactly consistent with those regions in the preceding scan that were not fluorescent. At the most negative potential excursion (E = -1.400 des  V/SCE) no changes in fluorescent intensity are observed (e) due to the complete removal of all thiol from the metal. This is consistent with the measured capacitance resembling a bare electrode. Interestingly, at the end of the experiment, fluorescence was observed at the electrolyte surface, strongly suggesting the segregation of the desorbed thiolates to the gas/solution interface. The analysis of capacitance and fluorescence intensity on the same S A M coated electrode using different negative potential limits in each scan revealed a progressive removal of thiol molecules from different regions of the metal surface. The heterogeneous structure ofthe desorbed 168  layers is due to different energies of the polycrystalline surface of the metal. From the images acquired, a selective reductive desorption process can be imagined. This process is more clearly observed through some simple image analysis.  8.3  Selective Reductive Desorption Outlining the fluorescent features within the first image acquired at each negative potential  limit was accomplished in order to reveal the relationship between the metal surface morphology and potential at which desorption was observed. The images upon which the analysis was carried out are shown in the first column of Figure 8-3. The features of each image are selected using an intensity segregation procedure and are displayed in the second column of the figure. These images were then applied to the -1.300 V/SCE image, displayed in the third column. Clearly, the extent of the negative potential excursion is related to the fluorescent features. Fluorescence from elliptic feature observed on the bottom right of image (c) occurred at the least negative potential. When applying the outlines of this image onto the -1.300 V/SCE image, it becomes clear that this region is not fluorescent on the subsequent potential scan. Using a surface analysis method of Electron Backscattered Diffraction (EBSD) this oblong surface feature was determined to be a (111) facet. This was determined by the Metals and Materials Engineering Department of the University of British Columbia with the help of Dr. Chad Sinclair. Energetically, the thiols would be expected to desorb from this face at the least negative potential [ 10,162]. The lack of fluorescence demonstrates the irreversible loss of the surfactant due to desorption from this region of the electrode. The selective desorption is more clear when overlaying the outlines of (a) onto the -1.300 V/SCE image. Here the regions that were not fluorescent at -1.300 V/SCE become fluorescent at -1.350 V/SCE. The features fit exactly into the dark regions of the -1.300 V/SCE image. The sensitivity toward the value of the potential driving these changes is clear and can be understood in terms of the energetics of the electrode/electrolyte interface. Because bead electrodes are typically composed of regions of differing surface crystallography [221 ] etching the bead in a solution of aqua regia should enhance these different features on the metal. The etched bead with the surface features revealed is represented in Figure 8-4. A brightfield image of the bead before acquisition of the fluorescence images is shown in image (a). Here the elliptical (111) facet is shown 169  a) Edes = -1.350 V / S C E  Outlines  Overlay onto Edes = -1.300 V / S C E  b) Edes = -1.300 V / S C E  c) Edes = -1.250 V / S C E  Figure 8-3 Delineation of features in the desorption images at (a) -. 1350, (b) -1.300, and (c) -1.250 V / S C E with the outlines displayed in the middle column. The application of the outlines to the desorption image at -1.300 V / S C E is displayed in the third column.  170  (a) SEM image of the etched bead revealing surface heterogeneity, facets and grain boundaries, (b) Optical image of the etched bead positioned as in the experiment with the surface features labeled, (c) Optical image with the grain boundaries highlighted. Image (c) with the fluorescent features of E = -1.250 (d), -1.300 (e) and -1.350 V/SCE (f) applied. Figure 8-4  des  171  at the bottom right of the image. A Scanning Electron Microscope (SEM) image and an optical image of the etched bead (etched after the experiment) are shown in image (b) and (c). These images were positioned similar to that during the experiment by using the oblong orientation of the facet as a guide. This can be done with confidence because SEM images of the complete bead surface did not reveal any other large oblong facet. By orientating the etched bead in this manner, the outlined features can be placed onto the image to observe any correlation between the fluorescent features and surface features on the metal. The SEM of etched Au bead (b) reveal grain boundaries and various facets of the heterogeneous electrode surface. These features are also present in the optical image (c), with some of the features outlined. The elliptical facet that was determined to be (111) is labelled as region I. It is from this region that fluorescence was observed for the least negative desorption potential. Desorption from other regions of the electrode surface at this potential are most likely also dependent on surface crystallography, but EBSD was unable to determine the surface normals for curved portions of the surface. The feature at the top of the image resembling an elongated triangle is labelled region II. A grain boundary labelled III curves downward toward a crescent feature labelled IV. This boundary continues on the bottom of the image toward the facet and is labelled as region V. Upon close inspection, some striated features just above the grain boundary V, are observed and extend towards the facet (I). These features will be the main features for comparison when overlaying of the fluorescence outlines onto the optical image. Image (d) is the optical image of the etched bead with the grain boundaries outlined. The optical image of the etched electrode was created by overlaying the in-focus regions from a number of images with differing focal planes to facilitate comparison with the SEM image. This was not done for the fluorescence images and fluorescence will be measured from regions not in the focal plane, so a direct comparison is qualitative but still very useful. The feature masks presented in Figure 8-3 were filled (cross hatch) and placed onto the optical image of the etched bead with grain boundaries outlined. The filled feature mask for the fluorescent image at -1.250 V/SCE was overlaid onto image (c) and shown in (d). The fluorescent feature protruding from the top of the desorption image matches well with feature II in the optical image. Unfortunately, the circular feature in the center of the image does not seem to correlate well with any surface irregularities however, the fluorescent feature at the bottom right corresponds 172  exactly with the (111) facet. Overlaying the feature mask of the -1.300 V/SCE image onto the optical image is shown in (e). Clearly, the facet is no longer fluorescent, but the regions immediately around the facet do fluoresce, allowing a clear fit with the optical image. The grain boundary on the bottom of the image (V) correlates with the observed fluorescence as well as the region outlined at the top left correlating with area outlined by the grain boundaries of II and III. Finally, overlaying the feature mask created from the -1.350 V/SCE image is shown in (f). The curved region outlined correlates well with the striations in VI. The small circular region next to the triangular feature is encompassed in the outlined area, and the grain boundaries III and V is also captured in the outlined region. Overall, the correlation of the fluorescence features to the electrode surface features is quite close, and, although an exact correspondence of fluorescence to surface features was not possible, a number of regions of fluorescence were 'assigned' to surface features. The potential-dependent fluorescence features seem to match well with the electrode surface structure from which desorption took place.  8.4  General Observations and Conclusions  The selective reductive desorption of a S A M from various regions of a polycrystalline Au bead was demonstrated.  One S A M coated electrode was progressively stripped of adsorbed  molecules with increasing negative potential excursions. The (111) facet is expected to have the lowest work function, therefore desorption from this face was observed to occur at the least negative potentials. While desorption from other regions of the electrode surface also occurred at this negative potential, they are most likely surface regions with similar work functions or surface energies as the (111) facet. Diffusion of the desorbed thiolates rendered these molecules unavailable for re-adsorption onto the void region. The selective removal of the S A M was most clearly observed through image analysis when comparing the distribution of desorbed fluorescing species. The structure of the desorbed molecules were compared to surface features of the Au bead electrode, and a correlation observed. The reductive desorption of a S A M from the electrode surface can be parallelled to the desorption of physisorbed surfactants. Results from Chapters 5 and 6 have demonstrated that the physically adsorbed surfactants are displaced from the electrode surface at negative potentials due 173  to the potential control over interfacial tension. This is similar to the method by which the SAM is desorbed although this process requires the transfer on electrons. Since the work function and energies of the various surface regions of the Au bead electrode are different, the electric variable is capable of changing the interfacial energy such that only portions of the thiol are removed at a given potential. For the physically adsorbed surfactants studied in Chapters 5 and 6 were investigated on a single crystal with only one face exposed, the desorbed molecules, although aggregated, are removed from the electrode surface uniformly. The re-adsorption of these water insoluble molecules onto the Au( 111) electrode was complete upon cycling in the positive direction. This occurs because the desorbed molecules remain in the vicinity of the electrode surface from adhesive interactions between the desorbed surfactant and structured water at the electrode. While the desorbed thiolates exist near the electrode surface, the re-adsorption of the S A M coated electrode was not observed because of diffusion of desorbed thiolates. The oxidative re-adsorption could proceed along the same lines as the physically adsorbed molecules if this diffusion were not occurring. For long-chained thiol molecules that do not diffuse as readily, oxidative re-adsorption observed can be explained through the existence of desorbed molecules that remain near the electrode surface. While this was not directly shown in the results presented, the slight diffusion of desorbed thiol molecules made the observation of selective desorption possible.  174  Chapter 9 9  Summary and Conclusions The experimental results presented in this thesis describe a process by which the physisorbed  surfactants undergo repeatable desorption and re-adsorption onto the Au(l 11) surface. This chapter combines the results to form a general mechanism for this process. This general picture is linked with the reductive desorption of chemisorbed SAMs presented in Chapter 8. The chapter begins with a summary of the electro-fluorescence results followed by a description of the general mechanism by which the surfactants undergo the desorption/re-adsorption phenomena.  9.1  Summary The in situ electro-fluorescence characterization of organic modified electrode surfaces was  fully realized. We have developed a system where electrochemical techniques are combined with fluorescence microscopy. This technique was found to be capable of monitoring the desorption/readsorption process of surfactants through the addition of a small amount of fluorescent dye molecules.  9.1.1  Electrochemical  Characterization  of Physisorbed  Alcohols  Initially, C180H and OLA were studied using electrochemical methods in both pure and mixed proportions. Both of the organic compounds were observed to respond to electric potential in a very similar manner. At potentials close to the pzc, the molecules were adsorbed onto the metal surface lowering the capacitance of the interface. At the extreme negative end of potentials, the molecules were displaced from the electrode surface increasing the capacitance to values coincident with those for a water covered interface. This demonstrated that both organic molecules were replaced by adsorbed water molecules at the negative potential limit. During the positive potential scan, the organic molecules re-adsorbed onto the metal surface re-establishing a low value of capacitance. This desorption/re-adsorption was attributed to the control over interfacial tension by the electrical variable. The potential range over which the desorption/re-adsorption occurred was between 0.150 V/SCE and -0.800 V/SCE. Only slight differences between the capacitance curves  175  of pure OLA and pure C180H were observed. The minimum capacitance for adsorbed OLA was slightly lower than that for C180H which indicated that OLA forms a more well-ordered and defect free monolayer on the electrode surface. This was further confirmed through the slight negative shift in potentials where the OLA molecules undergo phase transitions manifested as pseudo-capacitance peaks. Mixtures of the two alcohols adsorbed on the electrode surface showed features in the capacitance curves that were characteristic of both pure OLA and C180H and the measured film pressure increased with the inclusion of OLA. Some minor non-ideal mixing was observed. Layers of C180H, O L A and 25 mol% OLA/C 180H were reproducible and reliably adsorbed onto the electrode surface. An estimation of the degree of disturbance that the DilC 18(5) inflicted on the adsorbed layer of C180H was measured electrochemically. The most notable change was in the decrease ofthe maximum film pressure with increasing dye content. The results obtained in these investigations showed that small concentrations of dye would accurately represent the potential induced behaviour of the adsorbed surfactants without significant disruption to the electrochemical response.  9.1.2 Electro-fluorescence Characterization of Physisorbed Alcohols Initial electro-fluorescence studies were conducted on a 3 mol% DiIC18(5)/C180H layer adsorbed onto A u ( l l l ) and the desorption/re-adsorption process revealed.  When adsorbed,  fluorescence intensity was reduced due to metal-mediated quenching. The quenching efficiency became less during the negative sweep in potentials due to the separation of the surfactant from the metal surface. When desorbed, the surfactant molecules were directly observed to exist in an aggregated structure that did not diffuse from the region near the electrode surface. Furthermore, a second imaging investigation on the same organic coated electrode and on the same region ofthe metal surface revealed a desorbed structure that was identical with the first study. This suggests that the mechanism by which the molecules are displaced from the metal surface is reproducible for a given adsorbed surfactant and does not manipulate or perturb the organic layer on the metal surface. These were the first results revealing the structure and morphology of the desorbed layer and confirmed that the desorbed molecules exist near the electrode surface and remain in this vicinity allowing for the re-adsorption event. On the positive scan of potentials, the fluorescence intensity 176  was slightly greater than that for the negative scan due to the different mechanisms involved in the desorption and re-adsorption process. Re-adsorption proceeded by way of contact adsorption of the desorbed aggregates followed by spreading to reform the adsorbed layer. The region of contact adsorption was directly confirmed by a small number of features present within the fluorescence images at these potentials. This was clear evidence of contact adsorption since some molecules would exist far enough away from the metal surface to avoid total fluorescence quenching. The observation of fluorescence intensity decreased below zero for the re-adsorbed layer was attributed to the continual photobleaching of crystallites present in the adsorbed layer that were far enough away from the electrode surface to avoid fluorescence quenching by the metal. This was the first observation that the adsorbed layers are not fully uniform as previously thought. Thus, the ability of the electro-fluorescence technique to probe the local character of the adsorbed surfactant has shed new light into the transfer of these molecules from a floating layer formed at the equilibrium spreading pressure on the electrolyte surface. The creation of desorbed aggregates was dependent on the potential perturbation used to displace the adsorbed organic layer from the electrode surface. When the interface was exposed to a large potential stepping procedure, the layer was disrupted and slightly fluorescent even at the adsorption potentials. Furthermore, the images revealed desorbed features that were not present in the slow potential sweep investigations. This showed that the mechanism by which the organic molecules are desorbed can be modified with the potential perturbation. To confirm the ability of DilC 18(5) to report on C180H, the surfactant was mixed two dye molecules and the fluorescence from each dye was measured separately on the same layer. The desorbed structure was unchanged when comparing the fluorescence images measured for each dye. Both DilC 18(5) and 5-octadecanoylaminofluorescein reported a heterogeneous and aggregated desorbed structure. This demonstrated that the fluorescent dye molecules used were suitable for accurately reporting on the potential induced response of the organic molecules. Mixing OLA with the two dye molecules allowed for an electro-fluorescence investigation of the O L A / A u ( l l l ) interface and the results differed only slightly from the observed behaviour of C180H due to the similarity of the two alcohols. The OLA molecules desorb and re-adsorb in a similar fashion to C180H with only a slight negative shift in potentials. The influence of the fluidity of OLA on the  177  fluorescence response was most notably observed with an efficient decrease in fluorescence intensity for the DilC 18(5) dye at desorption potentials. This decrease in fluorescence intensity occurred more efficiently than for DilC 18(5) in C180H. Since the decrease in fluorescence intensity was not observed for 5-octadecanoylaminofluorescein in the same layer, the decrease in fluorescence intensity for DilC 18(5) was proposed not to be the result of a slow drift of the desorbed molecules back to the metal surface where metal-mediated fluorescence quenching would dominate. The ability of the electro-fluorescence technique to distinguish between OLA and C180H was tested in the investigations of OLA/C 180H mixed monolayers.  The electrochemical  characterization in Chapter 5 determined that the molecules were mostly miscible. Since the desorbed structures of OLA and C180H were very similar, the electro-fluorescence method could not unambiguously separate the response of one alcohol from the other. This agrees with the miscible character found by the electrochemical results. Some miscible character was observed when measuring the fluorescence intensity of DilC 18(5) in the mixed layer. The fluorescence intensity with potential showed an efficient decrease in the measured intensity at the negative limit (characteristic of OLA) that was shifted slightly in the positive direction (characteristic of C180H). This was attributed to the unique photophysical response of DilC 18(5) in the fluid OLA phase. This was also consistent with the observed miscibility suggesting that the presence of OLA in C180H makes the mixed monolayer more fluid. This observation was only made available since the DilC 18(5) response is unique when desorbed. All of the imaging investigations prove that the developed technique of electro-fluorescence microscopy is useful for monitoring the adsorption/desorption process of surfactants adsorbed onto the electrode surface. This technique has extended the description of these unique systems since it is not restricted to measuring average properties of the interface. We now know that the repeatable nature of the desorption/re-adsorption process is due to the ability of the desorbed molecules to remain in the vicinity of the electrode surface. We also have shown that these molecules transferred from a floating monolayer at the ESP are slightly defective and contain small regions of crystallites transferred from the gas/solution interface during deposition. This result was not previously observed because many methods of characterization including electrochemistry, PM-FTIRRAS, and scattered light can only probe average properties of the adsorbed layer. Therefore, because these  178  crystallites make up less than 10% of the adsorbed surfactant, their presence may go unnoticed when using typical probes used in characterizing the interface. Using brightfield illumination, surface irregularities were shown not to influence the structure desorbed structure and therefore the desorption process. While small defects exist on the surface of the electrode due to mechanical and electrochemical polishing, the surface is considered energetically homogeneous and therefore desorption does not preferentially occur at defect sites.  9.1.3  Electrochemical  Control  over Radiative  Decay  At the negative potential limit of the imaging/potential scan investigations, a unique decay of fluorescence intensity of the DilC 18(5) molecules spurred further investigations into the method by which the decay occurs. Initial characterization was conducted on an adsorbed layer containing 3 mol% DilC 18(5) in C180H. When desorbed into the electrolyte, a rapid decrease in fluorescence intensity was observed with time. Unexpectedly, the fluorescence intensity was reproducibly recovered to values greater than 80% following a subsequent adsorption/desorption cycle. This decay occurred more rapidly for DilC 18(5) in OLA consistent with the imaging/potential scan investigations. Conversely, a recovery of fluorescence intensity for 5-octadecanoylaminofluorescein was not observed proving that the decay was unique to DilC 18(5) and not from a slow drift of the desorbed surfactant back to the electrode surface with time. The fluorescence decay of DilC 18(5) was attributed to two parallel processed which were irreversible photobleaching and reversible fluorophore aggregation when separated from the metal. While not shown spectroscopically, it was proposed that the dye aggregate was a non-absorbing complex that became destabilized when readsorbed onto the metal surface. The monomer was capable of fluorescing at the initiation of a new desorption event accounting for the observed recovery in fluorescence. A kinetic model was used to quantify the decay and recovery process. The dye aggregation rate constant was roughly 10-40 times greater than the rate constant for the photodecomposition process. The fluorophore aggregation process was a dominant factor in the observed decay and recovery consistent with experiment. When DilC 18(5) was present in the fluid OLA phase, the observed decay and recovery was at least an order of magnitude more rapid in comparison to DilC 18(5) in C180H. This again was consistent with the image/potential scan experiment and  179  attributed to the fluidity of the OLA phase allowing the DilC 18(5) molecules to aggregate more quickly. This explained the rapid decrease in fluorescence intensity at negative potentials for the DiIC18(5)/OLA layers.  The recovery of fluorescence at the onset of the positive scan for  DiIC18(5)/OLA could then be credited to the de-stabilization of the aggregated complex due to the initial stages of the re-adsorption process.  These unique investigations demonstrated the  electrochemical control over radiative characteristics of a fluorophore. Of interest to the research in this laboratory, was the determination that the desorbed molecules did not diffuse back to the electrode surface with time. This has for some time been an open question for these desorbed layers. This was further evidence that the adsorbed water molecules at the negative potential limit were a rigid species.  9.1.4 Selective Reductive Desorption of a SAM Electro-fluorescence microscopy was used to show that the reductive desorption of a thiol bearing a fluorescent moiety occurred from selective regions of a polycrystalline Au bead electrode at different negative potential limits. The reductive desorption occurred at the least negative potentials from a facet determined to be (111). At incrementally more negative potential limits, the reductive desorption occurred from different surface regions of the bead electrode. Diffusion ofthe desorbed thiolates rendered these molecules unavailable for re-adsorption through an oxidative process. While the A u ( l l l ) surface for the physically adsorbed surfactants was energetically homogeneous, the pzc and work function of the various surface regions of the Au bead electrode are different resulting in the ability to selectively remove the thiol from various surface regions. The selective removal of the S A M was shown using image analysis through a comparison of the desorbed structure and distribution of surface features on the Au bead.  9.2  Proposed Mechanism for the Desorption and Re-adsorption Events A general picture describing the desorption/re-adsorption of insoluble molecules from/to the  electrode surface can be proposed based on the presented data. A unified picture ofthe process by which the molecules undergo these events is given in Figure 9-1. This figure depicts the desorption/re-adsorption of DilC 18(5)/C180H from the first investigation in Chapter 6. In this case 180  Reference  20 A CM  £ o  LL  a.  10  o  c  D  A.  Electrode Surfactant  —  -10 -1.0  -0.8  -0.6  -0.4  -0.2  0.0  0.2  0.4  E / V vs S C E  Figure 9-1 Proposed mechanism for the desorption/re-adsorption of organic molecules from/to the electrode surface. Fluorescence images taken from experiment are placed at unique potential regions around a typical C180H/Au(l 11) capacitance curve. The fluorescence images are cropped and enlarged according to the reference image. An enlarged schematic depiction of the interface is shown beneath the fluorescent images. In the schematic representation, the shading or gray value represent the fluorescence intensity. Black represents non-fluorescent and white represents fluorescent.  181  the images are cropped to expand a region of interest (compare reference image in Figure 9-1) and separately contrast enhanced to reveal more features. The fluorescence images are placed around a typical capacitance plot at unique potentials. Beneath each image is a schematic representation of the surfactant molecules which are not proposed to correlate exactly with the fluorescence images. In the schematic representation, the shading or grey value of the surfactant represents the fluorescence intensity. Only this C180H/Au(l 11) system is presented in the figure but the mechanism is generally applicable to the bulk of results presented in this thesis. At potentials, close to the pzc for Au(l 11) in contact with KC10 (0.255 V/SCE), the molecules strongly adsorbed onto 4  the electrode surface and the fluorescence is quenched (image A). When scanning potential in the negative direction, the adsorbed layer became disrupted (porous) through the penetration of water into the adsorbed layer. This resulted in the pseudo-capacitance peaks during the creation of these defects. The electro-fluorescence characterization of this process was not possible due to quenching. Even the contrast enhanced image at this potential (image B) does not reveal any structure demonstrating that these defects occur in the adsorbed layer. Therefore, the increase in capacitance at this potential is due to water penetrating into the interface. The defective layer is schematically presented in image B. The creation of these pores in the adsorbed layer was a crucial part to the initial stages of desorption which maybe due to a de-wetting process described by Young's equation. As water continues to collect and organize on the metal surface during the negative potential scan, the organic molecules become organized in aggregates and are displaced from the electrode surface. Image C in reveal the onset of this desorption where the molecules are still close enough to the metal surface to be slightly quenched. At further negative potentials, the organic molecules are completely desorbed from the electrode surface resulting in the increased fluorescence (image D). These desorbed were proposed to exist near enough to the metal to experience some fluorescence quenching (Chapter 7). This was consistent with the fluorescence intensity measured from the stepped surface in Chapter 6. Therefore, the heterogeneous features were brought out by the contrast in the desorbed aggregates near the electrode surface due to the gradient of quenching through features (schematic D). Close inspection of images C and D show a slight differences in structure (outline in C and D) suggesting that molecules are not necessarily removed from the electrode surface at the same time. The desorbed features in D are consistent with E because the molecules 182  do not diffuse or drift back near the electrode with time. Therefore, the decrease in fluorescence intensity (schematic E), at the negative potential limit results from photobleaching or non-fluorescent dye aggregation. As the molecules become contact adsorbed as shown in F some differences in image structure are observed when compared to the desorbed image E . This suggests that some material is far enough away from the electrode surface to avoid total fluorescence quenching and confirms a region of contact adsorption. Once in the contact adsorbed state, the spreading of the molecule may proceeded via wetting as described by Young's equation. This is a general mechanism that happens reproducibly for a given adsorbed layer. The results are paralleled to those observed for the OLA/Au(l 11) and OLA/C180H/Au(l 11) mixtures, and the slight negative shift in potential with increasing OLA content can be attributed to a slight negative shift in the electrocapillary curve for these adsorbed layers. This would result in more negative potentials required to induce defects and desorption. While the reductive desorption of a S A M coated electrode proceeds via a charge transfer reaction, this proposed mechanism may still apply to the re-adsorption of the thiol molecules through an oxidative process. While it is not clear whether the faradaic transfer of electrons occurs at the E  des  potential, the removal of these species from the electrode is driven by the electrocapillary equation. When electron transfer has occurred, the thiolate molecules may still be physisorbed onto the metal surface lowering the interfacial tension. When the potential is continually scanned in the negative direction, the desorption of this physisorbed molecule can be thought to occur in the same manner as depicted for the physically adsorbed surfactants which is driven by changes in the interfacial tension.  9.3  Suggestions for Future Study and Modification to the Developed Technique The results obtained with the electro-fluorescence method has greatly extended our  knowledge of the electric potential driven changes in adsorbed organic layers. In the present setup, some experiments can be conducted without significant changes to the apparatus. For example, the fusion of uni-lamellar vesicles with a previously adsorbed layer can be directly observed on solid electrodes similar to that on Hg electrodes [110]. This may require the use of a fluid adsorbed layer for which OLA or mixed OLA/C180H is a candidate. These investigations would not require 183  significant modification to the existing setup and would expand the functionality of the technique. However, through the course of this thesis project, limitations in the developed technique were noted and some new phenomena are in need of investigation. This requires further experimentation that will require some modification to the existing setup. A large body of work could be focussed on determining an accurate measurement of the distance separating the molecules and the metal at desorption potentials. In literature, the typical method for separating the fluorescent molecules from the metal surface is done through the use of non-fluorescent spacer molecules and fluorescence lifetimes are measured rather than fluorescence intensity [52,65,68,70,71,114]. The electro-fluorescence technique could be modified to incorporate a pulsed laser and time-correlated photon counting system for measuring the fluorescence lifetime rather than the fluorescence intensity. Initial investigations would require calibrating a typical lifetime versus distance curve of fluorescent molecules separated from a thick metal through the use of L B deposited non-fluorescent spacer molecules. The stepped template described in Chapter 6 could also be used for this calibration. This calibration curve itself would be a challenge and many experiments would be required to obtain an accurate calibration curve. Once this is achieved, the surfactant/electrode interface could be characterized using lifetime versus potential investigations. An accurate determination of the distance separating the surfactant from the metal would give further insight into the structure of the electric double layer. This is not the only method to determine the distance separating the desorbed molecules from the metal. Recently an excellent review has been published on the interaction of thin films with electromagnetic waves in a technique known as surface plasmon resonance spectroscopy ( S P R ) [222]. The technique of S P R involves the total internal reflection of light at a boundary between two media of different optical properties. When light moves from a medium of high refractive index (glass hemi-spherical prism, n = 1.515) to a low refractive index (water n = 1.333), refraction will occur according to Snells Law. At some critical angle of incidence from the normal, the light can be totally reflected from the second medium with a small amount of energy exponentially dissipating into solution as an evanescent wave. If a thin film of metal is present at the second medium (Au (50nm)), the reflected light at the critical angle drops to a minimum due to the incoming radiation coupling with the collective plasma oscillations of the nearly free electron gas 184  in a metal. The coupling of electromagnetic modes then results in the absorption of incoming radiation at certain angles. The angle at which the minimum in reflected light occurs is then specific to the dielectric medium at the interface, and any changes that occur at the interface will be observed as a shift in the angle of minimum reflectance. Calculations based on Corn's program [223] indicate that a change in the angle of minimum reflectance of 4.7x10" degrees/A of separation from the 4  electrode surface is expected for a 25 A thick layer of organic molecules. This sensitivity can easily be achieved using Tao's highly sensitive S P R design [224] based on the same principles of a detector in atomic force microscopy  (AFM).  This design is miniaturized and will fit in the  developed electrochemical cell. The slight change in the angle of minimum reflectance with increasing dye content will give a more accurate estimate of the average separation. The microscope can be used in these experiments for the measurement of Total Internal Reflection Fluorescence. Since the plasmon resonance produces an electric field that decays exponentially into solution, fluorophores near the electrode can be excited. There are still effects of fluorescence quenching from the metal but the evanescent wave will penetrate to those molecules near but not on the electrode. This will give rise to a very narrow region of fluorescence activity near the electrode surface. One important factor that must be addressed in any of these investigations is the appropriate choice of fluorescent dye. Since unique photophysical properties can occur in these systems, the changing character of the excited state species could bias the results. Thus, the most accurate measurement for these investigations must combine the use of several dye molecules. As such, both of the above suggestions would each encompass a large body of work. Experiments outside the electro-fluorescence method are also required. The fluorescence decay and recovery described in Chapter 7 did not spectroscopically determine a non-absorbing dye complex. We had spent a lot of time trying to measure this using electro-reflectance spectroscopy but any appreciable signal was buried in noise. Typically, an electro-reflectance investigation involves directing monochromatic toward the electrode and measuring the reflected light as a function of potential. In our experiments, the reflected light was measured at the adsorption and desorption potentials with a time scale similar to the fluorescence imaging decay and recovery investigations. Therefore, 5 spectra were acquired at adsorption potentials for five minutes followed  185  by 20 spectra acquired at desorption potentials over 20 minutes. This was our major limitation to using this technique. In many electro-reflectance investigations, the potential is chosen such that there are dramatic changes in the structure of the interface, for example, oxide or reduction of the interface. If the interface is robust (as in the adsorption of CI" on Au(l 11)) high frequencies can be used and lock-in amplifiers greatly increase the signal to noise ratio. Many of these signal enhancing methods cannot be used to measure the aggregation of the dye molecules when desorbed because of the sensitivity of the system to the electrical perturbation as shown in Chapter 6. Therefore, measuring changes in reflected light from the desorbed molecules is restricted to Beer's law for close to monolayer thick desorbed layer containing a small concentration (3 to 5 mol%) dye. As such, the changes in reflected light (absorbed light) are very small and a stable light source is crucial to measure the signal over the 20 minutes spent at desorption. We spent a lot of time making some adjustments to a previous scattered light setup [16] to measure the changes in reflectivity. During these investigations, we did observe a change in the reflectivity between adsorption and desorption potentials due to the dramatic changes at the interface. However, only a small change in the absorbance at desorption potentials with time was observed. The signal however, required a significant amount of averaging and was only slightly above the noise level. This does suggest that the non-absorbing hypothesis is reasonable but a conclusive argument for this requires more electroreflectance investigations. This may be achieved by using a more stable light source and more sensitive detectors.  186  References 1.  D.L. Allara and R.G. Nuzzo, Langmuir, 1 (1985) 45.  2.  J.D. Swalen, Journal ofMolecular Electronics, 2 (1986) 155.  3.  J. Swalen, D. Allara, D. Andrade, E.A. Chandross, S. Garoff, J. Israelachvili, T.J. McCarthy, R. Murray, R.F. Pease, J.F. Rabolt, K.J. Wynne and H. Yu, Langmuir, 3 (1987) 932.  4.  J.D. Swalen, Surfactants in Solution, 11 (1991) 41.  5.  A. Ulman, Molecular Solid State, 2 (1999) 1.  6.  H. Kuramitz, K. Sugawara and S. Tanaka, Current Topics in Analytical Chemistry, 2 (2001) 1.  7.  S. Flink, F.C.J.M. van Veggel and D.N. Reinhoudt, Sensors Update, 8 (2001) 3.  8.  N.K. Chaki and K. Vijayamohanan, Biosensors & Bioelectronics, 17 (2002) 1.  9.  K. Kerman, Y. Morita, Y. Takamura, M . Ozsoz and E. Tamiya, Anal Chim. Acta. ,510 (2004) 169.  10.  C.J. Zhong, J. Zak and M.D. Porter, J. Electroanal. Chem., 421 (1997) 9.  11.  D. Bizzotto (1996): Characterization of the adsorption of insoluble surfactants onto an electrified interface. Ph.D. Dissertation, University of Guelph, Guelph, ON. 238 p.  12.  D. Bizzotto, J. Noel and J. Lipkowski, Thin Solid Films, 248 (1993) 69.  13.  D. Bizzotto, J.J. Noel and J. Lipkowski, J. Electroanal. Chem., 369 (1994) 259.  14.  D. Bizzotto and J. Lipkowski, Prog. Surf. Sci., 50 (1995) 237.  15.  J. Lipkowski, G. Szymanski, A. Chen, I. Burgess, D. Bizzotto, X . Cai, M . Hoon-Khosla and C. Jeffrey, NATO Science Series, Series C: Mathematical and Physical Sciences, 546 (2000) 155.  16.  Y.G. Yang and D. Bizzotto, J. Electroanal. Chem., 500 (2001) 408.  17.  D. Bizzotto and B. Pettinger, Langmuir, 15 (1999) 8309.  18.  S. Krasne, Biophys. J., 30 (1980) 415. 187  19.  S. Krasne, Biophys. J., 30 (1980) 441.  20.  P.J. Sims, A.S. Waggoner, C. H. Wang and J.F. Hoffman, Biochemistry, 13 (1974) 3315.  21.  D.E. Wolf in L . M . Loew(Ed.), Spectroscopic Membrane Probes, Vol. 1, CRC Press, Boca Raton, Florida, 1988.  22.  A.J. Bard and L.R. Faulkner, Electrochemical Methods. Fundamentals and Applications, John Wiley and Sons, New York, 2001.  23.  D.C. Grahame, Chem. Rev., 41 (1947) 441.  24.  T. Langkau and H. Baltruschat, Elec. Acta., 47 (2002) 1595.  25.  J. Goodisman, Electrochemistry: Theoretical Foundations, John Wiley and Sons, New York, 1987.  26.  W. Schmickler, Chemical Reviews (Washington, D. C), 96 (1996) 3177.  27.  J.O'M. Bockris and A.K.N. Reddy, Modem Electrochemistry. Interdisciplinary Area, Plenum Press, New York, 1970.  28.  R. Smoluchowski, Physical Review, 60 (1941) 661.  29.  G. Attard and C. Barnes, Surfaces, Oxford University Press, New York, 1998.  30.  J. Lecoeur, J. Andro and R. Parsons, Surf. Sci., 114 (1982) 320.  31.  I.N. Levine, Physical Chemistry, McGraw-Hill, New York, 1995.  32.  J. Lipkowski and L. Stolberg in J. Lipkowski and P.N. Ross(Eds.), Adsorption of Molecules at Metal Electrodes, VCH, New York, 1992, p. 171.  33.  A. van Den Eeden, J. Sluyters and J. van Lenthe, J. Electroanal. Chem., 171 (1984) 195.  34.  R. De Levie and A. Vogt, J. Electroanal. Chem., 278 (1990) 25.  35.  T. Pajkossy, J. Electroanal. Chem., 364 (1994) 111.  36.  N.A. Hampson, C. Lazarides, A.C. Tyrrell, R. Leek and J.P.G. Fair, Surface Technology, 15 (1982) 167.  37.  D.M. Mohilner in A.J. Bard(Ed.), Electroanalytical Chemistry, Vol. 1, Marcel Dekker, New 188  An Introduction to an  York, 1966, p. 241. 38.  B.B. Dmamaskin, O.A. Petrii and V.V. Batrakov, Adsorption of Organic Compounds on Electrodes, Plenum Press, New York, 1971.  39.  J. Richer and J. Lipkowski, J. Electrochem. Soc, 133 (1986) 121.  40.  J. Richer, L. Stolberg and J. Lipkowski, Langmuir, 2 (1986) 630.  41.  J. Richer and J. Lipkowski, J. Electroanal. Chem., 251 (1988) 217.  42.  L. Stolberg, J. Richer, D. Irish and J. Lipkowski, J. Electroanal. Chem., 207 (1986) 213.  43.  L. Stolberg, J. Lipkowski and D. Irish, J. Electroanal. Chem., 238 (1987) 333.  44.  I. Langmuir, J. Amer. Chem. Soc, 39 (1917) 1848.  45.  K.B. Blodgett, J. Amer. Chem. Soc, 57 (1935) 1007.  46.  G.L. Gains, Insoluble Monolayers at Liquid-Gas Interfaces, Interscience Publishers, New York, 1966.  47.  J.H. Fendler, Membrane Mimetic Chemistry. Characterizations and Applications of Micelles, Microemulsions, Monolayers, Bilayers, Vesicles, Host-Guest Systems, and Polyions, John Wiley and Sons, New York, 1982.  48.  N. Tillman, A. Ulman and T. Penner, Langmuir, 5 (1989) 101.  49.  G. Roberts (Ed.), Langmuir-Blodgett Films, Plenum Press, New York, 1990.  50.  A. Ulman, An Introduction to Ultrathin Organic Films, Academic Press, Boston, 1991.  51.  A. Ulman (Ed.), Characterization of Organic Thin Films, Manning Publications, Greenwich, 1995.  52.  I. Langmuir and V.J. Schaefer, J. Amer. Chem. Soc, 60 (1938) 1351.  53.  H. Kuhn, D. Mobius and H. Biicher in A . Weissberger and B.W. Rossiter(Eds.), Physical Methods of Chemistry. Part IIIB Optical, Spectroscopic, and Radioactivity Methods, Vol. 1, Wiley-Interscience, New York, 1972, p. 577.  54.  F.W.D. Rost, Fluorescence Microscopy, Cambridge University Press, New York, 1992.  189  55.  H. Buecher and H. Kuhn, Chemical Physics Letters, 6 (1970) 183.  56.  V. Czikkely, G. Dreizler, H.D. Foersterling, H. Kuhn, J. Sondermann, P. Tillmann and J. Wiegand, ZeitschriftfixerNaturforschung, Teil A: Astrophysik, Physik und Physikalische Chemie, 24(1969) 1821.  57.  E.E. Jelley, Nature, 138 (1936) 1009.  58.  V. Czikkely, H.D. Foersterling and H. Kuhn, Chem. Phys. Lett, 6 (1970) 207.  59.  L. Wolthaus, A. Schaper and D. Moebius, Chemical Physics Letters, 225 (1994) 322.  60.  M . Kasha, H.R. Rawls and M.A. El-Bayoumi, Pure Appl. Chem., 11 (1965) 371.  61.  T. Forster, Naturwiss., 33 (1946) 166.  62.  H. Kuhn, J. Chem. Phys., 53 (1970) 101.  63.  R.R. Chance, A.H. Miller, A. Prock and R. Silbey, J. Chem. Phys., 63 (1975) 1589.  64.  S. Astilean and W.L. Barnes, Applied Physics B: Lasers and Optics, 75 (2002) 591.  65.  R.R. Chance, A. Prock and R. Silbey in I. Prigogine(Ed.), Advances in Chemical Physics, Vol. 37, Interscience Publishers, New York, 1978, p. 1.  66.  G.W. Ford and W.H. Weber, Physics Reports, 113 (1984) 195.  67.  W.L. Barnes, Journal ofModern Optics, 45 (1998) 661.  68.  K.H. Drexhage, H. Kuhn and F.P. Schafer, Ber. Bunsenges. Phys. Chem., 72 (1968) 329.  69.  K.H. Drexhage, J. Lumin., 1/2 (1970) 693.  70.  K.H. Drexhage, Sci. Am., 222 (1970) 108.  71.  K.H. Drexhage in E. WolffEd.), Progress in Optics, Vol. 12, Interscience, Amsterdam, 1974, p. 163.  72.  W.H. Weber and C F . Eagen, Optics Letters, 4 (1979) 236.  73.  I. Pockrand, A. Brillante and D. Moebius, Chemical Physics Letters, 69 (1980) 499.  74.  O.S. Wolfbeis, M . Bohmer, A. Durkop, J. Enderlein, M . Gruber, I. Klimant, C. Drause, J. 190  Durner, G. Liebsch, Z. Lin, B. Oswald and M . Wu in R. Kraayenhof, A J.W.G. Visser and H.C. Gerritsen(Eds.), Fluorescence Spectroscopy, Imaging and Probes. New Tools in Chemical, Physical and Life Sciences, Springer, Berlin, 2001, p. 390. 75.  A.H. Herz, Photographic Science and Engineering, 18 (1974) 323.  76.  M.N. Slyadnev, T. Inoue, A. Harata and T. Ogawa, Colloids and Surfaces, A: Physicochemical and Engineering Aspects, 164(2000) 155.  77.  H.G. Kapitza (1994): Microscopy from the very beginning. Copywrite Carl Zeiss, Oberkochen.  78.  http://www.chroma.com/products/index.php?doPrintSet=yes&sSetID=41004.  79.  V.V. Arslanov, Advances in Colloid & Interface Science, 40 (1992) 307.  80.  G.G. Roberts, Advances in Physics, 34 (1985) 475.  81.  D.K. Schwartz, Surface Science Reports, 27 (1997) 241.  82.  A.V. Gorodetskaya and A. Frumkin, Akad. NaukSSSR, 18 (1938) 649.  83.  A.V. Gorodetskaya, Zh. Fiz. Khim., 14 (1940) 371.  84.  R. Parsons and LR. Miller, J. ColliodInt. Sci., 45 (1973) 126.  85.  LR. Miller, J. Rishpon and A. Tennenbaum, Bioelectrochem. Bioenerg., 3 (1976) 528.  86.  I.R. Miller in G. Milazzo(Ed.), Topics in Biochemistry and Bioenergetics, Vol. 4, John Wiley and Sons, New York, 1981, p. 161.  87.  A. Nelson, Marine Environmental Research, 17 (1985) 306.  88.  A. Nelson and A. Benton, J. Electroanal. Chem., 202 (1986) 253.  89.  A. Nelson and N . Auffret, J. Electroanal. Chem., 244 (1988) 99.  90.  F.A.M; Leermakers and A. Nelson, J. Electroanal. Chem., 278 (1990) 53.  91.  A. Nelson and H.P. van Leeuwen, J. Electroanal. Chem., 273 (1989) 183.  92.  A. Nelson and H.P. van Leeuwen, J. Electroanal. Chem., 273 (1989) 201.  93.  A. Nelson, N . Auffret and J. Readman, Anal. Chim. Acta., 207 (1988) 47. 191  94.  A. Nelson, N. Auffret and J. Borlakoglu, Biochim. Biophys. Acta., 1021 (1990) 205.  95.  A. Nelson, Biophysical Journal., 80 (2001) 2694.  96.  . A. Nelson and D. Bizzotto, Langmuir, 15 (1999) 7031.  97.  A. Nelson, Langmuir, 12 (1996) 2058.  98.  A. Nelson, J. Chem. Soc, Faraday Trans., 89 (1993) 2799.  99.  A. Nelson, Anal. Proc. (London), 28 (1991) 64.  100.  A. Nelson, J. Electroanal. Chem. Interfacial Electrochem., 303 (1991) 221.  101.  M.R. Moncelli, L. Becucci, R. Herrero and R. Guidelli, J. Phys. Chem., 99 (1995) 9940.  102.  M.R. Moncelli and L. Becucci, J. Electroanal. Chem., 385 (1995) 183.  103.  L. Becucci, M.R. Moncelli and R. Guidelli, Biophys. J., 82 (2002) 852.  104.  R. Guidelli, L. Becucci, A. Dolfi, M.R. Moncelli and F. Tadini Buoninsegni, Solid State Ion., 150(2002)13.  105.  M.R. Moncelli, L. Becucci and S.M. Schiller, Bioelectrochemistry, 63 (2004) 161.  106.  D. Bizzotto and A. Nelson, Langmuir, 14 (1998) 6269.  107.  J.O. Agak, R. Stoodley, U. Retter and D. Bizzotto, J. Electroanal. Chem., 562(2) (2004) 135.  108.  R. Stoodley, J. Shepherd, K . M . Wasan andD. Bizzotto, Biochimica etBiophysica Acta, 1564 (2002) 289.  109.  V. Stauffer, R. Stoodley, J.O. Agak and D. Bizzotto, J. Electroanal. Chem., 516 (2001) 73.  110.  R. Stoodley and D. Bizzotto, Analyst, 128 (2003) 552.  111.  I. Zawisza, I. Burgess, G. Szymanski, J. Lipkowski, J. Majewski and S. Satija, Elec. Acta., 49 (2004)3651.  112.  I. Zawisza, X . Cai, V. Zamlynny, I. Burgess, J. Majewski, G. Szymanski and J. Lipkowski, Polish Journal of Chemistry, 78 (2004) 1165.  113.  J.R. Lakowicz, Y . Sheh, S. DAuria, J. Malicka, J. Fang, Z. Gryczynski and I. Gryczynski, 192  Analytical Biochemistry, 301 (2002) 261. 114.  H. Kuhn and D. Moebius in B.W. Rossiter and R.C. Baetzold(Eds.), Physical Methods of Chemistry, Vol. IXB, Wiley Interscience, New York, 1993, p. 375.  115.  J.R. Lakowicz, Anal. Biochem., 298 (2001) 1.  116.  J.R. Lakowicz, J. Malicka, I. Gryczynski, Z. Gryczynski and C D . Geddes, Journal ofPhysics D: Applied Physics, 36 (2003) R240.  117.  R.K. Smith, P.A. Lewis and P.S. Weiss, Prog. Surf. Sci., 75 (2004) 1.  118.  J.M. Tour, Accounts of Chemical Research, 33 (2000) 791.  119.  J. Telegdi, T. Rigo and E. Kalman, Corrosion Engineering, Science and Technology, 39 (2004) 65.  120.  C M . Whelan, M . Kinsella, L. Carbonell, H.M. Ho and K. Maex, Microelectronic Engineering, 70(2003)551.  121.  S.G. Perm, L. He and M.J. Natan, Current Opinion in Chemical Biology, 7 (2003) 609.  122.  N . Battaglini, H. Klein, P. Dumas, C. Moustrou and A. Samat, Applied Surface Science, 212213 (2003)481.  123.  H. Klein, N . Battaglini, B. Bellini and P. Dumas, Materials Science & Engineering, C: Biomimetic and Supramolecular Systems, C19 (2002) 279.  124.  M . C. Bourg, A. Badia and R.B. Lennox, J. Phys. Chem. B, 104 (2000) 6562.  125.  A. Badia, R. Back and R.B. Lennox, Angew. Chemie Int. Ed., 33 (1994) 2332.  126.  E. Boubour and R.B. Lennox,  127.  E. Boubour and R.B. Lennox, Langmuir, 16 (2000) 7464.  128.  E. Boubour and R.B. Lennox, Langmuir, 16 (2000) 4222.  129.  F. Ma and R.B. Lennox, Langmuir, 16 (2000) 6188.  130.  R.G. Nuzzo, E.M. Korenic and L.H. Dubois, J. Chem. Phys., 93 (1990) 767.  131.  P. Fenter, P. Eisenberger and K.S. Liang, Physical Review Letters, 70 (1993) 2447.  Phys. Chem. B, 104 (2000) 9004.  193  132.  P. Fenter, A. Eberhardt, K.S. Liang and P. Eisenberger, J. Chem. Phys., 106 (1997) 1600.  133.  G.E. Poirier and E.D. Pylant, Science (Washington, D. C), 272 (1996) 1145.  134.  K.A. Peterlinz and R. Georgiadis, Langmuir, 12 (1996) 4731.  135.  M.R. Anderson, M.N. Evaniak and M . Zhang, Langmuir, 12 (1996) 2327.  136.  R. Yamada and K. Uosaki, Langmuir, 13 (1997) 5218.  137.  R. Yamada and K. Uosaki, Langmuir, 14 (1998) 855.  138.  M . Himmelhaus, F. Eisert, M . Buck and M . Grunze, J. Phys. Chem. B, 104 (2000) 576.  139.  X . Yang and S.S. Perry, Langmuir, 19 (2003) 6135.  140.  M.R. Shadnam, S.E. Kirkwood, R. Fedosejevs and A. Amirfazli, Langmuir, 20 (2004) 2667.  141.  F. Bensebaa, T.H. Ellis, A. Badia and R.B. Lennox, Langmuir, 14 (1998) 2361.  142.  M . M . Walczak, D.D. Popenoe, R.S. Deinhammer, B.D. Lamp, C. Chung and M.D. Porter, Langmuir, 7 (1991)2687.  143.  C A . Widrig, C. Chung and M.D. Porter, J. Electroanal. Chem., 310 (1991) 335.  144.  D.E. Weisshaar, B.D. Lamp and M.D. Porter, J. Amer. Chem. Soc, 114 (1992) 5860.  145.  D.E. Weisshaar, M . M . Walczak and M.D. Porter, Langmuir, 9 (1993) 323.  146.  N.L. Abbott, C.B. Gorman and G.M. Whitesides, Langmuir, 11 (1995) 16.  147.  D.F. Yang, C P . Wilde and M . Morin, Langmuir, 12 (1996) 6570.  148.  D.F. Yang and M . Morin, J. Electroanal. Chem., 429 (1997) 1.  149.  S. Imabayashi, D. Hobara, T. Kakiuchi and W. Knoll, Langmuir, 13 (1997) 4502.  150.  C. J. Zhong and M.D. Porter, J. Electroanal. Chem., 425 (1997) 147.  151.  D.F. Yang, H. Al-Maznai and M . Morin, J. Phys. Chem. B, 101 (1997) 1158.  152.  D.F. Yang, C P . Wilde and M . Morin, Langmuir, 13 (1997) 243.  194  153.  D.F. Yang and M . Morin, J. Electroanal. Chem., 441 (1998) 173.  154.  M . Byloos, H. Al-Maznai and M . Morin, J. Phys. Chem. B, 103 (1999) 6554.  155.  A. Badia, S. Arnold, V . Scheumann, M . Zizlsperger, J. Mack, G. Jung and W. Knoll, Sensors and Actuators, B: Chemical, B54 (1999) 145.  156.  T. Kawaguchi, H. Yasuda, K. Shimazu and M.D. Porter, Langmuir, 16 (2000) 9830.  157.  S. S. Wong and M.D. Porter, J. Electroanal. Chem., 485 (2000) 135.  158.  M . Byloos, H. Al-Maznai and M . Morin, J. Phys. Chem. B, 105 (2001) 5900.  159.  S. Imabayashi, D. Hobara and T. Kakiuchi, Langmuir, 17 (2001) 2560.  160.  M . Byloos, S. Rifai, H. Al-Maznai, M . Laferriere and M . Morin, Langmuir, 17 (2001) 2478.  161.  F. Loglio, M . Schweizer and D.M. Kolb, Langmuir, 19 (2003) 830.  162.  M . M . Walczak, C A . Alves, B.D. Lamp and M.D. Porter, J. Electroanal. Chem., 396 (1995) 103.  163.  K. Aoki and T. Kakiuchi, J. Electroanal. Chem.,452(1998) 187.  164.  D. Bizzotto, J.J. Noeel and J. Lipkowski, Thin Solid Films, 248 (1994) 69.  165.  T.W. Schneider and D.A. Buttry, J. Amer. Chem. Soc, 115 (1993) 12391.  166.  T. Kakiuchi, K. Sato, M . Iida, D. Hobara, S. Imabayashi and K. Niki, Langmuir, 16 (2000) 7238.  167.  H. Wano and K. Uosaki, Langmuir, 17 (2001) 8224.  168.  D. Hobara, K. Miyake, S. Imabayashi, K. Niki and T. Kakiuchi, Langmuir, 14 (1998) 3590.  169.  M.A. Schneeweiss, H. Hagenstrom, M.J. Esplandiu and D.M. Kolb, Applied Physics a (Materials Science Processing)., 69 (1999) 537.  170.  D. Hobara, T. Sasaki, S. Imabayashi and T. Kakiuchi, Langmuir, 15 (1999) 5073.  171.  T.D. Dunbar, M.T. Cygan, L.A. Bumm, G.S. McCarty, T.P. Burgin, W.A. Reinerth, L. Jones, J.J. Jackiw, J.M. Tour, P.S. Weiss and D.L. Allara, J. Phys. Chem. B, 104 (2000) 4880.  195  172.  T. Felgenhauer, H. T. Rong and M . Buck, J. Electroanal. Chem., 550-551 (2003) 309.  173.  C. Chung and M . Lee, J. Electroanal. Chem., 468 (1999) 91.  174.  S. Kuwabata, H. Kanemoto, D. Oyamatsu and H. Yoneyama, Electrochemistry (Tokyo), 67 (1999) 1254.  175.  M . Satjapipat, R. Sanedrin and F. Zhou, Langmuir, 17 (2001) 7637.  176.  D. Oyamatsu, H. Kanemoto, S. Kuwabata and H. Yoneyama, J. Electroanal. Chem., 497 (2001) 97.  177.  K. Shimazu, T. Kawaguchi and T. Isomura, J. Amer. Chem. Soc, 124 (2002) 652.  178.  B. Pauwels, G. Van Tendeloo, W. Bouwen, L. Theil Kuhn, P. Lievens, H . Lei and M . Hou, Physical Review B: Condensed Matter and Materials Physics, 62 (2000) 10383.  179.  X. Zhang and A.J. Bard, J. Phys. Chem., 92 (1988) 5566.  180.  Y.S. Obeng and A.J. Bard, Langmuir, 7 (1991) 195.  181.  F.R.F. Fan, A. Mau and A.J. Bard, Chemical Physics Letters, 116 (1985) 400.  182.  H.D. Abruna and A.J. Bard, J. Amer. Chem. Soc, 104 (1982) 2641.  183.  X . H. X u and A.J. Bard, Langmuir, 10 (1994) 2409.  184.  L. Wang, V. Silin, A.K. Gaigalas, J. Xia and G. Gebeyehu, Journal of Colloid and Interface Science, 248 (2002) 404.  185.  A.K. Gaigalas, L. Li and T. Ruzgas, Current Topics in Colloid & Interface Science, 3 (1999) 83.  186.  L. Li, T. Ruzgas and A.K. Gaigalas, Langmuir, 15 (1999) 6358.  187.  K.W. Kittredge, M.A. Fox and J.K. Whitesell, J. Phys. Chem. B, 105 (2001) 10594.  188.  S. Reese and M.A. Fox, J. Phys. Chem. B, 102 (1998) 9820.  189.  J.L. Shepherd, A . Kell, E. Chung, C.W. Sinclar, M.S. Workentin and D. Bizzotto, J. Amer. Chem. Soc, 126 (2004) 8329.  190.  J. Shepherd, Y. Yang and D. Bizzotto, J. Electroanal. Chem., 524-525 (2002) 54. 196  191.  J.L. Shepherd and D. Bizzotto, J. Phys. Chem. B, 107 (2003) 8524.  192.  D. Bizzotto, Y. Yang, J.L. Shepherd, R. Stoodley, J.O. Agak, V. Stauffer, M . Lathuilliere, A. Ahktar and E. Chung, J. Electroanal. Chem., (2004) .  193.  E. Chung, J.L. Shepherd, D. Bizzotto and M.O. Wolf, Langmuir, 20 (2004) 8270.  194.  J. Richer (1990): . Ph.D. Dissertation, University of Guelph, Guelph.  195.  J. Canullo, Y. Uchida, G. Lehmpfuhl, T. Twomey and D.M. Kolb, Surf. Sci., 188 (1987) 350.  196.  http://www.molecularprobes.com/servlets/product?item=307.  197.  http://www.molecularprobes.corn/servlets/product?item=1300.  198.  http://www.molecularprobes.corn/servlets/product?item=3822.  199.  A. Hamelin in B.E. Conway, R.E. White and J.O'M. Bockris(Eds.), Modern Aspects of Electrochemistry, Vol. 16, Plenum Press, New York, 1985, p. 1.  200.  http://www.sciscope.com/spot-rt-specs.php.  201.  L.J. van Vliet (2001): DIPLib, the http://www.ph.tn.tudelft.nl/DIPLib/DIPLib.html.  202.  S.B. Howell, Handbook of CCD Astronomy, Cambridge University Press, Cambridge, UK, 2000.  203.  D. Bizzotto, E. Wong and Y.G. Yang, J. Electroanal. Chem., 480 (2000) 233.  204.  D. Bizzotto and J. Lipkowski, Progress in Colloid & Polymer Science, 103 (1997) 201.  205.  V. Zamlynny, I. Zawisza and J. Lipkowski, Langmuir, 19 (2003) 132.  206.  T.K.L. Meyvis, S.C. De Smedt, P. Van Oostveldt and J. Demeester, Pharmaceutical Research, 16(1999) 1153.  207.  J.R. Lakowicz, Anal. Biochem., 298 (2001) 1.  208.  P. Tinnefeld, D. P. Herten and M . Sauer, Journal ofPhysical Chemistry A, 105 (2001) 7989.  209.  P. Tinnefeld, V. Buschmann, D. P. Herten, K. T. Han and M . Sauer, Single Molecules, 1 (2000) 215. 197  Delft  Image  Processing  Library.  210.  W. Cooper and N.B. Liebert, Photogr. Sci. Eng., 16 (1972) 25.  211.  D.A. Weitz, S. Garoff, C D . Hanson, T.J. Gramilaand J.I. Gersten, J. Lumin., 24-25 (1981) 83.  212.  J. Malicka, I. Gryczynski, J. Kusba, Y. Shen and J.R. Lakowicz, Biochemical and Biophysical Research Communications, 294 (2002) 886.  213.  A. Leitner, M.E. Lippitsch, S. Draxler, M . Riegler and F.R. Aussenegg, Appl. Phys. B, B36 (1985) 105.  214.  D. Moebius, Advanced Materials (Weinheim, Germany), 7 (1995) 437.  215.  R.F. Khairutdinov and N. Serpone, J. Phys. Chem. B, 101 (1997) 2602.  216.  A.A. Muenter, D.V. Brumbaugh, J. Apolito, L.A. Horn, F . C Spano and S. Mukamel, J. Phys. Chem., 96 (1992) 2783.  217.  M.L. Ferrer, F. del Monte and D. Levy, Langmuir, 19 (2003) 2782.  218.  F. del Monte and D. Levy, J. Phys. Chem. B, 103 (1999) 8080.  219.  H. von Berlepsch, C. Boettcher, A. Ouart, C. Burger, S. Daehne and S. Kirstein, J. Phys. Chem. B, 104(2000)5255.  220.  E . Nicol, A . Moussa, J. L. Habib-Jiwan and A . M . Jonas, Journal of Photochemistry and Photobiology, A: Chemistry, 167 (2004) 31.  221.  C L . Perdriel, A.J. Arvia and M . Ipohorski, Journal of Electroanalytical Chemistry and Interfacial Electrochemistry, 215 (1986) 317.  222.  W. Knoll, Annu. Rev. Phys. Chem., 49 (1998) 569.  223.  R.M. Corn (2000): N Phase Fresnel Calculation,  224.  N.J. Tao, S. Boussaad, W.L. Huang, R.A. Arechabaleta, and J. D'Agnese Review ofScientific Instruments. 70 (1999) 4656.  198  http://corninfo.ps.uci.edu/calculations.html  Appendix A l Image analysis programs used in the extraction of the average histogram and delineation of features within the images. Jeff Shepherd, Dec 26, 2004 Introduction  The programs in this appendix include the following: histogram_calc. m -main program that calls the subroutines below. read_in_image. m -reads in the images within the experiment. A median filter is applied to reduce noise in the image. -creates a 'hotspot' mask from the first and last adsorption create_mask_hist. m images in the experiment. -arranges all the images in the experiment as Al/Io create_delI_Io. m create_pallet.m -creates an image pallet based on the range of histogram values throughout the sequence of images. contrast_enhance.m -applies a linear contrast enhancement to all the images in the sequence. find_Jeatures.ni -delineates features within the images for measuring statistics, -calculates the average histogram of the images excluding hist_mean_stdev2. m zeros created from the 'hotspot' mask. -An Image Pro Plus 5 script that counts the delineated features count_features.scr within the images and outputs statistics such as number of particles and average area. Summary of histogram calc.m This program gathers the sequence of images from an electro-fluorescence experiment and extracts the average histogram of each image. Typically, the sequence of images begins at adsorption potentials, cycles to desorption and returns to adsorption. A l l images are flat field corrected by division by an I image. This is an image of the electrode surface before the introduction of the fluorescent dye. The image is taken with the filters in place and serves to correct for non-uniform illumination. Thefirstnormalized image in the sequence (adsorption image) is subtracted from all others rendering AI/I . 0  0  The program works as follows: 1. Five images are defined by the user and called upon by the read_in_image.m subroutine. These images are the I image, a dark noise image, an adsorption image, and two mask images. The dark image is an image with the same exposure time as the images in the experiment with no light reaching the detector. This image reveals the dark current within the images from the small amount of pixels that fire randomly. The adsorption image, is the first adsorption image in the experiment. The two mask images are the first and last adsorption images in the sequence and are called upon later for the creation of the 'hotspot' mask. 0  199  2.  The dark image is subtracted from the Io image. This modified image is then divided into the adsorption image for the creation of the I /I image. ads  0  3.  The 'hotspot' mask is created by calling on the create_hotspot_mask.m subroutine.  4.  All the images in the sequence are divided by the I image and have I /I subtracted from them. This is achieved in the create_delI_Io.m subroutine.  5.  An image pallet is created using the create_pallet.m subroutine.  6.  All the images in the sequence have the same linear contrast enhancement filter applied to them. This is achieved in the contrast_enhance.m subroutine.  7.  If requested by the user, the features in the sequence of images are delineated using the find_features.m subroutine.  8.  The average histogram of the image is calculated using the hist_mean_stdev2.m subroutine  0  200  ads  0  % histogram_calc.m % -9-9-9-9-9-9.S-S-9-S-2-S-9-S-  % % % % % % % o, o o.  % %  h i s t o g r a m _ c a l c . m i s a p r o g r a m w r i t t e n t o g a t h e r images f r o m a n e l e c t r o - f l u o r e s c e n c e e x p e r i m e n t and r e p r e s e n t t h e s e images a s D e l t a l / I o . The mean h i s t o g r a m o f e a c h D e l t a l / I o image i s c a l c u l a t e d a f t e r t h e a p p l i c a t i o n o f a h o t s p o t mask. The mean h i s t o g r a m d o e s n o t i n c l u d e t h e p i x e l v a l u e s o f t h e mask. T h i s r o u t i n e can a l s o d e f i n e the f e a t u r e s i n an (1) (2)  o, o g, o  % % g_ *S  (3) (4) (5)  % "o  (6)  % % % %  (7)  Q, "O  Io i s a f l u o r e s c e n t image o f t h e e l e c t r o d e b e f o r e any f l u o r o p h o r e i s d e p o s i t e d . I t s e r v e s t o c o r r e c t n o n - u n i f o r m i l l u m i n a t i o n o f t h e lamp. Dim i s a n image w i t h t h e same e x p o s u r e t i m e and g a i n as i n t h e e x p e r i m e n t w i t h o u t any l i g h t coming t o t h e camera. I t s e r v e s t o c o r r e c t f o r d a r k n o i s e due t o low l i g h t l e v e l s . Dim i s s u b t r a c t e d f r o m t h e I o i m a g e . A l l images i n one e x p e r i m e n t s e q u e n c e a r e d i v i d e d b y I o i m a g e . i . e . I ( E ) / I o where E i s a v a r i a b l e p o t e n t i a l . A l l t h e I ( E ) / I o images have I ( E a d s ) / I o s u b t r a c t e d f r o m them t o c r e a t e D e l t a l / I o . I ( E a d s ) i s a n image a t a d s o r p t i o n p o t e n t i a l s . Some a d s o r p t i o n images have i n t e n s e f l u o r e s c e n c e f e a t u r e s known a s 'hot s p o t s ' . T h e s e h o t s p o t s a r e e l i m a t e d f r o m t h e a n a l y s i s b y t h e a p p l i c a t i o n o f a mask. The h o t s p o t mask i s c r e a t e d f r o m t h e f i r s t and f i n a l images i n t h e s e q u e n c e . The mean h i s t o g r a m o f t h e masked D e l t a l / I o images i s c a l c u l a t e d e x c l u d i n g t h e z e r o s f r o m t h e mask  a. o  % Jeff  Shepherd,  September 4, 2004 9-9*9-9*2-2-5-9-9-9-'  o o o o  % c l e a r a l l memory f r o m p r e v i o u s  clear a l l  Matlab  Script  9-5.9-2-5-2-2-2-9-9-9-9-!2-^-y-!2-y-y-y-»-s-s-3-2-3-!i"b'6"5'S'S'&'o'6'o'6'6'o'o^'o'o"5"5"6"5'5"5'5"5"5"o  %% p a r a m e t e r s t o change %%  %% P a t h and f i l e n a m e d e t a i l s ImFilePath='d:\jeff\exp50\' ; DImFilePath='d:\jeff\dark_current\'; SavePath='d:\jeff\exp50\' rawimprefix = 'jeg'; rawimsufix = ; dimprefix = 'zzz'; dimsufix = ; Ioimprefix = 'jeg' ; Ioimsufix = ' ' ; I_Ioimsufix = 'I_Io'; p m a s k i m s u f i x = 'pmask' palletsufix = 'pallet'; t i f f e x t = ' . t i f ' ,• c a p _ p o t _ f i l e = 'jeg490-jeg567-caps.txt' n o _ h e a d e r l i n e s _ i n _ c a p _ f i l e = 11; hist_outputfilename='hist_jeg490.dat ; 1  1  1  1  1  %% image numbers and s t a r t i m n o = 490; endimno = 567;  sequence d e t a i l s  %  2-s-9-2-9-9-9-9* ' g. O. Q. Q. O. C "5 "5 "o "o "o "o o o1 0 0 * 0 0 0 1 N  raw image p a t h d a r k c u r r e n t image p a t h path t o save data raw image p r e f i x raw image s u f f i x d a r k c u r r e n t image p r e f i x d a r k c u r r e n t image s u f f i x Io image p r e f i x Io image s u f f i x I / I o image s u f f i x p a r t i c l e mask s u f f i x p a l l e t image s u f f i x image e x t e n s i o n f i l e from expt number o f h e a d e r s f i l e histogram data output  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % f i r s t image i n s e q u e n c e % l a s t image i n s e q u e n c e  201  % I o image number % d a r k c u r r e n t im number  Io_imno = 500 0; D imno = 2;  3-9^ ^  m r [ l^ i dool JS.  iJ.i T l lil sc ri gT O G  n n m > > oXr co IlLlllLDt^  3- o 3- o 3- o S- o S- o S- o 3- o 3- 9o3-o3-^3-o9-^Sr oS- ^3- o3- ^ o 9o 3o 3o 3o 9o 9o 3o 3o 8r ^ 9o 3o 3o 3o 3^ 9o 3o 3o 3o 9o 3o 9o 2o 3o 2o 3o 2o 9o 9o So So 3-3-3-9-3-2-3-9.9. o o o o o o o o o  mask_im_a = 491 mask_im_b = 567  3-3-  H o fi n o  % 1 s t image i n 1 s t s e q u e n c e % f i n a l image i n 1 s t s e q u e n c e  d i c t - n n r ^ i n  "K-i T-I c  9-9- 3-3-5-3-9-9-9-9-9.9.9-9-9-9-9-9-9-9-9-9-3-9-9-9-9- 9- 9-9-9- 3-3-3-3- 3-3-3-3-3-3-3-3-S-3-3-3- 3-9-2-2-3-3-  m i n _ b i n _ v a l u e = -1; m a x _ b i n _ v a l u e = 9; n o _ b i n s = 1 0 0 0;  % minimum b i n v a l u e % maximum b i n v a l u e % t o t a l number o f b i n s  %% image m o d i f i c a t i o n d e t a i l s gmagrad=1.5; -5-5  LlIiQ  p d l  find_particles  %% r e a d  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % g r a d i e n t magnitude f i l t e r r a d i u s  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^'o'o'o"S"5'5'5 5"5"5o'5"5ooo _  =1;  % 1 means u s e f i n d _ f e a t u r e s s u b r o u t i n e % 0 means d o n ' t u s e s f i n d f e a t u r e s s u b r o u t i n e  i n d e f i n e d images  s u b - r o u t i n e %%  [Ioim_med] = r e a d _ i n _ i m a g e ( I m F i l e P a t h , I o i m p r e f i x , . . . Io_imno,Ioimsufix, tiffext); [Dim_med] = read_in_image(DImFilePath,dimprefix,... D_imno,dimsufix, t i f f e x t ) ; [adsim_med] = read_in_image(ImFilePath,rawimprefix,... s t a r t i m n o , r a w i m s u f i x , t i f f e x t ) ;' [Maskim_a_med] = r e a d _ i n _ i m a g e ( I m F i l e P a t h , r a w i m p r e f i x , mask_im_a,rawimsufix, t i f f e x t ) ; [Maskim_b_med] = r e a d _ i n _ i m a g e ( I m F i l e P a t h , r a w i m p r e f i x , mask im b , r a w i m s u f i x , tiffext);  %% c r e a t e k e y images  i n making  dell/Io  % F i g u r e A l - 1 a) .% F i g u r e A l - 1 b) % F i g u r e A l - 1 c) % F i g u r e A l - 1 d) % F i g u r e A l - 1 e)  images %%  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% s u b t r a c t t h e D a r k C u r r e n t image Ioim_dksub = Ioim_med - Dim_med;  from t h e medain  filtered  I o image %%%%%%%%% % Figure Al-1 f)  %% c r e a t e t h e I ( a d s ) / I o i m _ d k s u b image % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % Subim = adsim_med/Ioim_dksub; % F i g u r e A l - 1 g) Subim_med = m e d i f ( S u b i m , 3 , ' r e c t a n g u l a r ' ) ; % F i g u r e A l - 1 h)  ^^^^^^^^^^^^^^^^^^^^^^^^^'o'o'o'o'o'o'o'o'o'o'o'o  %% c r e a t e h o t s p o t  mask s u b - r o u t i n e %%  ^^^^^^^^^^^^^^^^^^^^^'«"o"o"o"o"o"o'o"o'5"5'6'5'o  o "0  %% c r e a t e h o t s p o t mask s u b - r o u t i n e % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % [hsmask] = create_mask_hist(Maskim_a_med,Maskim_b_med,Ioim_dksub,Subim,... gmagrad,SavePath,rawimprefix,startimno,tiffext); % Figure Al-1 i )  202  ^^^^^^^^'o'o'o'o'o'o'ooooooo'o  %% c r e a t e  o'S'S'S'S'S'S'S'S'S'S o  D e l t a l / I o sub-routine  %%  ^"5'S'S'S'S'S'S'S'S'S'S'S'S o'S'o'o'S o "5 o "o "o "o "o "6 o "6 o "6 *6 "o "o  [delI_Ioarray,pallet_val,hist_mean,hist_stdev] = create_delI_Io(startimno,.. endimno,ImFilePath,rawimprefix,rawimsufix,tiffext,Ioim_dksub,Subim_med,.. hsmask,min_bin_value,max_bin_value,no_bins); % f o r image a t -800mV s e e F i g u r e A l - 1 j )  %% C r e a t e image P a l l e t s u b - r o u t i n e %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% [pallet,maxval_pal,maxval_pal_a] = c r e a t e _ p a l l e t ( p a l l e t _ v a l , h i s t _ m e a n , . . . SavePath,rawimprefix,startimno,palletsufix,tiffext),-  2. 5-9.4.4.Q-Q-Q-Q-£.£.£.£.9-9-5.2.2.9-9-9-2.9.9.2.2.9-S-9-2.£-9-9-9-9-2-9-2-9-&-9-S-S-S,  %% a p p l y  contrast  enhancement s u b - r o u t i n e  %%  o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o _ o , o , o , ^ . ^ ^•^'^'^'i'S'S'S'S'^'i^'S^'S'S'S'S^^'S'S^^^^^'S'S^^'S'G'S'S'S^^ o'o'o o o T»  [ce_image] = c o n t r a s t _ e n h a n c e ( s t a r t i m n o , e n d i m n o , d e l I _ I o a r r a y , m a x v a l _ p a l _ a , . . SavePath,rawimprefix,I_Ioimsufix,tiffext); % f o r image a t -800mV s e e F i g u r e A l - 1 k)  3c%%!:?r%?r?rSr3c?r3r3;%?r?rSc%Sr?;S;?r?;?r?;?r%?r%%% o V o o o O O O O o o o o o o o o o o o o o o o o o o o o o o  %% F i n d  features  sub-routine  %%  9-9-9-9-9-9-9-9-9-9-9-S-9-S-9-9-S-9-9-9-2-2-2-2-2-2-9-S-2-2-S'  if  f i n d _ p a r t i c l e s == 1; [features,max_im_no] = find_features(pallet_val,maxval_pal_a,startimno,.. endimno,hsmask,gmagrad, r a w i m p r e f i x , S a v e P a t h , I _ I o i m s u f i x , . . . pmaskimsufix,tiffext) ; % f o r image a t -800mV s e e F i g u r e A l - 1 1) else end  %% w r i t e  histogram values  t o f i l e %%  Q.£.^.9.£.4.2.9-9^ ^ ^ I J ^ ^ I S I S ^ ' o 'l.S2-S-2-9.4.5-2-S.9-9-2.2-S-2-9-9-9-9-9-4-a-9-9-9-S-S'o'o'o'o'o'S'o'o'o'o'o'S'b'o'o'o'o'o'o'o'o^'o'o'o <  cap_E_file  <  = strcat(ImFilePath,  cap_pot_file);  [E_col, Cap_col] = t e x t r e a d ( c a p _ E _ f i l e , ' % f % * f % f % * f ' , . . . h e a d e r l i n e s , n o _ h e a d e r l i n e s _ i n _ c a p _ f i l e ) ; % read p o t e n t i a l and c a p a c i t y E_col = rot90(E_col); % r o t a t e p o t e n t i a l column C a p _ c o l = r o t 9 0 ( C a p _ c o l ) ,% r o t a t e c a p a c i t y column 1  1  figure(startimno) ; plot(E_col,hist_mean); f i d = fopen(strcat(SavePath, hist_outputfilename), 'wt+'); f p r i n t f ( f i d , ' % s \ n ' , ' H i s t o g r a m o b t a i n e d from a n a l y s i s ' ) ; fprintf(fid, %15s\t%15s\t%15s\t%15s\n','Potential(mV/SCE)',... 'Capacitance(uF/cm2)','Avg H i s t ' , 'Hist s t d e v ' ) ; for i=l:length(hist_mean) 1  203  fprintf(fid,'%f\t%f\t%f\t%f\n',E_col(i), hist_stdev(i) ); end f c l o s e ( f i d ) ,-  %% end o f s c r i p t %% 9.9-2.9-9.9-9-9-9-9-9-9-9-9-9-9-9-5-9' 'S'S'S'O'D'O'O'O'O'O'O'O'O'O'O'S'O'O'O  204  Cap_col(i),hist_mean(  a)  b)  c)  Figure Al-1 Selected images from the histogram_calc.m routine. The letter legend corresponds to images detailed in the script. All images presented were contrast enhanced unless labelled with "no CE". 205  % read_in_image.m% '5'S"Sib"S"S"S'6"S"S"S"S'5"S"S'5  % % % % &% % %  o o "6 o o o "6 o o o o o o "6 o o o o o ' o ' S ' S ' S ' S ' S ' S ' 5 ' 5 ' 5 o o o "6 o " S o o o o o o o o "o o o "6 o ' o * o ' o ' S * 5 ' S ' 5  o  o  o'6'o  r e a d _ i n _ i m a g e . m i s a s u b r o u t i n e t o c r e a t e a f i l e n a m e o f an image b a s e d on % t h e i n p u t (Image F i l e P a t h , image f i l e n a m e p r e f i x , image number, image % f i l e n a m e s u f f i x and t h e image e x t e n s i o n . The image i s t h e n opened, and a % median f i l e t e r a p p l i e d . T h i s image i s t h e o u t p u t o f t h e s u b - r o u t i n e % I n an e x p e r i m e n t , t h e images a r e numbered w i t h f o u r v a l u e s h a v i n g i n f r o n t . T h i s s c r i p t w i l l f o r m a t t h e a p p r o p r i a t e number o f z e r o s f r o n t o f t h e i n p u t image number.  % Dan B i z z o t t o O O O O O O O O o  and J e f f o o o o o o  Shepherd.  zeros in  September 4, 2004  % % % %  o o o o o o o o o O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O o o t> o o o o o o o o O  f u n c t i o n [med_filt_image] = read_in_image(ImageFilePath, imgno, i m a g e f n a m e s u f f i x , imagetype); %% c r e a t e f i l e n a m e o f a g i v e n image by a d d i n g j =imgno; i f (j>999) [imgno,err] = s p r i n t f ( ' % 4 d ' , j ) ; front e l s e i f (j>99) [imgno,err] = s p r i n t f ( ' % 3 d , j ) ; front imgno = s t r c a t ( ' 0 , i m g n o ) ; e l s e i f (j>9) [imgno,err] = s p r i n t f ( ' % 2 d ' , j ) ; imgno = s t r c a t ( ' 0 0 ' , i m g n o ) ; else [imgno,err] = s p r i n t f ( % l d ' , j ) ; front imgno = s t r c a t ( 0 0 0 , i m g n o ) ; end 1  imagefnameprefix,...  # of zeros  i n front  %%%%%%%%%%%  % a d d no z e r o s i n  % a d d one  zero i n  1  1  1  % add two z e r o s  i n front  % add t h r e e z e r o s i n  1  %% c r e a t e an image f i l e n a m e c a l l e d f u l l _ i m a g e _ f n a m e % % % % % % % % % % % % % % % % % % % % % % % % % image_fname = s t r c a t ( i m a g e f n a m e p r e f i x , imgno, i m a g e f n a m e s u f f i x , imagetype); full_image_fname = strcat(ImageFilePath,image_fname); %% r e a d i n t h e image a n d a p p l y a m e d i a n f i l t e r % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % r e a d i n t h e image % image = r e a d i m ( f u l l _ i m a g e _ f n a m e , ' T I F F ' ) ; % a p p l y a median f i l t e r med_filt_image = medif(image,3,'rectangular');  206  % c r e a t e mask h i s t . m %  % c r e a t e _ m a s k _ h i s t . m i s a s u b r o u t i n e w r i t t e n t o c r e a t e a h o t s p o t mask f r o m % % t h e i n p u t s Masklm_a_med ( 1 s t image i n t h e s e q u e n c e ) a n d Masklm_b_med % % ( f i n a l image i n t h e s e q u e n c e ) . T h e s e images a r e o u t p u t f r o m % % r e a d _ i n _ i m a g e . m ' and t h u s a r e m e d i a n f i l t e r e d . This subroutine puts % % them i n t h e f o r m I / I o on w h i c h a g r a d i e n t m a g n i t u d e f i l t e r i s a p p l i e d t o % % enhance t h e e d g e s o f t h e h o t s p o t s . T h e s e edges a r e t h e n s e l e c t e d w i t h a % % t h r e s h o l d i n g p r o c e d u r e ( w i t h M a t l a b ' s volume f u n c t i o n ) , e r o d e d t o remove % % s m a l l f e a t u r e s and t h e n d i l a t e t o e n l a r g e t h e masked a r e s . The two masks % % a r e t h e n i n v e r t e d and m u l t i p l i e d t o g e t h e r t o a c c o u n t f o r any d r i f t i n t h e % % hot spots a r i s i n g from a p o s s i b l e d r i f t i n the e l e c t r o d e w i t h time. % 1  %  %  % Jeff  S h e p h e r d September 4, 2004  %  f u n c t i o n [hsmask] = create__mask_hist(Maskim_a_med Maskim_b_med,... Ioim_dksub Subim, gmagrad, S a v e P a t h , r a w i m p r e f i x , s t a r t i m n o , t i f f e x t ) ; /  /  %%%%%%%%%'&'%"&'%%%"&'"o''S'&''o''&''o''S'S'o'"o'  %%  start  with  1 s t image %%  %%%%%%%%%%"6%%%'6""o'"6''S'&"'&'"6''S'S"&*'6""&"  %% p u t s e l e c t e d images i n t h e f o r m I / I o % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % d i v i d e by Ioim_dksub MaskIm_a_Io = Maskim_a_med/Ioim_dksub; % F i g u r e A l - 2 a) &- S-  n > -  Q  . f Q  n  a  v  , ,  m  a  e  l ,  Q, Q, O. o. q. o, o, a a a a a. q. q. o. q. q. q. q. q. q. o. o. q. q. q. q. o. q. q. o. Q. a. g. g. a. g.  g. o. g. a. <^ g. o. g. o 5. 3. 0.5. ^ ^ 3. g.  creaue jjinary m a s K •6-&-5'5-5-5-5'5'5'5'5^^'6^^^'S^^^^^5^^^^^^^^^^^^^^^^^^^^^^^^^ o ^ ^ ^ ^ o o % gaussa i a bg n f=i l bt le ur r g a u s s ( M a s k I m _ a _ I o , 0 . 1 ) ; % F i g u r e A l -2 b) hsmask -s"S  % g r a d i e n t magnitude hsmask a gmag = gradmag(hsmask a b g , g m a g r a d ) ; % threshold features hsmask a t h r e s h = t h r e s h o l d ( h s m a s k a gmag, 'volume ,0.025) ; % erode f e a t u r e s hsmask a e r o s = b e r o s (hsmask a t h r e s h , 1,-2 ,1) ,% dilate features hsmask a d i l a = b d i l a ( h s m a s k a e r o s , 7 , - 2 , 0 ) ; % i n v e r t image hsmask a= -hsmask a d i l a ; 1  % F i g u r e A l -2 c) g.  F i g u r e A l -2 d)  q. o  F i g u r e A l -2 e)  q. o  F i g u r e A l -2 f)  q. o  F i g u r e A l -2 g)  S.9-3-q-2-2-9-2-2-3-3-S-9-S-S-S-2-S-9-S-2-S-2-2-9-9-9"5'o'5'5^'q'5'5'5"5"S'o'5"o'5 o o'S'5'5 o o o o o "b o  %%  repeat  for last  image %%  9.9-3-3-9.2.9.9-5-9-3-9-2-2-2-9-2-9-&-S-S-2-S-B-S-2-S. ^'S'S'o'S^S'S'S'o'o'o'o'o'S'o'o'o'o'o'o'o'o'o'o'o o"5  MaskIm_b_Io = Maskim_b_med/Ioim_dksub; hsmask_b_bg = b l u r g a u s s ( M a s k I m _ b _ I o , 0 . 1 ) ; hsmask_b_gmag = gradmag(hsmask_b_bg,gmagrad); hsmask_b_thresh = threshold(hsmask_b_gmag,'volume',0.025); hsmask_b_eros=beros(hsmask_b_thresh,1,-2,1) ; hsmask_b_dila = bdila(hsmask_b_eros,7,-2,0) ; hsmask_b= ~ h s m a s k _ b _ d i l a ; ' 9"-5 9"-5 9"-6 9"5 - 9o>  ;%  combine images and s a v e %%  5.9.9.5.9.9.9.9.9.9-9-9.9.%%%%%%%%%%%%%%%%  207  <  % F i g u r e A l - 2 h)  % multiply the masks together%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% hsmask = hsmask_a*hsmask_b; % Figure A l - 2 i) %% i f h s m a s k i s a l l z e r o s i f h s m a s k == zeros; hsmask = hsmask+1; end  make  i t  a l l  ones  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%  %% s a v e h o t s p o t m a s k %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% [outbinname, full_outbinname]=create_imfname(SavePath,rawimprefix,... (startimno), 'hotmask', t i f f e x t ) ; % create a save filename writeim(hsmask,full_outbinname, tiff ,0); % save images 1  1  208  a)  b)  c)  Figure A l - 2 Selected images from the create_mask_hist.m routine. The letter legend corresponds to images detailed in the script. All images presented were contrast enhanced unless labelled with "no CE".  209  create_delI_Io.m  %  U2-9-9-9-9-9'9-9'9'9*9-9'9-2-2-S*9-23'o'o'o'So'oo'o o o'S'S'S'S'S'S'o o ( 0  % % % % % % % % %  c r e a t e _ d e l I _ I o . m i s a s u b - r o u t i n e w r i t t t e n t o c r e a t e an a r r a y o f a l l t h e images i n a s e q u e n c e and p u t them i n t h e f o r m D e l t a l / I o . T h i s s u b - r o u t i n e a l s o f i n d s t h e maximum and minimum p i x e l v a l u e s i n t h e e n t i r e s e q u e n c e t h a t a r e l a t e r u s e d t o c r e a t e a l i n e a r c o n t r a s t enhancement t o t h e images t o i n c r e a s e t h e v i s u a l c o n t r a s t and a l s o t o be u s e d i n t h e c r e a t i o n o f an image p a l l e t . T h i s s u b - r o u t i n e a l s o c a l l s on a n o t h e r s u b - r o u t i n g t h a t c a l c u l a t e s t h e mean h i s t o g r a m v a l u e s o f e a c h masked image. Note!  The  image a t  -800mV w i l l  be  followed i n this  routine  % % % % % % % % % 9,  o  % J e f f S h e p h e r d . September 4, 2004 % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% f u n c t i o n [ d e l I _ I o a r r a y , p a l l e t _ v a l , h i s t _ m e a n , h i s t _ s t d e v ] =... create_delI_Io(startimno,endimno,ImFilePath,rawimprefix,rawimsufix,... tiffext,Ioim_dksub,Subim_med,hsmask,min_bin_value,max_bin_value,... no_bins); %% c r e a t e a b l a n k a r r a y w i t h enough s p a c e t o h o l d a l l images i n a s e q u e n c e delI_Ioarray=newimar(endimno-startimno +1) ; %% c r e a t e d e l l / I o a r r a y and f i n d f o r j = s t a r t i m n o : endimno ,% a progress indicator j % c l e a r the v a l u e of maxval c l e a r maxval; % c l e a r the v a l u e of minval c l e a r minval;  t h e max  and  min  hist  values  %%  %%%%%%%%%%%%%%%%  %% c r e a t e t h e d e l l / I o a r r a y % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % read_in_image sub-routine % [I_med] = r e a d _ i n _ i m a g e ( I m F i l e P a t h , . . . rawimprefix, j , rawimsufix, t i f f e x t ) ; % F i g u r e A l - 3 a) % d i v i d e by Ioiamge_dksub I_Ioim=(I_med/Ioim_dksub); % F i g u r e A l - 3 b) % m e d i a n f i l t e r I / I o im I_Ioim_med = m e d i f ( I _ I o i m , 3 , ' r e c t a n g u l a r ' ) ; % F i g u r e A l - 3 c) % s u b t r a c t t h e a d s o r p t i o n image d e l I _ I o = I_Ioim_med - Subim_med; % F i g u r e A l - 3 d) % a p p l y mask t o d e l l / I o delI_Io2 = delI_Io*hsmask; % F i g u r e A l - 3 e) % f i n d t h e max p i x e l v a l u e m a x v a l = max ( d e l I _ I o 2 ) ,% f i n d t h e a b s o l u t e min p i x e l v a l u e minval = min(abs(delI_Io2)); % p u t v a l u e o f m a x v a l i n j t h row 1 s t c o l o f new a r r a y p a l l e t _ v a l ( j - s t a r t i m n o + 1 , 1 ) = maxval; % p u t v a l u e o f m i n v a l i n j t h row 2nd c o l o f new a r r a y pallet_val(j-startimno+1,2) = minval; %% h i s t _ m e a n _ s t d e v 2 s u b - r o u t i n e %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% [h_mean, h _ s t d e v ] = h i s t _ m e a n _ s t d e v 2 ( d e l I _ I o , hsmask, m i n _ b i n _ v a l u e , . . . max_bin_value, no_bins) ; % p u t t h e v a l u e o f h i s t mean i n an a r r a y  210  hist_mean(j-startimno+l)=h_mean; % p u t t h e v a l u e o f h i s t s t d e v i n an a r r a y hist_stdev(j-startimno+l)=h_stdev; % make an a r r a y o f d e l l / I o images delI_Ioarray{j-startimno+l}=delI_Io; end  211  a)  b)  c)  Figure A l - 3 Selected images from the create_delI_Io.m routine. The letter legend corresponds to images detailed in the script. All images presented were contrast enhanced.  212  % create_pallet.m % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%.%%%%%%%%%%%%%%%%%%%%%%%%%%%%  % c r e a t e _ p a l l e t . m i s a s u b - r o u t i n e w r i t t t e n t o c r e a t e an image p a l l e t % r e p r e s e n t i n g t h e maximum and minimum v a l u e s i n t h e masked d e l l / I o image % array. The images a r e a l l 8 - b i t and r a n g e i n v a l u e s f r o m 0 t o 255. 0 is % c o n s i d e r e d b l a c k , and 255 i s c o n s i d e r e d as w h i t e . % % J e f f S h e p h e r d . S e p t e m b e r 4, 2004 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% function [pallet,maxval_pal,maxval_pal_a] = create_pallet(pallet_val,... hist_mean,SavePath,rawimprefix,startimno,palletsufix,tiffext); %% f i n d s e q u e n c e min and max v a l u e s f r o m a r r a y o u t p u t f r o m c r e a t e _ d e l I _ I o %° maxval_pal_a = m a x ( p a l l e t _ v a l ( : , 1 ) ) ; % f i n d max v a l u e i n t h e p a l l e t a r r a y minval_pal_a = min(pallet_val(:,2)); % f i n d min v a l u e i n t h e p a l l e t a r r a y maxval_pal = max(hist_mean); 1  %% c r e a t e an image w i t h enough s p a c e f o r 256 l i n e s n o _ b i n s _ f o r _ p a l l e t = 256; p a l l e t = newim(150, n o _ b i n s _ f o r _ p a l l e t + 3 2 ) + 2 5 5 ;  %%%%%%%%%%%%%%%%%%%%%%%%%"  %% c r e a t e p a l l e t b o a r d e r %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% p a l l e t ( 0 : 1 , ( 1 4 : n o _ b i n s _ f o r _ p a l l e t + 1 6 ) ) = 0; % l e f t c o l border p a l l e t ( 5 1 : 5 2 , ( 1 5 : n o _ b i n s _ f o r _ p a l l e t + 1 6 ) ) = 0; % r i g h t c o l border p a l l e t ( 0 : 6 0 , 1 4 : 1 5 ) = 0; % top l i n e border p a l l e t ( 0 : 6 0 , ( ( n o _ b i n s _ f o r _ p a l l e t + 16) : . . . ( n o _ b i n s _ f o r _ p a l l e t + 1 7 ) ) ) = 0; % bottom l i n e border %% f i l l p a l l e t w i t h a p p r o p r i a t e g r e y v a l u e f o r j = 0:255; pallet(2:50,j+16) = 255-j; end  %°  %% add max and m i n n u m e r i c a l v a l u e s t o t h e t o p and b o t t o m o f p a l l e t %%%%%%%%% [maxval_pallet,err] = s p r i n t f ( ' % 2 . I f , m a x v a l _ p a l _ a ) ; % c r e a t e max num string maxvalstr = s t r c a t ( ' ' , m a x v a l _ p a l l e t ) ; % a p p l y max number [minval_pallet,err] = sprintf('%2.Of',minval_pal_a); % c r e a t e min num string minvalstr = strcat('',minval_pallet); % a p p l y min num string 1  p a l l e t = text(65,15,maxvalstr,'FontSize',25,'FontWeight','bold'); p a l l e t = t e x t ( 6 5 , n o _ b i n s _ f o r _ p a l l e t + 1 7 , m i n v a l s t r , F o n t S i z e ' , 2 5 , .. . 'FontWeight','bold'); pallet % show p a l l e t 1  •s-s s a v e p a l l e t t o r u e ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^-s-s-s-s-s-s-s-s-s'o's-5-b-b [outbinname, full_outbinname]=create_imfname(SavePath,rawimprefix,... startimno,palletsufix,tiffext); saveas(gcf,full_outbinname,'tiff');  213  % contrast_enhance.m  % % % % % % %  %  contrast_enhance.ra i s a sub-routine w r i t t t e n t o a p p l y a c o n t r a s t % enhancement t o t h e i m a g e s . T h i s i s f o r v i s u a l c o n t r a s t o n l y a n d does n o t % d i r e c t l y c o r r e s p o n d t o t h e h i s t o g r a m mean v a l u e s . The l i n e a r c o n t r a s t % enhancement i s a c h i e v e d b y f i n d i n g t h e most i n t e n s e image f r o m t h e d e l I / I o % a r r a y , f i n d i n g t h e maximum v a l u e o f t h a t image a n d s e t t i n g t h i s v a l u e % t o 255. A l l o t h e r v a l u e s i n t h e a r r a y o f images a r e t h e n l i n e a r l y % s c a l l e d by t h i s f a c t o r . %  9o  9o  % N o t e ! The image a t -800mV w i l l be f o l l o w e d % % J e f f S h e p h e r d . September 4, 2004  % % %  f u n c t i o n [ce_image] = contrast_enhance(startimno,endimno,... delI_Ioarray,maxval_pal_a,SavePath,rawimprefix,I_Ioimsufix,tiffext); for  j = startimno:endimno; % a progress indicator j % open t h e image delI_Io_im = delI_Ioarray{j-startimno+1}; % F i g u r e A l - 4 a) % d i v i d e b y t h e max v a l u e ( m a x v a l _ p a l _ a ) a n d s c a l e t h e image delI_Io_im_ce=(delI_Io_im/maxval_pal_a)*255; % F i g u r e A l - 4 b) % r o u n d v a l u e s up delI_Io_ceil = ceil(delI_Io_im_ce) ; % make images 8 - b i t ce_image = d i p _ i m a g e ( d e l I _ I o _ c e i l , ' u i n t 8 ' ) ; % save f i l e n a m e [outbinname, full_outbinname]=create_imfname(SavePath,rawimprefix,... j, I_Ioimsufix, t i f f e x t ) ; % s a v e images writeim(ce_image,full_outbinname, ' t i f f ' , 0 ) ;  end  214  a)  b) no C E  Figure A l - 4 Selected images from the contrast_enhance.m routine. The letter legend corresponds to images detailed in the script. Image (a) was contrast enhanced after output from the routine, "no C E " stands for no contrast enhancement.  215  find  features.m  %  f i n d _ f e a t u r e s . m i s a s u b r o u t i n e w r i t t e n to d e f i n e the f e a t u r e s of a f l u o r e s c e n c e image. The images a r e t a k e n f r o m d e l I _ I o a r r a y and a g r a d i e n t magnitude f i l t e r i s a p p l i e d to d e f i n e the f e a t u r e s . These f e a t u r e s are then s e l e c t e d u s i n g a t h r e s h o l d . The t h r e s h o l d v a l u e i s d e t e r m i n e d by f i n d i n g t h e m e d i a n v a l u e o f t h e most i n t e n s e image i n t h e s e q u e n c e . T h i s t h r e s h o l d v a l u e i s t h e n a p p l i e d t o e v e r y image i n t h e s e q u e n c e . Dan B i z z o t t o and J e f f S h e p h e r d . Aug4, 2004 % N o t e ! The % Dan  image a t  B i z z o t t o and  -800mV w i l l  Jeff  be  % % % % % % %  followed  S h e p h e r d . September 4,  % 2004  %  o o o o o o 0,0,0,0,0,9,0,0,0,0,0,9,0,0,9,3,0,0,9,9,9,9,0,9,0^ ^ ^ ^ ^ ^ ^ o o o o o ^ o ^ o o ^ o o o ^ o o ^ ^ ^ ^ ^ ^ ^ o o o ^ o o  o"5  o"6  o o o o o  f u n c t i o n [features,max_im_no] = f i n d _ f e a t u r e s ( p a l l e t _ v a l , m a x v a l _ p a l _ a , . . . startimno,endimno,hsmask,gmagrad, rawimprefix,SavePath,I_Ioimsufix,... pmaskimsufix,tiffext); %% f i n d image w i t h h i g h e s t p i x e l v a l u e t o d e f i n e a t h r e s h o l d for j = startimno-startimno+1:endimno-startimno+1; % o u t p u t t h e j t h row o f f i r s t column t o v a l u e m m = pallet_val(j,1); % i f m i s e q u a l t o maxval i f m == m a x v a l _ p a l _ a ; % c a l l max_im_no t h a t v a l u e o f j max_im_no = j ; else end end % d e f i n e t h e most i n t e n s e image number as max_im_no = m a x _ i m _ n o + s t a r t i m n o ;  %%%%%%%%%%%%%%%%  max_im_no  %% l o a d t h e most i n t e n s e image and f i n d t h e h i s t o g r a m mean % % % % % % % % % % % % % % % % % % [maxim_med] = r e a d _ i n _ i m a g e ( S a v e P a t h , r a w i m p r e f i x , m a x _ i m _ n o , I _ I o i m s u f i x , . . . tiffext); % gaussian filter t h r e s h _ i m _ b g = blurgauss(maxim_med,0.1) ; % g r a d i e n t mag filter thresh_im_bg_gmag = g r a d m a g ( t h r e s h _ i m _ b g , g m a g r a d ) ; % mean o f t h e histogram t h r e s h o l d = mean(thresh_im_bg_gmag); %% f i n d t h e p a r t i c l e s i n t h e s e q u e n c e b a s e d on for j = (startimno):(endimno); % progress indicator  the  threshold value  j % open d e l l / I o images [im_med] = r e a d _ i n _ i m a g e ( S a v e P a t h , r a w i m p r e f i x , j , . . . I _ I o i m s u f i x , t i f f e x t ) ,% gaussian filter features_gauss = blurgauss(im_med,0.1); % g r a d i e n t magnitude features_gmag = gradmag(features_gauss,gmagrad); % threshold features features_thresh = thresh(features_gmag,threshold);  216  %%%%%%%%  % Figure Al-5  a)  % Figure Al-5  b)  % Figure Al-5  c)  % Figure Al-5  d)  !  % erode features features_eros=beros(features_thresh,1,-2,1); % % dilate features features_dila = bdila(features_eros,1,-2,0); % % c l o s e h o l e s i n image features_clos = bclosing(features_dila,2,-2,0); % % s e t O ' s a n d l ' s t o 0 ' s a n d 255 features_binary = features_clos*255; % % a p p l y t h e h o t s p o t mask features_w_mask = features_binary*hsmask; % % make i m a g e 8 b i t features_8bit = dip_image(features_w_mask,'uint8 ); % save images features = features_8bit; % % c r e a t e a save filename [outbinname, full_outbinname]=create_imfname(SavePath,... rawimprefix, j , pmaskimsufix, t i f f e x t ) ; % save image writeim(features_8bit, full_outbinname, ' t i f f ' , 0) ;  F i g u r e A l - 5 e) F i g u r e A l - 5 f) F i g u r e A l - 5 g) F i g u r e A l - 5 h) Figure A l - 5 i)  1  end  217  F i g u r e A l - 5 j)  a) no C E  b no CE)  c)  d) no C E  e) no C E  f) no C E  g) no C E  h no CE)  i) no C E  j)  no C E Figure Al-5 Selected images from the findJeatures.m routine. The letter legend corresponds to images detailed in the script. Image (c) was the only image in this group that was contrast enhanced after output from the routine, "no CE" stands for no contrast enhancement. 218  % hist  % % % %  o  mean s t d e v 2 . m %  hist_mean_stdev2.m i s a s u b - r o u t i n e w r i t t t e n t o c a l c u l a t e t h e average h i s t o g r a m o f t h e images w h i l e c o m p e n s a t i n g f o r t h e p r e - e x i s t i n g z e r o s i n t h e raw images i f a n y . F i n d o r i g i n a l number o f z e r o s i n t h e image by a d d i n g 1 t o t h e image a n d t h e n c o u n t i n g t h e number o f l ' s a f t e r m a s k i n g  % Dan B i z z o t t o a n d J e f f  S h e p h e r d . September 4, 2004  mask,...  % make s u r e n o t t o c o u n t t h e z e r o s i n t h e mask m a s k h i s t o g r a m = d i p h i s t ( m a s k , [0 1] ,2) ; % # z e r o s i n mask a r e t h e # p i x e l s n o t c o u n t e d i n t h e a n a l y s i s maskzeropixels = maskhistogram(1); image=image*mask; sum= 0; nopixels=0;  for  max_bin_value],...  i=l:no_bins %binning c a l c s shoudl use the mid-point value of the b i n sum = sum + h i s t o g r a m ( i ) * b i n _ c e n t e r s ( i ) ; nopixels = nopixels + histogram(i);  end %sum = h i s t o g r a m * b i n _ c e n t e r s ; i f nopixels > 0 hist_mean=sum/(nopixels-maskzeropixels) else h i s t _ m e a n = 0; end  ;  ^generate s t d e v v a l u e s f o r t h e h i s t o g r a m s and use t h i s as w e i g h t i n g sum=0; nopixels=0; for i=l:no_bins %binning c a l c s shoudl use the mid-point value of the b i n sum = sum + h i s t o g r a m ( i ) * ( b i n _ c e n t e r s ( i ) - h i s t _ m e a n ) ; nopixels = nopixels + histogram(i); end i f nopixels > 0 hist_stdev = sqrt(sum/(nopixels-l-maskzeropixels)); else h i s t _ s t d e v = 0; end  219  o,  %  f u n c t i o n [hist_mean, h i s t _ s t d e v ] = hist_mean_stdev2(image, min_bin_value, max_bin_value, n o _ b i n s ) ;  % make s u r e t o o n l y c o u n t t h e o r i g i n a l b l a c k [histogram, b i n _ c e n t e r s ] = d i p h i s t ( i m a g e , [ m i n _ b i n _ v a l u e no_bins);  % % % %  Image Pro Plus 5 Macro Option Explicit Sub count_features 'This script is used to take the particle masks output from histogram_calc.m and perform and segment the features with 'a watershed filter. The calibration is set to 50x2bin for the 50X objective with 2x2 binning. A counting feature is then 'applied to each image and some statistics measured. Since the resolution was found to be accurate for features with a 'minimum mean diameter of 4um, the features that have an area less than 12 are not counted in this routine. Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim Dim  startfilenumber As Integer stopfilenumber As Integer TempImageDirName As String CalcImageDirName As String WorklmageName As String * 255 WorklmageMaskName As String * 255 ImagePrefixName As String ImageSuffixName As String ImageExtension As String OutlinesSuffixName As String ModlmageSuffixName As String CountSettingsName As String CountSettingsExtension As String imagedocid As Integer i As Integer ' start file number in sequence ' end file number in sequence ' directory where images are ' save directory name ' image prefix ' image suffix - output from Matlab ' image extension ' saved outlined image statistics suffix name ' saved outlined image name count settings ' count settings extension  startfilenumber=490 stopfilenumber=567 TempImageDirName="d :\j eff\exp5 0\" CalcImageDirName="d:\jeff\\exp50\" ImagePrefixName="j eg" ImageSuffixName="pmask" ImageExtension=".tif' OutlinesSuffixName="_pmask_wshed_outl" ModImageSuffixName="_pmask_wshed" CountSettingsName="count_settings_thesis" CountSettingsExtension=".env"  1  " load the settings for counting features ret = IpBlbLoadSetting(TempImagedirName+CountSettingsName+CountSettingsExtension) ret = IpSCalSelect("50x2bin")  " begin the counting routine For i = startfilenumber To stopfilenumber " search for images in sequence ret IpStSearchDir(TempImageDirName,ImagePrefixName+Format(i,"0000")+ImageSuffixName+ImageExtension,0,W orklmageName) ' for image at -800mV see Figure A1-6 a)  220  Ifret>OThen " open a specified resolution of the image ret = IpWsLoadSetRes(LOAD_PROMPT) imagedocid = IpSeqOpen(WorkImageName, "tif ,0,1)  " perform a watershed segmentation on the image ret = IpFltWatershedEx(50, -1)  'for image at -800mV see Figure A1-6 b)  " count features ret = IpBlbCount() ret = IpBlbUpdate(O) " save outlines in the image created by counting routine ret = IpBlbSaveOutline(CalcImageDirName+ImagePrefixName+Format(i,"0000")+OutlinesSuffixName+".out") ' for image at -800mV with outlines applied Figure A1-6 c) " save the outline statistics in .txt format ret = IpBlbSaveOutline(CalcImageDirName+ImagePrefixName+Format(i,"0000")+OutlinesSuffixName+".txt") " save outline statistics in .cnt format (image pro feature) ret = IpBlbSaveData(CalcImageDirName+ImagePrefixName+Format(i,"0000")+OutlinesSuffixName+".cnt", S_APPEND+S_HEADER+S_Y_AXIS) " save outline statistics of the whole sequence by appending new data ret = IpBlbSaveData(CalcImageDirName+ImagePrefixName+Format(startfilenumber,"0000")+"-"+Format(stopfilenumbe r,"0000")+OutlinesSuffixName+".cnt", S_APPEND+S_STATS+S_HEADER+S_X_AXIS+S_Y_AXIS) " save the watershed image ret = IpWsSaveAs(CalcImageDirName+ImagePrefixName+Format(i,"0000")+ModImageSuffixName+".tif, "tif) " close the image ret = IpDocClose() End If Next i End Sub  221  a) no C E  b no CE)  c)  Figure A l - 6 Selected images from the countJeatures.scr routine. The letter legend corresponds to images detailed in the script. Image (c) was the only image in this group that was contrast enhanced after output from the routine, "no CE" stands for no contrast enhancement.  222  Appendix A 2 Kinetic model fitting routine for a fluorescence decay and recovery Jeff Shepherd, Dec 26, 2004 Introduction The programs in this appendix that are different from appendix A2 are the following: • freap.m. -main program that is similar to histogram_calc.m. calc_decay.m -calculated the mean histogram values from the Al/Io images from freap.m. ka_kp_Ap_do_no_wtJit_dblok.m -minimizes the sum of the square of the errors with constraints to negative values  223  % freap.m %  f r e a p . m i s a l m o s t e x a c t l y t h e same as h i s t o g r a m _ c a l c . m described i n appendix A l . The o n l y d i f f e r e n c e b e t w e e n t h i s p r o g r a m a n d h i s t o g r a m _ c a l c . m-s i s that the f i n d _ f e a t u r e s . m subroutine i s r e p l a c e d with calc_decay.m. % f r e a p . m r e q u i r e s a d e f i n i t i o n o f t h e number o f d e c a y s i n t h e e x p e r i m e n t . T h i s i s r e q u i r e d s i n c e each decay i s f i t s e p a r a t e l y . The mean h i s t o g r a m of e a c h D e l t a l / I o image i s c a l c u l a t e d a f t e r t h e a p p l i c a t i o n o f a h o t s p o t mask. The mean h i s t o g r a m does n o t i n c l u d e t h e p i x e l v a l u e s o f t h e mask. The h i s t o g r a m v a l u e s a r e t h e n s u b j e c t e d t o a f i t t i n g r o u t i n e b a s e d on a Nedler-Mead simplex ( d i r e c t s e a r c h ) . This i s accomplished i n the calc_decay.m subroutine. O u t p u t f r o m t h e f i t t i n g r o u t i n e a r e kp, k a , A p c a l c and a c o n s t a n t d . T h e s e r e p r e s e n t t h e r a t e c o n s t a n t s f o r t h e f i r s t order photobleaching, second order aggregation, decrease i n f l u o r e s c e n c e from p h o t o b l e a c h i n g and a s c a t t e r i n g c o n s t a n t r e s p e c t i v e l y . % J e f f S h e p h e r d , September 29, 2004 % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% clear a l l  % clear  a l l memory f r o m p r e v i o u s  Matlab  Script  %%%%%%%%%%%%%%%%%%%%%%%%%% %% p a r a m e t e r s t o change %% %%%%%%%%%%%%%%%%%%%%%%%%%% 9-9-9-9-9-9-9-9-9.9-9-9-9-9-9-9-9-S.9-9-9-9-^S.9-9-°^9.9-9-9-9-5'S.9.9.S.S.S-9.9. %% P a t h a n d f i l e n a m e d e t a i l s raw image p a t h ImFilePath='d:\jeff\exp96\'; d a r k c u r r e n t image p a t h DImFilePath='d:\jeff\dark_current\'; p a t h t o save data SavePath='d:\jeff\exp96\' raw image p r e f i x r a w i m p r e f i x = 'jgm'; raw image s u f f i x rawimsufix = ' ' ; d a r k c u r r e n t image p r e f i x dimprefix = 'zzz'; d a r k c u r r e n t image s u f f i x dimsufix = ' ; I o i m p r e f i x = 'jgm'; Io image p r e f i x Ioimsufix = '' ; I o image s u f f i x I_Ioimsufix = 'I_Io'; I / I o image s u f f i x palletsufix = 'pallet'; p a l l e t image s u f f i x tiffext = '.tif'; image e x t e n s i o n hist_outputfilename='hist_jeg490.dat' histogram data output rate_constant_outputfilename='tst.dat f i t t i n g parameters output 0  1  1  %% image numbers a n d s e q u e n c e d e t a i l s ^^^^^^^^^^^^^^^^^^^^^^^^o^o^s^s^s^s^s^^^a^s^s^s^s^a^^s s t a r t i m n o = 5; % f i r s t image i n s e q u e n c e endimno = 3 5; % l a s t image i n s e q u e n c e Io_imno = 2; % I o image number D_imno = 2; % d a r k c u r r e n t im number n o _ d e c a y s = 1; % t o t a l # of decays i n the experiment no_set_one=2; % t o t a l # o f S e t l ' s ( i . e . 5 a d s images) i n e x p e r i m e n t n o _ a d s _ i m _ i n _ s e t = 5 ,% t o t a l # o f a d s o r p t i o n images i n e a c h S e t 1 no d e s im a t e n d = l ; % number o f d e s o r p t i o n images a t t h e e n d o f e x p e r i m e n t  %% mask image numbers mask_im_a = 1 mask_im_b = 3 4  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 1 s t a d s o r p t i o n image i n s e q u e n c e % l a s t a d s o r p t i o n image i n s e q u e n c e  224  3-3-  HQ-FTT-IO  H-i a t - . - ^ v - a m  -6-6 Q e r m e H i s t min_bin_value max_bin_value no b i n s = 1 0 0  K i n c  ogram Dins = - 1 ; = 9; 0;  3- 3- 2- 3- 3. 3- 3- 2- 9- 9- 3- S- 9- 9- 8. 3- 8r a- a- 3- 3- 2- S- 9- 3- 2- 9- Sr 3- 3- 2r 2- 9- 2- 2-S-2-2-2-Sr 2r 3-3-3-9-9-9-9-3-9-9-9-  •s^'S'S^^^^^^^'S^'S-s^^'S^^'S^^'S'S'S^^^^^^^^^^^^^^^^^^^^^^^^^^ % minimum b i n v a l u e % maximum b i n v a l u e % t o t a l number o f b i n s 1-2-2-2.2.2-9-2-9-2-9.9-2-9-2-2-2-9-2-S-9-9-S-9-2-S-B-3-2-2-2-2-2-2-9-2-2-2.2.2.  gmagrad=l;  %% r e a d  % gradient  i n defined  % described  images  sub-routine  magnitude  f i l t e r  radius  %%  i n Appendix A l  [Ioim_med] = read_in_image(ImFilePath,Ioimprefix, Io_imno,... Ioimsufix, t i f f e x t ) ; [Dim_med] = r e a d _ i n _ i m a g e ( D I m F i l e P a t h , d i m p r e f i x , D _ i m n o , . . . dimsufix, t i f f e x t ) ; [adsim_med] = read_in_image(ImFilePath,rawimprefix, startimno,... rawimsufix, tiffext) ; [Maskim_a_med] = read_in_image(ImFilePath,rawimprefix, mask_im_a,... rawimsufix, t i f f e x t ) ; [Maskim_b_med] = read_in_image(ImFilePath,rawimprefix, mask_im_b, .. rawimsufix, t i f f e x t ) ;  o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o  %% c r e a t e k e y i m a g e s i n m a k i n g d e l l / I o i m a g e s %% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % described  i n Appendix A l  %% s u b t r a c t t h e D a r k C u r r e n t i m a g e Ioim_dksub = Ioim_med - Dim_med;  from  the medain  f i l t e r e d  Io  image  %%%%%%%%%  "8"*s c r e a t e t h e I (ads) / I o i m dksub image •S'S%%"S"ff%%'S%%%"o%%%%'S'S-S"o%%"6"o"S'S'S'S'S%'S'S"S"o'5%"ff"& Subim = adsim_med/Ioim_dksub; Subim_med = medif(Subim,3,'rectangular');  2-2-9-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-2-9-2-2-2-2-9-2-2-2-2-2-2-9-2-9-9-9-2"S o "S "6 o o ' S ' S ' S ' S ' S ' S o o o o o "S "6 o "5 o o o ~S o o o "6 "5 "5 o f f o i l  %% c r e a t e  hotspot  mask  s u b - r o u t i n e %%  3-3-9-3-3-3-3-3-° o o o o o o o o l  described  i n Appendix A l  %% c r e a t e h o t s p o t mask s u b - r o u t i n e %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%»- - - - - [hsmask] = create_mask_hist(Maskim_a_med,Maskim_b_med,Ioim_dksub,Subim, gmagrad,SavePath,rawimprefix,startimno,tiffext); 9  9-2-2-2-2-2-2-2-2-2-2-2-2-2-2-9-9-9-2-9-9-9-2-2-9-2-2-2-2-2-9-2-9-9o'b'o o o o o'o'S'S'o'o'S'o'S o'o'o o o'S'o'S o o o ' S ' 0 ' 0 ' 5 ' 5 ' 5 ' 5 ' 5  %% c r e a t e  D e l t a l / I o  *© o o O ' S ' O ' O ' O ' S ' O ' O ' D ' D " ©  sub-routine  o o o o ' o ' S ' o ' S o'o'S'5  %%  o'S'6'S'S'S'S'S  225  9  s  e  9  % described i n Appendix A l [delI_Ioarray,pallet_val,hist_mean,hist_stdev] = create_delI_Io(startimno, endimno,ImFilePath,rawimprefix,rawimsufix,tiffext,Ioim_dksub,Subim_med, hsmask,min_bin_value,max_bin_value,no_bins);  'o'o'o'o'o'o'o'o  o "o o o o o o o "o "o "o o"o  o o o 0*0 o  %% C r e a t e image P a l l e t 'o'o'o'o'o'o'o'o  o'o'o  "6 "S  o'o'o'o'o'o'o  s u b - r o u t i n e %%  o "o "6 "o o © " © " S o o o o o o o o " o " o " o o  o o "o "6 "6 "6  % d e s c r i b e d i n Appendix A l [pallet,maxval_pal,maxval_pal_a]=create_pallet{pallet_val,hist_mean,... SavePath,rawimprefix,startimno,palletsufix,tiffext);  o o o o o o ^ o o ^ o ^ o o o o o o ' o o o o o o o o o o ' o o ' o o o o o o o ' o ' o o ' o o ' o ' o  %% a p p l y c o n t r a s t enhancement s u b - r o u t i n e %% % d e s c r i b e d i n Appendix A l [ce_image] = c o n t r a s t _ e n h a n c e ( s t a r t i m n o , e n d i m n o , d e l I _ I o a r r a y , m a x v a l _ p a l _ a , SavePath,rawimprefix,I_Ioimsufix,tiffext); %% C a l c u l a t e d e c a y v a l u e s and f i t s u b r o u t i n e %% [Aexpt_mean] = c a l c _ d e c a y ( n o _ d e c a y s , n o _ d e s _ i m _ a t _ e n d , d e l I _ I o a r r a y , . . . no_ads_im_in_set,hsmask,min_bin_yalue, max_bin_value,no__bins,SavePath,. rate_constant_outputfilename,hist_outputfilename);  *© o o o "0 "0 "o  o "o "S "0 o 0 * 0 0 * 0  %% end o f s c r i p t o'o'S'So  o o "0 "S "0 "6 o o o  %%  o'S'So'S  226  %  calc_decay.m  "b'o'o'S'S'&'o'o'S'S'o'oo o ' S ' S ' o ' o o o  o o o o o  oooo^oo^^^^o^o^^^'S'S'o'o'o'o'oo'So'o'S'b'ooo'b'bo'b'b'S'oo'o'o'S'o  o o o o " 5 o o o  % c a l c _ d e c a y . m i s a s u b r o u t i n e i n f r e a p . m t h a t c a l c u l a t e s t h e mean h i s t o g r a m % % f r o m t h e D e l t a l / I o images o b t a i n e d f r o m f r e a p . m a n d e x p o s e d t h e s e v a l u e s t o % % a f i t t i n g r o u t i n e (FMINSEARCH) % % %  %  %  % Jeff  Shepherd,  September 4, 2004  %  o o o o ^ ^ ^ o ^ o o o o i  f u n c t i o n [Aexpt_mean] = c a l c _ d e c a y ( n o _ d e c a y s , n o _ d e s _ i m _ a t _ e n d , . . . delI_Ioarray,no_ads_im_in_set,hsmask,min_bin_value, max_bin_value,... n o _ b i n s , S a v e P a t h , r a t e _ c o n s t a n t _ o u t p u t f i l e n a m e , h i s t _ o u t p u t f i l e n a m e ) ,% c a l c u l a t e t h e mean h i s t o g r a m v a l u e s o f e a c h image i n t h e d e c a y if no_des_im_at_end==l n o _ d e c a y s = n o _ d e c a y s ,else no_decays=no_decays-l; end  experiment  % c a l c u l a t e t h e mean h i s t o g r a m u s i n g h i s t _ m e a n _ s t d e v 2 d e s c r i b e d i n a p p e n d i x A l for decay_no=l:no_decays for j=(decay_no-l)*25+l:(decay_no-l)*25+5 [h_mean, h _ s t d e v ] = h i s t _ m e a n _ s t d e v 2 ( d e l I _ I o a r r a y { j } , hsmask,... min_bin_value, max_bin_value, n o _ b i n s ) ; hist_mean(j)=h_mean; hist_stdev(j)=h_stdev; end  for  j=(decay_no-l)*25+6:(decay_no-l)*25+25 [h_mean, h _ s t d e v ] = h i s t _ m e a n _ s t d e v 2 ( d e l I _ I o a r r a y { j } , min_bin_value, max_bin_value, n o _ b i n s ) ; hist_mean(j)=h_mean; hist_stdev(j)=h_stdev;  hsmask,...  end f i g u r e (18) ; i f d e c a y _ n o == 1 hold o f f ; else h o l d on; end plot(hist_mean); % c a l c u l a t i o n o f Ap f r o m t h e e x p e r i m e n t a l d a t a w h i c h % two c o n s e c u t i v e d e s o r p t i o n images b e t w e e n d e c a y s  the difference  between  dec ay_no < no_de c a y s  if  decay_no<no_decays Apexplmage=newimar(1) ; Apexplmage = d e l I _ I o a r r a y { ( d e c a y _ n o ) * 2 5 + 6 } ; Apexplmage = d e l I _ I o a r r a y { ( d e c a y _ n o - l ) * 2 5 + 6 } - A p e x p l m a g e ; [expAp_mean, e x p A p _ s t d e v ] = h i s t _ m e a n _ s t d e v 2 ( A p e x p l m a g e , hsmask,...  227  min_bin_value, max_bin_value, n o _ b i n s ) ; Apexpt_stdev =expAp_stdev; Apexpt_mean = expAp_mean; elseif no_des_im_at_end==l Apexplmage=newimar(1); Apexplmage = d e l I _ I o a r r a y { ( d e c a y _ n o ) * 2 5 + 6 } ; Apexplmage = d e l I _ I o a r r a y { ( d e c a y _ n o - 1 ) * 2 5 + 6 } - A p e x p l m a g e ; [expAp_mean, e x p A p _ s t d e v ] = h i s t _ m e a n _ s t d e v 2 ( A p e x p l m a g e , hsmask,... min_bin_value, max_bin_value, n o _ b i n s ) ; Apexpt_stdev =expAp_stdev; Apexpt_mean = expAp_mean; end %% c a l c u l a t e k a , kp, Ap, and Ao f r o m D e l t a l / I o d a t a f r o m I m a g e a r r a y { j } %% f i t t h e d e c a y s t o t h e c o n s e c u t i v e 1 s t and 2nd o r d e r r a t e e x p r e s s i o n %% c a l l t h e f i t t i n g r o u t i n e - n o t f v a l i s m u l t by l e 6 a f t e r c a l c for  j=(decay_no-l)*25+6:(decay_no-l)*25+25 Aexpt_mean(j-((decay_no-l)*25+6)+1)= hist_mean(j); A e x p t _ s t d e v ( j - ( ( d e c a y _ n o - l ) * 2 5 + 6)+1) = h i s t _ s t d e v ( j ) ;  end % i n i t i a l guess v a l u e s ka_kp_Ao_do_o = [.1, .1, .1, . 1 ] ; % i n t i a l g u e s s v a l u e s o p t i o n s = o p t i m s e t ( ' M a x l t e r ' , 5000, M a x F u n E v a l s ' , 5000, ' D i s p l a y ' , ' i t e r ' ) ; %, ' D i s p l a y ' , 'iter'); 1  1  TolX',  10 -10,  % f i t t i n g r o u t i n e u s i n g t h e i n i t i a l g u e s s p a r a m e t e r s and c o n s t r a i n t s % by ' k a _ k p _ A p _ d o _ n o _ w t _ f i t _ d b l o k ' [k_A,fval,exitflag] = FMINSEARCH('ka_kp_Ap_do_no_wt_fit_dblok',... ka_kp_Ao_do_o, o p t i o n s , Aexpt_mean, Apexpt_mean);  A  define  ka = k _ A ( 1 ) ; kp = k _ A ( 2 ) ; Ao = k _ A ( 3 ) ; do = k _ A ( 4 ) ; t = 2 0; Apcalc = kp/ka*log((kp+Ao*ka-ka*Ao*exp(-kp*t))/kp); % calculate A(t) for i=l:20 t(i) = i-1; Acalc(i) = ((kp*Ao*exp(-kp*t(i))/(kp+Ao*ka-ka*Ao*exp(-kp*t(i))))+do); W(i) = ( ( A e x p t _ m e a n ( i ) '- A c a l c ( i ) ) / ( A e x p t _ s t d e v ( i ) ) ) 2 ; end %plot data &f i t f i g u r e (decay_no+100) ,s u b p l o t (2,1,1).; plot(t,Aexpt_mean,'b.'); h o l d on; plot(t,Acalc,'r-'); hold o f f ; subplot(2,1,2); p l o t ( t , W , 'b-. ) ; A  1  % write data to f i l e i f (decay_no == 1)  228  f i d = f o p e n ( s t r c a t ( S a v e P a t h , r a t e _ c o n s t a n t _ o u t p u t f i l e n a m e ) , 'wt+'); else f i d = fopen(strcat(SavePath, rate_constant_outputfilename), 'at+'); end f p r i n t f ( f i d , ' % s \ n ' , O u t p u t k i n e t i c f i t d a t a , u s i n g c o n s e c u t i v e 1 s t and 2nd o r d e r r a t e law and A p ' ) ; fprintf(fid,'%6s\t%15s\t%15s\t%15s\t%15s\t%15s\n','Decay#', ka/min1 , kp/min-1','Apcalc','Apexpt', d_contrib ); 1  1  1  1  1  1  fprintf(fid,'%6d\t%12.8e\t%12.8e\t%12.8e\t%12.8e\t%12.8e\n',decay_no,ka,kp,Ape ale,Apexpt_mean,do); fprintf(fid,'%15s\t%15s\t%15s\t%15s\n', time/min','Aexpt','Astdev','Acalc'); f o r i=l:length(Aexpt_mean) fprintf(fid,'%12.8e\t%12.8e\t%12.8e\t%12.8e\n',i1,Aexpt_mean(i),Aexpt_stdev(i),Acalc(i)); end fprintf(fid,'%s\t%12.8e\n','fval',fval); fclose(fid); end 1  %write histogram data to f i l e f i d = f o p e n ( s t r c a t ( S a v e P a t h , h i s t _ o u t p u t f i l e n a m e ) , 'wt+ ); f p r i n t f ( f i d , ' % s \ n ' , ' H i s t o g r a m o b t a i n e d from a n a l y s i s ' ) ; f p r i n t f ( f i d , ' % 1 5 s \ t % 1 5 s \ n ' , ' A v g H i s t ' , 'Hist stdev'); for i=l:length(hist_mean) fprintf(fid,'%12.8e\t%12.8e\n',hist_mean(i),hist_stdev(i)); end fclose(fid); 1  229  %  ka_kp_Ap_do_no_wt_fit_dblok.m > o o o o o o o o o o o o o o o o "6 o o ' S ' S  'S'i'i'S'o'S'o'S'S'S'S^'o'S'S'S'S'i'S'S'SS'S  % A subroutine within % the e r r o r . % Jeff  Shepherd,  function  calc_decay.m that  September 4,  the d i f f e r e n c e  %break out parameters ka = x ( l ) kp = x ( 2 ) Ao = x ( 3 ) do = x(4)  e x p t _ d a t a , Apexpt) d e c a y and  the expt'1  vector  %expt d a t a i s a r r a n g e d i n time: e x p t _ d a t a ( l ) s q s u m e r r o r = 0; for  o "6 "5 *S "5  % % %  between t h e c a l c ' d  from x  o "5 "S o ' S ' S ' 5 ' S o ' S ' S ' S ' 5  of the square of  2004  [sqsumerror] = k a _ k p _ A p _ f i t ( x ,  % calculate  m i n i m i z e s t h e sum  = Omin  j=l:20 t = j - 1 ; %minutes Acalc = ((kp*Ao*exp(-kp*t)/(kp+Ao*ka-ka*Ao*exp(-kp*t)))+do); sqsumerror = sqsumerror + ((expt_data(j) - A c a l c ) ) * 2 ;  end % e r r o r f o r Ap // n o t e l o g i s I n , l o g l O i s l o g t=2 0; Apcalc = (kp/ka*log((kp+Ao*ka-ka*Ao*exp(-kp*t))/kp)); sqsumerror = sqsumerror + ((Apexpt - A p c a l c ) ) ^ 2 ; % f l a g any n e g a t i v e v a l u e s i f (ka < 0) sqsumerror = sqsumerror*10 6; e l s e i f (kp <0) sqsumerror = sqsumerror*10*6; e l s e i f (Ao<0) s q s u m e r r o r = s q s u m e r r o r * 10*6 ,e l s e i f (do<0) sqsumerror = sqsumerror*10*6; end A  230  data  

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.831.1-0061078/manifest

Comment

Related Items