Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Spectroelectrochemical characterization of self-assembled monolayers on a single crystal Au bead electrode… Yu, Zhinan 2017

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

Item Metadata


24-ubc_2017_february_yu_zhinan.pdf [ 11.22MB ]
JSON: 24-1.0340627.json
JSON-LD: 24-1.0340627-ld.json
RDF/XML (Pretty): 24-1.0340627-rdf.xml
RDF/JSON: 24-1.0340627-rdf.json
Turtle: 24-1.0340627-turtle.txt
N-Triples: 24-1.0340627-rdf-ntriples.txt
Original Record: 24-1.0340627-source.json
Full Text

Full Text

SpectroelectrochemicalCharacterization of Self-AssembledMonolayers on a Single Crystal Au BeadElectrode: the Influence of SurfaceCrystallographybyZhinan YuB.Sc., Nanjing University, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate and Postdoctoral Studies(Chemistry)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)January 2017© Zhinan Yu 2017AbstractSelf-assembled monolayers (SAMs) are important structures commonly employed to function-alize metal surfaces. To optimize a metal-SAM construct, it is important to characterize theinfluence of the surface crystallography. In this thesis, a single crystal Au bead electrode wasemployed to investigate different types of SAMs, enabling studies on a variety of surfaces underidentical conditions and avoiding laborious experimental replicates on a large number of crystalorientations.The application of a single crystal Au bead electrode was demonstrated by investigatingthe reductive desorption process for two types of SAM: the alkanethiolate SAM and the α-aminoisobutyric acid (Aib) peptide thiolate SAM. Using in situ fluorescence imaging, the influ-ence of surface crystallography on reductive desorption was observed, reflected as a correla-tion between the density of broken bonds of a surface and the reductive desorption potential ofa SAM deposited on the surface. Besides the surface crystallography, intermolecular interac-tions also had a significant impact on determining the desorption potential.Aib peptide thiolate SAMs on a Au(111) facet were further investigated. It was found thatthe low packing density Aib peptide thiolate SAMs exhibited a potential-modulated fluorescenceresponse which was believed to be due to the orientational or structural change of the peptidemolecules in response to the applied potential.The potential-driven reorientation effect of the DNA SAMs has been intensively exploreddue to its application in biosensing. Characterization of the DNA SAMs with in situ fluorescencemethods suggested that surface crystallography exerts an influence not only on the formationof the DNA SAMs but also on the efficiency of the potential-driven response. Moreover, aspectroelectrochemical technique that couples electrochemistry, fluorescence microscopy andharmonic analysis was developed to explore the non-linearity of the fluorescence response toan applied AC potential. This technique could be potentially applied to detect changes in DNAiiAbstracthybridization state.The experimental results demonstrate the convenience and wide applicability of using asingle crystal Au bead electrode to investigate SAMs. On the other hand, applying existingand developing new spectroelectrochemical techniques give insights into creating SAMs withdesirable properties.iiiPrefaceAll the experimental work presented in this thesis was performed by the author in AdvancedMa-terials and Process Engineering Laboratory (AMPEL) at University of British Columbia (UBC).The experimental design and data analysis was performed in collaboration with Dr. Dan Biz-zotto (the research supervisor).Contents from two publications have been included in part of the thesis.A journal article (Yu, Z. L.; Casanova-Moreno, J.; Guryanov, I.; Maran, F.; Bizzotto, D.“Influence of Surface Structure on Single or Mixed Component Self-Assembled Monolayers viain Situ Spectroelectrochemical Fluorescence Imaging of the Complete Stereographic Triangleon a Single Crystal Au Bead Electrode” J. Am. Chem. Soc. 2015, 137, 276-288.) covers thecontents presented in Chapter 4 and a portion of Chapter 6. The thesis author performed allthe experiments in this publication and the majority of data analysis with the advice from Dr.Dan Bizzotto. Dr. Jannu Casanova-Moreno optimized the DNA SAM preparation procedurespreviously and these procedures were followed by the thesis author in this publication. Dr. IvanGuryanov and Dr. Flavio Maran at University of Padova synthesized the BODIPY fluorophorelabeled thiol-modified Aib peptide HS-Aib4-BODIPY. Dr. Dan Bizzotto structured themanuscriptwhich was edited by the thesis author and Dr. Jannu Casanova-Moreno.Part of a book chapter (Casanova-Moreno, J.; Yu, Z. L.; Massey-Allard, J.; Ditchburn, B.;Young, J. F.; Bizzotto, D. “In-situ Spectroelectrochemical Fluorescence Microscopy for Visu-alizing Interfacial Structure and Dynamics in Self-Assembled Monolayers” In Luminescence inElectrochemistry: Applications in Analytical Chemistry, Physics and Biology; Miomandre, F. andAudebert, P., Eds.; Springer International Publishing, 2017.) contains the content presented inChapter 5. The thesis author was responsible for experiments, data analysis and manuscriptcomposition for the section related to the Aib peptide thiolate SAMs, with advice and edits fromDrs. Dan Bizzotto and Jannu Casanova-Moreno.ivPrefaceThree important contributions to the work presented in this thesis are worth highlighting.Brian Ditchburn, the glassblower in Department of Chemistry, UBC made all the cells forelectrochemical and spectroelectrochemical measurements.Arnold Kell and Dr. Mark S. Workentin at University of Western Ontario synthesized theBODIPY fluorophore labeled alkanethiol, HS-C10-BODIPY, enabling the studies of the alkan-thiol SAMs presented in Chapter 4.Drs. Ivan Guryanov and Flavio Maran synthesized the HS-Aib4-BODIPY, enabling the stud-ies of the Aib peptide SAMs presented in Chapter 4 and Chapter 5. Dr. Flavio Maran also gavevaluable advice in these studies.vTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xNomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiiiAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiv1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Theoretical background and literature review . . . . . . . . . . . . . . . . . . . . 42.1 Surface crystallography of a face-centered cubic single crystal . . . . . . . . 42.2 Fundamentals of self-assembled monolayers . . . . . . . . . . . . . . . . . . . 212.3 Electrochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282.4 Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Experimental methodology and instrumentation . . . . . . . . . . . . . . . . . . 573.1 Adsorbates, reagents and materials . . . . . . . . . . . . . . . . . . . . . . . . . 573.2 Substrate fabrication and cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . 613.3 Electrochemical instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 63viTable of Contents3.4 Spectroelectrochemical instrumentation . . . . . . . . . . . . . . . . . . . . . . 654 Spectroelectrochemical investigation of the reductive desorption behavior ofself-assembled monolayers on a single crystal Au bead electrode due to theinfluence of surface crystallography . . . . . . . . . . . . . . . . . . . . . . . . . . 704.1 Reductive desorption of self-assembled monolayers . . . . . . . . . . . . . . . 714.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764.3 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764.4 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005 In situ fluorescence imaging characterization of the α-aminoisobutyric acidpeptide thiolate self-assembled monolayers on Au(111) surfaces . . . . . . . 1025.1 Aib peptides and Aib peptide SAMs . . . . . . . . . . . . . . . . . . . . . . . . . 1025.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1065.3 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1075.4 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1216 Spectroelectrochemical investigation of the potential-driven DNA reorientationon a single crystal Au bead electrode . . . . . . . . . . . . . . . . . . . . . . . . . 1226.1 The chemical and biophysical properties of DNA . . . . . . . . . . . . . . . . . 1236.2 Potential-driven DNA reorientation . . . . . . . . . . . . . . . . . . . . . . . . . . 1296.3 Harmonic analysis of nonlinear fluorescence response driven by AC potentialperturbation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1316.4 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336.5 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336.6 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1366.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1557 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1567.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156viiTable of Contents7.2 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162AppendicesA Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179A.1 Hg arc lamp spectrum and excitation filter bands . . . . . . . . . . . . . . . . . 179A.2 Spectra of the used fluorophores and their corresponding filter sets . . . . . 179viiiList of Tables2.1 The five types of two-dimensional Bravais lattices and their Lattice parameterrelationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.2 Step notations for the major crystallographic planes within the three side zonesof the stereographic triangle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.3 Work function and potential of zero charge for the three low-index crystallo-graphic planes of Au . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.1 Fluorophores and corresponding filter sets . . . . . . . . . . . . . . . . . . . . . 675.1 Comparison of 310-helix and α-helix. . . . . . . . . . . . . . . . . . . . . . . . . . 1055.2 AuS-Aib4-BODIPY SAMs prepared and their formation conditions . . . . . . . 1086.1 Comparison of the three dsDNA conformations: B form, A form and Z form. . 1246.2 The two categories of MCH-DNA SAMs prepared and their formation conditions 134ixList of Figures2.1 Unit cell of Au (created with CrystalMaker). . . . . . . . . . . . . . . . . . . . . . 62.2 Constructing the stereographic projection of the crystallographic planes by: (a)vectorial projection of the crystallographic planes onto a sphere and (b) stereo-graphically projection from the sphere onto the equator plane. . . . . . . . . . 102.3 (a) A circular stereographic projection map (created with WinWulff) and (b) thesteoreographic triangle of the cubic lattice created by perpendicular projection. 112.4 Comparison of (a) stereographic projection and (b) perpendicular projection. 122.5 (a) Top view and (b) side view of a lattice truncated at (511) direction. . . . . . 122.6 Top views of lattices truncated at a. (931) direction and b. (913) direction. . . 132.7 Top views of a lattice truncated at (a) (221) direction and (b) (310) direction. . 142.8 Top view of a lattice truncated at (531) direction. . . . . . . . . . . . . . . . . . . 152.9 Schematic of the reconstructed Au(111) surface. The open circles denote thetop layer of atoms and the crosses denote the second layer of atoms. . . . . . 192.10 Schematic of the (1×5) reconstructed Au(100) surface. The blue circles denotethe top layer of atoms and the green circles denote the second layer of atoms. 202.11 Schematic of the (1× 2) reconstructed Au(110) surface. . . . . . . . . . . . . 202.12 Schematic diagram of an ideal SAM structure. . . . . . . . . . . . . . . . . . . . 222.13 Schematic of an alkanethiol molecule immobilized on a Au surface with the threeangles describing the orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.14 Schematic diagram of arrangement of alkanethiol SAMs on Au(111) surface withthe S atoms located at the 3-fold fcc hollows. The (2p3× 3) unit cell is markedby short dash lines and the c(4× 2) unit cell is marked by the long dash lines. 262.15 Schematic of possible defects on a SAM formed on a polycrstalline Au surface. 27xList of Figures2.16 Experimental electrocapillary curves for Hg in various electrolyte solutions. . 292.17 The differential capacitance-potential curves of a Hg electrode in various con-centrations of NaF solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322.18 Schematic of the Gouy-Chapman-Stern model of the electric double layer neara metal surface with excess positive charge and the profile of the potential fromthe metal surface to the bulk solution. CH is the inner layer capacitance and CDis the diffuse layer capacitance. Note that the electric double layer eventuallymerges with the bulk solution which is represented by the solution resistance . 332.19 Schematic of the electron density profile near a metal surface based on jelliummodel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382.20 Schematic of the electron spill over and density smoothing effects near the metalsurface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392.21 Capacitance curve of a Au(100) electrode measured during the process of liftingthe (20× 5) reconstruction to (1× 1) in 0.01 M HClO4 solution. . . . . . . . . 422.22 Correlation between PZC and density of broken bonds for Au and Ag, whereopen circles represent the experimentally determined PZC in 0.01 M NaF at 25°C and pH 5.6 and crosses represent the calculated dbb of the crystallographicsurfaces studied. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432.23 Fluorescence life time of the Eu3+ complex fluorophore as a function of distancefrom a Ag surface with the filled circles representing the experimental data pointsand the solid line representing the theoretical prediction. . . . . . . . . . . . . . 472.24 The calculated decay rates as a function of distance from the metal surface ofthe major pathways near a 633 nm dipole emitter-Ag interface: photons, surfaceplasmons and lossy surface waves. . . . . . . . . . . . . . . . . . . . . . . . . . 492.25 The surface plasmons induced by a emitter in the near field: (a) the widelyspaced charge distribution induced by the emitter far away enables propaga-tion of the surface plasmons in space as redative decays; (b) the widely spacedcharge distribution induced by the emitter nearby leads to non-radiative decaysof the surface plasmons; (c) recovery of the non-escaping decays with the sub-strate of metal thin film placed on top of a high refraction index medium. . . . 52xiList of Figures2.26 Schematic of the inverted epifluorescence microscope used in this thesis. . . 532.27 Schematic of a filter set for the inverted fluorescence microscope. . . . . . . . 543.1 (a) Core structure of BODIPY fluorophores; (b) structure of BODIPY 493/503 (thebonds marked by an asterisk will be linked to the adsorbate terminal groups.);structures of (c) BODIPY H-dimer (DI) and (d) BODIPY J-dimer (DII). . . . . . 583.2 Structures of (a) Alexa Fluor® 488 and (b) Alexa Fluor® 647. The bonds markedby asterisks will be linked to the adsorbate terminal groups. . . . . . . . . . . . 593.3 Structures of (a) HS-C10-BODIPY and (b) HS-Aib4-BODIPY. . . . . . . . . . . 603.4 Schematic of the electrochemical setup. . . . . . . . . . . . . . . . . . . . . . . . 643.5 Schematic of the spectroelectrochemical setup. . . . . . . . . . . . . . . . . . . 663.6 (a) Potential profiles of the two potential stepping schemes and (b) events occur-ring during a potential step. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684.1 Cyclic voltammagrams measured during desorption of decanethiolate SAMs de-posited on (A) a polycrystalline Au bead electrode and (B) polished single crystalAu electrodes with the indicated crystallographic orientations. . . . . . . . . . 744.2 Brightfield optical image of the bottom of a gold bead electrode with (111) facetsencircled in red, (100) facet encircled in green and defect encircled in blue. . 784.3 Montage of selected fluorescence images taken from –1.14 V to –1.36 V (vs.Ag|AgCl) in –20 mV increments during the reductive desorption of the AuS-C10-BODIPY SAM created with a 15 min immersion time. . . . . . . . . . . . . . . . 804.4 Fluorescence images of the AuS-C10-BODIPY SAM (15 min immersion time)modified bead electrode acquired at: (a) −1.24 V (vs. Ag|AgCl) and (b) −1.30V (vs. Ag|AgCl) with overlay of crystallographic map showing the low-index andstepped surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824.5 Fluorescence intensity - potential for the three low-index and (210) surfacestaken from different quadrants around the Au bead electrode surface: (a) theAuS-C10-BODIPY SAM created with 15 min immersion time, (b) the AuS-C10-BODIPY SAM created with 18 h immersion time. . . . . . . . . . . . . . . . . . 84xiiList of Figures4.6 Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-C10-BODIPY SAM created with15 min immersion time, shown for the (100)-(111), (111)-(110) and (110)-(100)zones in the WNW stereographic triangle: (a) raw fluorescence intensity in log-arithmic scale false colored, with a white contour line drawn at an intensity of1 kcts/sec, similar to the dotted line in Figure 4.5; (b) the logarithm of the ratioof the fluorescence intensity to the maximum intensity for each pixel along thezones false colored, with the white contour line drawn at 10% of the maximumintensity. In both figures, the density of broken bonds calculated for each surfaceon the y-axis is included. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864.7 Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-C10-BODIPY SAM created withan 18 h immersion time, shown for the (100)-(111), (111)-(110) and (110)-(100)zones in the WNW stereographic triangle: (a) raw fluorescence intensity in log-arithmic scale false colored, with a white contour line drawn at an intensity of1 kcts/sec, similar to the dotted line in Figure 4.5; (b) the logarithm of the ratioof the fluorescence intensity to the maximum intensity for each pixel along thezones false colored, with the white contour line drawn at 10% of the maximumintensity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884.8 Montage of selected fluorescence images taken from –1.08 V to –1.30 V (vs.Ag|AgCl) in –20 mV increments during the reductive desorption of the AuS-Aib4-BODIPY SAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914.9 Fluorescence images of the AuS-Aib4-BODIPY SAM modified bead electrodeacquired at –1.18 V (vs. Ag|AgCl) with overlay of crystallographic map showingthe low-index and stepped surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . 924.10 Fluorescence intensity - potential for the three low-index and (210) surfacestaken from different quadrants around the AuS-Aib4-BODIPY SAM modified Aubead electrode surface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93xiiiList of Figures4.11 Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-Aib4-BODIPY SAM, shown forthe (100)-(111), (111)-(110) and (110)-(100) zones in the WNW stereographictriangle: (a) raw fluorescence intensity in logarithmic scale false colored, with awhite contour line drawn at an intensity of 1 kcts/sec, similar to the dotted linein Figure 4.10; (b) the logarithm of the ratio of the fluorescence intensity to themaximum intensity for each pixel along the zones false colored, with the whitecontour line drawn at 10% of the maximum intensity. In both figures, the densityof broken bonds calculated for each surface on the y-axis is included. . . . . 944.12 Maps of the interpolated reductive desorption potential (V (vs Ag|AgCl)) for theAuS-C10-BODIPY SAMs using a threshold of 1000 kct/sec (left column) or athreshold of 10% of the maximum intensity (right column). First row (a, b) for theAuS-C10-BODIPY SAM created with 15 min immersion time and second row (c,d) for the AuS-C10-BODIPY SAM created with 18 h immersion time. . . . . . 964.13 Maps of the interpolated reductive desorption potential (V (vs Ag|AgCl)) for theAuS-Aib4-BODIPY SAM using (a) a threshold of 1000 kct/sec or (b) a thresholdof 10% of the maximum intensity. . . . . . . . . . . . . . . . . . . . . . . . . . . . 974.14 The map of calculated density of broken bonds (dbb) for a fcc spherical surfaceobserved from the bottom with the 100 surface in the center. . . . . . . . . . . 975.1 The primary structure of the “+” thiol-modified Aib peptide series. . . . . . . . 1055.2 Cyclic voltammogram recorded during a reductive desorption measurement of ahigh-packing-density AuS-Aib4-BODIPY SAM deposited on a polished Au(111)electrode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1105.3 (a) Bright field image of a Au(111) facet on a single crystal bead electrode;(b) montage of selected fluorescence images from a high-packing-density AuS-Aib4-BODIPY SAM deposited on a single crystal bead electrode with a Au(111)facet in view, representing the overall fluorescence response of the layer from–0.925 V to –1.3 V (vs. Ag|AgCl); (c) the minimum projection image of the imagestack; (d) the maximum projection image of the image stack. . . . . . . . . . . 112xivList of Figures5.4 (a) In situ fluorescence intensity as a function of potential from a Au(111) facet forthe high-packing-density AuS-Aib4-BODIPY SAM deposited on an unpolishedsingle crystal bead electrode; (b) capacitance change per unit area as a functionof potential for the AuS-Aib4-BODIPY SAM modified electrode. . . . . . . . . 1145.5 Fluorescence images of a low-packing-density AuS-Aib4-BODIPY SAM taken at(a) 0 V (vs. Ag|AgCl) and (b) –0.55 V (vs. Ag|AgCl) with the (111) facet outlined. 1155.6 (a) Profile of the modulated potential steps; (b) the fluorescence response ofa low-packing-density AuS-Aib4-BODIPY SAM; (c) ordinate enlarged fluores-cence response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165.7 Percent fluorescence intensity change ((FEstep− FEbse)/FEbse) as a functionof step potential of a low-packing-density AuS-Aib4-BODIPY SAM. . . . . . . 1165.8 Schematic of the hypothesized structure or orientation changes in the Aib peptidethiolate responding to changes in the electrode potential. . . . . . . . . . . . . 1195.9 (a) Percent fluorescence intensity change ((FEstep− FEbse)/FEbse) as a func-tion of step potential and (b) fluorescence intensity as a function of potentialrecorded during reductive desorption, of the AuS-Aib4-BODIPY SAMs of vari-ous packing density and a AuS-C10-BODIPY SAM. . . . . . . . . . . . . . . . 1206.1 Structure of a 5’-nucleotide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236.2 Chemical structure of a dsDNA segment showing the complementary base pairs. 1256.3 Three-dimensional structure of a dsDNA segment with B form configuration. A,G, T, and C are colored in red, green, yellow and blue, respectively. . . . . . . 1256.4 Effective diameter of dsDNA as a function of ionic concentration. . . . . . . . 1276.5 The potential-dependent thickness of the DNA SAM immobilized on a Au(111)surface, determined by EC-AFM. The dashed linemarks the DNASAM thicknessat open-circuit potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1296.6 The potential-controlled fluorescence emitted from a fluorophore labelled dsDNAimmobilized on a Au surface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130xvList of Figures6.7 Fluorescence image of a MCH-ssDNA-AlexaFluor488 SAM on a single crystalAu bead electrode taken at −0.4 V (vs. SCE) indexed with (a) the low indexplanes shown for the full image with four quadrants; (b) low-index and steppedsurfaces in the NW quadrant; (c) kinked surfaces in the NW quadrant. The scalebar in each image is 200 μm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386.8 Fluorescence intensity profiles extracted from the WNW portion of Figure 6.7along the three crystallographic zones which connects the low-index planes: (a)(100)−(111), (b) (111)−(110), (c) (110)−(100) and the three zones from the alow-index surface to the turning point stepped surface of the opposite side: (d)(111)−(210), (e) (100)−(331), (f) (110)-(311). . . . . . . . . . . . . . . . . . . . . 1396.9 Fluorescence images: (a) a MCH-ssDNA-AlexaFluor647 SAM and (b) a MCH-dsDNA-AlexaFluor647 SAM on a single crystal Au bead electrode captured at−0.4 V (vs. SCE). A variety of ROIs with their assigned crystallographic orienta-tions are marked on the images. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1416.10 The potential-driven fluorescence response of a ROI inside a Au(111) facet (markedin Figure 6.9) for aMCH-ssDNA-AlexaFluor647 SAMand aMCH-dsDNA-AlexaFluor647SAM: (a) the profile of potential steps applied to drive the DNA reorientation; (b)the fluorescence response of MCH-ssDNA-AlexaFluor647 SAM; (c) the fluores-cence response of MCH-ssDNA-AlexaFluor647 SAM. The intensities at the steppotentials (Estep) and the based potentials (Ebase) are marked with filled circleand open circles, respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1436.11 The fluorescence intensity as a function of step potential of a ROI inside a Au(111)facet (marked in Figure 6.9) for a MCH-ssDNA-AlexaFluor647 SAM (y-axis onthe left) and a MCH-dsDNA-AlexaFluor647 SAM (y-axis on the right). Estep is aparticular step potential and Ebase is the base potential 0.35 V (vs. SCE) beforestepping to corresponding step potential. . . . . . . . . . . . . . . . . . . . . . . 1446.12 The fluorescence intensity as a function of step potential measured on ROIs from(a) (100) surface, (b) (910) surface and (c) (311) surface (marked in Figure 6.9)for a MCH-ssDNA-AlexaFluor647 SAM and aMCH-dsDNA-AlexaFluor647 SAM. 145xviList of Figures6.13 (a) Cyclic voltammograms of a MCH-ssDNA-AlexaFluor647 SAM and a MCH-dsDNA-AlexaFluor647 SAM and (b) fluorescence intensity of the two SAMsmea-sured with potential scanning from a ROI inside a (111) facet. The arrows in (b)depict the oscillation directions of the AC potential perturbation and the fluores-cence response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1476.14 Amplitudes of the first harmonic (solid line) and second harmonic (dashed line)signals of the current response driven by 50 mV rms potential perturbation ata series of DC potentials from 0.275 V (vs. SCE) to -0.35 V (vs. SCE) for theMCH-ssDNA SAM and (blue) the MCH-dsDNA SAM (red). . . . . . . . . . . . 1486.15 Measured and simulated amplitudes of the harmonics of the fluorescence re-sponse from a (111) featured ROI driven by 100 mV rms perturbation at a seriesof DC potentials from 0.2 V (vs. SCE) to -0.25 V (vs. SCE) for the MCH-ssDNASAM and (blue) the MCH-dsDNA SAM (red): (a) first harmonic, (b) second har-monic and (c) third harmonic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1496.16 Comparison of measured and simulated: (a) rms fluorescence intensity and (b)ratio of the first harmonic amplitude over the rms fluorescence intensity. . . . 1516.17 Second harmonic amplitude normalized with first harmonic amplitude as a func-tion of DC potential for a MCH-ssDNA SAM SAM, a MCH-dsDNA SAM and aurea treated MCH-DNA SAM measured from three ROIs: (a) a (111) ROI, (b) a(100) ROI and (c) a (910) ROI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153A.1 Hg arc lamp spectrum of an Excelitas X-Cite® exacte fluorescence illuminatorand the two excitation filter bands (Chroma ET470/40x and ChromaHQ620/60x).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179A.2 Spectra of BODIPY 493/503 (monomer) and its corresponding filter set (excita-tion: ET470/40x, dichromatic: T495LPXR, emission: ET525/50m). . . . . . . 180A.3 Spectra of AlexaFluor488 (monomer) and its corresponding filter set (excitation:ET470/40x, dichromatic: T495LPXR, emission: ET525/50m). . . . . . . . . . . 180A.4 Spectra of AlexaFluor647 (monomer) and its corresponding filter set (excitation:HQ620/60x, dichromatic: Q660LP, emission: HQ700/75m). . . . . . . . . . . . 181xviiNomenclatureα angular semi-apertureχM metal surface potential induced by dipoleϵ the dielectric constant of a mediumϵ0 the permittivity of free spacei surface excess of a speciesγ surface tension1/κ diffuse layer thickness, i.e., Debye lengthλ wavelengthμ¯ electrochemical potentialμi chemical potential of a species work functionϕ2 potential at OHPϕM potential at the electrode surfaceϕS potential at bulk solution4ϕ dipole barrierσ surface charge densityσD total charge density of diffuse layerxviiiNomenclatureσ total charge density of inner layerσM excess charge density on a metal surfaceσS the excess charge density on the solution side of a metal-solution interfaceθ fractional surface coverageθ1 angle of incidentθ2 angle of refractionτ measured lifetime of a fluorophoreτ0 intrinsic lifetime of a fluorophore surface packing densityA electrode areaC capacitanceCD diffuse layer capacitanceCD diffuse layer capacitanceCH inner layer capacitanceCH inner layer capacitanceCdl capacitance of the electric double layerCd differential capacitance of surface described by the simple capacitor modeld distance between a fluorophore and a metal surfaced distance between the two plates of a capacitordbb density of broken bondsdmin depth of focus or axial resolving powerxixNomenclaturedmin minimum resolved distanceE FRET efficiencyE electric potentiale elementary chargeE a positive potential at which a SAM is adsorbedEb base potential in a potential stepping processEd a negative potential at which a SAM is desorbedEf final potential in a potential stepping processEH(2)min potential at which the second harmonic intensity is minimumEi initial potential in a potential stepping processErms rms amplitude of an AC potential waveEs step potential in a potential stepping processF Faraday constantƒ aperture diameterƒ focal lengthƒ frequency of an AC perturbationFEbse fluorescence intensity extracted from the last image taken at the base po-tential immediately after the step potentialFEstep fluorescence intensity extracted from the last image taken at a step potentialh Planck constanthk  Miller indices cathodic currentxxNomenclaturem imaginary component of an AC currentRe real component of an AC currentk Boltzmann constantk wavevectorkr emission rate constantknr non-radiactive decay rate constantn charge numbern refraction index of a mediumn0 number concentration of a z:z type of electrolytep momentumQ total chargeR distance between a FRET donor-acceptor pairR0 Förster RadiusRS solution resistanceRS solution resistanceS0 electronic ground stateS1 first singlet electronic excited stateT absolute temperatureTm the temperature at which half of the DNA strands are in ssDNA stateV voltage applied across the two plates of a capacitor scan rate of a linear scan voltammetry measurementxxiNomenclature1 distance betwen IHP and the metal surface2 distance betwen OHP and the metal surface, i.e., inner layer thicknessAC alternating current strictly, broadly referring to a signal that changes period-ically with timeAib α-aminoisobutyric acidBODIPY the abbreviation of boron-dipyrromethene and the trade name for the seriesof fluorophores with the core structure of 4,4-difluoro-4-bora-3a,4a-diaza-s-indaceneCCD charge-coupled deviceCE counter electrodeDAQ data acquisition boardDC direct current strictly, broadly referring to a signal that does not change pe-riodically with timeDFT density functional theorydsDNA double-stranded DNAEC-AFM electrochemical atomic force microscopyEM gain electron-multiplying gainfcc face-centered cubicFLIM fluorescence-lifetime imaging microscopyFRET fluorescence resonance energy transferhcp hexagonal close packedHPLC high performance liquid chromatographyIHP inner Helmholtz planexxiiNomenclatureIPE ideal polarized (or ideally polarizable) electrodeLB film Langmuir–Blodgett filmLEED low-energy electron diffractionMCH 6-mercaptohexan-1-olNA numerical apertureOHP outer Helmholtz planePAGE polyacrylamide gel electrophoresisPCT potential of conformation transition during a DNA reorientation processPMT photomultiplier tubePZC potential of zero chargeQY quantum yieldRC circuit resistor–capacitor circuitRE reference electroderms root mean squareSAM self-assembled monolayerSCE saturated calomel electrodessDNA single-stranded DNASTM scanning tunneling microscropyTris 2-amino-2-(hydroxymethyl)propane-1,3-diolUV ultravioletWE working electrodexxiiiAcknowledgementsMy journey to Canada, to Vancouver and to University of British Columbia is a difficult butrewarding one. During this journey, so many people gave me helping hands, without whomthis work cannot be presented here. I am sincerely expressing my acknowledgments to to allof them.Dan, forgive me if I don’t call you Dr. Bizzotto in a formal document. I am so glad that I haveyou as my supervisor. Your sense of responsibility, your scope of knowledge, your passionin research, your openness to diversity and your easily approachable personality, guide methrough the difficulties and provide me with rewards during the journey.All previous and present members or joint-members of Bizzotto lab that overlap with me:Dr. Amanda Musgrove, Dr. Jannu Casanova-Moreno, Dr. Tony Yang, Dr. Nidal Ashwawreh,Issac Martens, Santa Maria Gorbunova, Jonas Pfisterer, Kaylyn Leung, Elizabeth Fisher, KamilKrawczyk, etc., I am very grateful to all of you for whatever helps you gave me. In particular,Amanda and Jannu, I enjoyed a lot when I was still a junior member of the group and you twohelp me so much, not only on work, but also out of work.The contributions from the staffs in Department of Chemistry or AMPEL may not always benoticeable but are always invaluable. In particular, the glass blower in Department of Chemistry,Brian Ditchburn, fabricated all the electrochemical and spectrochemical cells of incomparablequality, making possible all the experiments presented here.The valuable advice from experts in particular fields are highly appreciated: notably, Dr.Flavio Maran, who gave advice in Aib peptide, Dr. Hua-Zhong Yu and Dr. Thomas Doneux,who gave advice in DNA and Dr. Jeff Young and Jonathan Massey-Allard, who gave advice inmetal-mediated fluorescence quenching.Alumni from School of Chemistry and Chemical Engineering, Nanjing University, it is suchan unforgettable experience for me to continue to study and work with you in Department ofxxivAcknowledgementsChemistry, UBC. Ru Li, Zhibo Liu, Zhengyu Cheng, Hui Yang, Cheng Qian and Sanjia Xu, howmuch difficulty I would have without you.Zhaocheng Zeng, I have known you for almost twenty years and you are still one of my bestfriends. We both came to Canada to study and to work, albeit in different fields and in differentcities, but any time when one of us had difficulty, the other one delivered the the psychologicalsupport immediately. I will cherish this friendship forever. Yiwei Li, you have also been a fiendof mine for almost twenty years. It is true that we lost contact for many years since you movedto Canada, but what a good luck I have to meet you again in Vancouver.From a foreigner to a de-facto Vancouverite, this transition was indeed very smooth for mewith such supports. Aunt Koi-Chi Chan and family members, although you are only distantrelatives of me, you helped me settle down and often served me dinner, preventing me fromfeeling lonely. Two very friendly landlords, Eric Xu and Fluorence Tam, warmly hosted me, justas their family member, making me enjoy living in Vancouver and with Vancouverites.The time I spent in the UBC Table Tennis Club was really beneficial to me. It was not justabout relaxation and exercise. There, I not only made friends sharing the same hobby, but alsomade fiends from a variety of different fields. It might sound strange, but I am very grateful tothose who broaden my understanding in math, physics, computer science, medical science,even history and music. What is more, about 60% of this thesis was written in the club officeroom, which was a very nice room.Last but not least, my beloved parents and grandparents, you all have high expectations forme and uninterruptedly support me. I might not be able to satisfy all your expectations but I amsure your supports are not wasted. I am proud of having supports from you and I hope you willalso be proud of me eventually.xxvChapter 1Introduction1.1 RationaleSelf-assembled monolayers (SAMs) are organized surface structures formed when organiccompounds are spontaneously adsorbed onto a solid substrate. Such structures can modifythe properties of the solid substrate. Often the solid substrate is a metal (Au in this work) andthe SAM-metal construct can be further developed into an electrochemical biosensor, becausetwo fundamental elements of a biosensor, i.e., the bio-receptor and the transducer [1] can beconveniently incorporated. Therefore, investigating and engineering the SAM-metal constructhas developed into an important research topic in an effort to construct biosensing interfaceswith desirable performance [1–7]. There are many critical aspects of the SAM-metal constructin reaching the ultimate goal of optimizing the performance of biosensing interfaces, of whichan important one is the influence of surface crystallography on SAMs [5]. Intuitively, surfacecrystallography determines a number of surface physical or chemical properties, e.g., atomicarrangement, surface roughness, and surface energy, which in turn influence the performanceof the biosensing interface. The atomic arrangement can exert an impact on the structuresand surface concentrations of the immobilized molecules, further affecting the binding speci-ficity and sensitivity. Mixed-component SAMs are typically employed to control the density ofbio-receptors and improve the accessibility to analytes. However, it has been proven that thelocal density of bio-receptors in these mixed-component SAMs depends on the surface crys-tallography [8]. Furthermore, the stability of a biosensing interface is greatly dependent on theadsorption strength, which in turn depends on surface energy [4, 5]. Thus the influence ofsurface crystallography on SAMs is profound and requires detailed exploration.Studying the influence of surface crystallography on SAMs with single-crystal electrodes11.2. Scopehas been established and widely exploited [9–12]. However, the conventional approach in-volves preparing a series of single crystal electrodes with designated crystallographic orien-tations and conducting copious experiments on these electrodes. On one hand the types ofcrystallographic orientations were mainly limited to low-index ones, and on the other hand thecopious experiments required strict control of experimental conditions in order to achieve de-sirable reproducibility. Thus, in this thesis, a single crystal Au bead electrode with a familyof crystallographic surfaces is employed as the substrate for the SAMs in an attempt to self-consistently investigate the influence from the crystallographic orientations within the stereo-graphic triangle under identical conditions.Electrochemistry is a natural tool to characterize the properties of a SAM-metal constructpotentially as the platform for an electrochemical biosensor. However, conventional electro-chemical methods only inform on the average response of an interface. Electrochemical meth-ods can be coupled with various spectroscopy methods to achieve in situ spectroelectrochem-ical characterization with desirable resolution and to obtain information from different aspects[13]. Fluorescence microscopy, with a resolution down to the micrometer-to-submicrometerscale considering the Abbe resolution limit [14], is coupled with electrochemical methods toobtain such information as the lateral segregation of species on a surface and the longitudinalseparation of species from a surface. In addition, the in situ fluorescence techniques enablethe investigation of the potential-controlled properties of SAMs, which are essential for electro-chemical biosensing interfaces.1.2 ScopeThe single crystal Au bead electrode will be employed as the substrate for the SAMs of in-terest throughout this thesis. Several different types of SAMs will be explored and differentspectroelectrochemical techniques will be utilized to achieve a better understanding of theseSAMs. Since this thesis covers many disciplines, it is necessary to briefly review the back-ground theories behind these topics. The fundamental theoretical background combined with ageneral literature review will be presented in Chapter 2. Following the theoretical background,an overview of the general experimental methodology and instrumentation will be presented21.2. Scopein Chapter 3. In Chapter 4, a well-studied alkanethiolate SAM and a recently-developed α-aminoisobutyric acid (Aib) peptide thiolate SAM will be explored with an in situ fluorescenceimaging technique to demonstrate the influence of the surface crystallography on the reductivedesorption process and the convenience of using the single crystal Au bead electrode. SAMsformed with Aib peptides which have a special helical structure will be further explored andpresented in Chapter 5. Chapter 6 will include studies of an alkanethiolate-spaced DNA SAM.This SAM will be first investigated with the imaging technique to demonstrate the influence ofsurface crystallography on DNA formation. In addition, a harmonic analysis technique of non-linear fluorescence will be developed to explore the potential-driven reorientation effect of theDNA SAM, which can be potentially applied in DNA sensing. Finally the results of Chapter 4 to6 will be summarized in Chapter 7. In this chapter, the significance of using the single crystalAu bead electrode as the substrate will be discussed and future studies will be proposed.3Chapter 2Theoretical background and literaturereviewIn this chapter, a general review of theoretical background is presented in order to help un-derstand the disciplines covered in the thesis. The single crystal Au bead electrode will beemployed as the substrate for the SAMs of interest, so it is necessary to briefly discuss the sur-face crystallography of a face-centered cubic (fcc) single crystal. Fundamentals of SAMs arealso presented, highlighting the structures of SAMs on Au crystalline surfaces. Coupled electro-chemical methods and fluorescence techniques are the analytical tools used to investigate thesystems. General principles of electrochemistry are reviewed, with the focus on depicting theinterface between the SAM-modified Au surface and the electrolyte solution in electrochemicalterms. A brief review of fluorescence near a metal surface is given in an attempt to intuitivelydescribe the metal-mediated quenching effect. With the theoretical background reviewed, thischapter also aims at establishing a bridge between the systems studied and the methods em-ployed.2.1 Surface crystallography of a face-centered cubic singlecrystal2.1.1 OverviewCrystallography is the study of the periodic atomic structures in crystalline solids. Traditionally,crystallography mainly focuses on the three-dimensional arrangements of the atoms in bulkcrystals. As a consequence of the fast development of surface science and nano science wherenot all three dimensions have infinite atomic arrangements, surface crystallography becomes42.1. Surface crystallography of a face-centered cubic single crystalan important branch of crystallography. In bulk crystals, usually, three-dimensional periodicgeometric models simulate the systems with high accuracy. In surface crystallography, thesystems are two-dimensional in a macroscopic scale and three-dimensional in a microscopicscale. For simplification, the truncated bulk crystal periodic models can be used to describethe ideal crystal surfaces. However, since the periodicity breaks in one dimension, there aresome special structures and properties associated with the broken periodicity. Thus, somemodifications of the models are necessary to describe the system more accurately [15]. In thissection, what will be reviewed are the basic concepts of bulk crystals, the use of truncated bulkcrystal models for description of the ideal crystal surfaces and the necessary modifications ofthe models for description of the real crystal surfaces. Au is the metal of interest in this thesis,and has a face-centered cubic (fcc) structure, so the review of fcc crystallography will be themain focus.2.1.2 Basic concepts of bulk crystallographyThe key characteristic of a crystal is its three-dimensional periodic repetition in atomic arrange-ment. To describe this periodicity, a lattice of infinite size is introduced. Based on the symmetryof the crystal, three non-unitary lattice vectors , b and c can be chosen to form the basis ofthe coordinate. Within the lattice, any lattice point resumes the same environment after thetransitional operation of Equation 2.1 (,  and  can be any integers).r=+b+c (2.1)The smallest repeating unit of the lattice is the unit cell. The unit cell is a parallelepipedformed by varying ,  and  from 0 to 1. To describe a crystal more conveniently, usuallythere are atoms or molecules located in the vertexes of the parallelepiped. The relationshipof the lengths of the edges, , b and c and the angles between any of the two vectors α, βand γ, which signify the symmetry of the crystal, are called lattice parameters. Combining thecrystallographic restriction theorem, i.e., in a crystal lattice, only 1-fold, 2-fold, 3-fold, 4-foldand 6-fold rotation axes and rotoinversion axes (1-fold rotoinversion operation is equivalent toinversion operation, 2-fold rotoinversion operation is equivalent to mirror operation) are allowed,52.1. Surface crystallography of a face-centered cubic single crystalFigure 2.1: Unit cell of Au (created with CrystalMaker).in 1850, French physicist Auguste Bravais derived the 14 lattices categorized into 7 crystalsystems [16]. Thereafter, theses lattices are usually called Bravais lattices.For the type of lattice of interest, the fcc lattice (also called Cubic-F), the lattice parametersfollow: = b= c, α= β= γ= 90°, just as the other cubic lattices. The unit cell of Au is shownin Figure 2.1. As can be seen, the Au unit cell is the most fundamental type of fcc unit cell(as compared to: e.g., diamond). Each unit cell contains 4 Au atoms and the cell parameter equals 4.08 Å. The fcc lattice is highly symmetric and the fact that each element point is aAu atom even enhances the degree of symmetry in space. The space group F 4m 3¯2m (usuallysimplified as Fm3¯m) denotes the following symmetry elements: first direction - the 4-fold ro-tation axes along the cell edges and the mirror planes perpendicular to them; second direction- the 3-fold rotoinversion axes along the cell body diagonal direction; third direction - the 2-foldrotation axes along the cell face diagonal direction and the mirror planes perpendicular to them.2.1.3 Basic concepts of surface crystallographySurface crystallography can be derived from bulk crystallography. From Equation 2.1, a lat-tice can be described by the sum of spanning the three non-co-planar base vectors , b and62.1. Surface crystallography of a face-centered cubic single crystalc. Here, for example, if  is fixed, the three-dimensional vector r degenerates into a two-dimensional vector, which is used to describe a plane. Varying  results in a collection ofparallel planes. This suggests a lattice can be viewed as a stack of parallel planes.Miller indices (named after the Welsh mineralogist William Hallowes Miller) hk  have beenintroduced to denote a plane in a crystal lattice. Still considering the three-dimensional co-ordinate with the three base vectors , b and c, if a plane has intercepts r, s and t on thethree directions, then the plane can be expressed as r +s +t = 1 and the group of parallelplanes can be expressed as r +s +t = C. By multiplying both sides of the equation witha common multiple (this must be durable according to the Law of Rational Indexes [16]), thefinal expression can be written as Equation 2.2, where h, k, , are all integers and called Millerindices.h+ k+ = C (2.2)By convention, the greatest common divider of h, k,  is 1 since parallel planes experiencethe same environment in an infinite crystal.Miller indices are usually quoted in brackets. Indices in different types of brackets havedifferent physical meanings [15]. Indices in the round brackets (hk) denote a particular planewhere the indices can be negative. Indices in the curly brackets {hk} denote a family ofequivalent planes based on the symmetry of the lattice. Indices in these two types of brackets,namely (hk) and {hk} are used to denote crystallographic planes. The form of (hk) is seenmore often although strictly speaking in a lot of cases {hk} should be used. In this thesis,the form of (hk) is used for the highly symmetric fcc crystal as a routine without interrogatingthe strict definition. Moreover, for a cubic lattice, because of the high order of symmetry, thesequence of h, k,  is highly flexible, which is not true for the other crystal systems. For example,the plane of (210) and the plane of (120) are symmetrically equivalent in a cubic lattice. Thus,hereafter, h¾ k¾  is also followed as a routine. Worth mentioning is that indices in the squarebrackets [hk] denote a particular direction based on the base vectors a, b and c where theindices can also be negative. Likewise, indices in the angle brackets < hk > denote a familyof equivalent directions based on the symmetry of the lattice.72.1. Surface crystallography of a face-centered cubic single crystalThe crystallographic restriction theorem also applies to two-dimensional surfaces. As aresult, there are only five types of two-dimensional Bravais lattices (Table 2.1): oblique, rect-angular, centered rectangular, square and hexagonal.The space group of Au, F 4m 3¯2m not only denotes the symmetry elements in three funda-mental directions in space, but also denotes the type of two-dimensional lattice of an atomicmonolayer of the three directions. The first direction is the normal direction of the (100) plane,which belongs to the square lattice. The second direction is the normal direction of the (111)plane, which belongs to the hexagonal lattice, and the 3-fold rotoinversion axis becomes a ro-tation axis if an atomic monolayer is considered. The third direction is the normal direction ofthe (110) plane, which belongs to the rectangular lattice.2.1.4 The ideal fcc surface crystallographyAn atomic monolayer is simply a conceptual understanding. The more reasonable model isan infinite three-dimensional lattice truncated at a particular direction. However, the three-dimensional model is not always easy to present. Therefore, a stereographic projection of thefamily of crystallographic planes onto a two dimensional map is frequently used. To constructsuch a map, first, a lattice is encapsulated into an auxiliary sphere which shares the same cen-ter with the lattice. From the center of the lattice, all the crystallographic plane normal vectorscan find their intersections on the sphere surface (Figure 2.2a). If the lattice has a mirror plane,it is only necessary to project the planes above the mirror plane (usually a (100) plane thatcrosses the center for a cubic lattice) of the lattice onto a hemisphere. Connecting the pole ofthe blank hemisphere, and the plane intersections on the opposite hemesphere results in theintersections on the equator plane (Figure 2.2b). These intersections on the equator planesmake the stereographic projection map of the crystallographic planes (Figure 2.3a). Since thecubic lattice is highly symmetric, the circular stereographic projection map can be further sim-plified into a stereographic triangle with the three low-index planes (100), (111) and (110) as thevertexes. The stereographic projection can be extended to any parallel plane with respect tothe equator plane, as shown in Figure 2.4a. The points on the hemisphere representing crys-tallographic surfaces are sometimes perpendicularly projected onto this parallel plane (Figure82.1. Surface crystallography of a face-centered cubic single crystalTable 2.1: The five types of two-dimensional Bravais lattices and their Lattice parameter rela-tionshipTwo-dimensional Bravais lattice Lattice parameter relationship 6= b; γ 6= 90◦, 60◦or120◦ 6= b; γ= 90◦= b; γ 6= 90◦, 60◦or120◦= b; γ= 90◦= b; γ= 120◦92.1. Surface crystallography of a face-centered cubic single crystalFigure 2.2: Constructing the stereographic projection of the crystallographic planes by: (a) vec-torial projection of the crystallographic planes onto a sphere and (b) stereographically projectionfrom the sphere onto the equator plane. Adapted from [17] with permission from Springer BerlinHeidelberg.2.4b), because practically this is the projection realized during imaging. The stereographictriangle shown in Figure 2.3b is created by this perpendicular projection. Strictly speaking, itshould not be called a stereographic triangle any more, but this term has been widely used todescribe the crystallographic planes on a cubic crystal surface without specifying the projectionused.For the low-index planes such as (111) and (100), the atoms adopt the closest packingand the second closest packing in two-dimension, so the exposure of the underlying layers ofatoms is relatively small. However, for other crystallographic planes, even the third low-indexplane (110), there are significant “gaps” between close-packing rows of atoms. Thus, surfacecrystallography is dealing with three-dimensional systems where the z-axis is atomic scale.Herein (511) is used as an example to demonstrate the morphology of high-index planes.Figure 2.5a shows the top view of a lattice truncated at (511) direction. The red plane is theconceptual (511) plane with parallel rows of atoms above it and the rest below it. The gapsbetween parallel rows of close-packing atoms of the same height are relatively big so the lowerrows of close-packing atoms are exposed. Viewed from the side (Figure 2.5b) the crystallo-graphic plane looks like steps. Therefore, this type of surfaces are called stepped surfacescomposed of wide terraces and narrow steps. In the case of (511), the terrace plane is the102.1. Surface crystallography of a face-centered cubic single crystal554332553221331551755991533540211320210311310511410711610911910111110100975753431971321751531951421731931baFigure 2.3: (a) A circular stereographic projection map (created with WinWulff) and (b) thesteoreographic triangle of the cubic lattice created by perpendicular projection.112.1. Surface crystallography of a face-centered cubic single crystalFigure 2.4: Comparison of (a) stereographic projection and (b) perpendicular projection.Adapted from [8] with permission from American Chemical Society.(100) plane (indicated with the square in 2.5a) and the step plane is the (111) plane (indicatedwith the triangles in 2.5b). The terrace plane is usually one of the low-index planes with closepacking. The step width is usually one atom for an ideal macroscopic two-dimensional crystal-lographic plane simply because forming multi-atomic steps requires removing atoms below theconceptual plane by non-ideal truncation of the lattice. For a plane (hk) (h¾ k ¾ , excluding(111) and (100)), if h = k or k = , or if h · k ·  = 0, the plane is a stepped plane. Steppedplanes are located on the sides of the steoreographic triangle. (110) is sometimes viewed asa stepped plane composed of equal width of (111) terrace and (111) step.Figure 2.5: (a) Top view and (b) side view of a lattice truncated at (511) direction.A plane like (931) (Figure 2.6a) adopts a more complicated morphology. The step edgesthemselves are stepped. This type of surfaces are called kinked surfaces. The terrace planeis still one of the low-index planes with close packing. However, the step plane is one of thestepped surfaces described above. In the case of (931), the terrace plane is the (100) plane122.1. Surface crystallography of a face-centered cubic single crystal(indicated with the square in Figure 2.6a) and the step plane is the (331) plane. For a plane(hk) , if h 6= k 6=  and h · k ·  6= 0, the plane is a kinked plane. Kinked planes are locatedin the interior of the stereographic triangle. A kinked surface does not have a mirror plane.Consequently, changing the sequence of any two of the three indices h, k,  effectively yieldsthemirror image crystallographic plane. As is shown in Figure 2.6, the (931) plane and the (913)planes are mirror images of each other. Therefore, a kinked surface is chiral. Chiral surfacesare potentially useful in chirality recognition or separation [18–20]. However, this thesis does notfocus on chiral systems. Therefore, the mirror image crystallographic planes are not specificallydifferentiated and (hk) in which h¾k¾  is also used to represent the group of kinked surfacesregardless of their chirality.Figure 2.6: Top views of lattices truncated at a. (931) direction and b. (913) direction.To describe the morphology of stepped and kinked surfaces, Somorjai proposed the stepnotation [21, 22]. The Miller indices of a stepped surface can be decomposed into the combina-tion of the terrace plane Miller indices and the step plane Miller indices, expressed in Equation2.3.(hk) = nt(htktt)+ ns(hskss) (2.3)This is also called the additivity theorem which can be proved using the reciprocal latticevector calculations [15]. Note that the Miller indices here can be negative. The step notation isusually written in the form of Equation Surface crystallography of a face-centered cubic single crystal(hk)≡ (nt+ 1)(htktt)× ns(hskss) (2.4)The coefficients(nt+ 1) and ns have clear physical meanings: (nt+ 1) is the width of theterrace plane and ns is the width of the step plane. It can be seen from Figure 2.5b that the(511) plane is composed of the 3 atom wide (100) terrace and the 1 atom wide (111) step (theatom shared with the terrace and the step is usually assigned to the terrace), thus the stepnotation of (511) is 3(100)× (111).Worth mentioning is that the selection of terrace and step planes might not be unique.Taking (221) shown in Figure 2.7a as an example, (111) is the terrace plane, but either (110)(indicated with the rectangle in 2.7a) or (111) (indicated with the triangle in 2.7a) can be viewedas the step plane and the terrace width is dependent on the step plane chosen. Therefore, thestep notation of (221) is either 3(111)× (110) or 4(111)× (111) (strictly speaking, the stepplane here should be written as (11−1) or (111¯)). Since (110) itself is stepped and composedof equal widths of (111) planes, it is more natural to view (111) as the step when the long sideof the (110) rectangle is the step of (221). However, another plane (310) shown in Figure 2.7bdoes not have (111) as the step plane, because the short side of the (110) rectangle is the stepof (310) (indicated with the rectangle in 2.7b). Therefore, the step notation of (310) can be onlywritten as 3(100)× (110).Figure 2.7: Top views of a lattice truncated at (a) (221) direction and (b) (310) direction.Table 2.2 shows the step notations for the major crystallographic planes within the three142.1. Surface crystallography of a face-centered cubic single crystalside zones from (111) to (100), from (100) to (110) and from (110) to (111) in the stereographictriangle. The general trends are as below. The terrace plane is the closest low-index plane andthe step plane is the far end low-index plane of the same zone. (311), (210) and (331) are three“turning point” planes of the three zones where the two low-index planes of at the end of thezone can be viewed as either the terrace plane or the step plane. The group of (n+2, n+2, n)planes each has two step notations depending on the step plane chosen, the reason of whichwas given in the last paragraph.Figure 2.8: Top view of a lattice truncated at (531) direction.The kinked surfaces are more complicated surfaces. The step notation cannot express thethe kinked surfaces completely and unambiguously. Somorjai also proposed the microfacetnotation which also decomposes the kinked step into low-index planes [22]. The microfacetnotation is not discussed here. For a general understanding, similar to a step surface, a kinkedsurface that is close to a low-index plane has this low-index plane as the terrace plane. Usingthe additivity theorem, the kinked step plane can be found. From the stereographic triangle(Figure 2.3b), the kinked surfaces can be understood more intuitively. As the example, from(111) to (210) which is the turning point plane from (100) to (110), there exist a series of kinkedplanes. The kinked planes close to (111) should have (111) as the terrace plane and (210) andthe step plane until a turning point after which (210) should conceptually become the terraceplane. Since (210) is stepped, so the terrace plane now becomes either (100) or (110). Thesame should apply to the series of kinked planes from (100) to (331) and from (110) to (311).This three kinked plane zones intersect at (531) (Figure 2.8), which would be the turning pointplane of the three zones. From another aspect, in (531), the (100) terrace (indicated with a152.1.Surfacecrystallographyofaface-centeredcubicsinglecrystalTable 2.2: Step notations for the major crystallographic planes within the three side zones of the stereographic triangle (cited from andderived based on [22]).Miller indices Step notations(111)(322) 5(111)× (100)(533) 4(111)× (100)(211) 3(111)× (100)(311) 2(111)× (100)(n+2, n, n) (n+ 1)(111)× (100)(2n+1, 1, 1) (n+ 1)(100)× (111)(311) 2(100)× (111)(511) 3(100)× (111)(711) 4(100)× (111)(911) 5(100)× (111)(100)Miller indices Step notations(100)(510) 5(100)× (110)(410) 4(100)× (110)(310) 3(100)× (110)(210) 2(100)× (110)(n+1, 1, 0) (n+ 1)(100)× (110)(n+1, n, 0) (n+ 1)(110)× (100)(210) 2(110)× (100)(320) 3(110)× (100)(430) 4(110)× (100)(540) 5(110)× (100)(110)Miller indices Step notations(110) 2(111)× (111)(991) 5(110)× (111)(771) 4(110)× (111)(551) 3(110)× (111)(331) 2(110)× (111)(2n+1, 2n+1, 1) (n+ 1)(110)× (111)(n+2, n+2, n) (n+ 1)(111)× (110) or (n+ 2)(111)× (111)(331) 2(111)× (110) or 3(111)× (111)(221) 3(111)× (110) or 4(111)× (111)(553) 4(111)× (110) or 5(111)× (111)(332) 5(111)× (110) or 6(111)× (111)(111)162.1. Surface crystallography of a face-centered cubic single crystalyellow quadrilateral), (111) terrace (indicated with a blue triangle) and (110) terrace (indicatedwith a green quadrilateral) can be found.2.1.5 The real fcc surface crystallographyThe model of an infinite three-dimensional lattice truncated at a particular direction assumesthe atoms on the surface stay at the same positions as they were in the bulk. However, for areal crystal surface, this is not a reliable assumption because the atoms on the surface expe-rience uneven forces from the two sides as soon as the crystal is truncated. To account forthis environment change, typically, the atoms on the surface undergo necessary and orderedrearrangement to lower the surface energy.The top layers of the atoms are usually not located at the assigned positions as those inthe bulk crystal. A widely observed phenomenon is the top layers as a whole shift, typicallyvertically either closer to or farther away from the bulk crystal. This effect is defined as surfacerelaxation. The shift towards the bulk crystal is defined as inward relaxation, and the oppositeway is defined as outward relaxation. The surface relaxation of a number of metals has beenexperimentally measured with low-energy electron diffraction (LEED) or theoretically calculated[23–32]. The results are not always compatible among each other. The experiment conditionsand the calculation assumptions can always affect the results.The surface relaxation of the metal of interest, Au, is not easy to measure or calculateaccurately, because it is usually complicated by another surface effect: surface reconstruc-tion. Surface reconstruction can occur at many exposed crystallographic surfaces. The re-construction occurring on the three low-index crystallographic surfaces of Au has been wellknown and studied [33–43]. However, for Au, little or no reconstruction was observed on most(with some exception) stepped crystallographic surfaces with scanning tunneling microscropy(STM) [44, 45]. There is a lack of study on reconstruction on kinked crystallographic surfaces.Therefore, the basic concepts are introduced and reconstructions of the three low-index crys-tallographic surfaces of Au, (111), (100) and (110) will be reviewed.Surfacial atoms can reconstruct into a different structure as compared to the bulk structure.Surface reconstruction primarily occurs on the topmost layer of a truncated lattice. After sur-172.1. Surface crystallography of a face-centered cubic single crystalface reconstruction, this topmost layer resumes a structure with periodicity different from thelayers beneath it. The reconstructed layer and the unreconstructed layers (viewed as the ideallytruncated lattice) can be described with the Wood notation [46]:S(hk)− i(b11× b22)Rα− ηA (2.5)The wood notation treats the unreconstructed layers as the substrate S with the Miller in-dices (hk) and the reconstructed layer as the adsorbate A. For a crystal with a single type ofatom, the Wood notation is typically simplified asS(hk)− i(b11× b22)Rα (2.6)The symmetry of the reconstructed layer is indicated with i, which is either “p” (usually omit-ted) for “primitive” or “c” for “centered”. The reconstructed layer has a different two-dimensionalBravais lattice from the layers beneath. The unit cell base vectors are b1, b2 for the recon-structed layer and 1, 2 for the unreconstructed layers. If the unit cell for the reconstructedlayer is rotated with respect to the unit cell for the unreconstructed layers, the angle need to benoted in α (R simply means rotation).The reconstruction of the Au(111) surface was discovered with LEED [33]. It was generallyaccepted as a (22×p3) reconstruction [34] before detailed location of the atoms on the re-constructed layer by STM [37, 38]. Thereafter, the reconstruction of the Au(111) surface hasbeen also written as (23×p3). The schematic of the reconstructed Au(111) surface is shownin Figure 2.9. The mismatch of the top layer atoms and the second layer atoms is a resultof contraction of the top layer in the surface normal direction in order to reduce the surfacetension. Because of this surface contraction, the Au atoms at the two proximal positions tothe second layer, C and A end up in two different types of close-packing structure hollows: Cis the fcc type of packing, similar to the packing scheme in the bulk and A is the hexagonalclose-packed (hcp) type of packing. It takes 23 atoms to finish one period along the direction ofthe long side of the reconstructed unit cell (the [111¯] direction), which lines up with 22 atomsalong the same direction on the second layer. This model is a local structure, the actual surfacehas more complicated long range periodicity [38].182.1. Surface crystallography of a face-centered cubic single crystalFigure 2.9: Schematic of the reconstructed Au(111) surface. The open circles denote the toplayer of atoms and the crosses denote the second layer of atoms. Reprinted from [37] withpermission from the American Association for the Advancement of Science.The reconstruction of the Au(100) surface is more complex. When it was studied with LEEDin the early stage, depending on the resolution of the measurements, a variety of reconstructedunit cells were found: (1× 5), (20× 5), (34× 5), (26× 48), (26× 48) [34, 36]. The simpli-fied scheme of the reconstruction is the (1×5) reconstruction shown in Figure 2.10. Before thereconstruction, the two directions of the unit cell are symmetrically equivalent. The reconstruc-tion involves the contraction in one direction (the [ 1¯10] direction in Figure 2.10) and translationof atoms in the other direction (the [110] direction in Figure 2.10), which as a consequence,resumes quasi-hexagonal symmetry [43]. The actual reconstruction of the unit cell in eitherdirection is more subtle, also the reconstruction is dependent on the temperature [40], so thereexists a variety of reconstructed unit cells.The (1× 2) reconstruction of Au(110) has been determined with a number of techniques[35, 39, 41, 42]. The schematic is shown in Figure 2.11. The intuitive way of understandingthis reconstructed unit cell is the every other “missing row” of atoms along the direction of thelong side of the (110) unit cell (the [100] direction in Figure 2.11). The ideal (110) surfacecan be viewed as a stepped surface with (111) terrace and (111) step of equal widths. An idealsurface only allows one-atom steps. This reconstructed surface still has a (111) terrace and a(111) step of equal widths, but the step is a three atom step.192.1. Surface crystallography of a face-centered cubic single crystalFigure 2.10: Schematic of the (1× 5) reconstructed Au(100) surface. The blue circles denotethe top layer of atoms and the green circles denote the second layer of atoms. Adapted from[43] with permission from American Physical Society.Figure 2.11: Schematic of the (1×2) reconstructed Au(110) surface. Reprinted from [42] withpermission from Elsevier.202.2. Fundamentals of self-assembled monolayers2.2 Fundamentals of self-assembled monolayers2.2.1 OverviewAn interface is the boundary between two immiscible phases, and typically if one phase is thegas phase, the interface can be called a surface. As a routine, the term “surface” is frequentlyused in scientific documents without strictly referring to its definition, so it is also employedhere to describe an interface. Surface chemistry has become an intensely studied domainin material science as a result of the fast development of nano materials where surfaces aredominant compared to large scale materials.The most evident characteristic of a surface is the large gradients due to the different com-positions and properties of the two bulk phases between which the interface is formed. Thusit is proposed by some researchers that surfaces be viewed as a fourth state of matter [4].Moreover, the atoms or molecules located at the surfaces are experiencing unequal forces, sothey are more chemically active compared with those in the bulk phases. For instance, sur-faces of some metals can spontaneously adsorb certain organic species. This forms the basisof self-assembled monolayers (SAMs). For SAMs, organic species adsorbed on the surfacesof substrates from a gas phase or a solution phase tend to organize into crystalline structuresand assemble into arrays based on the arrangements of surface atoms, which in turn lower thesurface energy and stabilize the surface atoms.Figure 2.12 is a schematic diagram of an ideal SAM structure. Usually the adsorbateshave head groups which chemically bind to the surface with considerable stability. A widelyused head group is thiolate because it has a high affinity to a number of metals and semi-conductors. The other ends of the organic adsorbates are the terminal groups that can befurther linked to other structures to achieve future functionality. The head groups and terminalgroups are usually connected by repeated units defined as spacers. By adjusting the length ofthe spacers, the thickness and other monolayer characteristics can be tuned. Therefore, thephysical and chemical properties of surfaces modified with organic species are altered, or de-fined by the adsorbates, and this promotes surface functionalizations for designated purposes.In a sense, SAMs can act as platforms for surface process investigations. As an example,when the terminal functional groups are linked to bioreceptors, such as enzymes, nucleic acids212.2. Fundamentals of self-assembled monolayersFigure 2.12: Schematic diagram of an ideal SAM structure.or antibodies, the SAMs can be employed in biosensor applications [1]. A number of biosen-sors have been devised through the anchoring of different bioreceptors on the SAM platform.Among these, glucose sensors [47–51], DNA sensors [52–57] and immunosensors [58–64]have been extensively studied.2.2.2 SAMs based on Au-S interactionsSAMs based on Au-S interactions with Au as the metal substrate and thiolate as the headgroupare a class of SAMs widely studied and applied to many fields. The main advantage of Au-Sinteractions is its high stability (~200 kJ/mol [4]) which accounts for SAMs of good quality andhigh order. In addition, Au of different forms, such as colloids, thin films, polycrystals andsingle crystals are available to meet different needs. Furthermore, as a noble metal, Au isrelatively inert and not subject to oxidation under common conditions (e.g., being exposed inair at room temperature). Because of all the features above, SAMs based on Au-S interactionshave become a primary choice for surface science studies. Among all SAMs based on Au-Sinteractions, SAMs formed by depositing alkanethiols on Au(111) surfaces, the most stable Aucrystallographic surface, are the most intensively studied [4, 65, 66].Assembling SAMs on substrates is usually performed in the gas phase or via the solutionphase. Although the mechanism of assembly from gas phase is better understood theoretically,assembly from solution phase is practically speaking more convenient because not all thiolshave an appreciable vapor pressure. This method simply requires immersion of substrates intoadsorbate solutions for a certain period of time [4].222.2. Fundamentals of self-assembled monolayersThe mechanism of alkanethiol adsorption onto a gold surface is complex and sometimescontroversial. There is evidence to show a two-step mechanism [66–68]. The first step is aphysisorption process during which the alkanethiol molecules form a highly disordered stateon the surface (Equation 2.7). The second step is a chemisorption during which the Au-S bondis formed and the H atom is released. It is generally believed that the second step followseither of the two routes below (Equation 2.8 and Equation 2.9, [O] in Equation 2 representsO2 or other types of oxidizing agents present in the surroundings). It is also debatable whathappens to the H atom [4, 66, 69–72]. Note that these proposed steps have been based onstudies of alkanethiols on Au(111), but to date there is a lack of literature on whether they aregeneric to any SAMs based on the Au-S interaction.RSH+A(srfce)→ RSH(phys)A (2.7)2RSH(phys)A→ 2RS−A(SAM)+H2(g) (2.8)2RSH(phys)+ [O]→ 2RS−A(SAM)+H2O (2.9)In theory, the formation of SAMs is a process driven by stabilization of active substrate sur-faces. In Section 2.1.5, the reconstruction of a crystallographic surface was discussed. Thereconstruction is also a process to lower the surface energy. Therefore, upon adsorption of aSAM, the surface tension is relaxed and the surface reconstruction is lifted. The process oflifting the reconstructed Au(111) surface by alkanethiol adsorbates has been observed experi-mentally [68, 73]. It is generally believed that highly dense alkanethiol SAMs with the relativelyelectronegative thiolate head group should withdraw the excess negative charge on the recon-structed surface and revert the reconstructed surface back to its (1×1) symmetry (also termedde-reconstruction) [74]. However, lifting the reconstructed surface does not result in a topmostlayer becoming identical to the underlying layers. Take Au(111) as an example, the (23×p3)reconstruction results in a more dense surface with one additional atom per (23×p3) unitcell. Upon adsorption of thiol molecules, the reconstructed surface is relaxed and some atoms232.2. Fundamentals of self-assembled monolayersare lifted on top of the top layer so that the density can be lowered to be similar to the bulk.This process leads to adatoms and vacancies. The adatoms tend to aggregate into big islands,and at the same time the vacancies also grow. As a whole, the surface becomes rough anddefective [75].When an adsorbate molecule is immobilized onto a surface, the alkyl chain orients at aparticular direction. This orientation can be described using three angles (Figure 2.13): α isthe tilt angle with reference to the surface normal; β is the angle between the plane of all-transchains and the plane defined by surface normal vector and chain axis vector i.e., the angle ofrotation; and χ is the angle of precession which defines the tilt direction with reference to theplane defined by surface normal vector and unit cell base vector [65, 66, 76]. For alkanethiolateSAMs on Au surfaces, these angles are dependent on the chain length of the adsorbate andthe substrate crystal structure. On Au(111) surface, the tilt angle α is close to 30° [4, 77, 78]whereas for the Au(100) surface, it is only around half of that for the Au(111) surface [4, 77].The value of β is usually close to 45° for the Au(111) surface and 70° for the Au(100) surface[4, 77]. The general trend for χ is that χ decreases as the chain becomes longer. Chainlengths of n¶ 14 and n¾ 16 fall into two different domains: 13°¶ χ¶ 18° and 4°¶ χ¶ 10°,respectively [78].The arrangement of adsorbed thiolates greatly depends on the exposed crystalline sur-faces. The Wood notation (2.5) is also used to describe the periodicity of the adsorbates. Thewidely accepted pattern of alkanethiolates on Au(111) surface is the (p3×p3)R30° structure[4, 79]. Despite this well-established pattern, the adsorption site for the S atom is still underintense debate and not yet resolved. There is evidence from both experiments and calculationsdemonstrating that the S atom is located at the 3-fold hollow of the gold lattice (Figure 2.14,a = 2.88 Å) [4, 80–83]. Disagreement arises with density functional theory (DFT) studies pre-dicting the adsorption of the S atoms on the bridge site [84–87], but these studies have beenfocused on the adsorption of CH3SH or dissociated CH3SSCH3. Some recent experimentalstudies have also shown the atop adsorption site preference [88–90]. Furthermore, from thedynamic aspect, two types of adsorption sites or adsorption states may co-exist as a result ofdisulfide formation [91], progressing through different adsorption stages [67], or reorganizationdriven by two competing forces [92]. The (p3×p3)R30° unit cell only takes into account the242.2. Fundamentals of self-assembled monolayersFigure 2.13: Schematic of an alkanethiol molecule immobilized on a Au surface with the threeangles describing the orientation. Reprinted from [65] with permission from Royal Society ofChemistry.position of the S atoms. If the orientation of the alkane chain is also considered, the adsorbatesadopt a (2p3× 3) structure (Figure 2.14). An equivalent unit cell denoted as c(4× 2) super-lattice is often used in early literature [4, 76, 93–95]. Although it is commonly accepted thatthe superlattice is a consequence of alkane chains of two different orientations (i.e., differentχ), some evidence also implies the position of the S atoms also plays a role in the superlattice[76]. Apart from the Au(111) surface, there is a lack of systematic investigation of alkanethiolSAMs deposited on other gold surfaces. Alkanethiolate SAMs deposited on Au(100) have beenstudied, but the results are with much less agreement. Vapor-deposited methanethiol SAMsseem to prefer a c(2× 2) structure [4, 77], and medium to long chain alkanethiol SAMs seemto prefer a c(2× 8) structure with (1× 4) Au missing row [96, 97], whereas alkanethiol SAMsprepared from solution phase can adopt a distribution of different structures, presumably dueto different chain lengths [98–102]. These differences do not seem to be resolved yet. DFTstudies have also predicted the favorable adsorption sites for a variety of surfaces, howeverthey are not always in agreement with each other [83, 87].In theory, the formation of SAMs is a process driven by stabilization of active surfaces,so SAMs tend to reorganize into structures patterned by exposed crystalline structures with252.2. Fundamentals of self-assembled monolayersFigure 2.14: Schematic diagram of arrangement of alkanethiol SAMs on Au(111) surface withthe S atoms located at the 3-fold fcc hollows. The (2p3× 3) unit cell is marked by shortdashed lines and the c(4× 2) unit cell is marked by the long dashed lines. Reprinted from [4]with permission from American Chemical Society.high order and few defects. However, in practice, SAMs still possess large numbers of de-fects due to any non-ideal conditions involved in the processes of assembling SAMs. If theSAMs are further applied as building blocks for functionalization, the defects are inevitably af-fecting the performance. Although to achieve a perfect defectless SAM is almost impossible,the number of defects can be minimized by realizing the sources of defects [4, 65, 66]. Figure2.15 depicts the possible sources of defects in a SAM formed on a polycrystalline Au surface.Take assembling SAMs from solution phase as an example: there might be some impuritieseither in the solution or on the substrate, which subsequently lead to defects in SAMs. Alsothe substrate surfaces are far from ideally depicted crystalline faces of high order, so defects262.2. Fundamentals of self-assembled monolayersFigure 2.15: Schematic of possible sources of defects in a SAM formed on a polycrstalline Ausurface. Reprinted from [4] with permission from American Chemical vacancy islands, step edges or grain boundaries on the surfaces would eventually be in-herited by SAMs adsorbed on them. Even if the substrate is a carefully fabricated crystallinesurface with well-defined structure (e.g., Au(111)), a considerate number of defects can still beobserved [65, 66]. One important type of defects is related to the intermolecular interactionsbetween the adsorbate molecules. For alkanethiolate SAMs, the main intermolecular interac-tions is the van der Waals forces between chains. A large number of pinholes or pinstripesfeaturing missing molecules exist on short-chain (R¶C6) alkanethiolate SAMs due to weakintermolecular interactions. [103, 104], resulting in a disordered liquid-like state [105]. On thecontrary, long chain alkanethiolate SAMs (R¾C16) appear to be less defective, approaching acrystalline-like state [104, 105]. However, even long chain alkanethiol SAMs tend to be in a dy-namic gel-like state below the temperature of order-disorder (gel-liquid) transition. Even thoughtypically the transition temperature for alkanethiol SAMs on planar substrates is above roomtemperature, the annealing and organizing process is usually too long to finish, so the gel-likestate has a high motional gradient along the chain [106]. Consequently, alkanethiolate SAMscan undergo various phase transitions on substrate surfaces. For example, an unsaturated de-canethiol monolayer assembled on Au(111) surface can undergo transitions between as manyas six phases depending on the surface coverage [107]. Thus explains why alkanethiolateSAMs tend to be defective even on ordered crystalline surfaces.Forming mixed-component SAMs via the thiol-exchange mechanism (i.e., partially displac-ing a surface-bound adsorbate with another) has become important in recent years, especiallyin SAM-based biosensing interfaces (e.g., DNA SAMs) where a molecular diluent is typically272.3. Electrochemistryused to prevent the non-specific adsorption and control the density of the bioactive adsorbate[108–111]. The presence of defects on a SAM indeed facilitates this process, because the dis-placement of the adsorbedmolecules occurs rapidly on defective regions and slowly on orderedones [112–115]. Moreover, the thiol-exchange efficiency is dependent on the intermolecular in-teractions between adsorbate molecules. In general, an adsorbate with weak interactions canbe easily displaced by another one with strong interactions. For alkanethiols, since the chainlength mostly determines the extent interactions, a mixed-component alkanethiolate SAM canbe formed by displacing a short alkanethiol with a long one [4, 116].In most cases, SAMs can be removed from the substrates to regenerate unmodified sur-faces. This can be achieved by conventional thermal treatments, mechanical polishing andchemical oxidation methods. But these methods may not only remove the layers, but alsodamage the substrates as well. For thiolate SAMs deposited on Au, one alternative method isthrough photo-oxidation by exposing the SAMs to ultraviolet (UV) light to oxidize thiolate groupsto sulfonate groups [4, 117]. Another widely studied method of removing SAMs is the reductivedesorption, which is based on the reduction reaction below (Equation 2.10) [118]. Thus whena sufficiently negative potential is applied to a modified surface, the adsorbed thiolate layer canbe removed from the substrate. Note that the reductive desorption process is reversible, sore-adsorption via oxidation can take place if the applied potential is less negative [12, 119]. Re-ductive desorption is an important process which can be used to quantify the adsorbate densityor study the adsorbate-substrate interaction. This will be further reviewed and investigated inChapter 4.RS-A+ e−→ RS− +A(0) (2.10)2.3 Electrochemistry2.3.1 OverviewElectrochemistry is a branch of chemistry employed to study the interface between a metalphase and a solution phase. SAMs are structures concerning modified solid surfaces, so elec-282.3. Electrochemistry240280320360400440-1.5-1.2-0.9-0.6-γ (mN m-1 )E - PZCθ=0 (V)KOHCa(NO3)2NaClKCNSNaBrKIFigure 2.16: Experimental electrocapillary curves for Hg in various electrolyte solutions. Re-produced from [120] with permission from American Chemical Society.trochemical methods are valuable tools to investigate this type of system. In this section, basicconcepts and the electric double layer model to describe the metal (unmodified or modified)| solution interface will be reviewed. Based on the electric double layer model, the system ofSAMs can be characterized. However, for the actual SAMs deposited on a substrate with cer-tain crystal structure, the crystal structure inevitably influences the electrochemical propertiesof the SAMs. This will also be reviewed in this section.2.3.2 Basic conceptsThe invention of Hg dropping electrode opened the field of detailed electrochemical studies onan interface. Hg is the only liquid metal at room temperature and its surface tension can beeasily measured. Using the dropping mercury electrode, the surface tension of the Hg drop incontact with an electrolyte solution at various electrochemical potentials was measured. Thecurve of surface tension as a function of potential is called the electrocapillary curve. Figure2.16 shows the electrocapillary curves obtained in various electrolyte solutions [120]. Thesimilarity of these curves is that they resemble the shape of a parabola with a maximum surfacetension at a particular potential.292.3. ElectrochemistryThe electrocapillary curve is the experimental expression of the electrocapillary equationshown in Equation 2.11 [13, 120, 121]. In the electrocapillary equation, γ is the surface tension,σM is the excess charge density on the metal surface, E is the electric potential, i is the surfaceexcess of a species, which measures the amount of the species from phase A (the solution)partitioning into phase B (the metal surface) and μi is the chemical potential of the species.− dγ= σMdE+∑idμi (2.11)The electrocapillary curves converge at negative potential and diverge at positive potential.This is can be rationalized with the electrocapillary equation. At negative potential, the cationsare non-specifically attracted onto the metal surface purely due to electrostatic interaction andthe surface excess i for all cations is similar; at positive potential, some anions can be specif-ically adsorbed onto the metal surface through chemical interaction and the surface excess ifor different cations varies.From the electrocapillary equation, it is clear that the slope of the electrocapillary curve isthe excess charge density on the metal surface (Equation 2.12).σM = −(∂γ∂E)μ (2.12)When the surface tension reaches the maximum, the excess charge density on the metalsurface σM equals zero. The potential at which this occurs is called the potential of zero charge(PZC) of the metal. Deviating from PZC, there is excess amount of charge on the surface,which induces repulsion. This repulsion counterbalances the trend of shrinking the Hg dropby surface tension and thus decreases the surface tension. Although the PZC is discoveredfrom experiments with Hg electrodes, it is a generic parameter for any conductor in contact withelectrolytes.2.3.3 Electric double layer model of an electrode | solution interfaceIn many electrode | solution interfaces, species at the interfaces are relatively redox inactive,either thermodynamically or kinetically, with no electron transfer through the interface over a302.3. Electrochemistryrange of potential. An electrode of this type is called an ideal polarized (or ideally polarizable)electrode (IPE) [13]. The Hg electrode, which is used widely, behaves similarly to an IPEover a potential range of nearly 2 V. Electrodes modified with SAMs can be viewed as a classof IPEs. However, the range of potentials at which they are stable and considered as IPEsmay be restricted because they undergo reduction at negative potentials, oxidation at positivepotentials and sometimes potential-induced structural transformation before being reduced oroxidized.An IPE resembles a capacitor of which the most common form is two parallel metal platesseparated by dielectric media, since both systems do not allow charge transfer through the in-terfaces within certain potential ranges. Thus, an IPE can be described with a simple capacitormodel devised by Helmholtz [13]. In this Helmholtz Model, an IPE is modeled as a parallel-platecapacitor described by Equation 2.13 where σ is the surface charge density, V is the voltageapplied across the two plates, ϵ is the dielectric constant of the medium, ϵ0 (≈ 8.854 × 10–12Fm-1) is the permittivity of free space, and d is the distance between the two plates.σ =ϵϵ0dV (2.13)Consequently, the charge stored in the capacitor is linearly proportional to the potential and thecoefficient is the differential capacitance Cd (Equation 2.14).Cd =dσdV=ϵϵ0d(2.14)This simple model suggests that Cd is constant and independent of potential V, which iscontradictory to experimental results on IPE systems. Figure 2.17 shows the experimentalresults of the differential capacitance of a Hg electrode as a function of potential in variousconcentrations of NaF solutions [120]. An important feature is that at low electrolyte concen-trations, a minimum capacitance value can be observed at PZC. Therefore, a more complexmodel was proposed by Gouy and Chapman independently and later extended by Stern. ThisGouy-Chapman-Stern model is detailed in [13] and summarized here.In the Gouy-Chapman-Stern model (Figure 2.18), the electrode has an excess amount of312.3. Electrochemistry04812162024283236-2.0-1.5-1.0- d (µF cm-2 )E - PZC (V)1.0 M NaF0.1 M NaF0.01 M NaF0.001 M NaFFigure 2.17: The differential capacitance-potential curves of a Hg electrode in various con-centrations of NaF solution. Reproduced from [120] with permission from American ChemicalSociety.charge with a density σM, while the solution side with the total excess charge density of σS, ismodeled as consisting of two theoretical layers made up of ions in the solution to compensatefor the excess charge. The layer closest to the electrode surface is the inner layer, sometimesalso called compact, Helmholtz or Stern layer, with the total charge density σ. The inner layermainly contains solvent molecules and sometimes some specifically adsorbed species with theelectrical centers located at a distance of 1 away from electrode surface at inner Helmholtzplane (IHP). The smallest distance to which solvated non-specifically adsorbed ions can ap-proach the electrode is defined as 2, the distance of outer Helmholtz plane (OHP) from theelectrode surface. Non-specifically adsorbed ions are not concentrated close to the surface,instead, they are distributed from OHP to bulk solution in the layer called diffuse layer with thetotal charge density σD. In this model, the charge densities of the two layers follow the relationof Equation 2.15 at equilibrium.σS = σ+ σD = −σM (2.15)The main goals of this model are to derive the potential profile from ϕM at the electrode322.3. ElectrochemistryFigure 2.18: Schematic of the Gouy-Chapman-Stern model of the electric double layer near ametal surface with excess positive charge and the profile of the potential from the metal surfaceto the bulk solution. CH is the inner layer capacitance and CD is the diffuse layer capacitance.Note that the electric double layer eventually merges with the bulk solution which is representedby the solution resistance .332.3. Electrochemistrysurface to ϕS at the bulk solution and the capacitance of the electric double layer Cdl. It is notattempted here but some of the most important points are highlighted in order to help describethe electrode | solution interface.The potential at the electrode | solution interface decays from ϕM to ϕS following the profiledepicted by Figure 2.18 if only non-specifically adsorbed ions are considered. To be specific,in the inner layer, the potential decays linearly from ϕM on electrode surface to ϕ2 at OHPwhereas in the diffuse layer, the potential decays non-linearly from ϕ2 at OHP to ϕS in the bulksolution. The potential decays non-linearly following Equation 2.16 and Equation 2.17 in thediffuse layer for z : z type of electrolyte, where e is the elementary charge, k is the Boltzmannconstant, T is the temperature and n0 is the concentration of the electrolyte.tnh(zeϕ/4kT)tnh(zeϕ2/4kT)= exp(−κ(− 2)) (2.16)κ=√√√(2n0z2e2ϵϵ0kT) (2.17)Two parameters are worth emphasizing here: 2 is the inner layer thickness and 1/κ canbe viewed as the diffuse layer thickness (also termed Debye length). The inner layer thickness,in ångström scale, is usually 1.5 to 2.0 times the thickness of a monolayer of solvent molecule[122]. While it is independent of the solid surface, the inner layer thickness becomes smallerwith an increase of electrolyte concentration [123]. The diffuse layer thickness is strongly in-fluenced by the electrolyte concentration as can be seen from Equation 2.17. With a highelectrolyte concentration, the diffuse layer can reach single molecular layer thick while witha low electrolyte concentration, the diffuse layer can extend to hundreds of nanometers thick[13, 122]. The roughness of themetal surface also has a strong impact on the diffuse layer thick-ness. Generally, the diffuse layer thickness decreases with increasing roughness [124, 125].The capacitance of the electric double layer in this model can be calculated with Equation2.18.1Cdl=2ϵϵ0+1(2ϵϵ0z2e2n0/kT)1/2cosh(zeϕ2/2kT)(2.18)342.3. ElectrochemistryThe capacitance of the electric double layer Cdl can be interpreted as the expression of thetotal capacitance of two capacitors in series depicted in Equation 2.19 where CH correspondsto the inner layer capacitance and CD corresponds to the diffuse layer capacitance.1Cdl=1CH+1CD(2.19)The double layer capacitance Cdl is dominated by the layer with the lowest capacitance.The diffuse layer capacitance, CD is dependent on ϕ2 which is in turn dependent on the ex-cess charge held on the electrode σM. In a dilute electrolyte solution, the concentration ofcounter ions in the diffuse layer is very low, so CD is very low as well. Cdl reaches the mini-mum at the electrode’s PZC where σM equals zero and consequently ϕ2 equals zero. In thiscondition, there is no excess charge at the electrode surface which requires no excess chargein the diffuse layer, so CD is a minimum. As a result, Cdl also reaches a minimum because itmainly depends on the low CD. Changing potential from the PZC in either direction leads toan increase in Cdl. This matches the experimental results shown in Figure 2.17. However, ina concentrated electrolyte solution, the effect of diffuse layer capacitance becomes less evi-dent due to the increase in CD. The inner layer capacitor behaves similarly to a conventionalparallel-plate electric capacitor. Its capacitance, CH, is a function of the dielectric constant ofthe medium ϵ and the thickness of the inner layer, i.e., 2, but not dependent on potential.The value of CH can be calculated with Equation 2.14 in which d is replaced by 2. However,the experimental result shown in Figure 2.17 suggests even at high electrolyte concentrationswhere the contribution from the diffuse layer capacitor is negligible, the double layer capaci-tance Cdl, mainly contributed from the inner layer capacitance, is still a function of potential.Further development of the electric double layer model involves the introduction of the dipolemoment of the solvent molecules in the inner layer [126–128]. The orientation of the solventmolecules is influenced by the potential via dipole-charge interaction, which in turn influencesthe dielectric constant.For an electrode surface modified with a SAM, the electric double layer model can still beapplied to describe the electrode | solution interface. However, since the inner layer is nowoccupied by an organic layer instead of solvent molecules, some important changes of the pa-352.3. Electrochemistryrameters in Equation 2.18 need to be considered. First, ϵ of the organic layer is much lowerthan the solvent. For an alkanethiol, ϵ is typically less than 7 as compared to 78 for H2O atroom temperature, and it decreases as the carbon chain becomes longer [129]. Second, 2for an organic layer covered surface is much higher than that for a solvent covered surface.The thickness of an alkanethiolate layer can be estimated using 1.3 Å per CH2 assuming ver-tical orientation [105]. Thus for a long alkanethiolate layer covered surface, 2 should be ona nanometer scale, which is about one order of magnitude greater than that of a solvent cov-ered surface. Considering the two contributions, the inner layer capacitance CH of the SAMmodified electrode is sufficiently small. Therefore, it is reasonable that the total capacitance ofdouble layer capacitance Cdl is dominated by the inner layer capacitance CH for SAM modifiedelectrodes and the contribution from diffuse layer capacitance CD is further reduced in highconcentration electrolyte solution. Experimental results showed that the reciprocal of doublelayer capacitance 1/Cdl for an alkanethiol coated Au electrode is linearly related to the num-ber of CH2 units for long chain (no less than 10 CH2 units) alkanethiol layers where ϵ is smallenough (the layer is even impermeable to adsorbing anions) and the change of ϵ as a resultof chain length change is negligible [105]. In this case, the Gouy-Chapman-Stern model is ap-proximated into the Helmholtz capacitor model where Equation 2.14 can be used to calculatethe capacitance of the electrode | solution interface. Worth mentioning is that the Helmholtzcapacitor model is applicable only within the potential range in which the organic layer is stableand devoid of any redox reactions.The capacitance of a SAM modified electrode can be used to calculate the fractional cov-erage of the SAM, which, in turn informs on the quality of the SAM. Since an ideally packedmonolayer results in the lowest capacitance, Cdl(θ=1), the number defects in the SAMs in-evitably lead to an increase in capacitance. Assuming the covered portion of the surface isideally packed with a capacitance of Cdl(θ=1) and the defects are covered with solvent with acapacitance of Cdl(θ=0). Thus based on this simple model, the fractional surface coverage, θ,can be calculated based on Equation 2.20 [130, 131].Cdl = θCdl(θ=1)+ (1− θ)Cdl(θ=0) (2.20)362.3. Electrochemistry2.3.4 Electrochemistry on crystalline substratesThe electric double layer model was established mostly based on experimental results with Hgelectrodes, where isotropic performance on a Hg drop surface would be expected. However, fora solid electrode with multiple crystallographic surfaces exposed, anisotropic electrochemicalresponses can be observed from different surfaces.The electrons near a metal (any metal, not exclusively in solid state) surface become looselybound because of the missing atoms on one side and tend to spread out of the crystal towardsthe vacuum. A jellium model is used to describe this spillover of electrons near the metalsurface [132]. The jellium model assumes the positively charged ions resume their positionbased on the surface crystallography, therefore, the positive charge is bounded. If the positionof this plane is referred to as zero, the positive charge only appears less than zero and treatedas a uniform background. The spillover of electrons results in an excess amount of negativecharge just outside the crystal and an excess amount of positive charge just inside the crystal.The electron density profile near the surface is represented in Figure 2.19. Because of theseparation of the positive and negative charges, a dipole is formed near the metal surface.Therefore, to bring an electron out from the bulk, this dipole needs to be overcome. The workfunction  is the physical parameter to measure the minimum energy required to bring anelectron from the bulk to a position just outside (a position where the electrostatic interactionwith the crystal approaches zero) [132, 133]. It can be calculated with Equation 2.21, where μ¯is the electrochemical potential and 4ϕ is the dipole barrier. The electrochemical potential μ¯measures the energy of the electron in the bulk with the reference to vacuum, so it is not relatedto the electron density profile near the surface. However, the dipole barrier 4ϕ, as its namesuggests, measures the barrier created by the spillover of electrons near the surface, which isdependent on the surface crystallography.= −μ¯+4ϕ (2.21)The jellium model discussed above is a one-dimensional model applicable to an ideally flatsurface. For a rough two-dimensional surface, an intuitive model proposed by Smoluchowskidemonstrates two effects which contribute to the electron density near the surface, represented372.3. Electrochemistry0Electron densityElectron density+-xFigure 2.19: Schematic of the electron density profile near a metal surface based on Figure 2.20 [134]. The first effect is suggested by the jellium model, the electrons spread outtowards the vacuum due to the weak binding to the bulk. This effect results in a dipole (directedfrom negative to positive) which points to the bulk. This dipole impedes the movement of anelectron from crossing the surface. The electron spillover is accompanied by the other effectnear a rough surface. The electron density tend to smooth out to avoid the sharp edges of highenergy instead of following the surface morphology. The sharp edges with positive charge arenow exposed which form a layer of positive charge above the smoothed electron layer. The netresult is a dipole which points away from the bulk. This dipole promotes an electron crossingthe surface. Due to the density smoothing effect, the rougher is the surface, the stronger is thedipole resulted from smoothing and thus the smaller is the work function . The conclusiondrawn from here is that there is anisotropy in electrochemistry performed on crystalline elec-trodes because the work function depends on the atomic roughness of crystallographic planesexposed.The work functions of the three low-index crystallographic planes of Au are listed in Table2.3. As can be seen, the work function follows the order of (111) > (100) > (110),which is consistent with the order of increasing atomic roughness of these three crystallographicplanes.The use of work function is more convenient if the metal is in vacuum. In an electrochem-ical system, the metal is in a solution, and what is typically measured is the PZC. The linear382.3. ElectrochemistryElectron spilloverDensity smoothing+ +-+ ++-- -Figure 2.20: Schematic of the electron spill over and density smoothing effects near the metalsurface.Table 2.3: Work function and potential of zero charge for the three low-index crystallographicplanes of AuCrystallographic plane  / eV PZC / V vs. SCE***(111) 5.26±0.04* [135], 5.15** [32] 0.23†, 0.32†† [136](100) 5.22±0.04* [135], 5.10** [32] 0.08†,0.30†† [136](110) 5.20±0.04* [135], 5.04** [32] –0.02†, –0.04†† [136]*Measured with angle-resolved photoemission spetroscopy.**Calculated based on density functional theory.***Measured in 0.01 M HClO4.†Unreconstructed.††Reconstructed.392.3. Electrochemistrycorrelation between the two parameters was detailed in [137] and summarized here.Equation 2.21 shows the correlation between work function  and dipole barrier 4ϕ, and4ϕ can be further calculated with Equation 2.22, where χM is the metal surface potential in-duced by dipole.4ϕ= eχM (2.22)When the metal is in contact with the solution, there is a small change δχM in the surfacedipole, resulting in a new surface potential gdipole across the metal | solution interface.gdipole = χM+ δχM (2.23)This is the change considering the metal side. However, the surface potential gdipole acrossthe metal | solution interface should be also felt on the solution side and experimentally mea-sured as the PZC with some selected reference point. Thus, combining Equations 2.21, 2.22and 2.23 and assuming electrochemical potential μ¯ is constant, the linear correlation betweenPZC and work function is expressed in Equation 2.24.PZC=e+ δχM− gdipole+ K (2.24)Usually δχM is also viewed as a constant, but gdipole is strongly influenced by any eventon the solution side, e.g., adsorption. But the linear correlation between PZC and  is es-tablished in this equation. Since  is dependent on the crystal structure of the surface, PZCis inevitably dependent on the crystal structure of the surface. The PZC values for the threelow-index crystallographic planes (unreconstructed) of Au listed in Table 2.24 demonstrate thisdependence.On the other hand, if there is a structure change on the surface, electrochemistry should beable to detect the change. Surface reconstruction is one of the examples [136, 138]. Surfacereconstruction on Au(111), Au(100) and Au(110) was reviewed in Section 2.1.5. In the elec-trochemical environment, the surface reconstruction can only survive at potentials that are notvery positive. The adsorption of anions at positive potential replenishes the excess amount of402.3. Electrochemistryloosely bound electrons on the surfaces and thus reverts the reconstruction back to the (1×1)structure of the surface exposed. The potential of lifting the reconstruction is dependent of theextent of specific adsorption. For example, Cl– is more specifically adsorbed than ClO4–, soa less positive potential is required to lift the reconstructed surface with Cl– adsorption. Thechange of surface structure when lifting the reconstruction has been confirmed with in situ elec-troreflectance spectroscopy. The reconstructed surfaces have different atomic roughness ascompared to the unreconstructed ones, and consequently, the PZC should be shifted as well.The PZC of an electrode can be measured in dilute and non-adsorbing electrolyte solution. Theshift of PZC can be observed when the surface exposed reverts back to the (1× 1) structure.Figure 2.21 shows the capacitance curve of a Au(100) electrode measured during the processof lifting the reconstruction [136]. The first cycle (1 and 2) was limited at potential less than 0.4 V,thus the electrode retained the reconstructed (20×5) structure. The PZC of this reconstructedsurface is +0.30 V, which is comparable with that of a (23×p3) reconstructed Au(111) surface.This can rationalized by the structural resemblance of the reconstructed Au(100) surface withthe quasi-hexagonal symmetry and the reconstructed Au(111) surface. The positive limit of thesecond cycle (3) was increased to 0.6 V and thus lifted the reconstructed surface to the (1×1)structure. The PZC of this unreconstructed surface measured at the negative scan of the sec-ond cycle (4) is +0.08 V. Lifting the surface reconstruction can be also observed with cyclicvoltammetry (CV) measurements [136, 139–142]. A sudden change in surface structure re-sults in additional charging process, shown as a prominent current peak in the voltammogram.The electrochemistry method has been used to determine whether or not one particular crys-tallographic surface of Au undergoes reconstruction. The results obtained with electrochemicalmethods are mostly consistent with those measured with in situ STM [44, 45], especially for thefrequently studied crystallographic surfaces. Worth mentioning is that no reconstruction hasbeen observed on the two important stepped surfaces, (210) and (311), with either methods.Inconsistent results were obtained with the two methods, even within electrochemistry mea-surements done by different researchers, for some rarely studied crystallographic surfaces,i.e., (755) and (533) (note that these are surfaces with large (111) terrace), presumable due todifferent experiment conditions.Above the relationship between PZC, an important electrochemical parameter and surface412.3. Electrochemistry102030405060-0.4 -0.2 0.0 0.2 0.4 0.6 0.8C (µF cm-2 )E (V vs. SCE)1,3234Figure 2.21: Capacitance curve of a Au(100) electrode measured during the process of liftingthe (20× 5) reconstruction to (1× 1) in 0.01 M HClO4 solution, reproduced from [136] withpermission from Elsevier.crystallography has been discussed. However, the concept of atomic roughness of a crystal-lographic surface is quite empirical. In addition, the experimentally determined PZC valuesavailable are limited to those frequently studied planes, namely the three low-index planes anda few stepped planes. Therefore, it would be convenient if there is a numerical parameter whichcan represent the atomic roughness and surrogate PZC. The density of broken bonds has beenproposed as such a parameter [143].An ideal crystallographic plane (hk) is the result of perfectly truncating a lattice at a des-ignated direction. The atoms on the surface lose their nearest neighbor atoms above the trun-cation plane, which is analogous to bond breaking (strictly speaking there is no chemical bondbetween the atoms). The number of these “broken bonds” per unit surface cell is defined asdensity of broken bonds (dbb). The higher the dbb, the rougher the crystallographic surface.For a fcc crystal, the density of broken bond of an unreconstructed crystallographic surface(hk) (h ¾ k ¾ ) can be calculated with Equation 2.25. Figure 2.22 shows the correlationbetween dbb and PZC for two metals with fcc crystal structure, Au and Ag. Although some dis-crepancy exists between the experimentally determined PZC and the numerically calculateddbb for some surfaces, the general correlation is clear and enough for qualitative and semi-quantitative studies.422.3. ElectrochemistryFigure 2.22: Correlation between PZC and density of broken bonds for Au and Ag, where opencircles represent the experimentally determined PZC in 0.01 M NaF at 25 °C and pH 5.6 andcrosses represent the calculated dbb of the crystallographic surfaces studied. Reprinted from[143] with permission from Elsevier.dbb =8h+ 4kph2+ k2+ 2(2.25)Also considering the step notation for the stepped surfaces reviewed in Section 2.1.4, someimportant information about the roughness of the stepped surfaces can be drawn by comparingwith dbb and PZC. The (210) surface is the turning point of the (100)-(110) zone, it is also theroughest surface of this zone as well as of all stepped surfaces based on its highest dbb andmost negative PZC. The (311) surface is the turning point of the (111)-(100) zone, it is closeto the roughest surface of this zone and indeed with the most negative PZC of this zone. The(111)-(110) zone is special because the (110) surface is formed by equal width of (111) terrace432.4. Fluorescenceand (111) step. Although the (331) surface is the turning point of this zone, the (110) surfaceis the roughest surface with the most negative PZC.2.4 Fluorescence2.4.1 OverviewSAM-modified surfaces can be described by the electric double layer model, so electrochemicaltechniques are employed to characterize the SAMs. However, the electrochemical techniquesgive only the average response of SAMs under the influence of an electric field. In other words,the resolution for the electrochemical techniques is mostly on the scale of the electrode size.In this thesis, the influence of surface crystallography on SAM will be studied, which requirestechniques to visualize various crystallographic domains distributed on an electrode surface.Therefore, spectroelectrochemical techniques combining electrochemical methods and fluo-rescence microscopy are employed to investigate the SAM modified surfaces. Here, the basicprinciple of fluorescence will be first reviewed. Moreover, fluorescence quenching near a metalsurface, the fundamental principle that the in situ fluorescence techniques are relying on, willbe discussed. Finally, fluorescence microscopy will be briefly discussed by introducing theinstrumental setup.2.4.2 Basic principles of fluorescenceFluorescence is a luminescence process when molecules in electronically excited states returnto the ground state without change in spin orientation. This process generally starts with theabsorption of light to excite a molecule from the lowest vibrational level of electronic groundstate (S0, S denotes the singlet state of spin orientation) to various vibrational levels of anexcited state (e.g., S1). This is followed by internal conversion, in which an excited moleculerelaxes to the lowest vibrational level of the excited state. From the lowest vibrational level of theexcited state, the excited molecule further returns to various vabrational levels of the groundstate accompanied by photon emission. Due to the energy loss in internal conversion, theemitted photon undergoes red shift, called Stokes shift, as compared to the absorbed excitation442.4. Fluorescencelight. According to the Franck-Condon principle, the transition time for this process is too shortfor change in nuclei positions, so the vibrational levels for both the ground state and excitedstate are similar. Consequently, a mirror image symmetry is observed between the excitationspectrum, representing the probability of the transitions (dependent of the overlap in vibrationalwave functions) from the lowest vibrational level of ground state to various vabrational levelsof excited state, and the emission spectrum, representing the probability of the transitions fromthe lowest vibrational level of excited state to various vabrational levels of ground state [144].Besides returning to ground states accompanied by photon emission, molecules in theexcited states can also undergo such processes as collision, complex formation and energytransfer, all of which result in a decreased fluorescence intensity or more precisely, quenchingof fluorescence. One important process is fluorescence resonance energy transfer (FRET),also called Förster resonance energy transfer, in which energy transfer occurs between anexcited donor molecule to an acceptor molecule through non-radiactive dipole-dipole interac-tion. Note that non-radiative energy transfer means the donor does not emit a photon and theacceptor does not adsorb a photon, so FRET can be viewed as a quenching process for thedonor molecule. The occurrence of the FRET requires a significant overlap between the donoremission spectrum and the acceptor excitation spectrum and favorable relative distance andorientation for the transition dipoles of both molecules. The FRET efficiency (E) has a depen-dence of the sixth power of the donor-acceptor distance (R) described in Equation 2.26. R0,referred to as Förster Radius, is defined as the donor-acceptor distance distance at which theefficiency is 50% [145].E=11+ (R/R0)6(2.26)Other processes that can reduce the excited state population, leading to a decrease of flu-orescence intensity include photobleaching and phosphorescence processes. Photobleachingis the process of depleting the excited state through destruction of the excited molecule by in-tensive exposure to excitation light. This differs to quenching in that the chemically damagedmolecule can no longer fluoresce [146]. The occurrence of phosphorescence requires the in-tersystem crossing in which the excited molecule in the singlet excited state S1 undergoes spin452.4. Fluorescenceconversion, ending up in the triplet excited state T1. Emission of light upon transition to groundstate from triplet excited state is called phosphorescence. However, this transition is forbiddenby symmetry, so the emission rate is very low [144].As discussed, fluorescence is competing with a number of processes such as quenching, sonot all the absorbed photons are converted into fluorescence emission. The parameter quan-tum yield (QY), i.e., the ratio of the number of photons emitted over the number of photonsabsorbed, describes the efficiency of a fluorescence event or a fluorophore. QY can be calcu-lated with Equation 2.27, where kr is the emission rate constant and knr is the non-radiactivedecay rate constant.QY=krkr+ knr(2.27)The reciprocals of the numerator and the denominator, are the intrinsic lifetime of a fluo-rophore, τ0 and the measured lifetime under experimental conditions, τ. Substituting thesetwo parameters into Equation 2.27 gives Equation 2.28. This demonstrates the proportionalityof QY to measured lifetime. Experimentally, these lifetimes are often measured to study thenon-radiactive processes.QY=ττ0(2.28)2.4.3 Fluorescence quenching near a metal surfaceThe fluorescence near a metal surface was first studied experimentally by Drexhage and Kuhn[147–150] with a classic system of an Eu3+ complex fluorophore located near a metal surface(Ag or Au) spaced by Langmuir-Blodgett (LB) films of various thicknesses. The experiment ofEu3+ complex near Ag surface was reproduced by Amos and Barnes with small modifications[151]. Figure 2.23 shows the lifetime of this classic system of an Eu3+ complex as a function ofdistances from the Ag surface. Note that QY is directly proportional to the measured lifetime,so effectively this result can also be interpreted as the unquenched fluorescence near the metalsurface. Two different domains are observed. One is at distance longer than 50 nm, where thelifetime oscillates with distance and the amplitude of oscillation gradually decays with increasing462.4. FluorescenceFigure 2.23: Fluorescence life time of the Eu3+ complex fluorophore as a function of distancefrom a Ag surface with the filled circles representing the experimental data points and the solidline representing the theoretical prediction. Reprinted from [151] with permission fromAmericanPhysical Society.distance. The other is at distance shorter than 50 nm, where the lifetime decreases sharplywith decreasing distance. The phenomenon in the long distance domain has been satisfactorilyexplained as the alternating constructive and destructive interferences between the emissionradiation and the reflected radiation due to the phase change with distance [152–154]. Thephenomenon in the short distance domain is usually referred to as the near-fieldmetal-mediatedfluorescence quenching.Theoretical interpretations of the quenching process near a metal surface were detailedby Chance et al. based on classical electromagnetic theory [152, 155] and by Yeung andGustafson based on quantum theory [156]. While both theories are in good agreement, it isnot always easy to understand the physical meanings behind the theory. In simple words, flu-orescence is competing with a number of transition pathways with some pathways resulting inradiative decay and some pathways resulting in non-radiative decay, so to calculate QY (Equa-tion 2.27), only the general radiative decay rate and non-radiative decay rate are considered.Therefore the classical electromagnetic approach typically looks for the pathways leading toradiative decay and calculates the rate of radiative decays from these pathways. The emitter472.4. Fluorescence(fluorophore) is typically treated as an electric dipole oscillator because chemically most flu-orophores are electric dipoles, required for molecular transitions to produce light [154]. Theelectromagnetic interpretation is built on the surface plasmons induced by the emitter radiationfield. However, the surface plasmons excitation only occurs in near field, i.e., at a distancesmaller than wavelength of the emission radiation (depending on the dielectric constant of themetal, this distance can be varied) where the equipotential line density of the emitter radiationfield is higher than the intrinsic charge distribution on the metal surface [157]. The calculateddecay rate (normalized with the decay rate at an infinite distance) of radiation from the sur-face plasmons near a 633 nm dipole emitter-Ag interface is shown Figure 2.24 as a functionof distance from the Ag surface. The radiation from the surface plasmons only contributesto the measured lifetime of the emitter radiation by changing kr and knr in Equation 2.27 atdistance less than about half of the wavelength. The contribution from the surface plasmonsstill does not explain the quenching in the short distance domain shown in Figure 2.23. Whilethe theoretical treatments were solid, a new pathway, the lossy surface waves, was proposedby Ford and Weber in an attempt to explain the quenching [153]. As its name suggests, theemitter radiation decaying into the lossy surface waves result in non-radiative decays, i.e., thefluorescence is “lost”. As can be seen from Figure 2.24, the decay rate for the lossy surfacewaves is dominating in the short distance domain (less than tens of nanometers), explaining thequenching. Unlike the surface plasmons, the nature of the lossy surface waves is neverthelessunclear. Further rationalizations are needed to find the cause for the quenching.One intuitive interpretation proposed by Lakowicz explains the nature of the lossy surfacewaves by linking to the surface plasmons [157]. Lakowicz’s explanation of the the lossy sur-face waves starts by considering that emitter radiation field and surface plasmons are in formsof electromagnetic waves. A one-dimensional sine wave propagating in x direction can beexpressed in wave equation (Equation 2.29).ƒ (,t) = Acos(k+ωt+ ϕ) (2.29)While this basic real form is well-known, the more general form adopts the complex form ofEquation 2.30 [14].482.4. Fluorescence10-210-11001011021030 100 200 300 400 500 600 700 800Relative decay rateDistance (nm)PhotonsSurface plasmonsLossy surface wavesFigure 2.24: The calculated decay rates as a function of distance from the metal surface ofthe major pathways near a 633 nm dipole emitter-Ag interface: photons, surface plasmons andlossy surface waves. Reproduced from [157] with permission from Elsevier. .ƒ (,t) = Aexp[ i(k+ωt+ ϕ)] (2.30)Strictly speaking, for an electromagnetic field propagating in space, the generalized waveequation for three-dimensional waves should be considered. Nevertheless, for a general un-derstanding, the one-dimensional form is used. The most important parameter in the waveequations used to explain the nature of the lossy surface waves is k, the wavevector (shouldbe written as k strictly speaking). The amplitude of the wavevector is inversely proportional tothe wavelength λ.k =2piλ(2.31)The wavevector is further related to the momentum p of a wave through the de Brogliematter wave relation where h is the Planck constant.p=hλ=hk2pi= h¯k (2.32)Thus when an electromagnetic wave travels frommedium 1 to medium 2 at an angle, basedon the momentum conservation, in the interface plane, Equation 2.33 should be fulfilled whereθ1 and θ2 are the angle of incident and angle of refraction respectively. From a different aspect,492.4. FluorescenceEquation 2.33 suggests that the projection of the wavevectors of the incident wave and therefraction wave equal each other, which is also referred to as wavevector matching.k1 sinθ1 = k2 sinθ2 (2.33)Specifically for light, the famous Snell’s law describing the refraction phenomenon can bederived from here. The refraction index n of a particular medium is defined as the ratio of speedof light in vacuum over that in the medium.n=c=2picλω=ckω(2.34)Snell’s law (Equation 2.35) is obtained considering this proportionality of n to k.n1 sinθ1 = n2 sinθ2 (2.35)If a beam of light travels from a high refraction index medium 1 to a low refraction indexmedium 2, then θ1 is smaller than θ2. An increase of θ1 can lead to a critical point where θ2equals 90°, i.e., light is traveling in the interface plane. No θ2 value satisfies Equation 2.35 foreven larger θ1, so in this case all the light is reflected back to medium 2 without traveling intomedium 1. This is known as total internal reflection.The total internal reflection can be further generalized for any form of electromagnetic waveas not being able to satisfy the wavevector matching condition of Equation 2.33. For surfaceplasmons which stay near the metal surface, only the projection of the incidental wavevector k′1need to be considered. For a small k′1, an angle of refraction θ2 is able to satisfy the wavevectormatching condition. However, as k′1 increases, θ2 eventually reaches the critical point of 90°.At even larger k′1, total internal reflection occurs and the surface plasmons are not able topropagate into space as radiation. Considering the magnitude of the wavevector is inverselyproportional to wavelength, large k′1 effectively means closely spaced charge distribution onthe surface.The surface plasmons are induced by the emitter radiation field, so the distance betweenthe emitter influences the charge distribution on the surface. In Figure 2.25, the emitter (the502.4. Fluorescenceradiation field equipotential lines shown) is placed at different distances from a bulky metalor metal thin film surface. In the case of Figure 2.25a, the emitter is at a distance wheresurface plasmons can be induced but is still far enough from the surface (typically from tensof nanometers to wavelength). The radiation field equipotential line density is relatively low,resulting in widely spaced charge distribution on the surface where the wavevector matchingcondition can be satisfied, so the surface plasmons propagate in space as radiation. In thecase of Figure 2.25b, the emitter is at a small distance from the surface (typically less thantens of nanometers). The radiation field equipotential line density is high, resulting in closelyspaced charge distribution on the surface where the wavevector matching condition cannot besatisfied, so the surface plasmons are not able to propagate in space, ending up as evanescentwaves. In other words, this explains the nature of lossy surface waves mentioned above in thatthe fluorescence is lost because the surface plasmons induced by the emitter can not escapeout of the metal as radiation. Worth mentioning is the case of Figure 2.25c, when the substrateis a metal thin film placed on top of a third high refraction index medium (e.g., a glass prism),the wavevector matching can be conditionally accomplished at the second interface and thesurface plasmon can be observed as radiation again.To sum up for the mechanism of the near field metal mediated fluorescence quenching pro-posed by Lakowicz, because of the inability of achieving the wavevector matching, the surfaceplasmons induced by an emitter close to the metal surface cannot propagate in space as radia-tive decays. Thus, the lossy surface waves can be viewed as the trapped evanescent surfaceplasmons.One drawback of Lakowicz’s interpretation of the near field metal mediated fluorescencequenching is that there is no quantitative estimation of the quenching efficiency near a metalsurface. Another intuitive interpretation is present by realizing the resemblance of the metalmediated quenching to FRET. Therefore, the metal surface can be viewed as the acceptor inthe energy transfer process. This analogy was used by Kuhn and Chance in the early days[149, 152]. This interpretation still has its own merit since it is relatively easy to understand andsemi-quantitative. One important characteristic of FRET is that the energy transfer rate has aR−6 dependence on the distance between donor and acceptorR. Analogously, for fluorescencenear metal surfaces, the quenching depends on both the distance between the fluorophore512.4. FluorescenceFigure 2.25: The surface plasmons induced by a emitter in the near field: (a) the widely spacedcharge distribution induced by the emitter far away enables propagation of the surface plas-mons in space as redative decays; (b) the widely spaced charge distribution induced by theemitter nearby leads to non-radiative decays of the surface plasmons; (c) recovery of the non-escaping decays with the substrate of metal thin film placed on top of a high refraction indexmedium. Adapted from [157] with permission from Elsevier.and the metal surface d and the metal thickness. For energy transfer to a thick metal film orbulky metal, the rate is a function of d−3 and for thin metal films, the rate is a function of d−4[152, 154, 158].2.4.4 Fluorescence microscopyThe theory of fluorescence near a metal surface was briefly reviewed above. Experimentally,measuring the fluorescence signal from a metal surface can be accomplished by fluorescencemicroscopy. The fluorescence microscopy is a technique that combines exciting a fluores-cence signal from a specimen of interest and employing a microscope to amplify, view anddetect the fluorescence signal. The epifluorescence microscope is the most widely used typeof fluorescence microscope. The prefix “epi-” represents “episcopic”, which means exciting thefluorescence with an incident light and collecting the reflected or scattered light after filteringthe excitation light through the same objective [159]. Below, the inverted fluorescence micro-scope used in this thesis (Olympus IX 70, depicted in Figure 2.26) is taken as an example todemonstrate the epifluorescence microscopy.522.4. FluorescenceFigure 2.26: Schematic of the inverted epifluorescencemicroscope used in this thesis. Adaptedfrom [160] with permission from Olympus.The light source for fluorescence excitation is usually a Hg arc lamp or a Xe arc lamp whichcovers the entire UV and visible spectrum (a Xe arc lamp even covers the infrared). A Xe arclamp is probably preferred because of its much more uniform intensity across the visible spec-trum as compared to the multiple spikes in the Hg arc lamp spectrum [161] even though a Hgarc lamp is used in this thesis. After collimated with the collector lens and optionally apertureddown with a field diaphragm, the excitation light is sent into the filter set. In a fluorescencemicroscope, a filter set is one of the most important components. Typically, a filter set is a com-bination of a band-pass excitation filter of a short wavelength, a dichromatic mirror equivalentto a long-pass filter and a band-pass emission filter of a long wavelength. Selection of thesefilters are based on the fluorophore used. Ideally, the excitation filter allows the maximum ex-citation peak to pass through, the emission filter allows the maximum emission peak to passthrough and the dichromatic mirror reflects the excitation light but allows the emission light topass through. The light path inside the filter set is presented in Figure 2.27. The light is firstfiltered with the excitation filter, then reflected by the dichromatic mirror through the objectiveto the sample surface. While the reflected excitation light from the specimen is again reflected532.4. FluorescenceFigure 2.27: Schematic of a filter set for the inverted fluorescence microscope. Reprinted from[162] with permission from the dichromatic mirror back to the light source, the fluorescence emitted from the surfacetravels through the objective, passes through the dichromatic mirror and is finally directed in tothe eyepieces or the detectors by multiple optical elements, e.g., mirror, lens, prism and beam-splitter. Multiple ports are available for integrating various detectors such as charge-coupleddevice (CCD) camera and photomultiplier tube (PMT).It is important to select the proper objective for imaging. The typical parameters taken intoaccounts aremagnifying power and numerical aperture (NA). Themagnifying power defines thearea of the specimen in view and the ability to observe the subtle details. While it is impossibleto achieve both, the actual trade-off made depends on the system studied. NA is anotherimportant parameter to consider in that it largely determines the resolution and the depth offocus.The theoretical resolution of an optical microscopy is diffraction limited. Briefly, the lightpassing through an aperture (in this case, the aperture of the objective) ends up as a diffusecircular diffraction pattern known as Airy disk. The theoretical minimum resolved distance isgiven by a sufficient separation of two neighboring Airy disks. Different criteria can be chosento define the sufficient separation, of which a famous one is the Abbe resolution limit given in542.4. FluorescenceEquation 2.36 (also with the definition of NA) where dmin is the minimum resolved distance, λis the wavelength, n is the refraction index of the medium and α is the angular semi-aperture[14]. As can be seen, increasing NA is one way of improving the resolution. Therefore, wateror oil immersion objectives are frequently employed due to the high NA as a result of the highrefraction index working medium.dmin =λ2nsinα=λ2NA(2.36)The depth of focus is a parameter to estimate image deterioration compared to the focalplane. Theoretically, the image of an object deviated from the focal plane is blurred. However,if the distance between an image point off the focal plane and an image point on the focal planeis smaller than resolved distance, there will be no significant image deterioration between thetwo points. Therefore, the depth of focus is also termed the axial resolving power (Rx) [158].The depth of focus or axial resolving power can be estimated using Equation 2.37 where ƒ isthe focal length and D is the aperture diameter [14].Rx t2.4λƒ2D2(2.37)Rx can be further related to NA.NA= nsinα = nsin(rctn(D2ƒ))t nD2ƒ(2.38)Rx t0.6n2λ(NA)2(2.39)Thus, as is shown in Equation 2.39, a large numerical aperture results in a small depthof focus, i.e., the image becomes out of focus at a small distance away from the focal plane.Very often, the depth of field is interchangeably referred to the depth of field. Strictly speaking,the definition of the depth of field, i.e., the distance between the nearest and farthest objectsthat are sufficiently sharp in an image, is somewhat arbitrary and not commonly used in opticalphysics.Note that in an inverted microscope, the objective is below the specimen, so the fluores-552.4. Fluorescencecence is excited from the bottom of the specimen. It is also possible in the setup to incor-porate the transmitted illumination from the top of the specimen. In this case, if the speci-men is transparent, the fluorescence (so-called diascopic fluorescence) as well as diffractedand refracted light is transmitted through the objective. This technique is the transmitted-lightfluorescence microscopy, developed before the epifluorescence microscopy [159]. However,this transmitted-light fluorescence microscopy is not used in this thesis simply because non-transparent systems are studied.56Chapter 3Experimental methodology andinstrumentationThe general experimental methodology applied in all following chapters is described. Thisincludes the adsorbates studied, general reagents and materials used. Worth highlighting isthe fabrications of the single-crystal bead electrode because of its important role in facilitatingthe investigations of a variety of SAMs throughout the thesis. Additionally, electrochemicaland spectroelectrochemical methods are the common techniques employed to characterizethe SAMs investigated. Thus, a general overview of the instrumentation is given in this chapterwith details presented in the corresponding chapters.3.1 Adsorbates, reagents and materials3.1.1 Adsorbates studiedThree main classes of adsorbates will be studied in this thesis: alkanethiol, thiol-modified Aibpeptide and thiol-modified DNA. These adsorbates are further labeled with fluorophores tofacilitate the spectroelectrochemical measurements. Three types of fluorophores are used:BODIPY, AlexaFluor488 and AlexaFluor647.BODIPY is the abbreviation of boron-dipyrromethene and the trade name for the seriesof fluorophores with the core structure of 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (Figure3.1a) [163]. The BODIPY fluorophores tend to be UV adsorbing but the excitation and emis-sion wavelengths can be tuned by derivatization. The derivative used to label the adsorbatesstudied is the BODIPY 493/503 (Figure 3.1b, the bond marked by asterisk will be linked to theadsorbate terminal groups). The BODIPY fluorophore can form two types of dimers: H-dimer573.1. Adsorbates, reagents and materialsaNF2BNBODIPY bNF2BNBODIPY 493/503*cNF2BNBODIPY H-dimer (DI)BF2NNdNF2BNNF2BN55°BODIPY J-dimer (DII)Figure 3.1: (a) Core structure of BODIPY fluorophores; (b) structure of BODIPY 493/503 (thebonds marked by an asterisk will be linked to the adsorbate terminal groups.); structures of (c)BODIPY H-dimer (DI) and (d) BODIPY J-dimer (DII).(also called DI, Figure 3.1c) and J-dimer (also called DII, Figure 3.1d). The H-dimer formedby two monomers in two parallel planes stacking together with a distance of 0.49 nm is non-fluorescent, whereas the J-dimer, formed by two monomers with a distance of 0.38 nm betweentwo mass centers and an angle of 55° between two long diagonal axes in one plane is fluores-cent, exhibiting red shifts in both the excitation band and the emission band as compared tothe monomer spectra [163–165].Alexa Fluor® is a family of fluorophores developed by Molecular Probe Inc. featuring highbrightness, high photostability and insensitivity to its environment [166, 167]. The structures ofthe Alexa Fluor® 488 (AlexaFluor488 hereafter) and Alexa Fluor® 647 (AlexaFluor488 here-after) are shown in Figure 3.2 a & b respectively.The selected representative of fluorophore labeled alkanethiol is the HS-C10-BODIPY (Fig-ure 3.3a), which is a ten carbon alkyl chain with a thiol group modifying end and a BODIPY493/503 fluorophore modifying the other end. This adsorbate was custom synthesized byArnold Kell and Mark S. Workentin at University of Western Ontario as reported in [168].The α-aminoisobutyric acid (Aib) peptide is a class of synthetic peptide forming the 310-helixstructure (Further discussed in Chapter 5). The selected representative of fluorophore labeledthiolated Aib peptide is the HS-Aib4-BODIPY (Figure 3.3b), which is a four Aib residue peptide583.1. Adsorbates, reagents and materialsaO NH2+NH2O-OSO3- SO3-O*Alexa Fluor® 488 b Alexa Fluor® 647N+ NSO3- -O3SO-O3S SO3-*Figure 3.2: Structures of (a) Alexa Fluor® 488 and (b) Alexa Fluor® 647. The bonds markedby asterisks will be linked to the adsorbate terminal groups.chain with the N-terminus modified by a thiol group spaced with two carbons and the C-terminusmodified by a BODIPY 493/503 fluorophore spaced with a benzene ring. This adsorbate wascustom synthesized by Ivan Guryanov and Flavio Maran at University of Padova as reported in[8]. Note that HS-Aib4-BODIPY was stored at –20 °C to prevent instant oxidation.The 30-mer DNA (5’-CTG-TAT-TGA-GTT-GTA-TCG-TGT-GGT-GTA-TTT-3’) adsorbates stud-ied were custom synthesized and dual high performance liquid chromatography (HPLC) puri-fied by Eurogentec North America. Few secondary structures are stable at room temperaturefor this DNA sequence. The 5’ end was modified with a thiol group spaced with a six carbonlinker and the 5’ end was labeled with an Alexa Fluor fluorophore (either AlexaFluor488 orAlexaFluor647). The complementary strand (5’-AAA-TAC-ACC-ACA-CGA-TAC-AAC-TCA-ATA-CAG-3’) was also custom synthesized by Eurogentec North America. All DNA stock samples(solid or solution) were store at –20 °C to prevent degradation.The last adsorbate is the 6-mercaptohexan-1-ol (MCH), which is a hydroxyl-terminated shortchain alkanethiol commonly used as the diluent for DNA SAMs [108–110]. This adsorbate waspurchased from Aldrich with a purity of ¾ 99%.593.1. Adsorbates, reagents and materialsaHSBN FFbHSBN FFNHHNNHHNOOOOONHFigure 3.3: Structures of (a) HS-C10-BODIPY and (b) HS-Aib4-BODIPY.3.1.2 ReagentsSurface chemistry is extremely sensitive to impurities, so the commercial reagents employedwere of the highest purity available from major chemical vendors (Sigma-Aldrich, Fisher Scien-tific, etc.) without further purification. These include the electrolyte salts to prepare the buffersolutions, organic solvents and acids or bases to adjust pH. Besides these directly used com-mercial reagents, some other important reagents are worth highlighting. The water was purifiedwith a Milli-Q® Integral 5 water purification system to an ultrapure level of > 18 MΩ•cm resis-tivity and ¶ 3 ppb total organic carbon. KClO4 was doubly recrystallized from the purchasedreagent using the ultrapure water prior to making electrolyte solutions. The gold wire used tofabricate the substrate electrodes was purchased from Alfa Aesar with a purity of 99.999%,which is a requirement for forming a single-crystal electrode. The gold wire (∅ = 1.0 mm ) wasfurther purified by immersing the part to be melted in aqua regia for 15 min. Ar gas used todeoxygenate the solutions in electrochemical measurements was purchased from Praxair witha purity of > 99.998%.3.1.3 MaterialsThe electrochemical and spectroelectrochemical cells were made by Brian Ditchburn, the UBCChemistry glassblower, of which the procedures are detailed in [169]. All glasswares were603.2. Substrate fabrication and cleaningimmersed in ~80 °C 1:1 volume of concentrated HNO3:H2SO4bath for 3 h, then rinsed andsoaked in ultrapure water for over 12 h prior to use. The pipette tips (from Gibson Diamond)used to measure small volumes of liquid samples were autoclaved at 120 °C for 40 min to re-move any contaminants potentially degrading the adsorbates. The siliconized microcentrifugetubes (Bio Plas G-Tube®) used to store small volumes of sample solutions were certified bythe manufacturer to be RNase and DNase free and human DNA free.3.2 Substrate fabrication and cleaning3.2.1 Single-crystal Au bead electrodeThe fabrication of a single-crystal Au bead electrode was based on the well-known proceduresof making Pt single crystal electrodes [170]. Briefly, the ultrahigh purity and further purifiedgold wire was heated and melted with a butane torch in an area protected from air currents.The bottom part of the gold wire was kept molten until a gold bead formed and grew to approxi-mately 2 mm in diameter. The bead was slowly cooled down and solidified by gently moving theflame away. Typically aqua regia cleaning was followed to further dissolve the superficial layerof Au atoms along with possible impurities. After this cleaning process, the melting and solidi-fying process was repeated with flame intensity and position well controlled to keep the moltenpart below the boundary between the uncrystallized wire and the crystallized bead, preventingfurther growth of the bead. This aqua regia cleaning followed by melting and solidifying wasusually repeated 2-3 times to guarantee a high quality single crystal bead electrode. Usuallybig (111) facets and small (100) facets could be seen on the surface through a microscope fol-lowing the fcc single crystal pattern, completely or partially. While incompletely formed singlecrystal (e.g. bi-crystal) with at least one complete stereographic triangle present was accept-able especially under high power magnification, the complete single crystal bead was achievedmost of the time in this thesis by carefully and slowly (sometimes repeatedly) melting a smallportion of uncrystallized wire so it recrystallizes.The single-crystal Au bead electrode was typically cleaned following four routines. First,each time after an electrochemical experiment where a SAM was desorbed, the electrode was613.2. Substrate fabrication and cleaningcleaned with cyclic voltammetry scans, −1.20 V to +0.85 V (vs. Ag|AgCl), at 200 mV/s in anewly prepared KClO4 solution (pH = ~12 adjusted with KOH) for 20 min. In addition, after thismild electrochemical cleaning, the electrode was heated with butane torch until the bead turnedorange (not melted) for approximately 20 s, followed by rinsing in the ultrapure water. This heat-ing and rinsing procedure was typically performed three times before assembling another SAMonto the electrode. Furthermore, a gentle electropolishing (modified from [171]) by polarizingthe gold bead at +2.45 V (vs. Au wire) in a 1 M HClO4 solution for 30 s followed by immersionin 1 M HCl solution for 10 s was conducted typically for an electrode with accumulated organicresidues left after repeated deposition of SAMs (typically no more than three times) especiallyless hydrophilic adsorbate (e.g. the thiolated Aib petide). The electropolishing treatment wasfollowed by cyclic voltammetry scans, −1.0 V to +1.0 V (vs. saturated calomel electrode (SCE)),at 200 mV/s in a newly prepared KClO4 solution for 10 min. Lastly, since electropolishing couldlead to roughening of the surface, this treatment was at most repeated twice. For an aged oroften used electrode, an aqua regia cleaning followed by melting and solidifying was performedto resume a good surface condition, but in order to keep the original surface crystallography,the melting did not exceed the wire - bead boundary.3.2.2 Polished single-crystal Au(111) electrodeThe polished single-crystal Au(111) electrode was prepared from a single-crystal Au bead elec-trode and used for electrochemical measurements in Chapter 5. This Au bead electrode wasfabricated by carefully adjusting the angle of melting so that a (111) facet was formed very closeto the bottom of the bead. The orientation of the electrode was aligned to be along the normaldirection of the (111) facet with a laser and fixed with a purposely designed holder. Mechanicalpolishing was conducted first with sandpapers (600 grit and 1200 grit) and then with diamondpolishing materials (LECO premium grade aerosol diamond spray 6 μm, 3 μm and 1 μm se-quentially) sprayed on polishing felts. After being cleaned by sonication in ethanol and watereach for 15 min to remove excess polishing materials embedded on the surface, the resul-tant polished electrode was annealed at 700 °C for 24 h. The final treatments were the gentleelectropolishing followed by cyclic voltammetry scans described above. The area of the planar623.3. Electrochemical instrumentationsurface were measured with two methods. First, an image of the planar surface was takenand analyzed with ImageJ to obtain the geometric area. Second, the electrochemical area wascalculated by dividing the capacitance of the planar surface (hanging meniscus configuration,described below) in ClO−4 solution (acidic or basic) at –0.8 V (vs. SCE) by 20 μF cm–2, the unitarea capacitance in the absence of specific adsorption [172, 173]. The similarity of the tworesults (roughness factor = 1.04) demonstrated the smoothness of the polished electrode.The polished Au(111) electrode was also cleaned with the same approaches as the singlecrystal Au bead electrode, except for aqua regia cleaning and remelting.3.3 Electrochemical instrumentation3.3.1 Electrochemical setupThe electrochemical measurements without coupling to spectroscopic techniques were onlyperformed for a polished Au(111) electrode in Chapter 5. The schematic of the electrochemicalsetup is shown in Figure 3.4. The electrode was fit into a holder and placed into the electro-chemical cell as the working electrode (WE). The counter electrode (CE) was a Au wire with acoil at the end and the reference electrode (RE) was SCE (BeckmanCoulter 511100) connectedthrough the salt bridge. The electrolyte solution was deoxygenated with Ar gas bubbling for 15min before measurements and the Ar flow was maintained above the solution during measure-ments. The WE was gently lowered so that the flat surface of the electrode just touches withthe supporting electrolyte surface creating a hanging meniscus [169, 174, 175].All electrochemical experiments were conducted with a HEKA PG590 potentiostat and aPrinceton Applied Research Model 175 universal programmer which generated the potentialscans. An EG&G Model 5210 lock-in amplifier was also used to generate AC potential per-turbation in capacitance measurements. All data were digitized with a National InstrumentsPCI-6052E data acquisition board (DAQ) and further transferred to a computer interface coor-dinated by a LabVIEW program.633.3. Electrochemical instrumentationFigure 3.4: Schematic of the electrochemical setup.3.3.2 Electrochemical techniquesCyclic voltammetry between two set potentials was the main technique used to study the SAMmodified electrode. For a SAM-modified electrode, the scan was limited in the potential range(typically -0.10 V to 0.30 V (vs. SCE)) where the SAM was stable without the occurrence ofreductive or oxidative desorption. Usually, more than 20 potential range limited scans wereconducted until no significant change was observed. The purpose of repeating scans was tominimize the wetting effect. As a matter of fact, in the cyclic voltammetry and capacitancemeasurements, charging current and capacitance signals greatly depended on the area of theelectrode covered with solution. The hanging meniscus configuration to the utmost guaranteesthat only the flat surface at the bottom of the electrode was covered, but the electrolyte solutionwas still able creep up the sides, covering part of the electrode above the edges. Nevertheless,with potential scans, the wetted area gradually decreased as a result of drying. Consequently,the charging current and capacitance decreased as well, approaching a constant level. There-fore, it was necessary to minimize the wetting issue by repeating potential scans until no changeof current or capacitance observed.Differential capacitance was also employed to check the quality of the SAM formed on theelectrode surface. The capacitance measurements were conducted by adding an AC sinu-soidal potential perturbation (200 Hz and ~5 mV) from the lock-in amplifier to the DC potentialapplied by the potentiostat. The output AC current was then fed into the lock-in amplifier and643.4. Spectroelectrochemical instrumentationseparated into real and imaginary components. Assuming the electrode | solution interfacecan be modeled as a capacitor (double layer capacitance) and a resistor (solution resistance)in series (see Section 2.3.3), the capacitance was calculated with Equation 3.1 derived fromimpedance theory [13].C=2Re+ 2m2piƒErmsm(3.1)Here C is the capacitance of the interface, Re is the real component of the AC current, m isthe imaginary component of the AC current, ƒ is the frequency of the AC perturbation and Ermsis the root mean square (rms) amplitude of the AC potential wave. Note that this capacitancecalculation is only valid for systems behaving as IPEs. For systems3.4 Spectroelectrochemical instrumentation3.4.1 Spectroelectrochemical setupThe schematic of the spectroelectrochemical setup is shown in Figure 3.5. All spectroelectro-chemical measurements were performed in a spectroelectrochemical cell featuring an opticalwindow of 250 μm thickness forming the base. Besides the optical window, the spectroelec-trochemical cell functioned similarly to an electrochemical cell with ports for WE (the polishedAu(111) electrode or the single crystal bead electrode), CE (a Pt wire with a coil at the end)and Ar and integrated salt bridge also acting as the RE (BASi® RE-6 Ag|AgCl or CHI150 SCE)housing. The potentiostat (FHI-ELAB 0599 potentiostat or FHI-ELAB G050-0298 potentiostat)and lock-in amplifier (EG&G 5208 lock-in amplifier or SRS SR830 lock-in amplifier) were em-ployed for electrochemical measurements.The spectroelectrochemical cell was further placed on top of an Olympus IX70 invertedfluorescence microscope. The type of objective was chosen as required for each specific ex-periment. Most of the optical measurements in the spectroelectrochemical experiments wereperformed with dry objectives. However, some of the experiments in Chapter 6 were performedwith water immersion objectives in an attempt to improve the resolution (according to Equation2.36, dmin of ~400 nm can be achieved using a 40× water immersion objective) and image653.4. Spectroelectrochemical instrumentationFigure 3.5: Schematic of the spectroelectrochemical setup. Adapted from [169] with permis-sion. Copyright (2017) Springer.quality. Since most of the surfaces imaged are curved, the depth of focus should also be con-sidered. According to Equation 2.39, the depth of focus is very small (~1 μm) when imagingis performed with an objective with a high numerical aperture (e.g., a 40× water immersionobjective). But typically in this type of measurements, a field diaphragm was closed down sothat a small region (~96 μm in diameter) was illuminated, so within the illuminated region, theimage was mostly in focus.A fluorescence illuminator equipped with a Hg arc lamp was employed to excite the fluo-rophores. The light from the fluorescence illuminator came from the back of themicroscope (notshown in Figure 3.5) and further traveled through the filter set which directs light for illumination,filtration and collection. Depending on the fluorophore used to label the adsorbate studied, aparticular filter set was chosen. The same filter set fromChroma (excitation: ET470/40x, dichro-matic: T495LPXR, emission: ET525/50m) was used for both the BODIPY and AlexaFluor488,whereas another filter set from Chroma (excitation: HQ620/60x, dichromatic: Q660LP, emis-sion: HQ700/75m) was used for AlexaFluor647. The spectra of the fluorophores used in thisthesis and their corresponding filter sets are shown in Appendix A. The brief specifications forthe filter sets are listed in Table 3.1. Worth mentioning is that no prominent spectral spike (from663.4. Spectroelectrochemical instrumentationTable 3.1: Fluorophores and corresponding filter setsFluorophore Excitation filter Dichromatic mirror Emission filterBODIPY (monomer), AlexaFluor488 450-490 nm 495 nm 500-550 nmAlexaFluor647 590-650 nm 660 nm 662-738 nmthe Hg arc lamp) was present in the two excitation bands listed in Table 3.1 (see Figure A.1 inAppendix A).The microscope was further equipped with two types of detectors: charge-coupled device(CCD) cameras and photomultiplier tube (PMT). Two CCD cameras were utilized for imag-ing: Photometrics Evolve® 512 EMCCD camera with the image size of 512×512 pixels andSBIG ST-7XMEI CCD camera with the image size of 765×510. The PMT used was a Newport77348 model with the active area of 8×24 mm2 and the signal output from the PMT was furtherconverted into voltage with a Stanford Research SRS570 current preamplifier. Note that the flu-orescence images taken with the Photometrics Evolve® 512 EMCCD camera was illuminatedwith an Excelitas X-Cite® exacte fluorescence illuminator, while the fluorescence images takenwith the SBIG ST-7XMEI CCD camera and the fluorescence detected by the PMT were with anExcelitas X-Cite® 120PC fluorescence illuminator.3.4.2 Spectroelectrochemical techniques and data analysisThe spectroelectrochemical techiniques employedweremainly dependent on the detector used.When a camera was the detector, the in situ fluorescence imaging technique was used. Briefly,the WE was stepped to a series of potentials, while at each potential, one or more fluores-cence images were taken with the preset conditions (exposure time, illuminator light intensity,electron-multiplying (EM) gain for the Evolve® 512 EMCCD camera) and the capacitance wasmeasured. Two major potential stepping schemes (profiled in Figure 3.6a) were programmed.In the first potential stepping scheme, the potential started from an initial potential (Ei), sequen-tially stepped to a series of step potentials (Es) with the same step width (typically a negativevalue) between two neighboring Es and completed at a final potential Ef. This potential steppingscheme was used to study the reductive desorption process of a SAM (Chapter 4 and Chapter5). Compared with the first potential stepping scheme, Ei was also a base potential (Eb) in thesecond potential stepping scheme and after each Es, the potential was stepped back to Eb.673.4. Spectroelectrochemical instrumentationE (vs. RE)timeThe two potential stepping schemesEi/EbEfEs(n)Es(n+1)Es(n+2)......First potential stepping scheme Second potential stepping schemeaE (vs. RE)timeEvents during a potential stepEs(n-1) Es(n)Es(n+1)Imaging period 1 Imaging period 2Shutter on Shutter onShutter off Shutter offImaging ImagingTransfer TransferCapacitance Capacitance...bFigure 3.6: (a) Potential profiles of the two potential stepping schemes and (b) events occurringduring a potential step.This potential stepping scheme was used to study the potential modulated response of a SAM(Aib peptide SAM in Chapter 5 and DNA SAM in Chapter 6). All experimental events, includingthe potential stepping, illuminator shutter triggering, imaging, capacitance measurement anddata transfer (illustrated in Figure 3.6b) were coordinated with Labview programs.After each potential stepping measurement, an image stack was created from the fluo-rescence images, which was analyzed with ImageJ. The images were first despeckled andsmoothed with the built-in function of Accurate Gaussian Blur (standard deviation = 2) [176].Subtracting the background was also performed for each image. However, the backgrounddetermination was not always easy. Cameras have dark current noise as part of the mea-surements [177, 178]. Background resulting from the dark current noise was measured in theabsence of illumination and subtracted from each image. Stray light and light reflected from themetal surface that leaked through the filters were also important components of the background.683.4. Spectroelectrochemical instrumentationFor a flat substrate (i.e., a Au(111) facet on a single-crystal bead electrode), the backgroundwas measured with an unmodified electrode. For a curved substrate (i.e., a bead electrode),measuring the background with an unmodified electrode was not performed because it wasdifficult to guarantee the same the position and orientation of the electrode. If no fluorescencewas emitted at some potentials, the images taken at these potentials could be used as the back-ground. Nevertheless, for most of the studies on bead electrodes in this thesis, the accuratebackground was not critical, so only the dark current background was subtracted. ImportantImageJ functions used for image processing are worth highlighting here [176]. The most com-monly used ImageJ function to extract data from an image stack is the z-axis profile functionwhich calculates the average intensity of all pixels inside a selected region of interest (ROI) forall images in the image stack. The z-projection functions look for the maximum or minimum val-ues for each pixel through an image stack and construct the maximum or minimum projectionimages with the corresponding values.When a PMT was the detector (further coupled with the current preamplifier), the opticalcyclic voltammetry technique for DC optical signal and the harmonic analysis technique withthe lock-in amplifier for AC optical signal were used. These techniques will be discussed in-depth in Chapter 6.69Chapter 4Spectroelectrochemical investigationof the reductive desorption behaviorof self-assembled monolayers on asingle crystal Au bead electrode dueto the influence of surfacecrystallographyThis chapter discusses the influence of the surface crystallography on the reductive desorptionbehaviors of SAMs deposited on Au single crystal bead electrodes. Two systems are explored:a BODIPY fluorophore-labeled alkanethiol (HS-C10-BODIPY) and a BODIPY fluorophore-labeledthiol-modified α-aminoisobutyric acid (Aib) peptide (HS-Aib4-BODIPY). The single-crystal goldbead electrode has the family of crystallographic surfaces symmetrically distributed followingthe fcc crystal structure. As the substrate, this type of electrode enables exploration of a SAMdeposited on the surfaces representing the full fcc crystallographic triangle simultaneously andunder identical conditions. In situ fluorescence imaging is utilized to monitor the fluorescenceresponses of a SAM during the reductive desorption process on the variety of crystallographicsurfaces.704.1. Reductive desorption of self-assembled monolayers4.1 Reductive desorption of self-assembled monolayersAs was mentioned in Section 2.2.2, thiolate SAMs deposited on Au electrode can be reduc-tively desorbed by applying a sufficiently negative potential. The reductive desorption is animportant process that requires further understanding so as to be used for surface modificationin a controlled manner. Selective desorption of mixed-component SAMs [179, 180] or SAMsdeposited on surfaces with different crystal structures [168, 181, 182] has been widely usedin controlled functionalization of the surface, especially on the nanometer scale. Because ofthe fast-developing nanotechnology, where the surface crystal structures become prominent,selective desorption of SAMs deposited on surfaces with different crystal structures is moreand more widely employed, and thus requires an in-depth investigation of the influence of thesurface crystallography on the reductive desorption in order to achieve controlled surface func-tionalization.The study of the reductive desorption process was pioneered by Porter [118, 183–185],Morrin [10, 186–188] and their co-workers. One advantage of the reductive desorption processis that the reaction follows a stoichiometric route shown in Equation 2.10, so it can be usedto quantify the surface density of the monolayers [183]. Typically, a linear scan voltammetrymeasurement can be performed from a positive potential E (at which a SAM is adsorbed) toa negative potential Ed (at which the SAM is desorbed). The total charge of desorption canbe calculated by integrating the current of the reductive desorption peak between these twopotentials, described in Equation 4.1, where Q is the total charge,  is the cathodic current and is the scan rate.Q=∫ EdEdE (4.1)Then based on the Faraday’s Law, the surface packing density can be calculated followingEquation 4.2 where  is the surface packing density, n is the charge number (n = 1 for theAu-S system), F is Faraday constant and A is the electrode area.= QnFA (4.2)714.1. Reductive desorption of self-assembled monolayersThis simple approach fails to take an important point into account [189–192]. The currentof the reductive desorption peak consists of not only the desorption current, but also the dou-ble layer charging current. This causes an overcalculated charge even if the SAM remainsadsorbed on the electrode surface. More importantly, once the Au-S bond is reduced at suffi-ciently negative potential, water molecules and hydrated cations substitute the monolayer, sothe dielectric constant changes with potential. As a result, the charging current and thus theeffective charge number n is potential-dependent. Since water has much a higher dielectricconstant than the monolayer, the charging contribution increases with the negatively scanningpotential. Therefore the packing density calculated based on Equation 4.1 and Equation 4.2is higher than the actual value. A method to accurately quantify the surface packing densitywas developed by Lipkowski and co-workers [191, 192]. This method employs chronocoulom-etry and a LB-deposited thiolate film of known packing density to calculate the effective chargenumber at various potentials during a desorption process. It was found that the packing den-sity calculated by integration of current in a cyclic voltammetry measurement is overestimatedby ~20%. However, while this chronocoulometry method does prove to give more accurateresults, it definitely requires complicated steps and does not seem to be applicable to all adsor-bates because of the important LB-deposition step. Therefore the simple calculation based onEquation 4.1 and Equation 4.2 by arbitrarily setting a linear capacitive current baseline is still aconvenient and conventional way to estimate the packing density of SAMs.A kinetic model proposed by Mulder et al. [193] describes the mechanism of the reductivedesorption of alkanethiolate SAMs. An alkanethiolate SAM modified Au surface inevitably hasSAM-covered domains and defective domains. The defective domains can be vacancy islands,crystal grain boundaries or less ordered adsorbates, which are relatively permeable to chargedspecies. The potential on the Au-S interfaces is easily set to a sufficiently negative point toinitiate the reduction reaction near these defective domains. Consequently, the vacant domainsgrow radially as the reduction reaction progresses and the SAM-covered domains shrink aswell. Eventually, all the vacant domains merge and the SAM-covered domains disappear.The factors influencing the reductive desorption process have also been studied in detailwith electrochemical methods. Studies of reductive desorption of alkanethiolate SAMs showedthat the desorption potential was strongly influenced by the chain length of the alkanethiol [10,724.1. Reductive desorption of self-assembled monolayers118, 183, 185, 186, 188, 190]. To be specific, the desorption potential shifted to more negativepotentials with increasing chain length. This can be rationalized by more well-organized andwell-packed SAMs formed with longer chain alkanethiol. It was also found that the adsorbatestructure[190], the SAM deposition conditions (e.g., the solvent used [189], the deposition timespent [189] and the electrochemical deposition duration [194]), pH of the solution used forelectrochemical measurements [10, 186], and composition of the electrolyte [185, 190] canalso influence the desorption potential. All these factors, can actually be generalized as factorswhich can have an impact on the intermolecular interactions between adsorbates.While the studies above were primarily performed on a Au(111) surface, either Au filmson mica sheets or polished single-crystal electrodes, the influence of surface crystallographyon reductive desorption has also been investigated. The voltammetry measurements showeddifferent reductive desorption behaviors for the alkanethiol SAMs deposited on three differentAu films substrates: Au on mica, Au on Si and Au on glass [184]. Au on mica had predominant(111) feature, so alkanethiol deposited on it had a sharp desorption peak. For the other twosubstrates, there existed a large portion of non-(111) steps, so additional desorption peaksappeared at more negative potentials. Compatible results were obtained with polished single-crystal electrodes and polycrystalline electrodes [10–12]. Multiple desorption peaks were ob-served for an alkanethiol SAM deposited on a polycrystalline electrode which correspond tothe desorption peaks observed at the same potentials on polished single crystal electrodes(Figure 4.1). The desorption potentials of SAMs on different crystallographic surfaces havesome correlation with the PZC of the surfaces. For alkanethiol SAMs on Au, it is generally truethat the desorption potentials for various crystallographic surfaces and PZC for correspondingsurfaces follow the same order. Thus the reductive desorption of an alkanethiol SAM on a poly-crystalline electrode starts from the (111) surface, followed by the (100) surface, and furtherfollowed by (110) and (210) surfaces. Note that as reviewed in Section 2.3.4, the (210) surfaceis the roughest among all stepped surfaces and with the most negative PZC, so it is also widelyinvestigated.Combining the two major contributions discussed above: the substrate crystal structureand the intermolecular interactions between adsorbates, an elaborated form of Equation 4.3 ispresented to describe the reductive desorption process. The influence of the two contributions734.1. Reductive desorption of self-assembled monolayersFigure 4.1: Cyclic voltammagrams measured during desorption of decanethiolate SAMs de-posited on (A) a polycrystalline Au bead electrode and (B) polished single crystal Au electrodeswith the indicated crystallographic orientations. Reprinted from [12] with permission from Else-vier.744.1. Reductive desorption of self-assembled monolayerson the reductive desorption of SAMs has been detailed by Doneux et al [12] and summarizedhere. The substrate crystal structure can influence the binding with the adsorbate and thestrength of the Au-S interaction. When the adsorbate molecules are reductively removed, thesubstrate is coated with solvent, so the difference between the applied potential and PZC alsoacts as a driving force for the desorption process. Even though PZC is shifted negativelyupon adsorption of a SAM [192, 195], this driving force is still taking effect. The intermolecularinteractions between adsorbates largely determines the stability of a SAM. A SAM stabilizedby attractive interactions tends to have a more negative desorption potential and vice versa, aSAM destabilized by repulsive interactions tends to have a less negative desorption potential.Note that the intermolecular interactions can be influenced by a number of factors mentionedabove: solvent, electrolyte, etc. and even the substrate crystal structure.RS(srf,A(hk))+ H2O(q)+ ne−(A(hk))ŠRS−(q)+ H2O(srf,A(hk))(4.3)The work reviewed above was performed on electrodes with predominantly one crystal-lographic orientation (films or polished single-crystal electrodes) or polycrystalline electrodes.Thus to study the influence of surface crystallography on the reductive desorption process, aseries of experiments with different electrodes are required. However, it is very important toperform the series of experiments under consistent conditions since the experimental condi-tions have substantial influence on the intermolecular interaction of the SAMs. Small changesin the SAM assembly conditions (e.g., the assembly time, the solution pH, the temperature,etc.) for each replicate on a particular surface, can result in SAMs that may not represent thedesignated ones. Small variations in each measurement can also lead to a lack of data repro-ducibility. Therefore, ideally, all the replicates should be assembled under the same conditionsand measured at the same time. In the sections that follow in this chapter, a single crystalbead electrode with the family of crystallographic surfaces symmetrically distributed followingthe fcc crystal structure is used as the substrate for the SAMs studied. With the single crystalbead electrode, all identifiable crystallographic surfaces are analyzed self-consistently underidentical conditions for one single SAM in one single measurement.754.2. Objectives4.2 ObjectivesIn this chapter, the single crystalline Au bead electrode will be used as the substrate for twotypes of adsorbates: alkanethiol and thiol-modified Aib peptide. The in situ fluorescence imag-ing method will be used to study the reductive desorption process of the SAMs prepared. Withthis method, the desorption from all identifiable crystallographic surfaces will be monitored inone single measurement. After detailed analysis, the influence of surface crystallography onreductive desorption will be demonstrated. The alkanethiolate SAMs have been studied in-tensively, so the comparison with literature reports will verify the feasibility of this method inexploring a fluorophore-labeled SAM on various crystallographic surfaces. This approach willbe strengthened by studying the Aib peptide thiolate SAM with vastly different properties, alsohelping elucidate the influence of intermolecular interactions on reductive desorption.4.3 Experimental4.3.1 SAM preparationThe two types of adsorbates studied are: an alkanethiol (HS-C10-BODIPY, Figure 3.3a) and athiol-modified Aib peptide (AuS-Aib4-BODIPY, Figure 3.3b). Both thiol-modified adsorbates arelabeled with the BODIPY fluorophore in order to perform fluorescence imaging measurements.The assembly of these two types of SAMs onto the single crystal Au bead electrode followedthe procedures below. Note that the solvents were carefully chosen so that the two adsorbateshad appreciable solubility, ensuring high quality SAM formation.Assembly of the AuS-C10-BODIPY SAMs was performed by immersing a clean substrateelectrode in 1 mL of 1 mM HS-C10-BODIPY 1:1 MeOH-CHCl3 solution in a glass vial. TwoSAMs were prepared, one for 15 min and the other for 18 h. The modified electrode was thenrinsed with the solvent and then water before characterization.Assembly of the AuS-Aib4-BODIPY SAM was performed by immersing a clean substrateelectrode in 50 μL of 1 μMAuS-Aib4-BODIPY ethanolic solution in a sealedmicrocentrifuge tube(siliconized, from BioPlas) for 2 h. The modified electrode was then rinsed with the solvent andthen water before characterization.764.4. Results and discussion4.3.2 In situ fluorescence imaging characterization of the SAMsThe in situ fluorescence imaging performed by applying multiple potential steps to the workingelectrode and taking the images at each step was discussed in Section 3.4. Specific for this setof experiments, the spectroelectrochemical cell was placed on top the Olympus IX-70 invertedepifluorescence microscope. The modified gold bead electrode was used as a working elec-trode. The electrochemical measurements were performed with a Pt coil counter electrode anda Ag|AgCl reference electrode connected to the working solution via a salt bridge. The work-ing electrolyte solution was 50 mM KClO4 (pH = 12 (±0.5) adjusted with KOH). The electrolytesolution was deoxygenated with Ar in all experiments. The fluorescence images were acquiredwith the Evolve® 512 EMCCD camera through an Olympus LMPlanFl 5× objective (NA = 0.13)producing images that represent 1.58 mm × 1.58 mm regions of the surface. The profiles ofthe potential steps applied were similar for the two types of SAMs in that the potential started at0.0 V (vs. Ag|AgCl), and was stepped negatively by –20 mV increment. The duration at eachstep was 2 s for the AuS-C10-BODIPY SAMs and 4 s for the AuS-Aib4-BODIPY SAM. The finalpotential was –1.4 V (vs. Ag|AgCl) for the AuS-C10-BODIPY SAMs and –1.3 V (vs. Ag|AgCl)for the AuS-Aib4-BODIPY SAM.4.4 Results and discussion4.4.1 Reductive desorption of the AuS-C10-BODIPY SAMThe reductive desorption of an alkanethiol SAM occurs at the least negative potential on theAu(111) surface. This behavior has been well studied with electrochemical methods, as re-viewed in Section 4.1. In addition, the reductive desorption of fluorophore- labeled alkanethiolSAM on polycrystalline Au bead electrode has also been investigated with in situ fluorescenceimaging, further confirming the fact that the desorption occurring at the least negative poten-tial on the identified (111) domain [168]. The fluorescence signal can be observed only whenthe fluorophore is sufficiently separated from the electrode surface as a consequence of SAMdesorption. Thus when the potential was stepped to negative potentials, a detectable fluores-cence signal appeared on (111) surface first, with other regions of the surface following at more774.4. Results and discussionFigure 4.2: Brightfield optical image of the bottom of a gold bead electrode with (111) facetsencircled in red, (100) facet encircled in green and defect encircled in blue. Adapted from [8]with permission. Copyright (2014) American Chemical Society.negative potentials. This fact is relied upon for identifying the surface features on the singlecrystalline gold bead electrode imaged in the in situ experiments.A bright field image looking up at the bottom of the Au bead used is shown in Figure 4.2.Four large facets encircled in red are observed in the corners of the image. These facets are thewell-known (111) facets which have been employed as the substrates for STM studies [196–200]. In the center of the bright field image, a small facet encircled in green can be vaguely seen(with proper orientation and focal plane, this feature can be clearly seen). Based on the fccsingle crystal structure, this small facet is where the (100) surface locates. The crystallographicfacets are characteristic of carefully annealed single crystals, particularly prominent in Pt singlecrystal electrodes [201–203]. The dimple feature encircled in blue is a defect formed during thecooling process of electrode fabrication (seems unavoidable under present conditions). Thesecrystallographic features are more clearly observed in the fluorescence images taken duringthe reductive desorption of a AuS-C10- BODIPY SAM from the electrode bead surface. Figure4.3 shows the montage of selected fluorescence images taken from –1.14 V to –1.36 V (vs.784.4. Results and discussionAg|AgCl) in –20 mV increments during the reductive desorption of the AuS-C10-BODIPY SAMformed with a 15 min immersion time. The fluorescence intensity is false colored (as indicatedwith the calibration bar). A noticeable amount of fluorescence was observed at –1.14 V (vs.Ag|AgCl) coming from four regions at the four corners of the image, symmetrically orientedwith respect to the center of the image. These four regions coincide with the four big (111)facets observed in the bright field image. As the potential was stepped negatively, starting from–1.22 V (vs. Ag|AgCl), fluorescence was observed from the center of the image, revealing a4-fold symmetric cross with each arm of the cross directed in between two neighboring (111)facets. The center of the cross coincides with the the (100) facet in the bright field image, eventhough the presence of the defect results in a distortion of the NE portion of the cross. Withmore negative potential applied, at –1.28 V (vs. Ag|AgCl) the areas between two neighboringarms of the central cross started to fluoresce and further evolved into another cross with thearms pointing towards the (111) factets. At potential more negative than –1.30 V (vs. Ag|AgCl),fluorescence was observed from the rest of the surface, including the identifiable (110) surfacesbetween two neighboring (111) facets and (210) surfaces between (100) facet and each (110)surface, according to the fcc crystal structure.The images present a highly symmetric potential-dependent behavior suggesting that thebead is indeed single crystalline in nature. The (111) facets located at the corners of the imageare oriented in a 4-fold symmetry around the (100) surface. In addition, between two neigh-boring (111) facets, the (110) surfaces are present, also oriented in a 4-fold symmetry aroundthe (100) surface, rotated by 45° with respect to the (111) facets. This pattern resembles thestereographic projection of an fcc crystal centered about (100), as reviewed in Section 2.1.4.However, the stereographic projection cannot be used for mapping the images here becausethe surface features are directly projected onto the image plane. The proper crystallographicprojection map was created using a Perl script by assuming the bead was a perfect sphere.Since the (100) facet shown in Figure 4.2 was not exactly at the bottom of the electrode surface,small adjustments of rotating around the x- and y-axes (in plane of the page) and the z-axis (outof the page) had to be made to compensate. The adjustments were properly made by over-laying with a fluorescence image taken at a negative potential with the single crystal patternclearly present. Figure 4.4a and 4.4b show the fluorescence images of the bead electrode ac-794.4. Results and discussionFigure 4.3: Montage of selected fluorescence images taken from –1.14 V to –1.36 V (vs.Ag|AgCl) in –20 mV increments during the reductive desorption of the AuS-C10-BODIPY SAMcreated with a 15 min immersion time.quired at −1.24 V (vs. Ag|AgCl) and −1.30 V (vs. Ag|AgCl) with overlay of crystallographic mapshowing the low-index and stepped surfaces. With easily recognizable features from the (111)facets and the (100) facet, the map was adjusted to the corresponding position and orientation.This also allowed for the other surfaces to be located to subtle features in the fluorescencepattern (e.g., the cross in the center). The almost perfect symmetry of the fluorescence im-age and the excellent correlation with the crystallographic map for all quadrants on the surfacedemonstrates the single crystal nature of the Au bead. With this crystallographic map, furtheranalysis of the reductive desorption process can be performed. Note that the assignment ofcrystallographic surfaces near the (111) facets might not be accurate because of the big sizeof the facets. Nevertheless, this effect is not expected to be large because these surfaces withlarge (111) terrace probably behave similarly as the (111) facet. In addition, the defect furtherdistorts the locations of some crystallographic surfaces, but because of the high extent of sym-metry, analysis can be conducted in the quadrants with minimum influence from the defect.Another important aspect to consider is the depth of focus because a curved electrode surfacewas imaged. The depth of focus estimated with Equation 2.39 (using λ= 525nm, NA= 0.13804.4. Results and discussionand n = 1.0 for simplification) is about 19 μm. In reality, the depth of focus is much largerbecause of the multi-medium optical aberration [14]. The experimentally estimated depth offocus (strictly speaking, should be called the depth of field in this case) is about ~90 μm in thespectroelectrochemical setup. Assuming the (100) facet is located exactly at the bottom of thebead, the z-distance between the (100) facet and the center of a (111) facet is expected tobe ~290 μm, which is three times larger than the apparent depth of field. Therefore, blurringobserved in some parts of the image (e.g., the (111) facets) has a negligible influence on thechange of fluorescence intensity with potential since the photon collection efficiency for eachregion is independent of potential.Based on the crystallographic map, surfaces of interest can be selected to make the plots offluorescence intensity as a function of potential for surfaces of interest. Figure 4.5a is a collec-tion of the intensity-potential plots for the three low-index and (210) surfaces taken from differentquadrants around the Au bead electrode surface for the AuS-C10-BODIPY SAM created with a15 min immersion time. The intensity is in logarithmic scale due to the large changes. The po-tential at the start of desorption can be determined by choosing a threshold of the fluorescenceintensity (e.g., 1 kcts/sec, indicated with the dotted line). As can be seen, the desorption poten-tial is least negative from (111) surface at ~–1.1 V , followed by (100) at ~–1.2 V and then (110)and (210) both at ~–1.3 V. This order is compatible with those obtained from polished single-crystal electrodes with voltammetry [10–12]. The intensity of the fluorescence also roughlycorrelates to the surface density of the adsorbates. Therefore, these plots contain comparableinformation with the standard linear scan voltammogram recorded when reductively desorbinga SAM on a polished single crystal electrode. Worth to mention is that the desorbed moleculesdiffuse away, both vertically and laterally. Thus, they not only contribute to the fluorescenceintensity from the region where they originate, but also contribute to the fluorescence inten-sity from the neighboring regions [168, 204, 205]. Consequently, it is not easy to use intensityto quantify the density of SAM. However, considering the diffusion coefficient for thiolate isexpected to be relatively low (~10–6cm/s) [206], so the lateral drift within the duration of themeasurement (estimated to be a few micrometers) should be negligible as compared to theimage dimension. In addition, the movement of desorbed molecules is prominent when thereis significant H2 evolution [205] while here the solution was made basic (pH = ~12) to minimize814.4. Results and discussionFigure 4.4: Fluorescence images of the AuS-C10-BODIPY SAM (15 min immersion time) mod-ified bead electrode acquired at: (a) −1.24 V (vs. Ag|AgCl) and (b) −1.30 V (vs. Ag|AgCl) withoverlay of crystallographic map showing the low-index and stepped surfaces. Adapted from [8]with permission. Copyright (2014) American Chemical Society.824.4. Results and discussionthe H2 evolution. Thus, it is reasonable to use the intensity - potential plot of a region of intereston the image to represent the desorption of adsorbates from the underlying crystallographicsurface. For moderately negative potentials before significant lateral drift and H2 evolution, thefluorescence intensity can be an indication of the density of adsorbates on the surfaces. Fur-ther complication comes from the BODIPY fluorophore, because as reviewed in Section 3.1.1,high concentration of BODIPY fluorophore can form a fluorescent J-dimer, which decreasesthe fluorescence intensity from the monomer [163–165, 204]. The error associated with thiscontribution is dependent on not only the density of fluorophore, but also the orientation, so itis difficult to compensate. Nevertheless the main purpose here is not to accurately quantify thedensity of adsorbate on different surfaces, but to compare the differences between surfaces,which are large, so it is not totally invalid to estimate the adsorbate density using the fluores-cence intensity. As a rough estimation, the density on a (111) surface is 2-3 times higher thana (100) surface, which is slightly less than a (110) surface but similar to a (210) surface. As areference, the maximum coverage of a decanethiolate SAM on a (111) surface prepared fromsolution was reported to be 6.5× 1014 molecules/cm2 [207]. The intensity-potential profilesare similar for the same types of crystallographic surfaces from different quadrants, illustratingthe single crystal quality of the Au bead electrode used. Similar in situ fluorescence imagingmeasurement and analysis were performed for the SAM created with an 18 h immersion time.The collection of the intensity-potential plots for this SAM is shown in Figure 4.5b. The intensi-ties measured for the two SAMs are very similar from corresponding crystallographic surfaces.However, the desorption potentials for the SAM created with long immersion time shifted tomore negative potentials except for (111) where the shift was not significant. This is compati-ble with previous reports showing that the adsorption of alkanethiol on Au follows two step: afast step with more than 80% coverage achieved within a few minutes and a slow step of an-nealing and organizing the resultant SAM spanning several hours without significant increaseof coverage [208, 209]. The annealed and organized SAMs stabilized by higher extent of inter-molecular interactions have been observed to have more negative desorption potentials [194].The rate for annealing and organizing the SAM is probably surface crystallography-dependentand this process was mostly completed on (111) within 15 min.It is also important to analyze the reductive desorption from the stepped surfaces along the834.4. Results and discussion 0.1 1 10 100-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0Fl. Int. (kcts/sec)E (V / Ag|AgCl)15 min HS-C10-BODIPY (1 mM)aStepping direction(111)NW(111)SW(100)(110)W(110)S(210)W(210)S 0.1 1 10 100-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0Fl. Int. (kcts/sec)E (V / Ag|AgCl)18 h HS-C10-BODIPY (1 mM)bStepping direction(111)NW(111)SW(100)(110)W(110)S(210)W(210)SFigure 4.5: Fluorescence intensity - potential for the three low-index and (210) surfaces takenfrom different quadrants around the Au bead electrode surface: (a) the AuS-C10-BODIPY SAMcreated with a 15 min immersion time, (b) the AuS-C10-BODIPY SAM created with an 18 h im-mersion time. Adapted from [8] with permission. Copyright (2014) American Chemical Society.844.4. Results and discussion(100)-(111), (111)-(110) and (110)-(100) zones on the stereographic triangle. According to thestep notations for the stepped surfaces reviewed in Section 2.1.4, a stepped surface is formedwith terrace of the nearest vertex low-index surface and step of the other vertex low-indexsurface of the zone. Thus along each zone, the stepped surfaces progress with systematicchanges in the step density. Analyzing the reductive desorption from the stepped surfacesalong three zones demonstrates the systematic influence of surface crystallography on reduc-tive desorption of the SAM. The one-dimensional analysis of plotting the intensity as a functionof potential discussed above can not only be performed on well-studied low-index and stepped(i.e., (210)) crystallographic surfaces, but also on other less-studied stepped or kinked sur-faces on the crystallographic map. However, since these surface features are all representedon the electrode surface and a more inclusive two-dimensional analysis can be performed byrepresenting the intensity change as a function of potential along three zones (in the order of(100)-(111)-(110)-(100)) with false color to create the line scan plots. Two types of line scanplots are shown. In Figure 4.6a, the false color represents the raw fluorescence intensity inlogarithmic scale while in Figure 4.6b, the false color represents the logarithm of the ratio ofthe fluorescence intensity to the maximum intensity for each pixel along the zones. The whitecontour lines are drawn at the intensity of 1 kcts/sec in Figure 4.6a and 10% of the maximumfluorescence intensity in Figure 4.6b (both calculated after subtracting the minimum intensity).The latter is to compensate for the differences in the coverage on different stereographic sur-faces. These are used as the thresholds above which the SAM is defined as desorbed. Thereare multiple complete stereographic triangles realized on the electrode surface in the mea-surements and shown is the analysis of the one in the west-northwest (WNW) half quadrant.Equivalent results can be obtained from the other complete stereographic triangles due to thesymmetry of the single crystal bead electrode. The density of broken bonds (dbb, reviewed in2.3.4), as the semi-quantitative numerical surrogate for PZC is also calculated and presentedfor the surfaces indicated in the line scan plots. The line scan plots demonstrate the stronginfluence of surface crystallography on the reductive desorption potential. It is true that apply-ing two different thresholds yields slightly different desorption potentials for the crystallographicsurfaces investigated, but the overall trends are similar. Worth to mention is that- the line scanplots can be generated for any crystallographic zone with systematic terrace-step transition,854.4. Results and discussionFigure 4.6: Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-C10-BODIPY SAM created with a 15 min im-mersion time, shown for the (100)-(111), (111)-(110) and (110)-(100) zones in the WNW stere-ographic triangle: (a) raw fluorescence intensity in logarithmic scale false colored, with a whitecontour line drawn at an intensity of 1 kcts/sec, similar to the dotted line in Figure 4.5; (b) thelogarithm of the ratio of the fluorescence intensity to the maximum intensity for each pixel alongthe zones false colored, with the white contour line drawn at 10% of the maximum intensity. Inboth figures, the density of broken bonds calculated for each surface on the y-axis is included.Adapted from [8] with permission. Copyright (2014) American Chemical Society.not restricted to the three sides of the stereographic triangle.Along the (100)-(111) zone (top panel in Figure 4.6), the desorption potentials for the sur-faces in this zone mostly correlate with dbb. The least negative desorption potentials are ob-served for the two low-index planes (111) and (100) with lowest dbb while the most negativedesorption potentials is observed for the (311) surface. The (311) surface can be expressedas 2(100)× (111) or 2(111)× (100) in step notation, which is the turning point of the tran-sition from (100) terrace to (111) terrace in this zone. Although it is not the roughest surfaceaccording to its dbb, its PZC is most negative in this zone (see Figure 2.22). The dbb decreasesdramatically away from both low-index planes, however, a smooth transition in the desorptionpotential is observed. Near the (111) surface, this smooth transition can be explained by the864.4. Results and discussionlarge (111) facet overwhelming a number of crystallographic surfaces in the map. However,the (100) facet is very small, and only slightly more negative desorption potentials are observedfrom the (911) and (711) surfaces well away from the (100) facet. Therefore, the desorptionpotential does not always follow the dbb, evidently demonstrate that PZC is only one of the ma-jor factors that control the desorption potential, and the intermolecular interactions also plays acrucial role, as reviewed in Section 4.1. The (100) terrace probably provides a favorable struc-ture for alkanethiol adsorption and organization, resulting in the smooth transition in desorptionpotential near the (100) facet.Similar correlation between desorption potential and dbb is observed along the (111)-(110)zone. A sharp change in desorption potential near the (111) surface is again not seen presum-ably because of the large (111) facet. Away from the (111) facet, desorption potential and dbbshare comparable trend until (331) which is the turning point of the transition from (111) terrace(3(111)× (111) or 2(111)× (110)) to (110) terrace (2(110)× (111)) in this zone. Nearconstant desorption potential is observed from (311) to (110) which suggests the (110) terracedominates the behavior of the alkanethiol adsorption.Along the (110)-(100) zone, a smooth transition in desorption potential is again observednear the (110) surface and the (100) surface, similar to the other two zones. The most neg-ative desorption potential is observed in (210), which is the the turning point of the transitionfrom (110) terrace (2(110)× (100)) to (100) terrace 2(100)× (110)) in this zone. This isnot unexpected considering its highest dbb and most negative PZC. However, considering allstepped surfaces in the three zones, although the (210) surface has the highest dbb and themost negative PZC, the desorption potential observed from (210) is not significantly more neg-ative than that from any other stepped surfaces. The smooth transition of desorption potentialfrom (100) surface to neighboring stepped surfaces with large (100) terrace is also observedin this zone, extending to (610).The line scan plots for the AuS-C10-BODIPY SAM created with an 18 h immersion timeare also shown in Figure 4.7 for comparison. The overall influence of surface crystallogra-phy on reductive desorption of this SAM created with a longer immersion time is similar tothat of the SAM created with short immersion time discussed above. It has been found with theone-dimensional intensity-potential plots that longer immersion time results in similar maximum874.4. Results and discussionFigure 4.7: Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-C10-BODIPY SAM created with an 18 h im-mersion time, shown for the (100)-(111), (111)-(110) and (110)-(100) zones in the WNW stere-ographic triangle: (a) raw fluorescence intensity in logarithmic scale false colored, with a whitecontour line drawn at an intensity of 1 kcts/sec, similar to the dotted line in Figure 4.5; (b) thelogarithm of the ratio of the fluorescence intensity to the maximum intensity for each pixel alongthe zones false colored, with the white contour line drawn at 10% of the maximum intensity.884.4. Results and discussionintensity and more negative desorption potential presumably because of the adsorbate anneal-ing and organizing. This is also observed with the line scan plots for all surfaces investigated.However, the negative shift of desorption potential is not the same for all surfaces. The ~20 mVnegative shift is minimal for the (111) surface. The desorption potentials for the stepped sur-faces with (111) terrace also do not shift too much. On the contrary, the ~70 mV negative shiftfor the (100) surface is the maximum shift observed. A large negative shift is also observedon surfaces with (100) terrace. The ~50 mV negative shift for the (110) is also relatively big.This big negative shift also applies to surfaces with (110) terrace. The difference in this shiftin the onset of desorption suggests high extent of molecular reorganization and stabilizationoccurring on the (100) and (110) terraces during the long immersion process.Analyses performed on the two AuS-C10-BODIPY SAMs with short and long immersiontimes demonstrate the general correlation of desorption potential with dbb or PZC of the un-derlying crystallographic surfaces and the progression of the SAM formation on different crys-tallographic surfaces. However, there are discrepancies between the trends of desorption po-tential and dbb or PZC, most notably for surfaces with a large (100) terrace or a large (110)terrace. Coincidentally, they are surfaces with the largest negative shift of desorption potentialafter hours of adsorbate annealing and organizing to achieve a higher extent of intermolecularinteraction. These observations from the analyses of the in situ fluorescence imaging mightoriginate from the energetics of the particular adsorption sites on the (100) and (110) terraces.However, as reviewed in Section 2.2.2, even the adsorption site on (111) surface is under con-stant debate, so further studies are required to elucidate the details of adsorption on the (100)and (110) terraces. Moreover, although it is has been observed that lifting the reconstructed(111) surface is achieved within 15min upon adsorption of alkanethiol at a slight negative poten-tial (–0.38 V (vs. SCE)) [68], there is a lack of in situ study on lifting the reconstruction of otherreconstructed surfaces by alkanethiol adsorption, so the possibility of lifting the reconstructed(100) and (110) (see Section 2.1.5) over long immersion time is not excluded. As a conclud-ing remark for this section, while this in situ fluorescence imaging performed on single crystalbead electrode cannot reveal all details, it provides direction to the in-depth investigation of theinteresting surfaces using methods to analyze molecular orientation or structure to understandthe adsorbate-surface and adsorbate-adsorbate interactions at the specific adsorption site.894.4. Results and discussion4.4.2 Reductive desorption of the AuS-Aib4-BODIPY SAMThe Aib peptides are a class of synthetic peptides with special structure and potentially usefulin biosensing applications [210–212]. They form the 310-helix induced by the intramolecularhydrogen bonds, which in turn induces a strong dipole moment along peptide backbone. Whenthe Aib peptides are thiol-modified and immobilized on a Au surface, not only does the helicalstructure retained, but an intermolecular hydrogen bond network is formed as well [210, 211].The Aib peptides and Aib peptide thiolate SAMs will be further discussed in Chapter 5. Here, aAib peptide thiolate SAM is studied as a contrast to the alkanethiolate SAM to demonstrate thewide applicability of the influence of surface crystallography and the influence of intermolecularinteractions on reductive desorption.The BODIPY fluorophore-labeled thiol-modified Aib peptide HS-Aib4-BODIPY was used tocreate a SAM, and with the fluorophore, similar in situ fluorescence imaging measurement wasperformed to study the reductive desorption process. One difference in the measurement ascompared to that for the AuS-C10-BODIPY SAM was that at each potential step, the waitingtime was slightly longer. The AuS-C10-BODIPY SAM here was created from a dilute solutionof HS-Aib4-BODIPY over a relative short period of immersion time as compared with highlypacked SAMs reported in literature [211]. In addition, the low solubility of the HS-Aib4-BODIPYprevented the desorbed molecules from diffusing quickly. These two factors combined resultedin lower fluorescence signal than the AuS-C10-BODIPYSAM. In order to achieve better imagingquality, the waiting time was longer (the camera exposure time was longer as well) at each stepto allow for larger separation between the fluorophore and the electrode.The substrate electrode used for the AuS-Aib4-BODIPY SAM was another single crystalbead electrode with similar orientation to the one used for the AuS-C10-BODIPY SAM (i.e.,the (100) facet in the center and the 4 (111) facet close to the four corners). Figure 4.8 showsthe montage of selected fluorescence images taken from –1.14 V to –1.36 V (vs. Ag|AgCl)in –20 mV increments during the reductive desorption of the AuS-Aib4-BODIPY SAM. Thereductive desorption was observed first from the 4 (111) facets, similar to what was observedfor the AuS-C10-BODIPY SAM. The reductive desorption from the (100) facet followed (111)at –1.12 V (vs. Ag|AgCl), but a square shape was seen instead of the cross shape around the904.4. Results and discussionFigure 4.8: Montage of selected fluorescence images taken from –1.08 V to –1.30 V (vs.Ag|AgCl) in –20 mV increments during the reductive desorption of the AuS-Aib4-BODIPY SAM.(100) facet for the AuS-C10-BODIPY SAM. At –1.16 V (vs. Ag|AgCl), fluorescence intensitystarted to increase quickly for the rest of the surface. The symmetry of the fluorescence imagesagain demonstrates the single crystal character of the bead electrode surface with the intensitydistributed according to the fcc single crystal pattern in the four quadrants equally (except forthe northeast quadrant distorted by a defect similar to that observed in Figure 4.2).Mapping of the electrode surface was performed with the fluorescence image taken at –1.18V (vs. Ag|AgCl). With the crystallographic map (Figure 4.9), the one-dimensional intensity-potential analysis for selected crystallographic surfaces and two-dimensional line scan anal-ysis for selected zones were also conducted and plotted in Figure 4.10. The desorption ofAuS-Aib4-BODIPY from various surfaces follows a similar sequence to that observed for theAuS-C10-BODIPY SAMs except for three major differences. First, reductive desorption occursat less negative potentials for the AuS-Aib4-BODIPY SAM (applying the 1 kcts/sec thresholdprobably even overestimate the negative desorption potentials considering the low coverageof the AuS-Aib4-BODIPY SAM). Moreover, a unique two-step desorption is observed from the(111) facets but not from any other surfaces studied. This two-step desorption will be furtherdiscussed in Chapter 5. Furthermore, the potential separation from (100) surface and the other914.4. Results and discussionFigure 4.9: Fluorescence images of the AuS-Aib4-BODIPY SAM modified bead electrode ac-quired at –1.18 V (vs. Ag|AgCl) with overlay of crystallographic map showing the low-index andstepped surfaces. Adapted from [8] with permission. Copyright (2014) American Chemical So-ciety.two surfaces, (100) and (210) is smaller than that observed for the AuS-C10-BODIPY SAM.The low adsorbate concentration and short immersion time resulted in a layer that may not beas well organized as reported ones on extended gold surfaces [211]. In this low coverage case,it might be assumed that the intermolecular hydrogen bonds may not play such a significantrole in influencing the desorption potential. However, the retaining dipole moment (in the caseof AuS-Aib4-BODIPY, the positive pole is proximate to the surface while the negative pole isdistal from the surface) can still have an impact on the desorption process. The maximumfluorescence intensity, which correlates to the coverage, on the (111) and (110) surface areroughly double that of the (100) and (210) surfaces. Considering the low packing density of theSAM, this surface coverage difference possibly reflects the initial stages of adsorption processas dependent on the coordination with surface.The line scan plots for the (100)-(111), (111)-(110) and (110)-(100) zones are presentedin Figure 4.11, showing the influence of the surface crystallography on the reductive desorp-tion for the AuS-Aib4-BODIPY SAM. Because of the low coverage, the desorption potentials924.4. Results and discussion0. -1.0 -0.8 -0.6 -0.4 -0.2 0.0Fl. Int. (kcts/sec)E (V / Ag|AgCl)2 h HS-Aib4-BODIPY (1 µM)Stepping direction(111)NW(111)SW(100)(110)N(110)W(210)N(210)WFigure 4.10: Fluorescence intensity - potential for the three low-index and (210) surfaces takenfrom different quadrants around the AuS-Aib4-BODIPY SAM modified Au bead electrode sur-face. Adapted from [8] with permission. Copyright (2014) American Chemical Society.extracted with the 1 kct/sec threshold (Figure 4.11a) are slightly more negative than those ex-tracted with the 10% of themaximum intensity threshold (Figure 4.11b). However, similar trendsare observed with the two thresholds. The comparison of desorption potential as a functionof underlying crystallographic surfaces with dbb of corresponding surfaces shows significantcorrelation between the two. The major discrepancy lies in the (111) surface with two-step des-orption. The two-step desorption for the (111) surface is also observed for the (755) surfacein the (100)-(111) zone and (553) surface in the (111)-(110) zone, which may be due to inac-curate indexing inside the big (111) facets. However, from the crystallographic map in Figure4.9, the (553) surface is clearly off the facet, suggesting the two-step desorption also occurson surfaces with large (111) terrace. Other than the (111) surface, despite the broad similar-ity between the desorption potential trend and the dbb trend, the Aib peptide SAM reductivedesorption potentials determined for various crystallographic surfaces are similar and do notvary too much (largest difference of ~50 mV). To date, only Aib peptide SAMs deposisted onplanar (111) surfaces have been systematically studied [211–214], with no reports on any othercrystalline surfaces. Considering the big difference in the desorption behaviors between (111)surface and other surfaces for this Aib peptide SAM, it would be valuable to further explore thestructures and properties of Aib peptide SAMs on other surfaces to elucidate the difference.Significant differences have been observed in the influence of surface crystallography onthe reductive desorption for the AuS-Aib4-BODIPY SAM and the AuS-C10-BODIPY SAMs.934.4. Results and discussionFigure 4.11: Influence of surface crystallography on the fluorescence intensity changes withpotential during reductive desorption of the AuS-Aib4-BODIPY SAM, shown for the (100)-(111),(111)-(110) and (110)-(100) zones in the WNW stereographic triangle: (a) raw fluorescenceintensity in logarithmic scale false colored, with a white contour line drawn at an intensity of 1kcts/sec, similar to the dotted line in Figure 4.10; (b) the logarithm of the ratio of the fluorescenceintensity to the maximum intensity for each pixel along the zones false colored, with the whitecontour line drawn at 10% of the maximum intensity. In both figures, the density of brokenbonds calculated for each surface on the y-axis is included. Adapted from [8] with permission.Copyright (2014) American Chemical Society.944.4. Results and discussionThe reductive desorption potential for the AuS-Aib4-BODIPY SAM seems to follow the dbb orPZC more closely than the AuS-C10-BODIPY SAMs except for the unique (111) surface. Theinfluence of intermolecular interactions on the reductive desorption has been demonstratedby comparing the two AuS-C10-BODIPY SAMs created with different immersion time. Whilethe hydrogen bonds between Aib peptide molecules might not be strong enough to impact thereductive desorption process on the whole bead surface for this low coverage SAM, locally theremight be densely-packed adsorbates connected by the hydrogen bonds, or the orientation of thedipole moment might be strongly dependent on surface crystallography via the packing density.Most importantly, comparing these two distinct types of SAMs shows the wide utility of this in situfluorescence imaging approach on the Au single crystal electrodes in exploring the influence ofsurface crystallography on various types of SAMs in a self-consistent and comparable fashion.4.4.3 Mapping the desorption potential for the alkanethiolate and Aib peptidethiolate SAMs on the whole single crystal Au bead surfaceThe influence of surface crystallography on the reductive desorption potentials for selected two-dimensional zone can be observed with the line scan plots described above. It is also importantand convenient to obtain the desorption potentials across the whole surface for one particularSAM. This desorption potential map can be achieved by interpolating the value of the potentialat which the fluorescence intensity crosses a threshold going through one reductive desorptionimage stack (background subtraction was conducted by subtracting the first image taken at 0V (vs. Ag|AgCl)). The analysis was performed at each pixel in the fluorescence images forall three SAMs (the AuS-C10-BODIPY SAM formed with short deposition time, the AuS-C10-BODIPY SAM formed with long deposition time and the AuS-Aib4-BODIPY SAM) using either 1kct/sec or 10% of the maximum fluorescence intensity (for that particular pixel) as the threshold.The latter threshold aims at accounting for the difference in coverage on difference surfaces.The maps of the interpolated reductive desorption potential for the AuS-C10-BODIPY SAMsand for the AuS-Aib4-BODIPY SAM are shown in Figure 4.12 and Figure 4.13, respectively.The maps in the left column were generated by applying the 1 kct/sec threshold while the themaps in the right column were generated by applying the 10% of maximum intensity threshold.954.4. Results and discussionFigure 4.12: Maps of the interpolated reductive desorption potential (V (vs Ag|AgCl)) for theAuS-C10-BODIPY SAMs using a threshold of 1000 kct/sec (left column) or a threshold of 10%of the maximum intensity (right column). First row (a, b) for the AuS-C10-BODIPY SAM createdwith 15 min immersion time and second row (c, d) for the AuS-C10-BODIPY SAM created with18 h immersion time. Adapted from [8] with permission. Copyright (2014) American ChemicalSociety.964.4. Results and discussionFigure 4.13: Maps of the interpolated reductive desorption potential (V (vs Ag|AgCl)) for theAuS-Aib4-BODIPY SAM using (a) a threshold of 1000 kct/sec or (b) a threshold of 10% of themaximum intensity. Adapted from [8] with permission. Copyright (2014) American ChemicalSociety.Density of Broken Bonds (fcc) 110110110110111111111111755211311411511611911755211311411511611911755211311411511611911755211311411511611911910610410310210320540910610410310210320540910610410310210320540910610410310210320540991551331221332554991551331221332554991551331221332554991551331221332554991551331221332554991551331 221 332 554991551331221332554991551331221332554765765765765765765765765432432432432432432432432321321321321321321321321531531531531531531531531951951951951951951951951761761761761761761761761541541541541541541541541431431431431431431431431532532532532532532532532954954954954954954954954421421421421421421421421521521521521521521521521721721721721721721721721921921921921921921921921Figure 4.14: The map of calculated density of broken bonds (dbb) for a fcc spherical surfaceobserved from the bottomwith the (100) surface in the center. Adapted from [8] with permission.Copyright (2014) American Chemical Society.974.4. Results and discussionThe map of calculated dbb for a fcc spherical surface projected to a plane with the (100)surface in the center is plotted in Figure 4.14 with color coding to match the desorption potentialtrend, facilitating a direct comparison. The (111) has the highest dbb of ~7 (yellow) and the(210) has the lowest dbb of ~9 (dark blue). Note that the one major assumption here is that thereconstructed crystallographic surfaces are reverted back to unreconstructed (1×1) structuresupon adsorption of thiol-modified molecules. This has been observed for the three low-indexsurfaces: (111) [68, 73], (100) [96, 97] and (110) [215] surfaces, even though the kinetics forlifting the reconstruction has been explored only for the (111) surface [68]. It has to be admittedthat the possibility of partial lifting of the reconstructed surface exists and has been observedexperimentally [74, 99–101]. However, from the fundamental point of view, to complete the de-reconstruction process, the SAM has to be highly-packed and well-organized (typically seen onhigh density long alkanethiolate SAMs) so that the relatively electronegative thiolate head groupcan withdraw the excess amount of negative charge on the surface [74]. Therefore, it is not un-reasonable to assume that the AuS-C10-BODIPY SAMs here fall into this category. As for theAuS-Aib4-BODIPY SAM, considering the dipole moment from the Aib peptide with the positivepole proximate to the surface, the assumption of de-reconstruction by the AuS-Aib4-BODIPYSAM, despite its low density, is probably valid. In addition, the highly stepped surfaces, (210),(320), (311) and (511) have been shown not to reconstruct and those reconstructed surfacesappear to be less stepped and have large low-index plane terrace, which might have a largeextent of resemblance to the terrace plane [44, 45, 141, 142]. Therefore, all surfaces are as-sumed to retain their prinstine unreconstructed structures and the dbb map for unreconstructedsurfaces is used for comparison with the desorption potential maps. The general influence ofsurface crystallography is similar for the two types of adsorbates compared with the surfacedbb map: the lower the dbb, the more negative the desorption potential. However, very sig-nificant differences can be observed between the desorption potential map and the dbb map,which again suggests dbb or PZC is not the only factor that determines the desorption po-tential.Only small differences are observed when comparing the desorption potential maps foreach AuS-C10-BODIPY SAM generated with the two different thresholds shown in Figure 4.12.The major discrepancy between the desorption potential maps for these alkanethiol SAMs andthe dbb map is the central cross observed in the desorption potential maps instead of a simple984.4. Results and discussionsquare with 4-fold symmetry. This central cross covers the (100) facet and a series of steppedand kinked surfaces with large (100) terrace. This smooth transition of desorption potentialnear the (100) facet has been shown with the line scan plots. The line scan plots were createdfor low-index and stepped surfaces, whereas the desorption potential maps here also show thecharacteristics of the kinked surfaces. One arm of the cross region around the (100) surfaceextends to (610) along the (100)-(110) zone and almost (911) along the (100)-(111) zone. Thestep notations for (610) and (911) are 6(100)× (110) and 5(100)× (111), respectively. Thissuggests that there might be a boundary between surfaces with 6 atom wide (100) terrace andthose with 5 atom wide (100) terrace, beyond which the adsorption environment for alkanethiolis completely different. Adsorption of alkanethiol on Au(100) adopts a c(2× 8) structure with(1× 4) Au missing row [97, 98]. The (1× 4) Au missing row requires at least 6 atom wide(100) terrace, which might explain the two different adsorption environments. The most neg-ative desorption potential is observed from the (210) surface for the AuS-C10-BODIPY SAMsas expected considering the lowest dbb of (210). However, this negative desorption potential isalso observed from stepped and kinked surfaces around the central cross, which is not agree-ing with the dbb trend. The desorption potential maps for the two AuS-C10-BODIPY SAMsshow similar features, but the desorption potential is more negative across the whole surfacefor the long immersion SAM. After extending the immersion time, a small negative shift (<20mV) of desorption potential is observed from the four (111) facets, whereas this negative shiftis considerably larger (~40-60 mV) from other surfaces.The desorption potential maps for the AuS-Aib4-BODIPY SAM generated with the two dif-ferent thresholds are considerably different presumably due to the low packing density andlarge difference among surfaces of this SAM, so it is reasonable to discuss based on the des-orption map generated with the 10% of the maximum fluorescence intensity threshold. Thedesorption potential map for the Aib peptide thiolate SAM shows closer correspondence to thedbb map. The two-step desorption for the (111) surfaces is hidden here. A square feature isseen in the center around the (100) surface in the desorption map, which is highly similar tothe central region in the dbb map. The large (100) terrace in the stepped or kinked surfacesdoes not seem to provide similar adsorption environment to the (100) surface, presumably be-cause of the larger footprint of the thiol-modified Aib peptide molecules [211]. There is also994.5. Conclusionsa similarity in the regions around the (210) surface, which resembles a round valley, betweenthe desorption potential map and the dbb map. However, the lowest desorption potential is notobserved from the (210) surface, but from the (511) to (211) in the (100)-(111) zone featuring2-3 atoms wide (111) or (100) terrace. The dbb for these surfaces ((511), (311) and (211))are somewhat comparable to each other but are not at the extreme value of the (210) surface.Thus the highly-stepped surfaces with narrow terracemight be the favorable adsorption sites forthe AuS-Aib4-BODIPY molecules where adsorbates of this size might coordinate well betweenstep edges or stable hydrogen bond networks might form on these surfaces.Large differences can be observed in the reductive desorption behaviors for different SAMsand significant discrepancies can be found between the desorption potential maps and thedbb map. These comparisons represent useful starting points for further in-depth molecularstudy of adsorbates on these various surfaces to understand these differences. The majority ofexperimental studies have focused on low-index planes for alkanethiol SAMs which correlatewell with the dbb map. There are significant discrepancies between the desorption potentialmap and the dbb map on the stepped or even kinked surfaces, which are worth exploring tounderstand these differences. For the Aib peptide thiolate SAM, so far the only crystallinesubstrate studied is the (111) surface. However, considering the big difference in desorptionbehaviors between on the (111) surface and on the other surfaces, pursuing the studies of Aibpeptide thiolate SAMs on surfaces other than (111) might be rewarding. Therefore, the use ofa single-crystal Au bead and the in situ imaging approach can guide the more detailed studyof the adsorption onto the unusual and understudied crystallographic surfaces.4.5 ConclusionsThe work discussed in this chapter introduces the single crystalline Au bead electrode withthe family of crystallographic surfaces symmetrically distributed following the fcc crystal struc-ture as the substrate electrode for SAMs. With this type of electrode, the influence of surfacecrystallography on the reductive desorption of SAMs has been investigated with the in situ fluo-rescence imaging method. The fluorescence images displaying the expected symmetry of thefcc crystal surface enable the indexing of the surface. More importantly, all identifiable crys-1004.5. Conclusionstallographic surfaces have been analyzed self-consistently under identical conditions for onesingle SAM in one single measurement. This not only reduces the number of experimentalreplicates, minimizing the error associated with them, but also enables the study of SAMs ona series of crystallographic surfaces in the stereographic triangle with self-consistency.Three SAMs prepared from two different adsorbates were characterized to demonstrate thewide applicability of this method to a variety of systems and to explore the influence of inter-molecular interaction on reductive desorption: a AuS-C10-BODIPY SAM created with shortimmersion time, a AuS-C10-BODIPY SAM created with long immersion time, and a AuS-Aib4-BODIPY SAM. Based on the intensity-potential plots generated with one-dimensional analysison selected crystallographic surfaces, line scan plots generated with two-dimensional analysison selected crystallographic zones crossing many related surfaces and desorption potentialmaps for the whole Au bead surface in view, the general correlation between desorption poten-tial and dbb as the surrogate for PZC of the surface studied has been realized. However, thereare large discrepancies between the desorption potential and dbb and in the reductive desorp-tion behaviors between the three types of SAMs. Comparison of the two AuS-C10-BODIPYSAMs informs on how the SAM annealing process influences on the reductive desorption pro-cess due to different crystallographic surfaces. The differences between the AuS-C10-BODIPYSAMs and AuS-Aib4-BODIPY SAM and the discrepancies between the desorption potentialtrend and the dbb trend are indicative of the influence of the generalized intermolecular inter-actions on the reductive desorption process. Thus the results from this method are paving theway for future in-depth explorations of various SAMs, unique interactions, and understudiedstepped or kinked surfaces.101Chapter 5In situ fluorescence imagingcharacterization of theα-aminoisobutyric acid peptidethiolate self-assembled monolayerson Au(111) surfacesThis chapter discusses the study of fluorophore-tagged α-aminoisobutyric acid (Aib) peptidethiolate SAMs deposited on Au electrodes using in-situ fluorescence imaging. Aib peptideshave a restrained conformation and an oriented dipole moment along the peptide backboneresulting from their stiff 310-helical structure and intermolecular hydrogen bond network. Thusthiol-modified Aib peptide have been used to make SAMs with high uniformity and stability,which are important in sensing applications.5.1 Aib peptides and Aib peptide SAMsDevising biosensors based on SAMs has developed into an intense multidisciplinary researchfield. SAMs are organized surface structures formed when organic compounds are sponta-neously adsorbed onto solid substrates. Therefore, SAMs are widely utilized to link the biore-ceptor and the transducer together, acting as the platform for biosensors, and most commonly,electrochemical biosensors. For example, a number of DNA sensors, a class of biosensors,have been developed, of which many are based on alkanethiolate SAMs (using alkanethio-1025.1. Aib peptides and Aib peptide SAMslate as the molecular spacer and diluent) [52–54, 108]. Despite the development of a varietyof electrochemical biosensors over the past decades, their low sensitivity and specificity stillhinders breakthroughs. One major reason for these issues is that the surface organizationsare still far from highly ordered [1]. In other words, the SAMs are formed with a considerablenumber of defects [4, 65, 66].As reviewed in Section 2.2.2, conventional alkanethiolate SAMs with van der Waals inter-actions tend to be highly dynamic and defective. Although alkanethiolate SAMs have beenintensely investigated and employed as the platform for biosensors, it is questionable whetherthey are the best candidates [65, 66]. Thus, to form SAMs with high order, one natural consider-ation is to replace the alkyl chains with other organic chain structures with favorable propertiesto promote the foundation of a SAM with less defects which result from spacer chains. Typicallythere are two aspects to improving spacer structure. First, since alkanethiolate SAMsmainly in-teracts with neighboring molecules through intermolecular van der Waals forces, alkanethiolateSAMs tend to be more flexible. Thus, if additional intermolecular interactions can be involved inthe system, the resultant SAMs would not only be more stable, but more densely packed withcorrespondingly fewer defects. Secondly, alkyl chains bear relatively flexible conformations,which accounts for the observed phase transitions on the surface [106, 107]. Consequently,good candidates for spacers in SAMs require a conformation constrained by intramolecular in-teractions. Combining these two aspects, Aib peptides, with the 310-helix secondary structure,becomes a class of attractive candidates for molecular spacers in SAMs [210–212].The general structure of the thiol-modified Aib(n+1) series is shown in Figure 5.1. Thepresence of the HS terminal group enables formation of Aib peptide SAMs on Au (as wellas other metals that can form a M-S bond), while the other end of the chain can be furtherfunctionalized to meet various application requirements. The Aib peptides bear the 310-helixsecondary structure induced by intramolecular C=O· · ·H-N hydrogen bonds [216, 217]. Then in the Aib(n+1) series denotes the number of hydrogen bonds. In the 310-helix structure,C=O· · ·H-N hydrogen bonds are formed between residues i and i+3. On average, a singleturn of the helix requires 3.24 residues, and the length of the peptide chain increases by 1.94Å per residue, resulting in a helix pitch of 6.29 Å. The steric structure of 310-helix is highlyconstrained at the α-carbon with backbone torsion angles of 310-helix being φ = 57° and ψ =1035.1. Aib peptides and Aib peptide SAMs30°. These parameters fall into the same conformational map (Ramachandran plot, or [φ, ψ]plot) as the those of the common α-helix which involves hydrogen bonds between residues iand i+4. The comparison of the two helices is shown in Table 5.1. The significant differencein bonding schemes of the two helices lead to considerable differences in stability. Generally,310-helix is not as stable as α-helix, but their conformational proximity might infer 310-helixas an intermediate involved in the process of folding α-helical proteins [217]. Although the α-helix is the dominant intramolecular secondary structure in peptides, more than 8 residues arerequired to form an α-helix. On the contrary, the length of 310-helix can range from one turn to7-12 residues, which gives a high degree of control over spacer length [210, 218]. The highlysterically constrained conformation of Aib peptides also induces a strong dipole moment alongpeptide backbone. Orientation of the dipole moment is maintained with increasing number ofAib residues due to the accompanied increase in the number of hydrogen bonds [210, 218,219]. As can be seen from Table 5.1, in the 310-helix structure, the angle of the carbonylgroup in the peptide bonds with respect to the helical axis is 28°. The projection of these polarcarbonyl groups onto the helical axis result in a dipole moment along the peptide backbone.Orientation of the dipole moment is maintained with increasing number of Aib residues due tothe accompanied increase in the number of hydrogen bonds [210, 211, 218, 220]. Here, thisHS-Aib(n+1) series is called the “+” series of peptide because the dipole moment orients fromthe positive pole on the N-terminus to the negative pole on the C-terminus. Likewise, the seriesof peptides with the HS- group on the other side is called the “–” series. A “+” thiol-modified Aibpeptide, HS-Aib4, is chosen here because the dipole moment orienting this way help stabilizethe SAM formed by this thiol-modified Aib peptide. The evidence for this is the more negativedesorption potential for a SAM form by a “+” thiol-modified Aib peptide than that formed bya “–” thiol-modified Aib peptide of the same number of Aib residues [221]. The maximumapparent packing density of the AuS-Aib4 SAM on planar Au(111) electrode was determined tobe ∼3× 1014 molecules cm–2 (0.33 nm2/molecule) using the conventional charge integrationtypically done after linear scan voltammetry measurement [211]. The simple charge integrationtends to provide an overestimate of the packing density (see Section 4.1). Even after taking thisinto account, the footprint calculated based on this measurement is significantly smaller thanthe ~0.8 nm2/molecule calculated from X-ray analysis result [222], which can be rationalized by1045.1. Aib peptides and Aib peptide SAMsHS NHHNNHOnOOFigure 5.1: The primary structure of the “+” thiol-modified Aib peptide series.Table 5.1: Comparison of 310-helix and α-helix. Summarized from [216] and [217].Parameter 310-helix α-helixBonding scheme i and i+3 i and i+4Backbone torsion angles (φ, ψ) (57°, 30°) (63°, 42°)C=O· · ·H-N bond angle 128° 156°Rotation per residue 111° 99°Axial translation per residue 1.94 Å 1.94 ÅNumber of residues per turn 3.24 3.63Pitch 6.29 Å 5.67 ÅAngle of C=O with respect to helical axis 28° 13°Dipole moment 4.57 D 5.16 DDiamter 9.7 Å (Aib) [222] Dependent on side chainsthe influence of the dipole moment on the desorption charge [211]. Interestingly, the calculatedfootprint for the HS-Aib(n+1) series further exhibits a periodicity based on n, and one periodis comprised of three types of peptides where n= 3k,3k+ 1,3k+ 2 (k represents integers).A slight increase of calculated footprint within one period was observed, whereas an overalldecrease of calculated footprint was observed from period to period with increasing k [221].This periodicity is probably due to a helical turn of three residues.SAMs formed by thiol-modified Aib peptide on gold nanoclusters [210, 218, 220] and goldplates [211] have been reported by Maran and co-workers. It was found that the peptidemolecules form well packed layers on both substrates. In addition to the intramolecular hy-drogen bonds, previous studies also show that a peptide molecule can form hydrogen bondswith adjacent molecules in a SAM, which creates a large hydrogen bond network on the surface.This hydrogen bond network accounts for the well-defined organization of this type of Aib pep-tide SAMs [210, 211, 218, 220]. In addition, the hydrogen bonds are also dipoles, which couldalso feel the influence from the electric field, so the conformation of the Aib peptide may also bemanipulated by an electric field. Venanzi and coworkers also studied SAMs on gold surfacesusing peptides with 310-helix Aib peptide as the core structure [72, 212–214, 223, 224]. Thesepeptide SAMs were also densely and homogeneously packed. Furthermore, the dipole mo-1055.2. Objectivesment along the peptide backbone could facilitate the formation of bi-component nano-structurethrough helix-helix interactions between opposite dipole moments on the surface [214]. Theseproperties have made Aib peptide SAMs promising platforms for bio-sensing [211] and photo-voltaic [223, 224] applications.As the platform for eletrochemical biosensors, the property of facilitating electron transfer isnecessary. Previous studies have shown that the Aib peptides can act as the electron tunnelingmediators. The Aib peptides display a special dependence of intramolecular electron transferrate on the number of Aib units [219, 225]. The intramolecular hydrogen bonds play a criticalrole in facilitating electron transfer. Thus the electron transfer mechanism through a Aib peptidebridge is believed to be a combination of the superexchangemechanism which greatly dependson the distance between the donor and the acceptor, and the hopping mechanism which greatlydepends on the number of intramolecular hydrogen bonds [225].5.2 ObjectivesAs an extension of the studies in Chapter 4, the Aib peptide thiolate SAMs will be exploredvia electrochemistry on a polished Au(111) electrode and spectroelectrochemically using aAu(111) facet on a single crystal Au bead electrode. It was shown in Chapter 4 that the low-packing-density Aib peptide SAM shows a two-step desorption from the (111) surface. In con-trast, high-packing-density SAMs similar to those reported in the literature [211] will be preparedand studied, thereby giving insights into Aib peptide SAMs.Additionally, the thickness of a SAM is not only dependent on the length of the adsorbate,but also the orientation. In the case of Aib peptides, the peptide backbone is featured with anoriented dipole moment induced by polar groups. This dipole moment may be influenced bythe electric field, which leads to the change of the orientation or structure. If the adsorbate isfurther labeled with a fluorophore, the change of SAM thickness can be reported as the changein fluorescence intensity. Therefore, in situ fluorescencemeasurements will be performed usingmodulated potential steps to investigate the potential induced orientation / structure change ofthe low-packing-density Aib peptide SAMs on a Au(111) facet.1065.3. Experimental5.3 Experimental5.3.1 The substratesTwo types of substrate electrodes were used in the formation of Aib peptide SAMs. Fabricationof both types involved melting an ultrapure gold wire with a torch and slowly cooling downto form a predominant single crystal bead. The first type of substrate was a planar Au(111)electrode accomplished by polishing a single crystal bead as described in Section 3.2. Thesecond type of substrate was a single crystal bead similar to that used in Chapter 4, exceptthat a (111) facet was formed very close to the bottom of the bead by carefully adjusting theangle of melting. For the second type of electrode, the orientation of the electrode was carefullyadjusted by twisting the electrode stem so that the (111) facet was perpendicular to the lightpath in subsequent spectroelectrochemical measurements.5.3.2 Preparation of AuS-Aib4-BODIPY SAMsThe thiol-modified Aib peptide is HS-Aib4-BODIPY the same as used in Chapter 4. The generalpreparation procedure for AuS-Aib4-BODIPY SAM is described as below. A clean substrateelectrode was immersed in the thiol-ethanol solution with designated concentration for a prede-termined period of deposition time. The modified electrode was then rinsed with and immersedin ethanol for 30 min to remove the non-specifically adsorbed species. The highest concentra-tion of the HS-Aib4-BODIPY solution was 60 μM and all the lower concentration solutions wereprepared from this stock. Table 5.2 lists all the AuS-Aib4-BODIPY SAMs prepared for furtherinvestigations with the conditions under which they were formed. Besides the categorizationbased on the substrate, they can be further distinguished based on packing density. The SAMsprepared with 60 μMsolution are categorized as high-packing-density SAMswhereas the SAMsprepared with 1-2 μM solutions are categorized as low-packing-density SAMs.1075.3. ExperimentalTable 5.2: AuS-Aib4-BODIPY SAMs prepared and their formation conditionsElectrode Solution concentration Deposition timePolished Au(111) 60 μM 1 hSingle crystal Au bead 60 μM 2 hSingle crystal Au bead 1 μM 0.5 hSingle crystal Au bead 1 μM 2 hSingle crystal Au bead 1 μM 18 hSingle crystal Au bead 2 μM 0.5 h5.3.3 Electrochemical and in situ fluorescence imaging characterization of theSAMsAll electrochemical and in situ fluorescence imaging measurements were conducted in 50 mMKClO4 solution (pH = 12 (±0.5) adjusted with KOH). The setup for these measurements hasbeen generally described in Section 3.3 and Section 3.4.A high-packing-density AuS-Aib4-BODIPY SAM deposited on a polished Au(111) electrodewas characterized electrochemically. A cyclic voltammetry measurement from 0.10 V to –1.30V (vs. SCE, converted to vs. Ag|AgCl in the subsequent analysis) at 20 mV/s was performed tostudy the reductive desorption process from the (111) surface and quantify the packing densityof this SAM.In situ fluorescence imaging measurements were conducted on the high-packing-densityAuS-Aib4-BODIPY SAM deposited on a single crystal Au bead electrode with the Evolve® 512EMCCD camera while capacitance was measured at the same time. A similar spectroelectro-chemical measurement of the reductive desorption process of a AuS-Aib4-BODIPY SAM wasdiscussed in Chapter 4. However, in this chapter, the electrode was oriented so that a Au(111)facet was in view through an Olympus LCPlanFL 20× objective (NA = 0.40).For the low-packing-density AuS-Aib4-BODIPY SAMs deposited on a single crystal Au beadelectrode, besides the potential stepping to study the reductive desorption, modulated potentialsteps were applied to study the potential induced response of the low-packing-density SAMsover the potential range before the optical onset of reductive desorption. Specifically, the po-tential was stepped from 0 V (vs. Ag|AgCl) to –0.7 V (vs. Ag|AgCl) with –25 mV intervals butthe potential was returned to the base potential (0 V) after each step potential. Five sequentialimages were taken at each potential step to monitor the gradual change of intensity. Note that1085.4. Results and discussionin the reductive desorption measurements, the fluorophore-substrate distance changed drasti-cally during the desorption process, so all imaging settings (illuminator intensity, exposure timeand EM gain) were set to low sensitivity levels to avoid signal saturation. For the potential step-ping measurements to study potential modulated response, all imaging settings were adjustedto high levels to magnify the signal by about ten times because the intensity change is expectedto be small as compared to that caused by desorption. The Evolve® 512 EMCCD Camera withthe desirable sensitivity and EM gain function was employed to study the low-packing-densityAuS-Aib4-BODIPY SAMs.5.4 Results and discussion5.4.1 High-packing-density AuS-Aib4-BODIPY SAM deposited on a polishedAu(111) electrodeAib peptide thiolate SAMs deposited on planar Au(111) surfaces have been studied. Reductivedesorption was one of the methods used to characterize the electrochemical stability and park-ing density of the SAMs. Here the thiol-modified Aib peptide is further tagged with a BODIPYfluorophore, facilitating the spectroelectrochemical characterizations. However, the presenceof the fluorophore noticeably changes the property of the adsorbate. The most obvious dif-ference is the solubility of the adsorbate HS-Aib4-BODIPY (60 μM), which is almost ten timeslower than that of the unlabeled HS-Aib4 used in [211]. Therefore, before further in situ flu-orescence imaging characterization, electrochemical characterization using cyclic voltamme-try was performed on the HS-Aib4-BODIPY SAM prepared on a polished Au(111) electrode.Figure 5.2 shows the cyclic voltammogram recorded during the reductive desorption of theHS-Aib4-BODIPY SAM. Compared with the voltammogram of the AuS-Aib4 SAM shown in[211], the peak potential of ~0.9 V (vs. Ag|AgCl) is comparable. The major difference is thebroader current peak for this AuS-Aib4-BODIPY SAM with the onset appearing at ~–0.6 V (vs.Ag|AgCl) which is less negative than the HS-Aib4 SAM. This might be an effect due to the pres-ence of the big BODIPY fluorophore, which perturbed the packing, considering the influenceof different terminal functional groups (e.g., the carbohydrate group in [211]) on the desorption1095.4. Results and discussion-5-4-3-2-1012-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2I (µA cm-2)E (V / Ag|AgCl)1 h HS-Aib4-BODIPY (60 µM)Forward scanReverse scanFigure 5.2: Cyclic voltammogram recorded during a reductive desorption measurement of ahigh-packing-density AuS-Aib4-BODIPY SAM deposited on a polished Au(111) electrode.potential. The packing density of this AuS-Aib4-BODIPY SAM can be estimated by integratingthe desorption current peak (the gray area in Figure 5.2 enclosed with the current peak and thelinear capacitive current baseline arbitrarily drawn for a rough estimation), giving ~1.4× 1014molecules cm–2 (0.71 nm2/molecule) assuming a one-electron reduction process. Despite thelower solution concentration, shorter deposition time and error associated with the calculation,this packing density is roughly comparable with the ∼3× 1014 molecules·cm–2 calculated inthe same way for the HS-Aib4 SAM. The approximate packing densities for the labeled andunlabeled SAMs suggest the even the AuS-Aib4-BODIPY SAM formed in this condition mightbe close to its maximum surface coverage.5.4.2 High-packing-density AuS-Aib4-BODIPY SAM deposited on anunpolished single crystal Au bead electrodeIn Chapter 4, the use of single crystal Au bead electrode with the full stereographic triangle hasbeen discussed. One important characteristic of the single crystal Au bead electrode is thatbig (111) facets can be observed on the surface and with a proper melting treatment and align-ment, one (111) facet can appear at the bottom of the electrode. This can be useful becausewith the imaging method used, the SAMs deposited on the (111) facet, the most stable and in-tensely studied surface, can be investigated spectroelectrochemically. There are advantagesof employing a facet on a single crystal bead electrode over a polished (111) electrode. First,a facet formed by proper annealing is close to atomically flat [201–203]. Additionally, it is rela-1105.4. Results and discussiontively easy to fabricate and clean the single crystal bead electrode compared to the laboriouspolishing needed to obtain a high quality planar electrode. Therefore, in this section, the AuS-Aib4-BODIPY SAM is prepared on an unpolished single crystal Au bead electrode and one ofthe (111) facets on the bead electrode is investigated. The high-packing-density AuS-Aib4-BODIPY SAM similar to those on the polished surface discussed above is first investigated todemonstrate the feasibility of this approach.The bright field image of the Au(111) facet is shown in Figure 5.3a. The facet is only occu-pying about half of the image and surrounding the facet are step-like features which are narrow(111) terraces. Montage of selected fluorescence images taken at potentials from –0.925 V to–1.3 V (vs. Ag|AgCl) shown in Figure 5.3b verifies the prediction above. At potentials positiveof –1.0 V (the fourth image of the first row), the fluorescence is from the facet. Stepping tomore negative potentials results in a fast increase of the intensity from the steps around thefacet which surpasses that from the facet at –1.175 V (the third image of the third row). It isclear the heterogeneity in intensity observed in the images is mainly due to the different surfacecharacteristics of the facet and the steps. Although the minimum projection of the image stack(Figure 5.3c) shows some bright spots indicative of non-uniformity of the SAM in adsorptionstate. However, the sources of these bright spots are unknown and they can either leave thesurface at negative potentials or stay after the complete potential stepping. The maximum pro-jection of the image stack (Figure 5.3d) demonstrates the two distinctive domains due to twodifferent types of surfaces. More importantly, the packing density within the (111) facet appearsto be quite uniform.The intensity as a function of potential is plotted for a ROI on the facet in Figure 5.4a. Fromthe starting potential 0 V to around –0.7 V, the intensity remains similar to the background level.This suggests that at this potential range, the layer remains adsorbed and the fluorophore isnear the substrate, so fluorescence is mostly quenched. The onset of a large intensity in-crease occurs at around –0.7 V. The intensity reaches a plateau at around –0.9 V but furtherincreases at around –1.1 V until another plateau at around –1.2 V. The two-step desorptionfeature has been observed for low-packing-density AuS-Aib4-BODIPY SAM from (111) sur-face and surfaces with large (111) terrace in Chapter 4. A sustained increase in capacitancestarts at ~–0.6 V and continues at more negative potentials (Figure 5.4b), suggesting the re-1115.4. Results and discussionFigure 5.3: (a) Bright field image of a Au(111) facet on a single crystal bead electrode; (b)montage of selected fluorescence images from a high-packing-density AuS-Aib4-BODIPY SAMdeposited on a single crystal bead electrode with a Au(111) facet in view, representing theoverall fluorescence response of the layer from –0.925 V to –1.3 V (vs. Ag|AgCl); (c) theminimum projection image of the image stack; (d) the maximum projection image of the imagestack.1125.4. Results and discussionduction of the Au-S bond and the uninterrupted displacement of the desorbed SAM with H2Oand hydrated electrolyte ions. However, the two-plateau profile of the intensity change dur-ing desorption suggests that the Aib peptide molecules do not diffuse away immediately afterthe reduction of the Au-S bond, but stay close to the surface instead. It was reported thatdesorbed Aib peptide thiolate molecules bound by intermolecular hydrogen bonds stay closeto the Hg substrate electrode surface at potential slightly more negative than the desorptionpotential [222]. The intermolecular hydrogen bond network observed in Aib peptide thiolateSAMs deposited on Au(111) surfaces [211], hypothetically would also be able to the stabilizethe desorbed SAM and hold the adsorbate molecules together, explaining the first plateau inthe intensity-potential curve. At more negative potential, this intermolecular hydrogen bondnetwork would presumably be destabilized by a large amount of H2O, hydrated electrolyte ionsand possibly the H2 evolution, leading to fast diffusion of Aib peptide molecules and thus drasticincrease of fluorescence intensity to form the second plateau. The two-step desorption appliesto both the high-packing-density SAM investigated in this section and the low-packing-densitySAM in Chapter 4, whichmay imply a relatively long active length of the intermolecular hydrogenbonds or tilted orientation of the peptide chains.The start of reductive desorption according to the capacitance-potential plot (Figure 5.3b) iscomparable to that obtained from a cyclic voltammetry measurement for a AuS-Aib4-BODIPYSAM on a polished Au(111) electrode (Figure 5.2). However, the capacitance is representingthe average property of the whole surface in the electrolyte solution. An unpolished singlecrystal bead electrode is featured with a variety of crystallographic surfaces including the (111)facet in view, so strictly speaking the capacitance-potential plot shown in Figure 5.4b and thecyclic voltammogram shown in Figure 5.2 inform on the reductive desorption processes fromdifferent types of surfaces.The area of the planar polished Au(111) electrode can be easilydetermined and used to normalize the capacitance. For an unpolished single crystal beadelectrode, due to the complexity of the surface, it is difficult to accurately measure the area of theelectrode. Nevertheless, for a rough estimation, at ~–1.0 V (vs. Ag|AgCl), the electrode surfaceis mostly covered with H2Omolecules and hydrated electrolyte ions, and the H2 evolution is notsignificant in such a high pH solution, so the unit area capacitance in the absence of specificadsorption for a number of crystallographic surfaces (~20 μF cm–2 [172, 173]) can be used to1135.4. Results and discussion01020304050Fl. Int. (kcts/sec)2 h HS-Aib4-BODIPY (60 µM)aStepping direction(111) Facet0481216202428-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0C avg (µF / cm2 )E (V / Ag|AgCl)bStepping directionFigure 5.4: (a) In situ fluorescence intensity as a function of potential from a Au(111) facet forthe high-packing-density AuS-Aib4-BODIPY SAM deposited on an unpolished single crystalbead electrode; (b) capacitance change per unit area as a function of potential for the AuS-Aib4-BODIPY SAM modified electrode.calculate the electrode area for the following normalization.5.4.3 Low-packing-density AuS-Aib4-BODIPY SAMs deposited on unpolishedsingle crystal Au bead electrodesIn the previous section, a high-packing-density AuS-Aib4-BODIPY SAM on a Au(111) facetwas investigated with in situ fluorescence imaging technique. Similar approach will be usedin this section to investigate the potential modulated response of the low-packing-density AuS-Aib4-BODIPY SAMs. The fluorescence intensity is dependent on the distance between thefluorophore and the substrate, so besides desorption, any orientational or structural changeof the peptide layer may result in a change of the fluorescence intensity. The orientational orstructural change was not observed on the high-packing-density SAMs presumably because ofa lack of space between adsorbate molecules. Here, modulated potential steps were appliedto the modified electrode to induce a change in the fluorophore-substrate separation. The1145.4. Results and discussiona bFigure 5.5: Fluorescence images of a low-packing-density AuS-Aib4-BODIPY SAM taken at(a) 0 V (vs. Ag|AgCl) and (b) –0.55 V (vs. Ag|AgCl) with the (111) facet outlined.potential profile applied is shown in Figure 5.6a, and note that the potentials applied werelimited so as to avoid the occurrence of reductive desorption. At each step, images were takento monitor the fluorescence intensity change.A typical low-packing-density AuS-Aib4-BODIPY SAM was created by immersing the singlecrystal bead electrode in 1 μM adsorbate solution for 0.5 h. Figure 5.5 shows the fluorescenceimages of this low-packing-density AuS-Aib4-BODIPY SAM taken at 0 V (vs. Ag|AgCl) and–0.55 V (vs. Ag|AgCl) with the perimeter of (111) facet drawn. Overall, higher intensity isobserved at 0 V than at –0.55 V. The fluorescence intensity in the center of the facet is mostlyuniform, so the average intensity of the outlined ROI is used to avoid further complication.The AuS-Aib4-BODIPY layer exhibits a fluorescence intensity which is modulated by the po-tential steps (Figure 5.6b & c). A decrease in intensity is observed at a negative step potentialwhile the intensity reverts to the original level when stepping back to the base potential. Notethat a slow and gradual change of intensity is observed from the five images taken sequentiallyat each potential step. This trend continues until about –0.6 V where the fluorescence intensityat the step potentials starts to increase significantly, eventually surpassing that at the base po-tential. The percent change in fluorescence can be calculated by following the formula below foreach step potential: (FEstep− FEbse)/FEbse, where FEstep and FEbse are the fluorescenceintensities extracted from the last images taken at a step potential and the base potential imme-1155.4. Results and discussion-0.7-0.6-0.5-0.4-0.3-0.2-0.10.0E (V / Ag|AgCl)0.5 h HS-Aib4-BODIPY (1 µM)a0. Int. (kcts/sec)bFl. Int. (kcts/sec)EstepEbase0. 400 800 1200 1600 2000 2400Fl. Int. (kcts/sec)t (sec)cFl. Int. (kcts/sec)EstepEbaseFigure 5.6: (a) Profile of the modulated potential steps; (b) the fluorescence response of alow-packing-density AuS-Aib4-BODIPY SAM; (c) ordinate enlarged fluorescence response.Adapted from [169] with permission. Copyright (2017) Springer.-0.6- -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0(FEstep-F Ebase)/FEbaseE (V / Ag|AgCl)0.5 h HS-Aib4-BODIPY (1 µM)Stepping directionFigure 5.7: Percent fluorescence intensity change ((FEstep − FEbse)/FEbse) as a functionof step potential of a low-packing-density AuS-Aib4-BODIPY SAM. Adapted from [169] withpermission. Copyright (2017) Springer.1165.4. Results and discussiondiately after the step potential. The fluorescence intensity as a function of the step potential andthe percent change as a function of the step potential are presented in Figure 5.7. The nega-tive percent fluorescence change from 0 V to –0.6 V suggests that at this potential range, thefluorescence intensity decreases when stepping to a negative potential. The percent changeslowly becomes more negative from almost 0% at 0 V to about –50% at –0.55 V whereafter itinflects to become less negative and eventually changes sign at more negative potentials. Thefluorescence intensity at the base potential remains constant before the inflection point exceptfor a slight decrease at the beginning which might be due to photo bleaching. Interestingly, thebase potential intensity decreases dramatically at the inflection point potential and stays at thislow level..From the capacitance change shown in Figure 5.4b, it is known that at –0.6 V (vs. Ag|AgCl),reductive desorption starts. However, for the reductive desorptionmeasurement, with low imag-ing sensitivity, the small fluorescence change at this potential cannot be detected. Here withimaging settings applied to achieve high sensitivity, the fluorescence signal is magnified byabout 10 times, so a pronounced increase of fluorescence intensity can be observed at steppotentials more negative than –0.6 V. It is reasonable to attribute this intensity increase to re-ductive desorption which leads to the increasing distance between the fluorophore and thesubstrate. From the percent fluorescence change as a function of the step potential shownin Figure 5.6c, an inflection point at –0.55 V can be observed which divides the fluorescenceresponse into two regions. This suggests that the actual start of the reductive desorption mightbe at about –0.55 V (vs. Ag|AgCl). Therefore, at potentials less negative than –0.55 V wherethe Au-S bond stays intact, there is another orientational or structural change which resultsin a negative percent fluorescence change. The drastic change of base potential intensity atthis inflection potential might be rationalized by the sudden change of SAM organization upondesorption.Experimentally, the fluorescence modulation for a Aib peptide thiolate SAM was highly re-producible within a limited potential range less negative than the onset of the desorption, orthe inflection point potential. Once the SAM experienced a potential negative enough to initiatedesorption, the potential modulated behavior was no longer observed. This suggests the Aibpeptide SAM responds to the potential modulation without the reduction of the Au-S bond.1175.4. Results and discussionTwo possible mechanisms can be proposed to explain the potential induced response (Fig-ure 5.8). First of all, the rigid 310-helix structure of the Aib peptide induces a dipole momentoriented along the backbone [210, 211, 218, 220]. This dipole moment could react to an elec-tric field. To be specific, in the case of AuS-Aib4-BODIPY, the positive pole is at the surface,so presumably the positive pole would respond more dominantly than the negative pole. Asa negative potential is applied to the surface, the peptide backbone is more inclined to be at-tracted because of the dipole moment, thus make the peptide molecule more tilted. As a result,the distance between the fluophore and the surface becomes smaller, which results in a de-creased fluorescence intensity. This potential induced response of peptide is analogous to theDNA reorientation reported by Rant et al. [7, 226–228]. However, DNA molecules have neg-atively charged backbones which strongly respond to electrode potential. The dipole momentalong the peptide backbone is the sum of a number of polar carbonyl bonds and may not re-spond to pottential as promptly. A second proposed mechanism originated from the “peptidespring” reported by Peggion et al. [229]. They showed that the secondary structure of a pep-tide molecule is dependent on the solvent. The 310-helix structure conformation is retained ina polar solvent while it converts to the fully extended conformation in a non-polar solvent. Thissolvent-mediated length change can be as big as 50%. It is hypothesized that the electrical po-tential applied to the substrate could change the polarity of the environment near the surface,which in turn changes the length of the peptide molecules, resulting in a modulated fluores-cence profile. The influence of the electric field on the Aib peptide thiolate SAMs deposited onHg substrate electrode has been investigated [222]. It was found that applying a potential wasable to dramatically change the structure of the Aib peptide thiolate SAM, reversing the dipolemoment which originally oriented against the electric field. Admittedly the mobile Hg surface issubstantially different from the Au surface, but it is evident that the dipole moment associatedwith the Aib peptide thiolate SAMs can be impacted by an electric field. However, further in-depth surface characterization techniques are needed to prove whether or not the modulatedfluorescence observed here is the consequence of either mechanism.AuS-Aib4-BODIPY SAMs with higher packing density were also investigated with the poten-tial stepping profiled in Figure 5.6a & b. The packing density was enhanced by either extendingthe deposition time or increasing the deposition solution concentration. As shown in Figure1185.4. Results and discussionδ+δ−SONBNF FSOδ−δ++ + + + + +_ _ _ _ _ _NBNF FSONBNFFδ+δ−_ _ _ _ _ _ _ _ _ _ _ _Figure 5.8: Schematic of the hypothesized structure or orientation changes in the Aib peptidethiolate responding to changes in the electrode potential. Adapted from [169] with permission.Copyright (2017) Springer.5.9a, fluorescence modulation is strongly dependent on the SAM packing density. Extendingthe immersion time to 2 h significantly reduces the percent fluorescence change per potentialstep, while further extending the immersion time to 18 h does not substantially change thispotential modulated response. Doubling the concentration of the deposition solution to 2 μMcompletely eliminates the fluorescence modulation at potentials more positive than the start ofdesorption, and the positive percent change is presumably due to diffusing away of the looselybound adsorbate molecules. The SAM packing densities are confirmed by the fluorescenceintensity during the reductive desorption measurements as shown in Figure 5.9b. As expected,the SAM of lowest packing density, created by immersion in 1 μM deposition solution for 0.5 h,shows the largest percent fluorescence intensity change. The two SAMs with similar percentchanges (immersion in 1 μM deposition solution for 2 h and 1 μM deposition solution for 18h) have similar packing densities, indicative of the highest packing density obtained with thisdeposition solution concentration. The packing density is significantly higher for the SAM cre-ated by immersion in 2 μM deposition solution for 0.5 h, which sets the upper packing densitylimit for AuS-Aib4-BODIPY SAMs with measurable potential modulated response. As a com-parison, a AuS-C10-BODIPY SAM of similar packing density (created with 0.5 h depositionin 0.1 μM HS-C10-BODIPY solution) has been tested in the same way also plotted in Figure5.9. The AuS-C10-BODIPY SAM exhibits no fluorescence intensity modulation before the startof desorption even though it has a coverage similar to the AuS-Aib4-BODIPY SAMs show-ing significant intensity modulation. This confirms that the potential modulated fluorescence1195.4. Results and discussion-0.6- -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0(FEstep-F Ebase)/FEbaseE (V / Ag|AgCl)aStepping direction0.5 h HS-Aib4-BODIPY (1 µM)2 h HS-Aib4-BODIPY (1 µM)18 h HS-Aib4-BODIPY (1 µM)0.5 h HS-Aib4-BODIPY (2 µM)0.5 h HS-C10-BODIPY (0.1 µM)02468-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0Fl. Int. (kcts/sec)E (V / Ag|AgCl)bStepping direction0.5 h HS-Aib4-BODIPY (1 µM)2 h HS-Aib4-BODIPY (1 µM)18 h HS-Aib4-BODIPY (1 µM)0.5 h HS-Aib4-BODIPY (2 µM)0.5 h HS-C10-BODIPY (0.1 µM)Figure 5.9: (a) Percent fluorescence intensity change ((FEstep − FEbse)/FEbse) as a func-tion of step potential and (b) fluorescence intensity as a function of potential recorded duringreductive desorption, of the AuS-Aib4-BODIPY SAMs of various packing density and a AuS-C10-BODIPY SAM. Adapted from [169] with permission. Copyright (2017) a characteristic of the Aib peptide SAM. The dependence of potential modulated responseon SAM packing density may infer that the orientation change of the Aib peptide molecule,which requires more space, predominantly accounts for the potential mediated fluorescencechange. Based on the fluorescence intensity during reductive desorption, the packing densityof the AuS-Aib4-BODIPY SAMs with potential modulated response can be roughly estimatedby comparing with the high-packing-density SAM (Figure 5.4a). Generally speaking, the po-tential modulated response requires a packing density that is roughly 10% of that for the high-packing-density SAM. Assuming the high-packing-density SAM on a (111) facet adopts similarpacking density to that on a polished (111) surface, 1×1013 - 2×1013 molecules·cm–2 is theestimated packing density required for the potential modulated response.The potential modulated response is thus a property of low-packing-density Aib peptidethiolate SAMs deposited on Au(111) surface. Typically, high-packing-density Aib peptide SAMs1205.5. Conclusionswith well-ordered intermolecular interactions are used in various applications [211, 212, 222],where the property of potential modulated response no longer exists. However, investigation ofthis property might still be valuable in e.g., the Aib peptide covered Au nano particles, where thesubstrate surfaces are in the same scale of the adsorbate molecule footprint and the adsorbatemolecules without many neighbors might exhibit the potential modulated response.5.5 ConclusionsIn this chapter, Aib peptide thiolate SAMs deposited on Au(111) surfaces have been stud-ied. The high-packing-density AuS-Aib4-BODIPY SAM deposited on a polished planar Au(111)electrode was first characterized electrochemically. Comparable desorption potential and pack-ing density to literature were obtained. In situ fluorescence imaging was employed to investi-gate the high-packing-density AuS-Aib4-BODIPY SAM deposited on an unpolished single crys-tal Au bead electrode. A two-step desorption was observed from a (111) facet for this high-packing-density SAM. With the (111) facet as the substrate, the potential modulated responseof low pack density AuS-Aib4-BODIPY SAMs was explored at potentials less negative than thedesorption potential spectroelectrochemically. The fluorescence intensity from the AuS-Aib4-BODIPY SAM decreased as the potential was stepped negatively, as a result of the change inorientation or structure of the Aib peptide chain in response to an electric field. This potentialmodulated response is a property of low-packing-density Aib peptide thiolate SAMs depositedon a Au(111) surface and thus increasing the packing density resulted in a decreased or evenzero modulated fluorescence.121Chapter 6Spectroelectrochemical investigationof the potential-driven DNAreorientation on a single crystal Aubead electrodeIn the two previous chapters, alkanethiolate SAMs and thiolated Aib peptide SAMs were dis-cussed. These two types of SAMs are mainly used as the molecular spacer and/or diluentto control the position and density of the bio-receptors in a biosensing interface. The DNASAMs which will be explored in this chapter, are directly related to biosensors capable of cap-turing specific targets (e.g., nucleic acids, proteins). The negatively-charged DNA strand canrespond to the electric potential by changing its orientation, which is an important and useful ef-fect in DNA sensing requiring in-depth investigation. A single-crystal Au bead electrode is alsoused in this chapter as the substrate to study the influence of surface crystallography on thepotential-driven DNA reorientation. Besides the in situ fluorescence imaging technique, a newspectroelectrochemical technique that couples electrochemistry, fluorescence microscopy andharmonic analysis is developed to explore harmonics of the non-linear fluorescence responseto an applied AC potential perturbation. Employing this technique aims to achieve a betterunderstanding of the DNA reorientation effect and develop a new sensing mechanism.1226.1. The chemical and biophysical properties of DNACOCC CBaseHHHHOHCH2HO1'2'3'4'5'P OOOFigure 6.1: Structure of a 5’-nucleotide.6.1 The chemical and biophysical properties of DNADNA (short for deoxyribonucleic acid) is a biological molecule encoding genetic information.The famous double-helix structure proposed byWatson and Crick [230] is one form (the B form)of double-stranded DNA (dsDNA) structure [231, 232]. In this chapter, DNA is an adsorbatesystem studied without considering its complicated behaviors in a physiological environment,so only the general chemical and biophysical properties of DNA are reviewed here.6.1.1 Structure of DNAThe structural motif of a nucleic acid chain can be expressed as [231]:For DNA, the sugar is a deoxyribose. In the sugar ring, the C1’ is connected to a baseand the C3’ and the C5’ are both connected to phosphate groups to form the polymeric single-stranded DNA (ssDNA). The monomeric unit, i.e., the nucleotide, is shown in Figure 6.1 (shownis the 5’-nucleotide). The five atoms in the sugar ring are not co-planar and usually one atomis out of the plane. The preferred conformations are C2’-endo and C3’-endo (endo refers topointing up, similar to the C5’) [231]. The conformation of the sugar ring has an impact on thedsDNA conformation.In Figure 6.1, the phosphate group has two “free” deprotonated hydroxyl groups. Indeed,in a ssDNA chain, one hydroxyl group is connected to the C3’ of another necleotide through anether bond, whereas the other hydroxyl is not bonded to any other group. Therefore, because1236.1. The chemical and biophysical properties of DNATable 6.1: Comparison of the three dsDNA conformations: B form, A form and Z form. Sum-marized from [231] and [232].Property B form A form Z formHelix chirality right-handed right-handed left-handedSugar conformation C2’-endo C3’-endo C: C2’-endo; G: C3’-endoRepeating unit 1 base pair 1 base pair 2 base pairsBase pair per turn 10.5 11 12Diameter 20 Å 23 Å 18 Å [233]Axial distance between base pairs 3.4 Å 2.6 Å 3.7 ÅPitch 35.7 Å 28.6 Å 44.6 ÅTorsion angle between base pairs 34.3° 32.7° CpG: -9°, GpC: -51°Axial tilt angle ~0° ~20° -7°of the low pK a(~1), at neutral pH, each necleotide is carrying one negative charge except forthe terminal ones, making DNA a polyelectrolyte [231, 232].In DNA, the base can be any of the four types: adenine (A), guanine (G), thymine (T) andcytosine (C). These bases are essential for the dsDNA helical structure where A and T formbase pair through two hydrogen bonds and C and G form base pair through three hydrogenbonds (Figure 6.2). The double-helix structure can be formed when these two requirementsare fulfilled: one helical chain is from 5’→3’ and the other is from 3’→5’; the bases from thefirst chain find their corresponding complementary bases from the second chain. Two fullycomplementary segments of DNA strands can further adopt different types of three-dimensionalstructures depending on the environment and the strand sequence. The most common formin vivo, proposed by Watson and Crick, is later termed B form [231] (Figure 6.3). B form DNAcan convert into A form DNA in solutions with low H2O activity [232]. Both B form and A formare right-handed, but a rare Z form is left handed. Z form mainly occurs in alternating poly-(dG-dC) and is adopted in the presence of high electrolyte concentration [233, 234]. The structuralproperties of the three forms are listed in Table 6.1. Note that all properties listed in Table 6.1are actually statistical properties. Although there are energetically preferred conformations foreach base pair, because of energy fluctuation or perturbation, it is not possible to keep eachbase pair in the preferred conformation. Therefore, the experimentally determined values arethe statistical averages.There exist other conformations, such as B’ form, triple helix, G-quadruplex, etc., but theseconformations are rare and require special conditions [232]. Worth mentioning is the hairpin1246.1. The chemical and biophysical properties of DNAONNNNNH OOOPO OANHNONOOOP OOCNOONOOOPOOTNNNONHNOOOP OOGONHNONOOOPO OCNNNNHNOOOP OOANNNONHNOOOPOOGNOONOOOP OOTHHHHHHHHHHFigure 6.2: Chemical structure of a dsDNA segment showing the complementary base pairs.Figure 6.3: Three-dimensional structure of a dsDNA segment with B form configuration. A, G,T, and C are colored in red, green, yellow and blue, respectively.1256.1. The chemical and biophysical properties of DNAstructure formed within one ssDNA chain with self-complementary segments. This structurehas often been exploited in bio-sensing applications [52, 53, 235, 236] in that opening up thehairpin structure with the full complementary strand dramatically changes the apparent lengthof the DNA chain.6.1.2 Physical propertiesTwo important physical properties are reviewed here: the persistence length and the effectivediameter. These are important physical properties especially in an electrolyte solution, whichis the environment for electrochemical characterization.The persistence length characterizes the flexibility of a polymer. It can be understood asthe maximum length for a polymer chain to behave as a rigid rod [231]. Considering the con-formations, dsDNA, with the double-helix structure, is significantly less flexible than ssDNA, sodsDNA has much higher persistence length than ssDNA. The characteristic persistence lengthfor dsDNA is ~50 nm (~147 base pairs for B form DNA) [237] while the characteristic persis-tence length for ssDNA is < 5 nm (~11 bases) [238]. Moreover, since the DNA backbone isnegatively charged, due to charge repulsion, a longer persistence length is expected with anincrease in the amount of unshielded charge carried by the DNA chain. Cations, especiallymultivalent cations, in a electrolyte solution counterbalance the DNA backbone charge andthus DNA has a low persistence length in a solution with high ionic strength[237–239]. Thisionic strength dependence is not significant for dsDNA in an solution with [Na+] of over 1 mMor [Mg2+] of over 0.2 mM, and the ~50 nm persistent length appears to be a minimum undercommonly used non-extreme conditions [237].The effective diameter of DNA characterizes the electrostatic interaction between neigh-boring DNA chains in an electrolyte solution. Charge repulsion due to the negative charge onthe DNA backbone, in addition to the steric hindrance, prevents a DNA chain from approach-ing the neighboring DNA chains. As expected, at high ionic strength, the backbone charge ismostly shielded, resulting in a small effective diameter and vice versa [240, 241]. The effectivediameter of dsDNA as a function of ionic concentration is demonstrated in Figure 6.4. For DNAtethered to a surface, effective diameter is further influenced by the bending of DNA strand. A1266.1. The chemical and biophysical properties of DNA061218240.001 0.01 0.1 1Effective diameter (nm)[Na+] (M)Without Mg2+ - experimentalWithout Mg2+ - calculated1 mM Mg2+5 mM Mg2+20 mM Mg2+100 mM Mg2+Figure 6.4: Effective diameter of dsDNA as a function of ionic concentration. Reproduced from[241] with permission from Oxford University Press.DNA strand longer than its persistence length has a larger effective diameter due to bendingthan a short DNA strand behaving as a rigid rod [242].6.1.3 DNA hybridization and dehybridizationThe process of hybridization between two complementary DNA single-strands is reversible.Traditionally, for experimental convenience, the DNAmelting process, i.e., thermal dehybridiza-tion of a dsDNA duplex is often discussed. From the thermodynamic point of view, considering4G◦ =4H◦−T4S◦, for the DNA melting reaction,4H◦ is positive and4S◦ is positive as well[243, 244]. Therefore, dsDNA is dehybridized into ssDNA at high temperature and the temper-ature at which half of the DNA strands are in ssDNA state is defined as the melting temperatureTm. From the microscopic aspect, the entropic change during DNA melting can be describedby the facts that one double strand duplex is separated into two single strands and that the sep-arated DNA single strands are structurally much more flexible than the double strand duplex.On the other hand, three factors are contributing to the dsDNA stability: the hydrogen bond andpi-stacking interactions formed between complementary bases which stabilize the dsDNA du-plex and the repulsion between negatively charged backbones which destabilizes the dsDNAduplex. These three factors are all closely related to the enthalpic change but the entropicchange is mostly independent of them [232]. As a consequence, high Tm is expected for long1276.1. The chemical and biophysical properties of DNADNA strands because of a large number of hydrogen bond and pi-stacking interactions. DNAstrands with a high fraction of GC base pairs is also expected to have high Tm. Additionally,high concentration of electrolyte ions which shield the DNA backbone charges results in highdsDNA Tm [232, 243, 244]. On the contrary, spontaneous dehybridization of dsDNA can occurin a solution with low ionic strength. Experimentally, the dsDNA melting process is usually de-termined by measuring the UV absorbance (typically at 260 nm) as the temperature increases.An increase of UV absorbance upon dehybridizing the dsDNA duplex (known as hyperchromic-ity) can be observed due to the breakdown of pi-stacked aromatic rings in the dsDNA duplex[245, 246].Dehybridization of dsDNA can be achieved not only by thermal treatment, but also by chem-ical treatment. Extreme acidic or basic conditions lead to dehybridization (often termed “de-naturation”) accompanied by bond break of the DNA chemical structure [245]. Urea treatment,which breaks the hydrogen bonds between complementary bases, exerts a more gentle de-naturing mechanism, effective in low ionic strength solutions [247]. Consequently, urea is thecommonly used denaturant for DNA polyacrylamide gel electrophoresis (PAGE) [248, 249].The review given above is based on hybridization in a solution phase. Hybridization on asolution-solid interface is also an important topic because this process is related to biosensing.Thermodynamically, surface hybridization is not as favorable as solution hybridization. Shield-ing backbone charges is not as effective on a surface because steric hindrance prevents ionsfrom moving into the layer of tethered DNA, so higher ionic strength is required for surface hy-bridization [244]. One key distinction from solution hybridization is that the density of ssDNAimmobilized on a surface (i.e., surface probe) plays a crucial role in the rate and efficiencyof the hybridization. Not surprisingly, higher surface probe density results in a lower rate andefficiency of hybridization. Typically, in order to accomplish 100% hybridization, a low surfaceprobe density of no more than 1012 molecules/cm2 is required [250] which is about 1% ofthe maximum density of thiolated DNA strand (depending on the length) immobilized on a Ausubstrate [110].1286.2. Potential-driven DNA reorientationFigure 6.5: The potential-dependent thickness of the DNA SAM immobilized on a Au(111)surface, determined by EC-AFM. The dashed line marks the DNA SAM thickness at open-circuit potential. Reprinted from [251] with permission from American Chemical Society.6.2 Potential-driven DNA reorientationThe negatively charged backbone of DNA is an interesting and useful feature particularly in DNASAMs because the orientation of the adsorbed DNA molecules can be controlled by electricpotential at the surface. To be specific, at a positive potential, the DNA backbone is attracted tothe surface, resulting in a “lying-down” configuration. At a negative potential, the DNA backboneis repelled by the surface, resulting in a “standing-up” configuration.The potential-driven DNA reorientation effect was first studied with electrochemical atomicforce microscopy (EC-AFM) by Kelly et al. [251]. It was observed that the thickness of the DNASAM immobilized on a Au(111) surface was potential-dependent, as shown in Figure 6.5.Further if the DNA adsorbate is labeled with a fluorophore, the orientation of a DNAmoleculeand thus the distance between the fluorophore and the surface is controlled by the potential.Due to the distance-dependent fluorescence quenching near a metal surface (reviewed in Sec-tion 2.4.3), the potential-driven DNA reorientation is observed as a potential-controlled fluores-cence response represented in Figure 6.6. Rant and co-workers have been investigating thepotential-driven DNA reorientation effect using a fluorophore labeled DNA, detailing the mech-anisms, kinetics and influencing factors for DNA reorientation [54, 226–228] and applying this1296.2. Potential-driven DNA reorientationFigure 6.6: The potential-controlled fluorescence response from a fluorophore labelled dsDNAimmobilized on a Au surface. Reprinted from [227] with permission from Elsevier.valuable reorientation effect in biosensing [7, 54, 55, 57]. A few important points are highlightedhere. First, the ssDNA and dsDNA have different backbone rigidity, resulting in distinctive reori-entation mechanisms. The dsDNA undergoes rigid-body rotation around the anchoring pointdue to the stiff molecular conformation. On the contrary, the motion of the flexible ssDNA oc-curs segment by segment. This movement is more prominent when the ssDNA is attracted tothe surface by a positive potential, reflected as a short transition time [227]. Moreover, when analternating potential is applied to drive the reorientation, as expected, above certain thresholdfrequency, the DNA SAM cannot completely respond to the alternating potential and thereforewill not reorient. It was shown that dsDNA can respond to AC potentials of higher frequencythan ssDNA of the same length and a short strand can reorient at higher frequencies than a longone [54]. Furthermore, the electrolyte concentration has a profound and critical influence onthe DNA reorientation [228]. Theoretically, the electric field across the interface, which drivesthe DNA reorientation, can only extend to the distance of Debye length (see Section 2.3.3).The Debye length is strongly dependent on the electrolyte concentration. At high electrolyteconcentrations (~1 M), the Debye length is only 1-2 nm. Thus, considering the the length ofthe alkyl linker (~1.3 Å per CH2 [105]), the electric field is only extended to four base pairs nearthe alkyl linker, so for a dsDNA with over 20 base pairs (assuming in B form, see Table 6.1),the motion driven by this electric field is small. On the other hand, for ssDNA, the persistencelength is on the same order as the Debye length, so the strand adopts a compacted state withlittle reorientation behavior. In an intermediate electrolyte concentration (~10-100 mM), both1306.3. Harmonic analysis of nonlinear fluorescence response driven by AC potential perturbationthe dsDNA and ssDNA exhibit the reorientation response to potential. As discussed in Section6.1.3, the dsDNA duplex is stable at a sufficiently high electrolyte concentration. Therefore, atlow electrolyte concentrations (~0.1-1 mM), the dsDNA duplex is not formed, but interestingly,the ssDNA strand may adopt an extended conformation (long persistence length) due to a lackof shielding of backbone charge and take up the rigid-body rotation mechanism. Besides theextent of reorientation, the potential at which the DNA changes its orientation between lyingdown and standing up (termed as the potential of conformation transition, or PCT in [228]) isalso dependent on the electrolyte concentration. A decrease of electrolyte concentration re-sults in a negative shift in PCT because of an increase in the amount of unshielded negativecharge on the DNA backbone. Lastly, Rant et al. exploited the potential-driven DNA reorienta-tion to create biosensing platforms, mainly based on three motifs: a change in the magnitudeof the reorientation response, a shift in the threshold reorientation frequency or a change in thereorientation transition time upon capturing the target (the complementary strand or biologicalanalytes)[7, 54, 55, 57]. Note that for the approach based on the frequency response, it isimportant to distinguish the charging process from the reorientation response because at highfrequencies (> 1 kHz), charging the electrode kinetically limits the DNA reorientation. The ac-tual reorientation rate can be determined by deconvoluting the electrochemical charging timefrom the measured frequency response [252, 253].6.3 Harmonic analysis of nonlinear fluorescence responsedriven by AC potential perturbationThe potential-driven reorientation of a DNA SAM monitored with fluorescence is intriguing inthat as shown in Figure 6.6, the shape of the fluorescence response closely resembles the driv-ing potential. Thus, a sinusoidal potential perturbation results in sine-wave like fluorescenceresponse. Therefore, the amplitude and frequency response of the fluorescence were often in-vestigated [7, 54, 55]. However, strictly speaking, the fluorescence signal does not necessarilyhave the sinusoidal waveform simply because the fluorescence intensity can be non-linearlyrelated to the applied potential.1316.3. Harmonic analysis of nonlinear fluorescence response driven by AC potential perturbationThe alternating potential applied can be described as a complex number:E= E0exp(i(ωt+ φ)) (6.1)The fluorescence intensity as a function of potential is generally expressed as:F= ƒ (E) (6.2)If the fluorescence intensity is linearly related to potential, thenF= E+C= E0exp(i(ωt+ φ))+C (6.3)In this case, the fluorescence response has the same form as the AC potential perturbation.If the fluorescence intensity is quadratically dependent on potential:F= 1E+ 2E2+C= 1E0exp(i(ωt+ φ))+ 2E20exp(i(2ωt+ 2φ))+C (6.4)Besides an alternating signal with an angular frequency of ω (fundamental frequency), an-other signal with an angular frequency of 2ω, i.e., the second harmonic, contributes to the ACfluorescence response. The second harmonic is the nonlinear component of the fluorescenceresponse. Equation 6.4 can be generalized for a polynomial function containing m terms:F=m∑n=1nEn+C=m∑n=1nEn0exp(i(nωt+ nφ)+C (6.5)Equation 6.5 is the sum of a series of signals and that with an angular frequency of nω isthe nth harmonic term. Therefore, if the fluorescence intensity-potential function F contains anEn term, this term is transformed into the nth harmonic in the AC fluorescence response. Inreality, F can be any non-linear function. A Taylor expansion performed around a DC potential(e.g., 0 V for simplification) gives a polynomial series (only the Maclaurin series is given here):F=∞∑n=0ƒ (n)(0)n!En =∞∑n=0ƒ (n)(0)n!En0exp(i(nωt+ nφ)) (6.6)1326.4. ObjectivesEquation 6.6 suggests that a general nonlinear fluorescence response can be separatedinto a series of harmonic terms. The amplitude of the harmonics indicates the extent of the non-linearity of the signal. Harmonic analysis is a method to extract the amplitude of the harmoniccomponents, which can be experimentally achieved using two techniques: Fourier Transformfrequency response analysis and phase sensitive detection using a lock-in amplifier. In thischapter, a lock-in amplifier is employed to analyze the harmonic components of the alternatingfluorescence response due to its high sensitivity and low noise as compared to the frequencyresponse analysis technique [254, 255].6.4 ObjectivesMCH-DNA SAMs formed by the thiol-exchange mechanism will be studied. The fluorescenceimage of a typical MCH-DNA SAM so prepared will be compared with the map of crystallo-graphic surfaces to study the influence of the surface crystal structure on DNA SAM formation.In situ fluorescence imaging will be performed to investigate the potential-driven DNA reorien-tation response as a function of the surface crystallography. Furthermore, a harmonic analysistechnique using a lock-in amplifier will be developed to characterize the non-linearity of thefluorescence response to an applied AC potential perturbation. This method will be employedto measure the higher harmonics of the fluorescence response for DNA SAMs of different hy-bridization states and from different crystallographic surfaces. A new DNA sensing mechanismwill be proposed based on the experimental results.6.5 Experimental6.5.1 Preparation of MCH-DNA SAMsCreating a MCH-DNA SAM by partial displacement (i.e., thiol-exchange) of a MCH SAM withthiolated DNA is used to achieve a low density of DNA spaced by MCH on a surface [111, 252,253, 256]. A variety of MCH-DNA SAMs were prepared for different purposes. In general, aclean single crystal Au bead electrode was first modified with 1 mM MCH solution for 1-2 h,followed by immersion in 0.5-1 μMHS-C6-DNA-AlexaFluor (see Section 3.1.1 for specification)1336.5. Experimentalsolution for 16-24 h. The solvent for MCH was either MeOH or a pH = 7.5 buffer solutioncomposed of 10 mM 2-amino-2-(hydroxymethyl)propane-1,3-diol (Tris) and 100 mM NaCl. Thesolvent for the thiolated DNA was a pH = 7.5 immobilization buffer solution composed of 10 mMTris, 100 mM NaCl and 50 mM or 500 mM MgCl2. The solvents used did not seem to notablyalter the properties of the SAMs, so they are not specified for each individual SAM. The DNAdensity was mainly controlled by adjusting the deposition time and the adsorbate concentration.Note that the DNA used is a 30-mer strand and the fluorophore at the distal end is positionedsufficiently far away from the Au surface so that fluorescence is not completely quenched. Thefluorophore was either AlexaFluor488 or AlexaFluor647 and each has distinctive advantages.AlexaFluor488 has a higher photo-stability. The intensity of AlexaFluor647 is more sensitive tofluorophore-metal separation in the regime of less than 10 nm (the length of a 30-mer dsDNA)away from the metal surface [169], favorable for studying the reorientation response. Overall,the SAMs prepared can be grouped into two categories, listed in Table 6.2.Table 6.2: The two categories of MCH-DNA SAMs prepared and their formation conditionsFluorophore MCH immersion DNA state DNA concentration DNA immersionAlexaFluor488 1 h ssDNA 1 μM 24 hAlexaFluor647 2 h ssDNA or dsDNA 0.5 μM 16 hBoth ssDNA SAMs and dsDNA SAMs were investigated. The hybridization was performedin solution before assembly on the surface. Hybridization in solution was performed by heatinga 1:2 mixture of the HS-C6-DNA-AlexaFluor and the corresponding unlabeled complementarystrand in a pH = 7.5 buffer solution of 10 mM Tris, 100 mM NaCl and 50mM MgCl2 to about 90°C and then slowly cooling it down to room temperature over 1 h. Dehybridization of dsDNASAM on the surface was achieved by soaking a MCH-dsDNA SAM-modified electrode into 8M urea solution for 2 min followed by rinsing with H2O.6.5.2 Spectroelectrochemical measurementsThe setup for spectroelectrochemical measurements was discussed in Section 3.4. All mea-surements were conducted in a pH = 7.5 buffer solution of 10 mM Tris and 10 mM KNO3. Anelectrode modified with a MCH-DNA SAM (except for one after urea treatment) had to be stored1346.5. Experimentalin the immobilization buffer solution for at least 12 h prior to further characterization.The DNA-AlexaFluor488 SAM was imaged with an Evolve® 512 EMCCD camera throughan Olympus LMPlanFl 5× objective at -0.4 V (vs. SCE) for comparison with the map of thecrystallographic surfaces. The potential-driven reorientation response was studied on the DNA-AlexaFluor647 SAMs either with a SBIG ST-7XMEI CCD camera or with a Newport 77348model PMT. In situ fluorescence imaging was performed with an Olympus UMPlanFl 10×(NA = 0.3) water immersion objective. The potential was changed by applying potential stepsfrom an initial/base potential to a final potential with stepping to the base potential (the secondscheme described in Figure 3.6a) to induce DNA reorientation and measure the fluorescenceresponse. The initial potential (or base potential) was 0.35 V (vs. SCE), the final potential was-0.40 V (vs. SCE) and the interval was -25 mV.For measurements with a PMT as the detector, the fluorescence signal was collected withan Olympus LUMPlanFl 40× (NA = 0.8) water immersion objective and a field diaphragm wasclosed down so that a small region (~96 μm in diameter) was illuminated. Either linear potentialscanning or potential stepping with the AC potential perturbation was applied to induce thereorientation response. The linear potential scanning was analogous to the electrochemicalcyclic voltammetry scanning between 0.35 V and -0.40 V, except that the fluorescence signalamplified by the PMT (and further converted to voltage with a current preamplifier) was alsorecorded. For the harmonic analysis measurements, potential stepping was employed (withoutstepping to the base potential) so that the harmonics at various DC potentials were measured.The modified electrode was held at a step potential applied from the FHI-ELAB G050-0298potentiostat and perturbed with an AC potential sent from the SRS SR830 lock-in amplifierto induce the alternating fluorescence response detected by the PMT. A Stanford ResearchSRS570 current preamplifier was used to convert the PMT amplified signal into voltage andfurther pass it through a 1 kHz low-pass filter. The rms amplitude of this processed signalwas measured with a data acquisition board (DAQ) and the harmonic response was analyzedwith the lock-in amplifier. The frequency of the AC potential was selected to be 163 Hz fortwo reasons. First, 163 Hz and its harmonics are not close to the power line frequency of 60Hz and its harmonics. Second, this frequency is relatively low and the DNA reorientation willnot be kinetically impeded by the electrode charging process [253]. The rms amplitude of the1356.6. Results and discussionAC potential was typically 50 mV, 100 mV or 150 mV. Care must be taken when choosing theamplitude of the AC potential so that the most positive and negative applied potentials do notexceed the stability limits of the SAMs, e.g., oxidation at positive potentials (i.e., more positivethan 0.35 V) or reduction at negative potentials (i.e., more negative than 0.40 V).6.6 Results and discussion6.6.1 Influence of surface crystallography on DNA SAM formationAs reviewed in Section 6.2, the potential-driven DNA reorientation is a useful motif in biosens-ing. One important requirement is to create SAMs with low DNA density (< 1011 strands/cm2),which is necessary for reorientation [54]. The standard approach to make DNA SAMs on aAu surface is to deposit thiolated DNA on the Au surface first, followed by displacing non-specifically adsorbed DNA with MCH [108–110]. Low DNA density SAMs can be achievedeither by short exposure to thiolated DNA [226, 227] or partial reductive desorption of DNAin a solution containing MCH at negative potentials [54, 57, 228]. Another approach to formlow DNA density SAMs is to deposit MCH first, followed by partial displacing MCH with thio-lated DNA [111, 252, 253, 256]. MCH-DNA SAMs formed by this competitive thiol-exchangemechanism have a relatively low DNA density (typically below 5× 1010 strands/cm2 [111]),enabling a high extent of DNA reorientation. However, a significant amount of non-uniformitywas observed on a MCH-DNA SAM deposited on a polycrystalline electrode with DNA pref-erentially adsorbed onto various regions [111]. Here, MCH-DNA SAM deposited on a singlecrystal Au bead electrode is investigated, in an attempt to understand the influence of surfacecrystallography on DNA SAM formation.Investigated in this section is the DNA-AlexaFluor488 SAM listed in Table 6.2. A fluores-cence image of this MCH-ssDNA SAM on a single crystal Au bead electrode was taken at −0.4V (vs. SCE) and shown in Figure 6.7. It is assumed that at −0.4 V (vs. SCE), all the DNAmolecules are repelled from the surface to the most extended conformation (but not reductivelydesorbed), so the fluorescence intensity is mainly dependent on the DNA density. The fluo-rescence image displays a symmetric pattern strongly correlated to the fcc crystal structure,1366.6. Results and discussionfacilitating proper indexing of surface.A high fluorescence signal is observed the (100) surface as well as surrounding surfaceswith large (100) terrace (wider than 5 atoms), forming a cross-shape feature in the center ofthe image. This feature is quite familiar, recalling a similar structure in the desorption poten-tial maps of AuS-C10-BODIPY SAMs (Figure 4.12) presented in Chapter 4. Considering theDNA is modified with a hexanthiol linker with a similar structure to MCH, in the linker portion,the coordination to the surface and the intermolecular interaction should be similar to the AuS-C10-BODIPY SAMs. However, in the case MCH-DNA SAM, the fluorescence intensity from aspecific region is indicative of the extent of thiol-exchange, which is dependent on the moleculararrangement of the MCH SAM, the availability of defects on the MCH SAM and the intermolec-ular interaction between DNA strands.The thiol-exchange process involves competition between an organized SAM and incomingadsorbates, so the molecular arrangement of the SAM has a strong impact on the tendencytowards or against adsorbate replacement. The hexanethiol (expected to be similar to MCH)covered (111) and (100) surfaces have been characterized with STM [102, 104]. On a (111)surface, the hexanthiolate SAM is featured with a large fraction of well-ordered domain adopt-ing the widely-observed (2p3× 3) (also termed c(4× 2), see Section 2.2.2) structure anddispersed vacancy defects [104]. As a comparison, the molecular arrangement of a hexanethi-olate SAM deposited on a (100) surface is more complicated [101, 102]. Fully lifting a recon-structed (100) surface does not seem to be achievable solely by adsorption of a hexanethiolateSAM. The molecular arrangements on a (1× 1) (100) surface and a reconstructed (100) sur-face are different (termed α phase and β phase respectively) and either one is featured with asignificant amount of domain segregation. Therefore, a MCH covered (100) surface is likelyto be highly heterogeneous and defective, which is favorable for the thiol-exchange process.Moreover, a compact alkanethiolate SAM is typically formed after a fast adsorption processand a slow annealing process [208, 209]. The number of defects are significantly decreasedafter the annealing process [209], which lowers the possibility of thiol-exchange. It was shownin Chapter 4 that extending the AuS-C10-BODIPY SAM assembly time from 15 min to 18 hcaused a negative shift in desorption potential, which was indicative of the SAM annealing pro-cess. The negative shift was more significant on the cross-shape region around (100) (~-701376.6. Results and discussionFigure 6.7: Fluorescence image of a MCH-ssDNA-AlexaFluor488 SAM on a single crystal Aubead electrode taken at −0.4 V (vs. SCE) indexed with (a) the low index planes shown forthe full image with four quadrants; (b) low-index and stepped surfaces in the NW quadrant; (c)kinked surfaces in the NW quadrant. The scale bar in each image is 200 μm. Adapted from [8]with permission. Copyright (2014) American Chemical Society.mV) than on a (111) facet (-20 mV), suggesting a difference in the annealing rate. Thus, fora MCH (thiolate) SAM formed after 1 h deposition, a larger number of defects are expectedon the cross-shape region, allowing for a higher amount of thiol-exchange. All these factorsdiscussed above combined might be the cause for a higher DNA density in the cross-shaperegion containing the (100) surface and surfaces with a large (100) terrace than on a (111)facet.A more detailed correlation between the fluorescence intensity and the surfaces can be ob-served by plotting the intensity profiles extracted from the WNW portion of Figure 6.7 along thethree crystallographic zones which connects the low-index planes (Figure 6.8 a-c). Among nu-merous subtle features, the fluorescence intensities observed near two highly rough surfacesare worth highlighting. High intensity is observed on a region containing the (311) surface(marked with “∗” in Figure 6.8a). The (311) surface is the turning point of the (111)-(100) zoneand features equal widths of (111) step and (100) step. In contrast to the (311) surface, fluores-cence near the (210) surface (the turning point of the (111)-(100) zone) is the lowest intensityof the three crystallographic zones of (100)-(110)-(111) (marked with “#” in Figure 6.8c). Al-though these two surfaces are both featured with high density of step edges, the extents ofthiol-exchange are vastly distinctive, so the DNA intermolecular interaction might also have aninfluence, other than simply the atomic arrangement of the surface. The fluorescence inten-sity stays high from (311) to the region surrounding (111) facet, suggesting the (111) terrace1386.6. Results and discussionFigure 6.8: Fluorescence intensity profiles extracted from the WNW portion of Figure 6.7 alongthe three crystallographic zones which connects the low-index planes: (a) (100)−(111), (b)(111)−(110), (c) (110)−(100) and the three zones from the a low-index surface to the turningpoint stepped surface of the opposite side: (d) (111)−(210), (e) (100)−(331), (f) (110)-(311).Adapted from [8] with permission. Copyright (2014) American Chemical Society.might be favorable for the intermolecular interaction between adsorbed DNA molecules. Onthe contrary, the presence of (110) terraces on the stepped surfaces from (210) to (110) mighthinder the thiol-exchange. The intensity profiles for the three zones from the a low-index sur-face to the turning point stepped surface ((210), (311) or (331)) of the opposite side (Figure 6.8d-f) enables exploring the DNA formation on kinked surfaces. These three zones cross at the(531) surface where (111), (100) and (110) terraces and corresponding (210), (331) and (311)steps coexist (see Figure 2.8 in Section 2.1.4). Locally, a relatively high density of DNA on the(531) surface is observed in the (111)-(210) and (100)-(331) zones (marked with “§” in Figure6.8d and e), whereas in the (110)-(311) zone, a similar intensity level is observed from (531)(marked with “†” in Figure 6.8f) as well as the neighboring surfaces. This suggests that thepresence of (311) characteristic in the (110)-(311) zone might exert an influence in the DNAdensity on the kinked surfaces.Besides the major features on the important surfaces mentioned above, there are also a1396.6. Results and discussionnumber of interesting subtle features which cannot be predicted or rationalized by the sur-face crystal structures. Nevertheless, with this fluorescence imaging technique, the influenceof surface crystallography on DNA formation by thiol-exchange for this MCH-DNA SAM wasobserved. This may be an important factor which causes the non-uniformity in DNA densityobserved on a MCH-DNA SAM deposited on a polycrystalline electrode. More importantly,the results presented in this section may give insights into energetics of thiol-exchange as afunction of surface crystallography.6.6.2 Investigation of the potential-driven DNA reorientation with in situfluorescence imagingIn Chapter 5, in situ fluorescence imaging was employed to investigate the potential-controlledbehavior of a low packing density Aib peptide SAM. In a similar fashion, potential steps areapplied to induce the DNA reorientation, resulting in modulated fluorescence which is detectedusing a camera. This imaging technique and the use of a single crystal Au bead electrodeenables exploration of the potential-driven fluorescence response on various crystallographicsurfaces. Two MCH-DNA SAMs are characterized: a MCH-ssDNA-AlexaFluor647 SAM anda MCH-dsDNA-AlexaFluor647 SAM. Note that both SAMs were assembled by direct adsorp-tion from solution, with no hybridization or dehybridization performed on the surface, and thedsDNA was formed in solution before being deposited onto the electrode. In addition, the mea-surements were conducted for as-prepared SAMs so that the possibility of intensity change dueto photo-bleaching or adsorbate degradation is minimized. Thus the hybridization states of thetwo SAMs investigated in this section are assumed to be pure ssDNA and dsDNA.Figure 6.9 shows the fluorescence images for the MCH-ssDNA-AlexaFluor647 SAM and theMCH-dsDNA-AlexaFluor647 SAM, both taken at −0.4 V (vs. SCE). The fluorescence patternwhich correlates to the crystallographic orientation map is observed in both images, similar tothe MCH-ssDNA-AlexaFluor488 SAM shown previously. The lack of some subtle structures inthese fluorescence images as compared to Figure 6.7 can be rationalized by the lower DNAdensity and lower camera sensitivity. However, the major features are easily recognized, no-tably the three low-index surfaces, the cross-shape feature around (100), the high intensity1406.6. Results and discussionabFigure 6.9: Fluorescence images: (a) a MCH-ssDNA-AlexaFluor647 SAM and (b) a MCH-dsDNA-AlexaFluor647 SAM on a single crystal Au bead electrode captured at −0.4 V (vs. SCE).A variety of ROIs with their assigned crystallographic orientations are marked on the images.1416.6. Results and discussion(311) surface and the low intensity (210) surface. Several ROIs are selected based on thesefeatures for further analysis and marked in both fluorescence images (the (910) surface isgenerally referring to one arm of the cross-shape around the (100) surface). The overall flu-orescence intensities are evidently different for the two SAMs in that the dsDNA SAM has asignificantly higher intensity than the ssDNA SAM. This is similar to reported results where anincrease of fluorescence intensity was observed when a ssDNA SAM was hybridized with thecomplementary strand, due to the conversion of the DNA adsorbate from a flexible curly chainto a rigid extended rod [54, 226].When the potential steps profiled in Figure 6.10a are applied, the fluorescence intensityalternates in response due to DNA reorientation. The average fluorescence change measuredfrom a selected ROI inside a (111) facet (marked in Figure 6.9) is shown for the MCH-ssDNASAM and the MCH-dsDNA SAM in Figure 6.10b and Figure 6.10c, respectively. These dataare also shown as the fluorescence intensity as a function of step potential (Figure 6.11). Asexpected, the overall intensity of the MCH-ssDNA SAM is lower than that of the MCH-dsDNASAM. In addition, theMCH-dsDNASAMhas a significantly larger change in fluorescence duringa reorientation process. The increase in magnitude of fluorescence response upon hybridiza-tion was reported previously and was used in DNA detection [54, 226]. Higher sensitivity wasachieved by using longer DNA strand since a much larger change in fluorophore-metal sepa-ration is possible for the rigid dsDNA [54]. However, the absolute intensity cannot be used todetermine the hybridization state. Different samples of the same hybridization state or eventhe same sample measured on different days could have different static intensities. Photo-bleaching is one major reason for intensity variation. Technically, the fluorescence illuminationand collection efficiency on a particular ROI and the background intensity (which is not easy todetermine for a curved surface covered with a fluorescent SAM) are strongly dependent on theorientation of this curved electrode surface, which also contributes to intensity variation.The potential-driven fluorescence response measured from selected ROIs ((100), (910) and(311)) highlighted in Figure 6.9 are analyzed and shown in Figure 6.12. On the (111) surfacestudied above, it is evident that the dsDNASAMhas a largermagnitude of fluorescence reponsethan the ssDNA at negative potentials, which is not necessarily observed on the other surfaces.On the (100), (910) and (311) surfaces, the difference in magnitude is insignificant. Considering1426.6. Results and discussion-400-2000200400E (V / SCE)aE (V / SCE)225235245255265Fl. Int. (cts/sec)bMCH-ssDNA SAMFl. Int. (cts/sec)EstepEbase3303603904204500 90 180 270 360 450Fl. Int. (cts/sec)t (sec)cMCH-dsDNA SAMFl. Int. (cts/sec)EstepEbaseFigure 6.10: The potential-driven fluorescence response of a ROI inside a Au(111)facet (marked in Figure 6.9) for a MCH-ssDNA-AlexaFluor647 SAM and a MCH-dsDNA-AlexaFluor647 SAM: (a) the profile of potential steps applied to drive the DNA reorientation;(b) the fluorescence response of MCH-ssDNA-AlexaFluor647 SAM; (c) the fluorescence re-sponse of MCH-ssDNA-AlexaFluor647 SAM. The intensities at the step potentials (Estep) andthe based potentials (Ebase) are marked with filled circle and open circles, respectively.1436.6. Results and discussion220250280310340-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3330360390420450Fl. Int. (cts/sec) - ssDNA SAMFl. Int. (cts/sec) - dsDNA SAMEstep (V / SCE)(111)ssDNA SAM EstepssDNA SAM EbasedsDNA SAM EstepdsDNA SAM EbaseFigure 6.11: The fluorescence intensity as a function of step potential of a ROI inside a Au(111)facet (marked in Figure 6.9) for a MCH-ssDNA-AlexaFluor647 SAM (y-axis on the left) and aMCH-dsDNA-AlexaFluor647 SAM (y-axis on the right). Estep is a particular step potential andEbase is the base potential 0.35 V (vs. SCE) before stepping to corresponding step potential.the different reorientation mechanism for ssDNA and dsDNA described in [227] (as reviewedin Section 6.2), it might be reasonable to assume that ssDNA which reorients segment bysegment requires less space whereas dsDNA which rotates about the anchor point requiresmore space. Depending on the DNA density, the dsDNA SAM might not be able to reorientcompletely.Therefore, this potential-driven fluorescence response can be influenced by numerous fac-tors, including the density of DNA, the molecular conformation and the surface crystallography.Moreover, inconsistent results have been obtained from newly prepared SAMs and SAMs aftermeasurements presumably due to photo-bleaching of the fluorophore. From the sensing pointof view, detection of DNA purely based on the change in fluorescence intensity and magnitudeof fluorescence response may not be reliable unless all the conditions are properly controlled.1446.6. Results and discussion210230250270290310360380400420440460(100)assDNA SAM EstepssDNA SAM EbasedsDNA SAM EstepdsDNA SAM Ebase240270300330360390530560590620650680Fl. Int. (cts/sec) - ssDNA SAMFl. Int. (cts/sec) - dsDNA SAM(910)b470485500515530-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3920935950965980Estep (V / SCE)(311)cFigure 6.12: The fluorescence intensity as a function of step potential measured on ROIs from(a) (100) surface, (b) (910) surface and (c) (311) surface (marked in Figure 6.9) for a MCH-ssDNA-AlexaFluor647 SAM and a MCH-dsDNA-AlexaFluor647 SAM.1456.6. Results and discussion6.6.3 Harmonic analysis of the nonlinear fluorescence response in DNAreorientationDevelopment of a harmonic analysis technique for measuring the nonlinearfluorescence response driven by an AC potential perturbationThe imaging technique presented has limitations in sensitivity and speed. The fluorescenceresponse driven by the potential steps is a low frequency signal typically accompanied by a largeamount of 1/f noise [257]. In order to achieve a higher signal-to-noise ratio, an AC potentialperturbation is employed to induce a fluorescence response of higher frequency. The DNAreorientation-based biosensing platforms implemented by Rant et al. are mainly applying thefrequency-resolved or the time-resolved AC techniques [7, 54, 55, 57]. These measurementstend to be long and setups can be quite complicated and expensive. In this section, a differentapproach, namely, the nonlinear fluorescence response of DNA reorientation is analyzed togain a better understanding and develop a harmonic analysis technique for DNA detection.The nonlinear fluorescence response expected can be intuitively illustrated with Figure 6.11where the fluorescence intensity is non-linearly related the step potential. As discussed inSection 6.3, when a sinusoidal potential wave is applied, the nonlinear response function resultsin harmonics in the fluorescence signal. Moreover, since the ssDNA SAM and the dsDNA SAMeach follows a distinctive nonlinear function (in simple words, the shapes of the F-E curves aredifferent), the amplitudes of the corresponding harmonics should be different and can be usedin distinguishing the hybridization states. Below, the fluorescence response from the (111)surface will be taken as an example to demonstrate the harmonic analysis of the resultantnonlinear fluorescence signal.The fluorescence intensity from a small ROI inside a (111) facet was first measured with aPMT during a linear potential scan. A 40× water immersion objective was used to increasethe optical efficiency. In addition, the field diaphragm was reduced so that a small region wasilluminated, excluding signals from other regions. Figure 6.13 shows the current and fluores-cence intensity measured during potential scanning for the MCH-ssDNA-AlexaFluor647 SAMand MCH-dsDNA-AlexaFluor647 SAM investigated in last section. The cyclic voltammogramsare typical for SAM-solution interfaces devoid of redox reactions that can be modeled as resis-1466.6. Results and discussionFigure 6.13: (a) Cyclic voltammograms of a MCH-ssDNA-AlexaFluor647 SAM and a MCH-dsDNA-AlexaFluor647 SAM and (b) fluorescence intensity of the two SAMs measured withpotential scanning from a ROI inside a (111) facet. The arrows in (b) depict the oscillationdirections of the AC potential perturbation and the fluorescence response.tor–capacitor (RC) circuits. Different magnitudes of capacitive current are observed for the twoSAMs because the immersed electrode areas were different. A RC circuit is a linear system[258], so the current response driven by an AC potential potential perturbation only containsthe first harmonic of fundamental frequency. However, in the case of DNA SAMs, the opticalresponse is nonlinear within the potential range scanned. As depicted in Figure 6.13b, an ACpotential is applied at a constant DC potential. Depending on this DC potential and on the am-plitude of the perturbation, the fluorescence response can be linear or nonlinear. Typically, inelectrochemical techniques (e.g., AC voltammetry, electrochemical impedance spectroscopy),the amplitude of the AC potential perturbation is small to ensure that the response stays linearand thereby the equivalent circuit analysis of the system is valid [255, 259–261]. Applying alarge amplitude potential perturbation to study the nonlinear electrochemical response is be-ing used more often [262–265]. In this section, a large amplitude potential perturbation (e.g.,100 mV rms) is applied to induce the nonlinear fluorescence response. More importantly, for1476.6. Results and discussion05101520-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3Harm (µA)E (V vs. SCE)Harm(1) ssDNAHarm(1) dsDNAHarm (µA)Harm(2) ssDNA (×100)Harm(2) dsDNA (×100)Figure 6.14: Amplitudes of the first harmonic (solid line) and second harmonic (dashed line)signals of the current response driven by 50 mV rms potential perturbation at a series of DCpotentials from 0.275 V (vs. SCE) to -0.35 V (vs. SCE) for the MCH-ssDNA SAM and (blue)the MCH-dsDNA SAM (red).this RC-like system, even with this large amplitude, the current response remains linear. Fig-ure 6.14 shows the amplitude of the first and second harmonics of the electrochemical currentdriven by 50 mV rms perturbation. The response is essentially linear, with <1% in the secondharmonic. Therefore the nonlinear nature of the fluorescence response is driven by a linearelectrochemical perturbation.Figure 6.15 shows the amplitudes of the first (a), second (b) and third (c) harmonics of thefluorescence response driven by 100 mV rms perturbation at a series of DC potentials from 0.2V (vs. SCE) to -0.25 V (vs. SCE) for the MCH-ssDNA SAM and the MCH-dsDNA SAM. Theseharmonic measurements were conducted after the potential scanning measurements, on thesame (111) ROI as in Figure 6.13. The fluorescence modulation was measured with the PMTand analyzed by the lock-in amplifier.Clear differences are observed between the ssDNA SAM and the dsDNA SAM in first andsecond harmonic signals. According to Equation 6.6, the amplitude of the nth harmonic isproportional to the nth derivative of the fluorescence response function. Intuitively, the first har-monic informs on the slope of fluorescence response function at a particular DC potential whilethe second harmonic is indicative of the curvature at this DC potential. At the negative limit ofDC potential, both first and second harmonic amplitudes for the two SAMs converge. This canbe explained by a lack of fluorescence intensity oscillation at negative DC potentials becausethe DNA strands are at the most extended orientation (not necessarily perpendicular to the sur-1486.6. Results and discussion0. (µA)assDNA (measured)ssDNA (simulated)dsDNA (measured)dsDNA (simulated)0.0000.0040.0080.0120.016Harm(2) (µA)b0.00000.00060.00120.00180.0024-0.3 -0.2 -0.1 0.0 0.1 0.2Harm(3) (µA)E (V vs. SCE)cFigure 6.15: Measured and simulated amplitudes of the harmonics of the fluorescence re-sponse from a (111) featured ROI driven by 100 mV rms perturbation at a series of DC po-tentials from 0.2 V (vs. SCE) to -0.25 V (vs. SCE) for the MCH-ssDNA SAM and (blue) theMCH-dsDNA SAM (red): (a) first harmonic, (b) second harmonic and (c) third harmonic.1496.6. Results and discussionface). At 0 V (vs. SCE), the first harmonic of the ssDNA SAM reaches a maximum which alsocorresponds to a minimum in the second harmonic, indicating a drastic change in the shapeof the fluorescence response at this potential. The potential where the second harmonic inten-sity reaches minimum (EH(2)min) has some similarity to the potential of conformation transition(PCT) [228]. However, the definition of PCT is quite ambiguous in that a DNA strand does notchange its orientation from lying down to standing up (or vice versa) immediately at a partic-ular potential, but over a range of potentials as observed in Figure 6.13b. A maximum in thefirst harmonic and a minimum in the second harmonic are observed at a potential slightly morepositive than 0.2 V (vs. SCE) for the dsDNA SAM. In other words, the dsDNA SAM has a morepositive EH(2)min than the ssDNA SAM. The difference in EH(2)min is intriguing because thiscan be applied in DNA sensing, but measuring the EH(2)min using harmonic analysis is prob-ably not convenient. Monitoring the change in the amplitude of a harmonic (e.g., the secondharmonic) at a constant DC potential (e.g., 0 V (vs. SCE)) upon capturing of a DNA strandmight be more practical and feasible.Since the frequency used is well within the charging time of the interface, the intensity-potential curve (Figure 6.13b) could be used to simulate the AC measurements. An Octavescript was used to model the potential scanning fluorescence response as a nonlinear func-tion. The AC fluorescence response was determined based on this nonlinear function for eachDC potential. The harmonic responses of these fluorescence signals at different DC potentialswere calculated using Fourier transform analysis. The simulated amplitudes of the first, secondand third harmonics of the fluorescence response are presented in Figure 6.15 as dashed lines.Overall, the experimental results match the simulated results especially for the ssDNA SAM.The discrepancy shown in the dsDNA results was most likely caused by a change in lamp inten-sity occurring between the potential scanning measurement and the harmonic measurements.There exists some agreement between the experimental and the simulated results for the thirdharmonic suggesting that this signal is similar to noise. Increasing the amplitude of the ACpotential perturbation (limited by the potential range within which the SAM is stable) may allowdetection of the third harmonic more accurately, but higher harmonics may be challenging.Normalization of the harmonic singal with a reference signal measured simultaneously isneeded to account for the intensity variation due to lamp intensity drift, photo-bleaching or a1506.6. Results and discussion0. rms (µA)assDNA (measured)ssDNA (simulated)dsDNA (measured)dsDNA (simulated)F rms (µA)ssDNA (AC off)dsDNA (AC off) -0.2 -0.1 0.0 0.1 0.2Harm(1) / FrmsE (V vs. SCE)bFigure 6.16: Comparison of measured and simulated: (a) rms fluorescence intensity and (b)ratio of the first harmonic amplitude over the rms fluorescence intensity.1516.6. Results and discussionchange in photon collection efficiency between different samples. For an AC signal, the rmsintensity is a plausible reference. However, at the moment, accurate measurements of the rmsfluorescence intensity cannot be achieved with the present setup due to undetermined noiselimitation. As shown in Figure 6.16a, the rms fluorescence intensity is underestimated at pos-itive positive potentials for the dsDNA SAM and overestimated at negative potentials for bothSAMs. The overestimation at the positive potential for the dsDNA SAM was probably due tothe lamp intensity drift, similar to difference between the measured and calculated results in theharmonics (Figure 6.15). Besides the lamp intensity drift, an unknown source of noise which ismore prominent with increasing intensity is present in the system and clearly observed whenthe potential perturbation was turned off as shown in Figure 6.16a. According to Figure 6.16b,the simulated ratio of the first harmonic amplitude over the rms fluorescence intensity is a con-stant (~1.4). This is reasonable considering the fluorescence signal is dominated by the firstharmonic and the ratio of center-to-peak amplitude and the rms amplitude for a sine wave isp2. Experimentally, at positive potentials the measured ratio is slightly smaller thanp2 butacceptable. However, a large deviation is observed at negative potentials where the fluores-cence oscillation due to DNA orientation is small and contaminated by the noise. Therefore,normalizing the results using the rms fluorescence intensity is not yet reliable. An alternative isto use the first harmonic amplitude for normalization. As a phase-sensitive detector, the lock-inamplifier is only sensitive to a signal with the frequency of interest, so the first harmonic ampli-tude should be more reliable as it will not be contaminated by the noise which is assumed tobe smaller than the measured bandwidth. Strictly speaking, the first harmonic is not an idealreference simply because different harmonics were measured at different times. Since a har-monic measurement was fast, it is assumed that the first harmonic and the second harmonicwere measured under the same conditions.Measuring the normalized second harmonic response for different crystallographicsurfacesThe harmonic analysis of the potential-driven fluorescence response for a MCH-ssDNA SAMand a MCH-dsDNA SAM has been demonstrated. A MCH-DNA SAM prepared by treating a1526.6. Results and discussion0. / Harm(1)a(111)ssDNA SAMdsDNA SAMUrea treated DNA SAM0. / Harm(1)b(100) -0.2 -0.1 0.0 0.1 0.2Harm(2) / Harm(1)E (V vs. SCE)c(910)Figure 6.17: Second harmonic amplitude normalized with first harmonic amplitude as a functionof DC potential for a MCH-ssDNA SAM SAM, a MCH-dsDNA SAM and a urea treated MCH-DNA SAM measured from three ROIs: (a) a (111) ROI, (b) a (100) ROI and (c) a (910) ROI.MCH-dsDNA SAM with 8 M urea solution for 2 min was also studied with the harmonic analysistechnique. Figure 6.17 shows the normalized second harmonic amplitude as a function of DCpotential for the three DNA SAMs. The harmonic analysis technique was performed on threeROIs (a (111) ROI (a), a (100) ROI (b) and a (910) ROI (c)) to study the influence of surfacecrystallography on the fluorescence response.The normalized second harmonic intensity for the two SAMs for the (111) surface (Figure6.17a) not only converge, but also approach a constant at positive potentials, demonstratingthe lack of intensity oscillation and reliability of the normalization. The two SAMs started to1536.6. Results and discussionshow a difference in the normalized second harmonic intensity at ~0.05 V (vs. SCE). A fastdecrease is observed for the ssDNA SAM, reaching a minimum of ~0 at ~0 V (vs. SCE). Thiscorresponds to the largest slope and smallest curvature in the fluorescence-potential function.In contrast, the normalized second harmonic intensity for the dsDNA SAM gradually decreasesover the potential range with a minimum at ~0.2 V (vs. SCE). As a comparison, the secondharmonic response for the urea treated DNA SAM falls between the ssDNA SAM and the ds-DNA SAM. To be specific, the urea treated DNA SAM has the similar response as the other twoSAMs at positive potentials and its minimum second harmonic intensity is observed at ~0.08 V(vs. SCE), between those for the other two SAMs. This suggests the short urea treatment atroom temperature might result in a surface covered with a mixture of ssDNA and dsDNA. Moreimportantly, it might be possible to detect DNA binding by monitoring the change in normalizedsecond harmonic amplitude at 0 V or at 0.2 V. In addition, a larger amplitude potential pertur-bation can be applied to increase the sensitivity, but this is only applicable to detection at 0 V.However, a series of DNA SAMs with different hybridization states need to be analyzed to testthe feasibility of this DNA sensing mechanism.The harmonic responses from the (100) surface (Figure 6.17b) and the (910) surface (Fig-ure 6.17c) are similar to each other, but are different from that measured on a (111) surface.The most evident difference is the minimum second harmonic intensity of a SAM, which is ob-served at more positive potentials on (100) or (910) surfaces as compared to the same SAMon a (111) surface. This difference reflects the different shapes of fluorescence-potential func-tions on these crystallographic surfaces. Considering the (910) surface is featured with large(100) terraces, DNA strands adsorbed on (100) and (910) surfaces might organize in a sim-ilar structure, which explains the similar harmonic responses. On the other hand, the DNApacking structure on (111) surface might be significantly different from that on (100) and (910)surfaces, resulting in the distinctive harmonic response. Different hybridization states (ssDNA,dsDNA and amixture of ssDNA and dsDNA) can be also distinguished based on the normalizedsecond harmonic amplitude on (100) and (910) surfaces, but only at a potential close to thepositive limit of the stable DC potential (e.g., 0.15 V (vs. SCE)), which excludes the possibilityof increasing the amplitude of the potential perturbation. Therefore, to develop a DNA sensingplatform based on this harmonic analysis technique, substrates with a greater percentage of1546.7. Conclusions(111) surface would be preferred.6.7 ConclusionsThe MCH-DNA-AlexaFluor SAMs were assembled on the single crystal Au bead electrode bythiol-exchange mechanism. The fluorescence image of a typical MCH-DNA SAM showed ahigh heterogeneity which correlates to the map of crystallographic orientations, demonstratingthe influence of the surface crystallography on formation of a DNA SAM via the thiol-exchangemechanism. Moreover, in situ fluorescence imaging was employed to study the potential-drivenDNA reorientation for a MCH-ssDNA SAM and a MCH-dsDNA SAM. It was found that potential-driven fluorescence response measured with the imaging technique could be influenced by nu-merous factors. Thus, the fluorescence intensity and the magnitude of fluorescence responseare not reliable criteria in determining the hybridization state of the DNA SAM.A harmonic analysis technique has been developed to measure the non-linearity of thefluorescence response driven by AC potential perturbation with a lock-in amplifier. There wasa minimum intensity observed in the second harmonic response as a function of DC potential.This minimum in the second harmonic intensity for a dsDNA SAM occurred at more positivepotentials than that for a ssDNA SAM, which may be useful in distinguishing the hybridizationstates.DNA binding may be possibly detected by monitoring the change in second harmonicintensity. Different harmonic responses were observed from different crystallographic surfaces.A surface with significant fraction of (111) character would be preferred for developing a DNAsensing platform based on this harmonic analysis technique.155Chapter 7Concluding remarks7.1 SummaryFunctionalization of a metal surface with self-assembled monolayers has been routinely usedin constructing biosensing interfaces with desirable performance [2, 3, 5, 6]. The basic compo-nents of a SAM are the substrate and the adsorbate. Naturally, in order to optimize the perfor-mance of a SAM-based biosensor, it is necessary to investigate and tune the properties of boththe substrate and the adsorbate. In electrochemical biosensors, a metal substrate is used andtherefore an important aspect to study is the influence of the surface crystallography on theSAM. Traditionally, research related to this subject involved depositing a number of SAMs on aseries of electrodes with one designated crystallographic orientation [9–12]. Two major disad-vantages arose in the copious experiments on these electrodes: a lack of reproducibility dueto the variation in experimental conditions and a limitation in the crystallographic orientationsstudied. To overcome these disadvantages, a single crystal electrode was used throughout theexperimental works presented in this thesis. With this electrode, identifiable crystallographicorientations were investigated under the same conditions, achieving self-consistent analyses.Three types of SAMs based on the Au-S interaction were investigated: AuS-C10-BODIPYSAMs, AuS-Aib4-BODIPYSAMs, andMCH-DNA-AlexaFluor SAMs formed by the thiol-exchangemechanism, demonstrating the utility of the single crystal Au bead electrode in the researcharea of surface modification and characterization. The behavior of the SAMs was manipulatedby electrochemical methods. Visualization of the fluorophore-labeled SAMs on different crystal-lographic domains on the electrode was achieved with fluorescence microscopic techniques.The coupled spectroelectrochemical techniques enables in-depth investigation of the SAMsassembled on the single crystal Au bead electrode in an efficient and effective way useful for1567.1. Summaryexploring the influence of surface crystallography on the SAMs.The use of the single crystal Au bead electrode was first demonstrated by characterizingthe reductive desorption behavior of AuS-C10-BODIPY SAMs and AuS-Aib4-BODIPY SAMsassembled on this electrode with in situ fluorescence imaging. Reductive desorption mapscreated by detailed imaging analysis showed a strong correlation to the dbb (as the surrogatefor PZC) map of the surface. However, numerous differences were observed between differenttypes of SAMs as well as between the reductive desorption and dbb (or PZC), suggesting theinfluence of intermolecular interaction on the reductive desorption. The results might inspirein-depth investigation of SAMs on a number of stepped or kinked surfaces showing interestingproperties but rarely studied. As an example, the desorption potential on surfaces with a large(100) terrace was similar to that on (100) surface for a AuS-C10-BODIPY SAM but notablydifferent for a AuS-Aib4-BODIPY SAM. This might be an indication of the footprint of the adsor-bates or the effective distance of intermolecular interaction, which are important in controllingthe packing density of the adsorbate.AuS-Aib4-BODIPY SAMs were further studied with the single crystal Au bead electrodebecause of their intriguing properties. The two-step reductive desorption on a (111) surfacemight result from the intermolecular hydrogen bond network, which helps stabilize and organizethe Aib peptide SAMs even when desorbed. Since the conventional alkanethiol SAMs do nothave such unique intermolecular interaction, it could be beneficial to create highly robust andordered SAMs with Aib peptides and further apply them in constructing biosensing platforms. Inaddition, it was found that the orientation or structure of low packing density AuS-Aib4-BODIPYSAMs on a (111) surface could be manipulated with potential, attributed to response of thehelix-induced dipole moment to electric field. Although this property was not observed on highpacking density SAMs, it could be potentially applicable in a nano-scale system or useful indeveloping a potential-controlled molecular switch.Spectroelectrochemical characterization of MCH-DNA-AlexaFluor SAMs focused on inves-tigating the potential-driven reorientation behavior which is useful in biosensing. A significantamount of heterogeneity was observed on the fluorescence image from a MCH-DNA SAMformed via thiol-exchange mechanism, which strongly correlates to the surface crystallogra-phy. This heterogeneity enabled exploring the potential-driven reorientation of the DNA SAM on1577.1. Summaryidentified crystallographic surfaces. However, the changes in fluorescence intensity and mag-nitude of potential-driven fluorescence response detected by fluorescence imaging were notreliable in distinguishing the hybridization states of a MCH-DNA SAM. In an attempt to deepenthe understanding of the DNA reorientation and to develop a new sensing mechanism, a har-monic analysis technique was employed to measure the harmonic response of the alternatingfluorescence signal driven by an AC potential perturbation for MCH-DNA SAMs of different hy-bridization states. Aminimum in the second harmonic response was observed at a DC potentialwhich was dependent on the hybridization state of a SAM, so measuring the change in secondharmonic response can be used in DNA detection. Compared with other reported DNA de-tection techniques based on the potential-driven DNA reorientation [7, 54, 55, 57], the sensingmechanism proposed here is probably easy and inexpensive to implement. Different harmonicresponses were observed from (111), (100) and (910) surfaces, giving insight into selection ofa proper substrate for a biosensing interface. Although it was found that the (111) surface wassuitable for such a substrate, there are still a number of identifiable crystallographic surfaceson the single crystal Au bead electrode yet to explore, from which a better one might be found.Typically in actual biosensing, vapor-deposited polycrystalline Au substrates are widely used.On this type Au surface, there should be Au(111) domains which are favorable for sensing andother regions of varied crystallographic structures. More importantly, this flat surface can beeasily indexed to identify the crystallographic domains. Thus the results obtained these studiesusing a single crystal Au bead electrode can be borrowed to optimize the surface preparationfor biosensing.Overall, the single crystal Au bead electrode provides a valuable platform to conduct re-searches on SAMs. With the spectroelectrochemical techniques discussed, a large number ofphysical or chemical behaviors dependent on the surface crystallography can be visualized,measured or controlled, guiding future studies on SAMs. As a long-term goal, the findings ob-tained by employing the single crystal electrode can make contribution to rationally developingbiosensing interfaces with desirable performance.1587.2. Future work7.2 Future workA variety of investigations can be conducted as an optimization or an extension of the workpresented in this thesis. The general ideas are outlined here without the attempt to demonstratethe feasibility.7.2.1 Optimization of the single crystal Au bead electrodeThe single crystal Au bead electrode used throughout the experimental studies was fabricatedsimply by flame melting. The conditions (e.g., the temperature, the cooling rate, the angleof heating, etc.) were not well-controlled. Therefore, the size, the crystalline state (completesingle crystal or bi-crystal), and orientation (e.g., the locations of the crystallographyc surfaceswith respect to the bead growing direction) of the electrode could not be controlled. Fabricationof single crystal bead electrode using electron beam heating or direct current heating wasreported [202, 203], which gave better control over the conditions and thus reproducible singlecrystal bead electrodes with high quality. Thus, it would be rewarding to apply these methodsto fabricate single crystal Au bead electrodes for future studies.7.2.2 Further explorations of the thiolate-terminated Aib peptide SAMsThe length of a Aib peptide adsorbate and the direction of the dipole moment have stronginfluences on the physical and chemical properties, which in turn affect the electrochemicalresponse of this Aib peptide SAM [210, 211, 218, 222]. To gain a better understanding ofthe potential-controlled response of the thiolate-terminated Aib peptide SAMs, Aib peptidesof various lengths and the opposite dipole moment should be systematically studied. More-over, it would also be valuable to investigate the influence of surface crystallography on thepotential-controlled response. Longer Aib peptides would be preferred, lessening the fluores-cence quenching.1597.2. Future work7.2.3 Further explorations of the MCH-DNA SAMsThe DNA detection mechanism based on harmonic analysis of the AC potential-driven fluo-rescence response of a DNA SAM was proposed in Chapter 6. This mechanism requiresfurther tests with MCH-DNA SAMs of a series of hybridization states to determine its feasi-bility, sensitivity and selectivity. In addition, the DNA length is expected to have a profoundinfluence on the potential-driven fluorescence response. On one hand, a longer DNA strandleads to an increase in intensity, but on the other hand, it might also result in a change in thefluorescence-potential function and the harmonic response. Therefore, it would be necessaryto investigate the harmonic responses for SAMs formed with various DNA lengths. Moreover,although the minimum second harmonic intensity can be correlated to the sharpest change inthe fluorescence-potential function, it is still not clear about the corresponding orientation orstructure of the DNA. Electrochemical atomic force microscopy (EC-AFM) would be useful inproviding structural information for a DNA SAM at the DC potential at which the second har-monic intensity is minimum. Furthermore, the sensing target is not necessarily limited to thecomplementary strand, it would be important to test the applicability of the harmonic analysistechnique in detection of other types of bio-molecules such as proteins.7.2.4 Development of new spectroelectrochemical techniquesThe fluorescence microscopic characterization presented in this thesis was mainly based onmeasuring the intensity of fluorescence, where the density of fluorophore and the fluorophore-metal separation are convoluted. It would be important to develop new techniques to investigatethese two factors separately in order to understand different properties on a SAM. Spectro-scopic methods based on FRET inform on the distance between a donor fluorophore and anacceptor fluorophore. If a SAM is labeled with an evenly distributed mixture of a FRET pair,the experimentally determined FRET efficiency should indicate the average lateral adsorbateseparation. It would be beneficial to investigate the change in FRET efficiency as a functionof potential for Aib peptide SAMs or DNA SAMs which might provide information on structuralchanges driven by an applied potential. Fluorescence-lifetime imaging microscopy (FLIM) is auseful technique to measure the fluorophore-metal separation [169]. Thus characterizing the1607.2. Future workpotential-driven DNA reorientation using FLIM coupled with electrochemistry would be reward-ing in fundamentally exploring the change in orientation with potential, especially on differentcrystallographic surfaces.161References[1] Grieshaber, D.; MacKenzie, R.; Voros, J.; Reimhult, E. Sensors 2008, 8, 1400–1458.[2] Wink, T.; J. van Zuilen, S.; Bult, A.; P. van Bennekom, W. Analyst 1997, 122, 43R–50R.[3] Chaki, N. K.; Vijayamohanan, K. Biosens. Bioelectron. 2002, 17, 1–12.[4] Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. Rev.2005, 105, 1103–1170.[5] North, S. H.; Lock, E. H.; Taitt, C. R.; Walton, S. G. Anal. Bioanal.Chem. 2010, 397,925–933.[6] Gooding, J. J.; Darwish, N. Chem. Rec. 2012, 12, 92–105.[7] Rant, U. Bioanal. Rev 2012, 4, 97–114.[8] Yu, Z. L.; Casanova-Moreno, J.; Guryanov, I.; Maran, F.; Bizzotto, D. J. Am. Chem. Soc.2015, 137, 276–288.[9] Lipkowski, J.; Stolberg, L.; Yang, D.-F.; Pettinger, B.; Mirwald, S.; Henglein, F.; Kolb, D.Electrochim. Acta 1994, 39, 1045–1056.[10] Yang, D.-F.; Wilde, C. P.; ; Morin, M. Langmuir 1996, 12, 6570–6577.[11] Zhong, C.-J.; Zak, J.; Porter, M. D. J. Electroanal. Chem. 1997, 421, 9–13.[12] Doneux, T.; Steichen, M.; Rache, A. D.; Buess-Herman, C. J. Electroanal. Chem. 2010,649, 164–170.[13] Bard, A. J.; Faulkner, L. R. Electrochemical methods: fundamentals and applications;John Wiley & Sons, Inc., 2001.162References[14] Lipson, A.; S. G. Lipson, H.; Lipson, Optical Physics - Fourth Edition; Cambridge Univer-sity Press, 2011.[15] Hermann, K. Crystallography and Surface Structure: An Introduction for Surface Scien-tists and Nanoscientists; Wiley-VCH Verlag GmbH & Co. KGaA, 2011.[16] Schwarzenbach, D. Crystallography; John Wiley & Sons, Inc., 1996.[17] Borchardt-Ott, W. Crystallography: An Introduction; Springer Berlin Heidelberg, 2011.[18] Hazen, R. M.; Sholl, D. S. Nat. Mater. 2003, 2, 367–374.[19] Gellman, A. J.; Horvath, J. D.; Buelow, M. T. J. Mol. Catal. A: Chem. 2001, 167, 3–11.[20] Gellman, A. J. ACS Nano 2010, 4, 5–10.[21] Lang, B.; Joyner, R.; Somorjai, G. Surf. Sci. 1972, 30, 454–474.[22] Van Hove, M.; Somorjai, G. Surf. Sci. 1980, 92, 489–518.[23] Cunningham, S. L.; Chan, C. M.; Weinberg, W. H. Phys. Rev. B 1978, 18, 1537–1549.[24] Kirschner, J.; Feder, R. Surf. Sci. 1979, 79, 176–188.[25] Feder, R.; Kirschner, J. Surf. Sci. 1981, 103, 75–102.[26] Noonan, J. R.; Davis, H. L. Phys. Rev. B 1984, 29, 4349–4355.[27] Xu, M. L.; Tong, S. Y. Phys. Rev. B 1985, 31, 6332–6336.[28] Noonan, J. R.; Davis, H. L. J. Vac. Sci. Technol., A 1990, 8, 2671–2676.[29] Methfessel, M.; Hennig, D.; Scheffler, M. Phys. Rev. B 1992, 46, 4816–4829.[30] Barrett, S. D.; Dhesi, S. S.; Evans, M. P.; White, R. G. Meas. Sci. Technol. 1993, 4,114–119.[31] Wan, J.; Fan, Y. L.; Gong, D. W.; Shen, S. G.; Fan, X. Q. Modell. Simul. Mater. Sci. Eng.1999, 7, 189–206.[32] Singh-Miller, N. E.; Marzari, N. Phys. Rev. B 2009, 80, 235407.163References[33] Perdereau, J.; Biberian, J. P.; Rhead, G. E. J. Phys. F: Met. Phys. 1974, 4, 798–806.[34] Van Hove, M.; Koestner, R.; Stair, P.; Bibérian, J.; Kesmodel, L.; Bartos˘, I.; Somorjai, G.Surf. Sci. 1981, 103, 189–217.[35] Robinson, I. K. Phys. Rev. Lett. 1983, 50, 1145–1148.[36] Ercolessi, F.; Tosatti, E.; Parrinello, M. Phys. Rev. Lett. 1986, 57, 719–722.[37] Wöll, C.; Chiang, S.; Wilson, R. J.; Lippel, P. H. Phys. Rev. B 1989, 39, 7988–7991.[38] Barth, J. V.; Brune, H.; Ertl, G.; Behm, R. J. Phys. Rev. B 1990, 42, 9307–9318.[39] Vlieg, E.; Robinson, I.; Kern, K. Surf. Sci. 1990, 233, 248–254.[40] Wang, X.-Q. Phys. Rev. Lett. 1991, 67, 3547–3550.[41] Koch, R.; Borbonus, M.; Haase, O.; Rieder, K. H. Appl. Phys. A 1992, 55, 417–429.[42] dos Reis, D.; Negreiros, F.; de Carvalho, V.; Soares, E. Surf. Sci. 2010, 604, 568–573.[43] Bengió, S.; Navarro, V.; González-Barrio, M. A.; Cortés, R.; Vobornik, I.; Michel, E. G.;Mascaraque, A. Phys. Rev. B 2012, 86, 045426.[44] Gao, X.; Hamelin, A.; Weaver, M. J. Surf. Sci. 1992, 274, L588–L592.[45] Gao, X.; Edens, G. J.; Hamelin, A.; Weaver, M. J. Surf. Sci. 1994, 318, 1–20.[46] Van Hove, M. A.; Weinberg, W. H.; Chan, C.-M. Low-Energy Electron Diffraction: Exper-iment, Theory and Surface Structure Determination; Springer Berlin Heidelberg, 1986.[47] Jiang, L.; McNeil, C. J.; Cooper, J. M. J. Chem. Soc., Chem. Commun. 1995, 1293–1295.[48] Willner, I.; Heleg-Shabtai, V.; Blonder, R.; Katz, E.; Tao, G.; Buckmann, A. F.; Heller, A.J. Am. Chem. Soc. 1996, 118, 10321–10322.[49] Katz, E.; Bückmann, A. F.; Willner, I. J. Am. Chem. Soc. 2001, 123, 10752–10753.164References[50] Patel, J. N.; Kaminska, B.; Gray, B.; Gates, B. D. Effect of self-assembled monolayers(SAMs) in binding glucose oxidase for electro-enzymatic glucose sensor with gold elec-trodes. 29th Annual International Conference of the IEEE Engineering in Medicine andBiology Society. 2007; pp 2677–2680.[51] Anandan, V.; Gangadharan, R.; Zhang, G. Sensors 2009, 9, 1295–1305.[52] Fan, C.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9134–9137.[53] Xiao, Y.; Lai, R. Y.; Plaxco, K. W. Nat. Protoc. 2007, 2, 2705–2780.[54] Rant, U.; Arinaga, K.; Scherer, S.; Pringsheim, E.; Fujita, S.; Yokoyama, N.; Tornow, M.;Abstreiter, G. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 17364–17369.[55] Rant, U.; Pringsheim, E.; Kaiser, W.; Arinaga, K.; Knezevic, J.; Tornow, M.; Fujita, S.;Yokoyama, N.; Abstreiter, G. Nano Lett. 2009, 9, 1290–1295.[56] Xia, F.; White, R. J.; Zuo, X.; Patterson, A.; Xiao, Y.; Kang, D.; Gong, X.; Plaxco, K. W.;Heeger, A. J. J. Am. Chem. Soc. 2010, 132, 14346–14348.[57] Langer, A.; Hampel, P. A.; Kaiser, W.; Knezevic, J.; Welte, T.; Villa, V.; Maruyama, M.;Svejda, M.; Jähner, S.; Fischer, F.; Strasser, R.; Rant, U. Nat. Commun. 2013, 4, 2099–2106.[58] Spinke, J.; Liley, M.; Guder, H. J.; Angermaier, L.; Knoll, W. Langmuir 1993, 9, 1821–1825.[59] Mirsky, V. M.; Riepl, M.; Wolfbeis, O. S. Biosens. Bioelectron. 1997, 12, 977–989.[60] Briand, E.; Salmain, M.; Compère, C.; Pradier, C.-M. Colloids Surf., B 2006, 53, 215–224.[61] Mantzila, A. G.; Maipa, V.; Prodromidis, M. I. Anal. Chem. 2008, 80, 1169–1175.[62] Takoh, K.; Horie, M.; Someya, H.; Ishida, M.; Kamijo, K. Label-free immunosensor usinga gold electrode covered with conductive self-assembled monolayer. IEEE Sensors 2010Conference. 2010; pp 411–414.165References[63] Ahmad, A.; Moore, E. Analyst 2012, 137, 5839–5844.[64] Wan, J.; Ai, J.; Zhang, Y.; Geng, X.; Gao, Q.; Cheng, Z. Sci. Rep. 2016, 6, 19806–19811,Article.[65] Vericat, C.; Vela, M. E.; Salvarezza, R. C.Phys. Chem. Chem. Phys. 2005, 7, 3258–3268.[66] Vericat, C.; Vela, M. E.; Benitez, G.; Carro, P.; Salvarezza, R. C. Chem. Soc. Rev. 2010,39, 1805–1834.[67] Torrelles, X.; Vericat, C.; Vela, M. E.; Fonticelli, M. H.; Millone, M. A. D.; Felici, R.; Lee, T.-L.; Zegenhagen, J.; Muñoz, G.; Martín-Gago, J. A.; Salvarezza, R. C. J. Phys. Chem. B2006, 110, 5586–5594.[68] Zhang, J.; Chi, Q.; Ulstrup, J. Langmuir 2006, 22, 6203–6213.[69] Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733–740.[70] Kondoh, H.; Kodama, C.; Sumida, H.; Nozoye, H. J. Chem. Phys. 1999, 111, 1175–1184.[71] Hasan, M.; Bethell, D.; Brust, M. J. Am. Chem. Soc. 2002, 124, 1132–1133.[72] Pace, G.; Venanzi, M.; Castrucci, P.; Scarselli, M.; Crescenzi, M. D.; Palleschi, A.;Stella, L.; Formaggio, F.; Toniolo, C.; Marletta, G. Mater. Sci. Eng., C 2006, 26, 918–923.[73] Yeganeh, M. S.; Dougal, S. M.; Polizzotti, R. S.; Rabinowitz, P. Phys. Rev. Lett. 1995, 74,1811–1814.[74] Darling, S. B.; Rosenbaum, A. W.; Wang, Y.; Sibener, S. J. Langmuir 2002, 18, 7462–7468.[75] Poirier, G. E. Langmuir 1997, 13, 2019–2026.[76] Schreiber, F. Prog. Surf. Sci. 2000, 65, 151–257.[77] Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Chem. Phys. 1993, 98, 678–688.166References[78] Fenter, P.; Eberhardt, A.; Liang, K. S.; Eisenberger, P. J. Chem. Phys. 1997, 106, 1600–1608.[79] Chidsey, C. E. D.; Loiacono, D. N. Langmuir 1990, 6, 682–691.[80] Sun, L.; Crooks, R. M. Langmuir 1993, 9, 1951–1954.[81] Grönbeck, H.; Curioni, A.; Andreoni, W. J. Am. Chem. Soc. 2000, 122, 3839–3842.[82] Yourdshahyan, Y.; Zhang, H. K.; Rappe, A. M. Phys. Rev. B 2001, 63, 081405.[83] Masens, C.; Ford, M.; Cortie, M. Surf. Sci. 2005, 580, 19–29.[84] Hayashi, T.; Morikawa, Y.; Nozoye, H. J. Chem. Phys. 2001, 114, 7615–7621.[85] Vargas, M. C.; Giannozzi, P.; Selloni, A.; Scoles, G. J. Phys. Chem. B 2001, 105, 9509–9513.[86] Gottschalck, J.; Hammer, B. J. Chem. Phys. 2002, 116, 784–790.[87] Barmparis, G. D.; Honkala, K.; Remediakis, I. N. J. Chem. Phys. 2013, 138, 064702.[88] Kondoh, H.; Iwasaki, M.; Shimada, T.; Amemiya, K.; Yokoyama, T.; Ohta, T.; Shimo-mura, M.; Kono, S. Phys. Rev. Lett. 2003, 90, 066102.[89] Roper, M.; Skegg, M.; Fisher, C.; Lee, J.; Dhanak, V.; Woodruff, D.; Jones, R. G. Chem.Phys. Lett. 2004, 389, 87–91.[90] Chaudhuri, A.; Odelius, M.; Jones, R. G.; Lee, T.-L.; Detlefs, B.; Woodruff, D. P. J. Chem.Phys. 2009, 130, 124708.[91] P. Fenter, P. E., A. Eberhardt Science 1994, 266, 1216–1218.[92] Cossaro, A.; Mazzarello, R.; Rousseau, R.; Casalis, L.; Verdini, A.; Kohlmeyer, A.; Flore-ano, L.; Scandolo, S.; Morgante, A.; Klein, M. L.; Scoles, G.Science 2008, 321, 943–946.[93] Camillone, N.; Chidsey, C. E. D.; Liu, G.; Scoles, G. J. Chem. Phys. 1993, 98, 3503–3511.[94] Fenter, P.; Eisenberger, P.; Liang, K. S. Phys. Rev. Lett. 1993, 70, 2447–2450.167References[95] Poirier, G. E. Chem. Rev. 1997, 97, 1117–1128.[96] Li, J.; Liang, K. S.; Camillone, N.; Leung, T. Y. B.; Scoles, G. J. Chem. Phys. 1995, 102,5012–5028.[97] Poirier, G. E. J. Vac. Sci. Technol., B 1996, 14, 1453–1460.[98] Yamada, R.; Uosaki, K. Langmuir 2001, 17, 4148–4150.[99] Schweizer, M.; Hagenström, H.; Kolb, D. Surf. Sci. 2001, 490, L627–L636.[100] Loglio, F.; Schweizer, M.; Kolb, D. M. Langmuir 2003, 19, 830–834.[101] Schweizer, M.; Manolova, M.; Kolb, D. Surf. Sci. 2008, 602, 3303–3307.[102] Grumelli, D.; Cristina, L. J.; Maza, F. L.; Carro, P.; Ferrón, J.; Kern, K.; Salvarezza, R. C.J. Phys. Chem. C 2015, 119, 14248–14254.[103] Hagenström, H.; Schneeweiss, M. A.; Kolb, D. M. Langmuir 1999, 15, 2435–2443.[104] Benítez, G.; Vericat, C.; Tanco, S.; Lenicov, F. R.; Castez, M. F.; Vela, M. E.; Sal-varezza, R. C. Langmuir 2004, 20, 5030–5037.[105] Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109,3559–3568.[106] Badia, A.; Lennox, R. B.; Reven, L. Acc. Chem. Res. 2000, 33, 475–481.[107] Poirier, G. E. Langmuir 1999, 15, 1167–1175.[108] Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 8916–8920.[109] Levicky, R.; Herne, T. M.; Tarlov, M. J.; Satija, S. K. J. Am. Chem. Soc. 1998, 120, 9787–9792.[110] Steel, A.; Levicky, R.; Herne, T.; Tarlov, M. Biophys. J. 2000, 79, 975–981.[111] Murphy, J. N.; Cheng, A. K. H.; Yu, H.-Z.; Bizzotto, D. J. Am. Chem. Soc. 2009, 131,4042–4050.168References[112] Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M. J. Am. Chem. Soc.1990, 112, 4301–4306.[113] Collard, D. M.; Fox, M. A. Langmuir 1991, 7, 1192–1197.[114] Hostetler, M. J.; Templeton, A. C.; Murray, R. W. Langmuir 1999, 15, 3782–3789.[115] Yang, G.; Amro, N. A.; Starkewolfe, Z. B.; Liu, G.-y. Langmuir 2004, 20, 3995–4003.[116] Laibinis, P. E.; Fox, M. A.; Folkers, J. P.; Whitesides, G. M. Langmuir 1991, 7, 3167–3173.[117] Huang, J.; Hemminger, J. C. J. Am. Chem. Soc. 1993, 115, 3342–3343.[118] Widrig, C. A.; Chung, C.; Porter, M. D. J. Electroanal. Chem. Interfacial Electrochem.1991, 310, 335–359.[119] Gorman, C. B.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1995, 11, 2242–2246.[120] Grahame, D. C. Chem. Rev. 1947, 41, 441–501.[121] Grahame, D. C.; Whitney, R. B. J. Am. Chem. Soc. 1942, 64, 1548–1552.[122] Sato, N. Electrochemistry at metal and semiconductor electrodes; Elsevier B.V., 1998.[123] Brown, M. A.; Goel, A.; Abbas, Z. Angew. Chem. Int. Ed. 2016, 55, 3790–3794.[124] Daikhin, L. I.; Kornyshev, A. A.; Urbakh, M. Phys. Rev. E 1996, 53, 6192–6199.[125] Daikhin, L.; Kornyshev, A.; Urbakh, M. Electrochim. Acta 1997, 42, 2853–2860.[126] Mott, N.; Watts-Tobin, R. Electrochim. Acta 1961, 4, 79–107.[127] Bockris, J. O.; Devanathan, M. A. V.; Muller, K. Proc. R. Soc. London, Ser. A 1963, 274,55–79.[128] Trasatti, S. InModern Aspects of Electrochemistry: No. 13; Conway, B. E., Bockris, J. O.,Eds.; Springer US: Boston, MA, 1979; pp 81–206.[129] Handbook of Chemistry and Physics, the 96nd Edition; CRC Press, 2015-2016.[130] Laitinen, H. A.; Mosier, B. J. Am. Chem. Soc. 1958, 80, 2363–2366.169References[131] Sabatani, E.; Rubinstein, I.; Maoz, R.; Sagiv, J. J. Electroanal. Chem. Interfacial Elec-trochem. 1987, 219, 365–371.[132] Lang, N. D.; Kohn, W. Phys. Rev. B 1970, 1, 4555–4568.[133] Lang, N. D.; Kohn, W. Phys. Rev. B 1971, 3, 1215–1223.[134] Smoluchowski, R. Phys. Rev. 1941, 60, 661–674.[135] Hansson, G. V.; Flodström, S. A. Phys. Rev. B 1978, 18, 1572–1585.[136] Kolb, D.; Schneider, J. Electrochim. Acta 1986, 31, 929–936.[137] Trasatti, S. J. Electroanal. Chem. 1971, 33, 351–378.[138] Kolb, D.; Schneider, J. Surf. Sci. 1985, 162, 764–775.[139] Hamelin, A.; Sottomayor, M.; Silva, F.; Chang, S.-C.; Weaver, M. J. J. Electroanal. Chem.1990, 295, 291–300.[140] Hamelin, A. J. Electroanal. Chem. 1996, 407, 1–11.[141] Hamelin, A.; Martins, A. J. Electroanal. Chem. 1996, 407, 13–21.[142] Silva, A. F.; Martins, A. In Interfacial Electrochemistry: Theory: Experiment, and Appli-cations; Wieckowski, A., Ed.; Marcel Dekker, Inc, 1999; pp 449–461.[143] de Levie, R. J. Electroanal. Chem. Interfacial Electrochem. 1990, 280, 179–183.[144] Lakowicz, J. R. Principles of fluorescence spectroscopy - Third Edition; Springer Sci-ence+Business Media, LLC, 2006.[145] Clegg, R. M. Curr. Opin. Biotechnol. 1995, 6, 103–110.[146] Song, L.; Hennink, E. J.; Young, I. T.; Tanke, H. J. Biophys. J. 1995, 68, 2588–2600.[147] Drexhage, K. H. Sci. Amer. 1970, 222, 108–119.[148] Drexhage, K. J. Lumin. 1970, 1, 693–701.[149] Kuhn, H. J. Chem. Phys. 1970, 53, 101–108.170References[150] Kuhn, H.; Möbius, D. Angew. Chem. Int. Ed. Engl. 1971, 10, 620–637.[151] Amos, R. M.; Barnes, W. L. Phys. Rev. B 1997, 55, 7249–7254.[152] Chance, R. R.; Prock, A.; Silbey, R. In Advances in Chemical Physics; Prigogine, I.,Rice, S. A., Eds.; John Wiley & Sons, Inc., 1978; pp 1–65.[153] Ford, G.; Weber, W. Phys. Rep. 1984, 113, 195–287.[154] Barnes, W. L. J. Mod. Opt. 1998, 45, 661–699.[155] Chance, R. R.; Prock, A.; Silbey, R. J. Chem. Phys. 1974, 60, 2744–2748.[156] Yeung, M. S.; Gustafson, T. K. Phys. Rev. A 1996, 54, 5227–5242.[157] Lakowicz, J. R. Anal. Biochem. 2005, 337, 171–194.[158] Bizzotto, D.; Shepherd, J. L. In Advances in Electrochemical Science and Engineering:Diffraction and Spectroscopic Methods in Electrochemistry; Alkire, R. C., Kolb, D. M.,Lipkowski, J., Ross, P. N., Eds.; Advances in Electrochemical Sciences and Engineering;Wiley-VCH Verlag GmbH, 2008; Vol. 9; pp 97–126.[159] Reichman, J. Handbook of Optical Filters for Fluorescence Microscopy. Chroma Tech-nology Corp, 2012.[160] Fester, W. K.; Abramowitz, M.; Davidson, M. W. Olympus IX70 Fluorescence MicroscopeCutaway Diagram.[161] Aswani, K.; Jinadasa, T.; Brown, C. M. Microscopy Today 2012, 20, 22–28.[162] Olympus IX71/IX81 Brochure.[163] Loudet, A.; Burgess, K. Chem. Rev. 2007, 107, 4891–4932.[164] Bergström, F.; Mikhalyov, I.; Hägglöf, P.; Wortmann, R.; Ny, T.; Johansson, L. B.-A. J.Am. Chem. Soc. 2002, 124, 196–204.171References[165] Mikhalyov, I.; Gretskaya, N.; Bergstrom, F.; Johansson, L. B.-A. Phys. Chem. Chem.Phys. 2002, 4, 5663–5670.[166] Panchuk-Voloshina, N.; Haugland, R. P.; Bishop-Stewart, J.; Bhalgat, M. K.; Millard, P. J.;Mao, F.; Leung, W.-Y.; Haugland, R. P. J. Histochem. Cytochem. 1999, 47, 1179–1188.[167] The Molecular Probe Handbook: A Guide to Fluorescent Probes and Labeling Technolo-gies. 2010.[168] Shepherd, J. L.; Kell, A.; Chung, E.; Sinclar, C. W.; Workentin, M. S.; Bizzotto, D. J. Am.Chem. Soc. 2004, 126, 8329–8335.[169] Casanova-Moreno, J.; Yu, Z. L.; Massey-Allard, J.; Ditchburn, B.; Young, J. F.; Biz-zotto, D. In Luminescence in Electrochemistry: Applications in Analytical Chemistry,Physics and Biology; Miomandre, F., Audebert, P., Eds.; Springer International Publish-ing, 2017.[170] Clavilier, J.; Faure, R.; Guinet, G.; Durand, R. J. Electroanal. Chem. Interfacial Elec-trochem. 1979, 107, 205–209.[171] Verlinden, J.; Celis, J.; Roos, J. In Passivity of Metals and Semiconductors; Froment, M.,Ed.; Elsevier, 1983; pp 341 – 346.[172] Hamelin, A. J. Electroanal. Chem. Interfacial Electrochem. 1982, 142, 299–316.[173] Hamelin, A.; Vitanov, T.; Sevastyanov, E.; Popov, A. J. Electroanal. Chem. InterfacialElectrochem. 1983, 145, 225–264.[174] Shepherd, J. L. Electro-fluorescence characterization of insoluble surfactants adsorbedon solid electrodes. Ph.D. thesis, The University of British Columbia, 2005.[175] Casanova Moreno, J. R. Spectroelectrochemical characterization of ultrathin organicfilms deposited on electrode surfaces. Ph.D. thesis, University of British Columbia, 2014.[176] Ferreira, T.; Rasband, W. ImageJ User Guide - IJ 1.46r. 2012.[177] User Manual for Evolve 512.172References[178] Operating Manual for CCD Camera Models ST-7XE/XME, ST-8XE, ST-9XE, ST-10XE/XME, and ST-2000XM/XCM with High Speed USB Interface.[179] Imabayashi, S.-i.; Hobara, D.; Kakiuchi, T.; Knoll, W. Langmuir 1997, 13, 4502–4504.[180] Satjapipat, M.; Sanedrin, R.; Zhou, F. Langmuir 2001, 17, 7637–7644.[181] Strutwolf, J.; O’Sullivan, C. K. Electroanalysis 2007, 19, 1467–1475.[182] Henry, O.; Maliszewska, A.; O’Sullivan, C. Electrochem. Commun. 2009, 11, 664–667.[183] Walczak, M. M.; Popenoe, D. D.; Deinhammer, R. S.; Lamp, B. D.; Chung, C.;Porter, M. D. Langmuir 1991, 7, 2687–2693.[184] Walczak, M. M.; Alves, C. A.; Lamp, B. D.; Porter, M. D. J. Electroanal. Chem. 1995,396, 103–114.[185] Zhong, C.-J.; Porter, M. D. J. Electroanal. Chem. 1997, 425, 147–153.[186] Yang, D.-F.; Wilde, C. P.; ; Morin, M. Langmuir 1997, 13, 243–249.[187] Yang, D.-F.; Morin, M. J. Electroanal. Chem. 1997, 429, 1–5.[188] Yang, D.-F.; Morin, M. J. Electroanal. Chem. 1998, 441, 173–181.[189] Schneider, T. W.; Buttry, D. A. J. Am. Chem. Soc. 1993, 115, 12391–12397.[190] Kawaguchi, T.; Yasuda, H.; Shimazu, K.; Porter, M. D. Langmuir 2000, 16, 9830–9840.[191] Kunze, J.; Leitch, J.; Schwan, A. L.; Faragher, R. J.; Naumann, R.; Schiller, S.; Knoll, W.;Dutcher, J. R.; Lipkowski, J. Langmuir 2006, 22, 5509–5519.[192] Laredo, T.; Leitch, J.; Chen, M.; Burgess, I. J.; Dutcher, J. R.; Lipkowski, J. Langmuir2007, 23, 6205–6211.[193] Mulder, W. H.; Calvente, J. J.; Andreu, R. Langmuir 2001, 17, 3273–3280.[194] Doneux, T.; Nichols, R.; Buess-Herman, C. J. Electroanal. Chem. 2008, 621, 267–276.173References[195] Ramírez, P.; Andreu, R.; Cuesta, Á.; Calzado, C. J.; Calvente, J. J. Anal. Chem. 2007,79, 6473–6479.[196] Kunitake, M.; Batina, N.; ; Itaya, K. Langmuir 1995, 11, 2337–2340.[197] Sawaguchi, T.; Mizutani, F.; Taniguchi, I. Langmuir 1998, 14, 3565–3569.[198] Ohira, A.; Ishizaki, T.; Sakata, M.; Taniguchi, I.; Hirayama, C.; Kunitake, M.Colloids Surf.,A 2000, 169, 27–33.[199] He, Y.; Borguet, E. Angew. Chem. Int. Ed. 2007, 46, 6098–6101.[200] Wen, R.; Yan, C.-J.; Yan, H.-J.; Pan, G.-B.; Wan, L.-J. Chem. Commun. 2011, 47, 6915–6917.[201] Komanicky, V.; Fawcett, W. Electrochim. Acta 2004, 49, 1185–1194.[202] Voigtländer, B.; Linke, U.; Stollwerk, H.; Brona, J. J. Vac. Sci. Technol., A 2005, 23,1535–1537.[203] Voigtländer, B.; Cherepanov, V.; Elsaesser, C.; Linke, U. Rev. Sci. Instrum. 2008, 79,033911.[204] Musgrove, A.; Kell, A.; Bizzotto, D. Langmuir 2008, 24, 7881–7888.[205] Casanova-Moreno, J. R.; Bizzotto, D. Langmuir 2013, 29, 2065–2074.[206] Jung, L. S.; ; Campbell, C. T. J. Phys. Chem. B 2000, 104, 11168–11178.[207] Sumi, T.; Uosaki, K. J. Phys. Chem. B 2004, 108, 6422–6428.[208] Debono, R. F.; Loucks, G. D.; Manna, D. D.; Krull, U. J. Can. J. Chem. 1996, 74, 677–688.[209] Forouzan, F.; Bard, A. J.; Mirkin, M. V. Isr. J. Chem. 1997, 37, 155–163.[210] Fabris, L.; Antonello, S.; Armelao, L.; Donkers, R. L.; Polo, F.; Toniolo, C.; Maran, F. J.Am. Chem. Soc. 2006, 128, 326–336.174References[211] Kaplan, J. M.; Shang, J.; Gobbo, P.; Antonello, S.; Armelao, L.; Chatare, V.; Ratner, D. M.;Andrade, R. B.; Maran, F. Langmuir 2013, 29, 8187–8192.[212] Gatto, E.; Venanzi, M. Polym. J. 2013, 45, 468–480.[213] Venanzi, M.; Pace, G.; Palleschi, A.; Stella, L.; Castrucci, P.; Scarselli, M.;Crescenzi, M. D.; Formaggio, F.; Toniolo, C.; Marletta, G. Surf. Sci. 2006, 600, 409–416.[214] Gatto, E.; Porchetta, A.; Scarselli, M.; Crescenzi, M. D.; Formaggio, F.; Toniolo, C.; Ve-nanzi, M. Langmuir 2012, 28, 2817–2826.[215] Hauptmann, N.; Robles, R.; Abufager, P.; Lorente, N.; Berndt, R. J. Phys. Chem. Lett.2016, 7, 1984–1990.[216] Shin, Y.-g. K.; Newton, M. D.; Isied, S. S. J. Am. Chem. Soc. 2003, 125, 3722–3732.[217] Toniolo, C.; Crisma, M.; Formaggio, F.; Peggion, C.; Broxterman, Q. B.; Kaptein, B. Pept.Sci. 2004, 76, 162–176.[218] Holm, A. H.; Ceccato, M.; Donkers, R. L.; Fabris, L.; Pace, G.; Maran, F. Langmuir 2006,22, 10584–10589.[219] Antonello, S.; Formaggio, F.; Moretto, A.; Toniolo, C.; Maran, F. J. Am. Chem. Soc. 2003,125, 2874–2875.[220] Perera, N. V.; Isley, W.; Maran, F.; Gascon, J. A. J. Phys. Chem. C 2010, 114, 16043–16050.[221] Gobbo, P. Stability of and Electron Transfer through Self-Assembled Monolayers of Con-formationally Constrained Peptides. M.Sc. thesis, University of Padova, 2010.[222] Becucci, L.; Guryanov, I.; Maran, F.; Guidelli, R. J. Am. Chem. Soc. 2010, 132, 6194–6204.[223] Gatto, E.; Venanzi, M.; Palleschi, A.; Stella, L.; Pispisa, B.; Lorenzelli, L.; Toniolo, C.;Formaggio, F.; Marletta, G. Mater. Sci. Eng., C 2007, 27, 1309–1312.175References[224] Gatto, E.; Stella, L.; Formaggio, F.; Toniolo, C.; Lorenzelli, L.; Venanzi, M. J. Pept. Sci.2008, 14, 184–191.[225] Polo, F.; Antonello, S.; Formaggio, F.; Toniolo, C.; Maran, F. J. Am. Chem. Soc. 2005,127, 492–493.[226] Rant, U.; Arinaga, K.; Fujita, S.; Yokoyama, N.; Abstreiter, G.; Tornow, M. Nano Lett.2004, 4, 2441–2445.[227] Rant, U.; Arinaga, K.; Tornow, M.; Kim, Y. W.; Netz, R. R.; Fujita, S.; Yokoyama, N.;Abstreiter, G. Biophys. J. 2006, 90, 3666–3671.[228] Kaiser, W.; Rant, U. J. Am. Chem. Soc. 2010, 132, 7935–7945.[229] Peggion, C.; Crisma, M.; Toniolo, C.; Formaggio, F. Tetrahedron 2012, 68, 4429–4433.[230] Watson, J. D.; Crick, F. H. C. Nature 1953, 171, 737–738.[231] Daune, M. Molecular biophysics: structures in motion; Oxford University Press, 1999.[232] Vologodskii, A. Biophysics of DNA; Cambridge University Press, 2015.[233] Wang, A. H.-J.; Quigley, G. J.; Kolpak, F. J.; Crawford, J. L.; van Boom, J. H.; van derMarel, G.; Rich, A. Nature 1979, 282, 680–686.[234] Thamann, T. J.; Lord, R. C.; Wang, A. H.; Rich, A. Nucleic Acids Res. 1981, 9, 5443–5457.[235] Jonstrup, A. T.; Fredsøe, J.; Andersen, A. H. Sensors 2013, 13, 5937–5944.[236] Wang, Q.; Gao, F.; Ni, J.; Liao, X.; Zhang, X.; Lin, Z. Sci. Rep. 2016, 6, 22441–22410.[237] Hagerman, P. J. Biopolymers 1981, 20, 1503–1535.[238] Tinland, B.; Pluen, A.; Sturm, J.; Weill, G. Macromolecules 1997, 30, 5763–5765.[239] Baumann, C. G.; Smith, S. B.; Bloomfield, V. A.; Bustamante, C. Proc. Natl. Acad. Sci.U.S.A. 1997, 94, 6185–6190.176References[240] Rybenkov, V. V.; Cozzarelli, N. R.; Vologodskii, A. V. Proc. Natl. Acad. Sci. U.S.A. 1993,90, 5307–5311.[241] Rybenkov, V. V.; Vologodskii, A. V.; Cozzarelli, N. R. Nucleic Acids Res. 1997, 25, 1412–1418.[242] Tsortos, A.; Papadakis, G.; Mitsakakis, K.; Melzak, K. A.; Gizeli, E. Biophys. J. 2008, 94,2706–2715.[243] Petruska, J.; Goodman, M. F. J. Biol. Chem. 1995, 270, 746–750.[244] Qiao, W.; Chiang, H.-C.; Xie, H.; Levicky, R. Chem. Commun. 2015, 51, 17245–17248.[245] Laland, S.; Lee, W.; Overend, W.; Peacocke, A. Biochim. Biophys. Acta 1954, 14, 356–365.[246] DeVoe, H.; Tinoco, I. J. Mol. Biol. 1962, 4, 518–527.[247] Herskovits, T. T. Biochemistry 1963, 2, 335–340.[248] Maniatis, T.; Jeffrey, A.; deSande, H. V. Biochemistry 1975, 14, 3787–3794.[249] Summer, H.; Grämer, R.; Dräge, P. J. Vis. Exp. 2009, e1485.[250] Peterson, A. W.; Heaton, R. J.; Georgiadis, R. M. Nucleic Acids Res. 2001, 29, 5163–5168.[251] Kelley, S. O.; Barton, J. K.; Jackson, N. M.; McPherson, L. D.; Potter, A. B.; Spain, E. M.;Allen, M. J.; ; Hill, M. G. Langmuir 1998, 14, 6781–6784.[252] Casanova-Moreno, J.; Bizzotto, D. Electrochim. Acta 2015, 162, 62–71.[253] Casanova-Moreno, J.; Bizzotto, D. Anal. Chem. 2015, 87, 2255–2263.[254] About Lock-In Amplifiers. Stanford Research Systems.[255] Lasia, A. Electrochemical Impedance Spectroscopy and its Applications; Springer NewYork: New York, NY, 2014; pp 67–84.[256] Josephs, E. A.; Ye, T. J. Am. Chem. Soc. 2012, 134, 10021–10030.177[257] Malmstadt, H. V.; Enke, C. G.; Crouch, S. R. Electronics and instrumentation for scien-tists; The Benjamin/Cummings Publishing Company, Inc., 1981.[258] Moura, L.; Darwazeh, I. Introduction to Linear Circuit Analysis and Modelling : From DCto RF ; Newnes, 2005.[259] Sluyters-Rehbach, M.; Sluyters, J. In Electroanalytical Chemistry: A Series of Advances;Bard, A., Ed.; Marcel Dekker, New York, 1970; Vol. 4; Chapter 1, pp 1–127.[260] Macdonald, D. D. Transient Techniques in Electrochemistry; Springer US: Boston, MA,1977; pp 229–272.[261] Sluyters-Rehbach, M.; Sluyters, J. In Comprehensive Treatise of Electrochemistry; Yea-ger, E., Bockris, J., Conway, B., Sarangapani, S., Eds.; Plenum Press, 1984; Vol. 9;Chapter 4, pp 177–289.[262] Engblom, S. O.; Myland, J. C.; Oldham, K. B. J. Electroanal. Chem. 2000, 480, 120–132.[263] Engblom, S. O.; Myland, J. C.; Oldham, K. B.; Taylor, A. L. Electroanalysis 2001, 13,626–630.[264] Kiel, M.; Bohlen, O.; Sauer, D. Electrochim. Acta 2008, 53, 7367–7374.[265] McDonald, T. J.; Adler, S. ECS Trans. 2012, 45, 429–439.[266] X-Cite exacte USER’S GUIDE.[267] Chroma Spectra Viewer.[268] Thermo Fisher Fluorescence Spectra Viewer. ASpectraA.1 Hg arc lamp spectrum and excitation filter bands04812162024400 450 500 550 600 650 700Power (mW)λ (nm)Power (mW)BODIPY (Monomer)AlexaFluor488Excitation AlexaFluor647ExcitationFigure A.1: Hg arc lamp spectrum of an Excelitas X-Cite® exacte fluorescence illuminator(data obtained from user’s guide provided by Excelitas [266]). The two excitation filter bandscorrespond to the ones listed in Table 3.1: Chroma ET470/40x and Chroma HQ620/60x.A.2 Spectra of the used fluorophores and their correspondingfilter sets179A.2. Spectra of the used fluorophores and their corresponding filter sets020406080100400 450 500 550 600 650 700Relative Intensity (%)λ (nm)Spectra of BODIPY 493/503 (monomer) and its filter setExcitation filterEmission filterDichromatic mirrorRelative Intensity (%)ExcitationEmissionFigure A.2: Spectra of BODIPY 493/503 (monomer) and its corresponding filter set (excita-tion: ET470/40x, dichromatic: T495LPXR, emission: ET525/50m). The BODIPY 493/503(monomer) spectra were measured with an Agilent Cary Eclipse Fluorescence Spectropho-tometer. The filter set spectral data were obtained from Chroma’s spectra viewer website [267].020406080100400 450 500 550 600 650 700Relative Intensity (%)λ (nm)Spectra of AlexaFluor488 and its filter setExcitation filterEmission filterDichromatic mirrorRelative Intensity (%)ExcitationEmissionFigure A.3: Spectra of AlexaFluor488 and its corresponding filter set (excitation: ET470/40x,dichromatic: T495LPXR, emission: ET525/50m). The AlexaFluor488 spectral data were ob-tained from Thermo Fisher’s fluorescence spectra viewer website [268]. The filter set spectraldata were obtained from Chroma’s spectra viewer website [267].180A.2. Spectra of the used fluorophores and their corresponding filter sets020406080100550 600 650 700 750 800 850 900Relative Intensity (%)λ (nm)Spectra of AlexaFluor647 and its filter setExcitation filterEmission filterDichromatic mirrorRelative Intensity (%)ExcitationEmissionFigure A.4: Spectra of AlexaFluor647 and its corresponding filter set (excitation: HQ620/60x,dichromatic: Q660LP, emission: HQ700/75m). The AlexaFluor488 spectral data were obtainedfrom Thermo Fisher’s fluorescence spectra viewer website [268]. The filter set spectral datawere obtained from Chroma’s spectra viewer website [267].181


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items