Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electrochemical in situ investigation of thiolate DNA monolayers on gold with fluorescence imaging Murphy, Jeffrey N. 2008

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

Item Metadata

Download

Media
24-ubc_2008_fall_murphy_jeffrey.pdf [ 24.81MB ]
Metadata
JSON: 24-1.0061760.json
JSON-LD: 24-1.0061760-ld.json
RDF/XML (Pretty): 24-1.0061760-rdf.xml
RDF/JSON: 24-1.0061760-rdf.json
Turtle: 24-1.0061760-turtle.txt
N-Triples: 24-1.0061760-rdf-ntriples.txt
Original Record: 24-1.0061760-source.json
Full Text
24-1.0061760-fulltext.txt
Citation
24-1.0061760.ris

Full Text

ELECTROCHEMICAL IN SITU INVESTIGATION OF THIOLATE DNA MONOLAYERS ON GOLD WITH FLUORESCENCE IMAGING  by  JEFFREY N. MURPHY B.SC., SIMON FRASER UNIVERSITY, 2005 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE  REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  JUNE 2008 © JEFFREY N. MURPHY 2008  Abstract DNA-modified surfaces have been widely studied for microarray and biosensor applications, in particular sequence-specific detection of DNA, for which electrochemical and optical signals can be produced. Variations in the organization and surface density of adsorbed DNA are known to affect the sensitivity and reliability of assays performed using such surfaces, however most measurements of such surfaces to date have little to no spatial resolution, limiting the information that can be gathered regarding the heterogeneity of the organization of adsorbed DNA molecules. We have applied in situ epi-fluorescence microscopic imaging in conjunction with electrochemical measurements to fluorescently labelled thiolate DNA, adsorbed on polycrystalline gold electrodes with a mercaptohexanol (MCH) passivation layer. Spatially resolved information on the organization of adsorbed DNA on the surface is gathered within an area measuring 520 by 730 micrometres with 0.96 micrometre resolution. The technique has enabled us to investigate "hotspots" (regions of anomalously bright fluorescence) and regional variation in fluorescence; since molecular fluorescence is quenched as a function of distance from the metal substrate, potential modulation with consequent DNA reorientation or layer desorption provide further means to study the distribution and response of DNA as well as the specificity of the adsorption. Furthermore, an alternative means to the conventional preparation of thiolate-DNA / MCH monolayers has been developed. In this new method, a gold substrate passivated with MCH is subsequently immersed in an aqueous solution of 5'hexylthiol modified DNA. Through a ligand exchange process, DNA is immobilized forming a mixed MCH / DNA monolayer. Samples prepared via the new method display fewer hotspots and improved fluorescence switching of the DNA during electromodulation for samples made with single stranded (ss) DNA and with double stranded (ds) DNA. Measurement of the DNA surface concentration using ruthenium (III) hexaammine chloride with cyclic voltammetry for self assembled monolayers (SAMs) prepared via the new method are on the order of 1% of the maximum grafting density obtainable for both ssDNA and dsDNA by conventional methods.  ii  Table of Contents ABSTRACT...............................................................................................................................ii TABLE OF CONTENTS...........................................................................................................iii LIST OF TABLES.....................................................................................................................vi LIST OF FIGURES..................................................................................................................vii LIST OF SYMBOLS AND ABBREVIATIONS...................................................................xviii ACKNOWLEDGEMENTS...................................................................................................xxiii DEDICATION.......................................................................................................................xxiv 1  INTRODUCTION....................................................................................................................1 1.1 Objectives.........................................................................................................................3 1.2 Rationale..........................................................................................................................4 1.3 Scope of Thesis................................................................................................................4  2 THEORETICAL BACKGROUND..........................................................................................6 2.1 Fluorescence Microscopy................................................................................................6 2.1.1 Electronic Transitions and the Franck-Condon Principle.........................................6 2.1.2 Fate of Excited State Molecule................................................................................7 2.1.3 Fluorescence Near Metal Surfaces...........................................................................9 2.1.4 Carbocyanine Dyes.................................................................................................10 2.1.5 Fluorescence Microscopy.......................................................................................11 2.1.6 Microscope Resolution...........................................................................................12 2.1.7 Epi-Fluorescence Microscope................................................................................12 2.2 Electrochemical Concepts and the Electric Double Layer.............................................16 2.2.1 Qualitative Description of the Electric Double Layer............................................16 2.2.2 Ideally Polarizable Electrodes................................................................................16 2.2.3 Simple Capacitors...................................................................................................16 2.2.4 Capacitance of the Electrode | Solution Interface..................................................18 2.2.5 The Gouy-Chapman-Stern Theory (Quantitative)..................................................20 2.3 Thiolate SAMs and Desorption......................................................................................23 2.3.1 Introduction............................................................................................................23 2.3.2 Preparation Methods...............................................................................................24 2.3.3 Gold Substrates......................................................................................................24 2.3.4 Metal-SAM Interface.............................................................................................24 2.3.5 Mixed Monolayers.................................................................................................26 iii  2.3.6 Displacement of SAMs by Exchange.....................................................................26 2.3.7 Electrochemical Desorption...................................................................................27 2.4 DNA Essentials..............................................................................................................28 2.4.1 Introduction............................................................................................................28 2.4.2 Structure of DNA...................................................................................................28 2.4.3 Hybridization and Melting.....................................................................................32 2.4.4 Information Content...............................................................................................37 2.5 DNA SAM Sensors........................................................................................................39 2.5.1 Introduction............................................................................................................39 2.5.2 DNA SAM Sensors................................................................................................39 2.5.3 Thiolate DNA/MCH SAMs....................................................................................41 2.5.4 Hybridization..........................................................................................................43 2.5.5 Surface Concentration............................................................................................44 2.5.6 Electrochemical Desorption of Thiolate DNA.......................................................45 2.5.7 Counterion Condensation and DNA Aggregation..................................................46 2.5.8 Experimental Evidence of Inhomogeneity.............................................................47 2.5.9 Potential-Controlled Switching of Fluorescently Labelled DNA SAMs...............49 3  EXPERIMENTAL METHODOLOGY..................................................................................51 3.1 Systems Studied.............................................................................................................51 3.1.1 Materials and Chemicals........................................................................................51 3.1.2 Oligonucleotides.....................................................................................................52 3.1.3 DNA Purification and Preparation.........................................................................53 3.1.4 Sample Preparation.................................................................................................53 3.2 Electrochemical Methods...............................................................................................54 3.2.1 Electrochemical Equipment...................................................................................54 3.2.2 Chemicals...............................................................................................................55 3.2.3 Electrochemical Instrumentation............................................................................55 3.2.4 Electrochemical Procedures...................................................................................56 3.2.5 Electrochemical Techniques...................................................................................56 3.2.6 Cyclic Voltammetry................................................................................................56 3.2.7 Differential Capacitance.........................................................................................57 3.2.8 DNA Surface Concentration...................................................................................58 3.2.9 Surface Area Measurements...................................................................................62 3.3 Spectroscopic Technique................................................................................................64 3.3.1 Epi-Fluorescence Microscopy................................................................................64 3.3.2 Image Analysis.......................................................................................................71  4  RESULTS AND DISCUSSION............................................................................................73 4.1 Introduction....................................................................................................................73 4.2 ss/MCH..........................................................................................................................74 4.2.1 ssDNA-CY5/MCH.................................................................................................74 4.2.2 ssDNA-CY3/MCH.................................................................................................80 4.2.3 Conclusions............................................................................................................87 4.3 ds/MCH..........................................................................................................................88 iv  4.3.1 dsDNA-CY3/MCH.................................................................................................88 4.3.2 dsDNA-CY5/MCH.................................................................................................95 4.3.3 Conclusions............................................................................................................96 4.4 MCH/ss........................................................................................................................100 4.4.1 MCH/ssDNA-CY5...............................................................................................100 4.4.2 MCH/ssDNA-CY3 ...........................................................................................................................................101 4.4.3 Conclusions..........................................................................................................107 4.5 MCH/ds........................................................................................................................109 4.5.1 MCH/dsDNA-CY3...............................................................................................109 4.5.2 Conclusions..........................................................................................................113 5  SUMMARY AND CONCLUSIONS....................................................................................115  6  FUTURE WORK..................................................................................................................119 6.1 Introduction..................................................................................................................119 6.2 Flat Substrates..............................................................................................................119 6.3 In situ Adsorption/Passivation with MCH...................................................................120 6.4 In situ Hybridization of DNA......................................................................................121 6.5 Investigations of Nanostructured Au and Pd Surfaces.................................................123 6.6 Surface Concentration of DNA ...................................................................................123 6.7 Creation of Distance-Quenching Curves.....................................................................124 6.8 Fluorescence Lifetime Imaging Microscopy...............................................................126 6.9 Particle Tracking..........................................................................................................126 REFERENCES.......................................................................................................................127 APPENDIX A: IMAGE STEP SEQUENCES........................................................................136 APPENDIX B: TIMING DURING IMAGE ACQUISITION...............................................142 APPENDIX C: PURIFICATION PROCEDURE FOR REDUCED DNA.............................146 APPENDIX D: DETERMINATION OF THE SURFACE AREA.........................................147 APPENDIX E: AUXILIARY EXPERIMENTS.....................................................................152 E-1 Effects of Ionic Strength............................................................................................152 E-2 Capacitance Measurements.......................................................................................155 E-3 Concurrent CV / Imaging..........................................................................................157 E-4 Comparison of Switching Characteristics.................................................................157 E-5 In situ MCH Adsorption............................................................................................159 E-6 Other Experiments.....................................................................................................160 APPENDIX F: CREATIVE COMMONS LICENCING........................................................161  v  List of Tables Table 1 The bases of DNA...................................................................................................................................30 Table 2 The number of unique sequence combinations possible for polynucleotides of a given length of DNA nucleotides. ............................................................................................................................................38 Table 3 Dependence of fraction of charges compensated (θ) on cation charge for ssDNA (b = 0.40 nm) and dsDNA (b = 0.17 nm). This minimum compensation fraction is independent of concentration in solutions of low ionic strength...............................................................................................................47 Table 4 Up and down switching time constants for 48-mer ssDNA and dsDNA in a buffered 10 mM Tris buffer (pH = 7.3, ionic strength = 10 mM monovalent electrolyte solution (κ-1 = 3.0 nm) measured upon changing of the electrode polarity, going from +400 mV to -200 mV and back to +400 mV. ....50 Table 5 Oligonucleotides sequences. All sequences used were all 30 units long. Complementary sequences were unlabelled.......................................................................................................................................52 Table 6 Electrochemically and optically determined bead surface areas for five samples (ACM-ACQ) and a single crystal reference sample of known area. ...............................................................................................................................................................147  vi  List of Figures Figure 1 A Jablonski diagram showing the electronic excitation and emission transition processes. Electronic states A and B have the same multiplicity but C does not. Incident radiation excites electrons from the ground state of A to excited states of B, which then relax to the ground state followed by fluorescence emission as they transition to state A, or they may undergo intersystem crossing, to a third state, C, which may undergo relaxation by phosphorescence. ...................................................9 Figure 2 Overlap between donor fluorescence emission spectrum and acceptor absorption spectrum, with the overlapping region giving the overlap integral J(λ)...........................................................................10 Figure 3 Molecular structures of CY3 and CY5. CY5 contains 5 methine groups in its linker, while CY3 contains 3. CY3 absorbs maximally at 550 nm, and emits maximally at 570 nm (φf = 0.15). CY5 absorbs maximally at 649 nm, and emits maximally at 670 nm (φf = 0.28).....................................11 Figure 4 An illustrative representation of an airy disc (left) and two overlapping Airy discs. Far apart, Airy discs are easily resolved, however when they begin to overlap closely, a limit, do, is reached, which is known as the diffraction limited resolution. Figure adapted from [59]. .......................................13 Figure 5 A diagram depicting the path taken by light in the epi-fluorescence microscope. The light source provides a broad spectrum of light, from which a particular band is selected as it passes through the excitation filter. The light is reflected by the dichroic filter/mirror and is then directed through the objective. The objective focuses the light on the sample surface containing fluorophores. Both fluorescent and reflected light is collected by the objective. The dichroic filter is a long wavelength pass filter, filtering out some of the shorter excitation light. The barrier filter removes any of the excitation wavelengths remaining. The light is then coupled to a detector, a CCD array camera....14 Figure 6 Transmission spectra for an Olympus U-MNG2 filter cube comprised of an excitation filter (BP530-550), an emission filter (BA590), and a dichroic mirror (DM570). This is designed for use with the CY3 fluorophore. Data taken from Olympus website [61]..................................................15 Figure 7 A circuit containing a battery and a parallel plate capacitor. The capacitor consists of two parallel plates of area A, separated by a distance d, filled with a dielectric medium with relative dielectric constant εr; the permittivity of free space, to which it is relative, is εo. When connected to the battery, charge builds up on the plates, producing capacity (equal to εrεoAd-1)..............................17 Figure 8 Model for the double-layer region including specifically adsorbed anions. The Inner Helmholtz Plane (IHP) and the Outer Helmholtz Plane (OHP) are shown in the figure with corresponding positions (x1 and x2) from the metal surface. Potential (Φ), and charge density (σ) are also shown. Below the graph shows variation in potential with distance from the electrode surface. Adapted from Bard [63] and Kolb [62]..............................................................................................................................19 Figure 9 A generic DNA nucleotide structure and the monomer (repeat) unit are shown on the left. To the right are shown the four bases found in DNA, with glycosidic bonds shown which anchor the base through the nitrogen atoms to the deoxyribose sugar. Uracil is also shown for comparison, however it is not present in DNA, but is found in ribonucleic acid (RNA). Hydrogen bonding between the base pairs is shown; guanine forms 3 hydrogen bonds with cytosine, whereas adenine forms 2 hydrogen bonds with thymine............................................................................................................29  vii  Figure 10 The possible permutations of the four base pairs. Note that since AT = TA, there are only six possible. Complimentary pairing is evident in (1) and (2) which represent the standard WatsonCrick base pairing of G-C and A-T. The total widths of the base pairs add up to the same distance and the interfaces are complementary. Mismatches (3) G-T and (4) A-C do not have complementary interfaces showing the inability to hydrogen bond properly. Two pyrimidines (5) are too distant to form hydrogen bonds and are non-complementary while two purines (6) do not have sufficient room to fit together and are non-complementary. ....................................................32 Figure 11 A small section of a DNA duplex containing two hydrogen bonded strands of DNA is shown. The strand on the left runs (top to bottom) from the 5' to 3' end, GACT - guanine, adenine, cytosine, thymine. The complementary strand is aligned opposite to the first strand, from the 3' to the 5' end, and read in 5' to 3' order, the bases are AGTC. .................................................................................34 Figure 12 A schematic depiction of the B-form of the DNA double helix. Base pairs are shown stacked along the axis of the duplex. The diameter of the double helix is 2.0 nm, with a rise of 3.32 nm, or 10 base pairs (bp) per turn.......................................................................................................................35 Figure 13 Three major interactions used with DNA-based sensors: DNA probe-target hybridization for genetic assays, DNA-protein interactions for protein sensing, and DNA interactions with small molecules (usually intercalants) and ions for sensing of smaller analytes and environmental sensing..............40 Figure 14 A schematic outline of the original preparation of ssDNA/MCH SAMs employed by Herne and Tarlov. First, thiol-derivatized ssDNA (1.0 µM) is adsorbed for a specified period of time (15 s to 22 h) in a phosphate buffer (1.0 M KH2PO4, pH 3.8). Adsorption of MCH (1 mM, 1 h) followed. Finally the monolayer was rinsed with deionized water to remove any remaining nonspecifically adsorbed ssDNA. (ssDNA probes shown are dual-labelled)..............................................................42 Figure 15 Hybridization of sufficiently dense DNA monolayers may have multiple modes: (A) the idealized 1 probe : 1 target hybridization. (B) Target bridging between two probe strands. (C) Target bridging between two probe strands induced by a mismatch. Adapted from Levicky (2005) [31].................44 Figure 16 Response of fluorescence to potential induced changes in DNA orientation. DNA is repelled by the negative surface charge resulting from application of a negative potentials to the electrode (left), causing it to stand near-normal to the surface. When a positive potential is applied (right), DNA is attracted to the surface, causing it to tilt toward the surface. As the fluorophore (shown as CY3) is on the terminal end, it will experience the greatest change in proximity to the surface; near the surface, fluorescence is mostly quenched, while further away the fluorescence is considerable.....50 Figure 17 The gold bead is immersed such that the surface of the water is tangent to the top of the bead. This is checked by looking upward through the water, to see the gold bead tangent to its mirror image. .............................................................................................................................................................58 Figure 18 Typical electrochemical cell with working solution and reference electrode in a separate saturated KCl solution. The cell used for ruthenium-DNA measurements with an electrolyte solution contains approximately 5 μM [Ru(NH3)6]3+ and 10mM TRIS buffer. Also used for surface area measurements with a working solution containing KClO4. WE is the working electrode, a DNAcoated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE) in saturated KCl; CE is the counter electrode, a platinum wire. .............................................................................................................................................................60 Figure 19 CV of a gold bead coated with an ssDNA/MCH SAM in a ruthenium solution ([Ru(NH3)6]Cl3, 5 μM in 10 mM Tris buffer). Two ruthenium peaks are observed in each direction (cathodic and  viii  anodic). Integrated areas of the oxidation (Aox) and reduction (Ared) are shown by the shaded areas; these give peak areas in units of charge (Coulombs), which can in turn be used to calculate the surface density of DNA................................................................................................................61 Figure 20 The original design of the spectro-electrochemical cell, used, unless otherwise specified, in the electro-fluorescence experiments. The optical window above the objective is made from a coverglass which is 0.17 mm thick. The 20X objective was used as the default, however a 1.5X internal magnification is possible to make a total 30X magnification if necessary. The filters are mounted in a rotating carousel to allow for interchange of filters during one experiment. Where both fluorescence and electrochemical measurements were preformed, the supporting electrolyte contained 10 mM Tris and 10mM NaBF4. WE is the working electrode, a DNA-coated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE); CE is the counter electrode, a platinum wire. ....................................................................................................................................65 Figure 21 The second generation of the spectro-electrochemical cell is shown schematically. The cell rests on a flange built into the 2.5 cm diameter tube which forms the body of the cell. An aluminum holder is used to station the cell in place above the objective and prevents contact between the glass window and any surfaces. This new cell has allowed for testing the adsorption of 1mercaptohexanol (MCH) in situ, and may be used to image the hybridization of complementary DNA to ssDNA coated surfaces in the near future; the smaller volume will be a particular asset due to the relatively high cost of DNA, and the concentration required for hybridization. WE is the working electrode, a DNA-coated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE) in saturated KCl; CE is the counter electrode, a platinum wire. ............................66 Figure 22 Quantum efficiency of the Kodak KAI-2092 CCD. Graph produced with data taken from SPOT RT specifications [148].............................................................................................................................67 Figure 23 A schematic description of the potential control during the acquisition of images using the image step program. The values depicted would be representative of a typical experiment with an initial potential (Ei) at -25 mV and a final step potential (Ef) at -400 mV with -25 mV increments (Einc) and a base potential (EBase) at 0 mV; the initial and final step potentials can be changed, as can the increment and the base potential; the base potential is independent of the initial and final potentials. In the sequence shown, 5 images (N) are taken at the step potential, and 5 images (M) are taken at the base potential each time; while 5×5 was the most widely used, other integer values of M×N are possible. Given these settings, 165 images were taken in the experiment. A secondary x-axis is shown for time. This is dependent on the interval time from one image to the next; the present scale is calculated based on a interval time of 2 seconds. ..........................................................................70 Figure 24 Diagram describing the coordinate systems used for image stacks. (A) Image stack 540 px by 760 px area and 3 images long. (B) Single pixel, of intensity Px,y,z, with coordinates (x,y,z). (C) The xy-coordinates define the image in two dimensions. (D) The z-coordinate provides the position of the image within the stack and in time, giving a third dimension......................................................71 Figure 25 Look up tables shown as gradients. Below: A greyscale look up table, where white corresponds to high intensity, and black corresponds to low intensity. Above: The “Royal” look up table defined in ImageJ maps values from greyscale to colour....................................................................................72 Figure 26 Images of ss-DNA-CY5/MCH#1 taken during a reductive desorption stepping sequence: 0 mV → -1100 mV, Ebase = 0 mV, 5×5 images, 50 ms exposure, 1000 ms interval time, 445 images. A shows a greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. The average (AVG) intensity of each pixel in the stack is shown in  ix  B, and a contrast enhanced version is shown in C. D and E show the standard deviation (STD) and median value (MED) of each pixel in the stack, which are used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. G and H are the maximum (MAX) and minimum (MIN) intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA, hence the difference, I, is taken to find areas where DNA desorbs. Using only those images for potentials in the fluoroelectromodulation region (0 mV → -500 mV) prior to desorption, values can be obtained for which areas are showing the greatest levels of switching, either by taking a MAX-MIN of the stack, K, to see an absolute change, or by taking the STD divided by the AVG to obtain a normalized switching map. K also shows an outlined ROI obtained by thresholding J to select for those areas which showed lower modulation relative to the AVG intensity. M shows a larger version of C, with chosen regions of interest (ROIs) highlighted in white boxes. Nine ROIs, each 48 μm x 48 μm, were chosen to investigate specific features of modulation and desorption: hotspots (3,4,5,7,8, & 9) of varying intensity were chosen, in addition to smooth (2,6) and extremely dark (1) areas for comparison..........................................................................................................................................76 Figure 27 Potential step sequence applied to ssDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1100 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 50 ms exp, 1000 ms interval. Other step sequences are shown in the Appendix.........................................78 Figure 28 Fluorescence intensities and capacitance for the reductive desorption of ss-DNA-CY5/MCH (0 mV → -1100 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages over each Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 26. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 50 ms exp, 1000 ms interval.............................................79 Figure 29 Images of ss-DNA-CY3/MCH taken during a reductive desorption stepping sequence: 0 mV → -1200 mV, Ebase = 0 mV, 5×5 images, 250 ms exposure, 583 ms interval time, 485 images. A shows a greyscale bright-field image (10 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. The average (AVG) intensity of each pixel in the stack is shown in B, and a contrast enhanced version is shown in C. D and E show the standard deviation (STD) and median value (MED) of each pixel in the stack, which are used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. G and H are the maximum (MAX) and minimum (MIN) intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA, hence the difference, I, is taken to find areas where DNA desorbs. Using only those images for potentials in the electro-fluoroescence modulation region prior to desorption, values can be obtained for which areas are showing the greatest levels of switching, either by taking a MAX-MIN of the stack, K, to see an absolute change, or by taking the STD divided by the AVG to obtain a normalized switching map. K and E also shows an outlined ROI obtained by thresholding J to select for those areas which showed lower modulation relative to the AVG intensity. M shows a larger version of C, with chosen regions of interest (ROIs) highlighted in white boxes. Nine ROIs, each 48 μm x 48 μm pixels, were chosen to investigate specific features of modulation and desorption: hotspots (3,4,6, & 9) of varying intensity were chosen, in addition to smooth (1,7) and darker (2,5) areas for comparison. .............81  x  Figure 30 Fluorescence intensities and capacitance for the reductive desorption of ss-DNA-CY3/MCH (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 29. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. The Estep capacitance curve is trunctuated due to excessive H2 evolution. Imaging settings: 200 ms exp, 853 ms interval..................................................................................83 Figure 31 Reslice image of the desorption the desorption sequence (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) for ss-DNA-CY3/MCH along line AB which is drawn on the MAX-MIN image on the right. Regions C and D display different desorption potentials, with C desorbing approximately 100 mV before region D. ...................................................................................................................85 Figure 32 Switching of ssDNA-CY3/MCH. Data from two sequential stepping sequences, first going negative (0 mV → -400 mV, Ebase = 0 mV, 5×5 images, 250 ms exp., 583ms/image), then going positive (0 mV → +400 mV, Ebase = 0 mV, 5×5 images); a dotted line marks the starting points. Change in fluorescence intensity (Δcps) is shown, measured from the minimum value; fluorescence intensity of the first sequence was decreased by 260 cps (average) to remove the offset, likely due to photobleaching. Capacitance shown calculated using an estimated area of 0.1 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right). ........................86 Figure 33 Images of dsDNA-CY3/MCH taken during a reductive desorption stepping sequence: +400 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 200 ms exposure, 1000 ms interval time, 695 images. (A) Greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. (B) AVG intensity of each pixel in the stack. (C) Reslice along line NO, shown in (M); desorption region shown. (D) STD of stack pixels and (E) MED value of each pixel in the stack; (F) Used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. (G) and (H) are the MAX and MIN intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA; (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN using only those images for potentials in the electro-fluorescence modulation region (+400 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD divided by the AVG for the same region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 48 μm x 50 μm, were chosen to investigate specific features of modulation and desorption: hotspots (1, 2, 5) of varying intensities, moderately intense (4, 6, 7, 8, 9) and extremely dark (3) areas for comparison. .............................89 Figure 34 Fluorescence intensities and capacitance for the reductive desorption of ds-DNA-CY3/MCH (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 33.Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 200 ms exp, 1000 ms interval................................................91 Figure 35 Images showing a substrate-dependent shift in the desorption potential. A shows a fluorescence image before desorption (-725 mV) shows that although in the region near the yellow line GH does show an inhomogeneous distribution of fluorescence, the "blotches are continuous, not dependent on the underlying surface, as image B (-775 mV) shows that the area in the red triangle has been  xi  fully desorbed while adjacent regions have not been fully desorbed. C and D show contrast enhanced versions of A and B, which clearly identify the angular region which has been completely desorbed. The line GH perpendicularly intersects the edge of the region; a reslice taken along the line is shown in E. The desorption region is expanded in F. It is clear from the intensities that one side immediately desorbs, while the other requires more time. ........................................................92 Figure 36 Switching of dsDNA-CY3/MCH. Data for a sequential stepping sequence (+400 mV → -400 mV, Ebase = 0 mV, 5×5 images). Change in fluorescence intensity (Δcps) is shown, measured from the minimum value (400 mV); fluorescence at Ebase changes with time, likely due to loss of DNA or photobleaching, and has been adjusted here to give a constant fluorescence at Ebase; original data is shown in Figure 37. Capacitance shown calculated using an estimated area of 0.1 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right). ........93 Figure 37 Original data for dsDNA-CY3/MCH; Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown in Figure 36.....................................................................................................................94 Figure 38 Analysis of the time dependence of the fluorescence at Ebase; Estep values not shown. Graph shows fluorescence intensity relative to the initial average (first 5 measurements) of fluorescence intensity. Fluorescence in dark areas (A, open diamonds) shows a parabolic change in intensity, with the final intensity greater than the initial. Fitting with a parabolic curve gave y = 6.823×10-7t2 – 2.019×10-4t+0.9955 (R2 = 0.617). Bright areas (B, black circles) show a more linear decrease in intensity with time, fit linearly with y = 5.882×10-5t + 0.9990 (R2 = 0.916). Initial intensities were: Io,A = 14148 cps; Io,B = 163759 cps. For areas of intermediate intensity, the Ebase intensity curve tended to intermediate shapes; more intense regions producing more linear curves. Regions (A) and (B) are shown (right) on a MAX-MIN, ce, image of the surface................................................94 Figure 39 Images of dsDNA-CY5/MCH taken during a reductive desorption stepping sequence: 400 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 750 ms exposure, 1095 ms interval time, 695 images. (A) Greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. (B) AVG intensity of each pixel in the stack and (C) contrast enhanced AVG. (D) STD of stack pixels and (E) MED value of each pixel in the stack; (F) Used to find areas which show high intensity during desorption by dividing STD by MED. (G) and (H) are the MAX and MIN intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA; (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN, gaussian blur 2px, ce, using only those images for potentials in the electrofluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD divided by the AVG for the same region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 50 x 50 pixels, were chosen to investigate specific features of modulation and desorption: hotspots (1,3,5,7,8) of varying intensity, moderately intense (2,4,5,9) and extremely dark (6) areas for comparison. ...........................................................................................................97 Figure 40 Fluorescence intensities and capacitance for the reductive desorption of ds-DNA-CY5/MCH (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 39. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 750 ms exp, 1095 ms interval. ..............................................99  xii  Figure 41 Images of MCH/ssDNA-CY3 taken during a reductive desorption stepping sequence: 0 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 2500 ms exposure, 2835 ms interval time, 525 images. (A) Greyscale bright-field image (15 ms exp, ce) of the electrode surface. All other images are colourmapped using the "Royal" look-up table shown in (L) with the analogous linear greyscale gradient, and have been background subtracted to remove light leakage. (B) AVG intensity of each pixel in the stack; (C) shows the same image contrast enhanced with a 2 px gaussian blur (gb) applied to reduce speckling. (D) STD, ce of stack pixels and (E) MED value of each pixel in the stack, smoothed by taking the mean of every pixel in a 2 px radius; (F) Used to find areas which show high intensity during desorption by dividing STD by MED; appears rather noisy but homogeneous. (G) and (H) are the MAX, ce, and MIN, ce, intensities of each pixel in the stack; large MIN values result from background subtraction resulting in greater relative noise in dark regions. (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN, ce, using only those images for potentials in the electro-fluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD ÷ AVG, ce, for the same potential region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 48μm x 48μm, were chosen to investigate specific features of modulation and desorption: hotspots (5, 6, 7) of varying intensity were chosen, in addition to moderately intense (1, 2, 8, 9) and extremely dark (3, 4) areas for comparison. .....102 Figure 42 Fluorescence intensities and capacitance for the reductive desorption of ds-DNA-CY5/MCH (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 41. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 2500 ms exp, 2835 ms interval. An artifact at -275 mV resulted from a momentary write-delay by the computer, resulting in overexposure, hence the data point was interpolated; refer to the Appendix for the complete data set..........................................104 Figure 43 Analysis of a desorbed hotspot moving. (A) shows a MAX-MIN of the MCH/ssDNA-CY3 sample; a bright, diagonal line extends from the middle top to the middle left edges of the image. This line shows the path taken by one desorbed hotspot as it moves across the surface during desorption. (B) shows line FG, in red, superimposed along the path of the hotspot shown in (A); this line in the runs antiparallel to the electric field vector, shown in green. (C) shows the stack resliced along line FG; yellow arrows point to the path traced along the line by the desorbed hotspot, moving with time/image number and potential. An expansion of the desorption region is shown in (D), to emphasize the movement that the particle makes with time. E shows the look-up table, "Cool Blue", used to colour the resliced image in order to highlight the particle's path. .........................105 Figure 44 Switching of MCH/ssDNA-CY3. Data for a sequential stepping sequence (+400 mV → -400 mV, Ebase = +400 mV, 5×5 images). Change in fluorescence intensity (Δcps) is shown, measured from the minimum value (400 mV); the eighth data point was removed and interpolated to correct for a large overexposure due to computer error. Capacitance shown calculated using an area of 0.1153 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right). Exposure time = 1500 ms; interval time 1845 ms...............................................................106 Figure 45 An illustration of three comparative free energy curves for adsorbed 5'hexylthiol DNA and MCH on gold. The transition state (‡) for the purpose of the experiment represents either the free site with no thiolate and both DNA and MCH in solution, as would be found at desorption, or some transition state through which the thiol(ate)-DNA replaces the thiolate MCH with ligand exchange. Both processes would have similar coordinates, however the magnitude of ΔG between the transition state and the adsorbed states would differ. The values for the associated reduction  xiii  potentials are merely representative of the change..........................................................................108 Figure 46 Images of MCH/dsDNA-CY3 taken during a reductive desorption stepping sequence: 0 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 1500 ms exposure, 1835 ms interval time, 525 images. (A) Greyscale bright-field image (15 ms exp, ce) of the electrode surface. All other images are colourmapped using the "Royal" look-up table shown in (L) with the analogous linear greyscale gradient, and have been background subtracted to remove light leakage. (B) AVG intensity of each pixel in the stack; (C) shows the AVG in the pre-desorption region (0 mV → -500 mV). (D) STD of stack pixels and (E) MED value of the stack. (F) STD ÷ MED, despeckled to reduce noise, ce. (G) and (H) are the MAX and MIN intensities of the stack; a broad range of MIN values result from background subtraction resulting in greater background noise at low intensities. (I) MAX minus MIN, ce, shows areas where DNA desorbs. (K) MAX minus MIN, ce, using only those images for potentials in the electro-fluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; an ROI is superimposed on this. (J) STD ÷ AVG, ce, for the same potential region (0 mV → -500 mV) to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 50 x 50 pixels, were chosen to investigate specific features of modulation and desorption: hotspots (1, 2, 3, 4) of varying intensity, moderately intense (5, 8, 9) and extremely dark (6, 7) areas for comparison. ...........................................................................................................................................................111 Figure 47 Fluorescence intensities and capacitance for the reductive desorption of MCH/ss-DNA-CY3 (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 46. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 1500 ms exp, 1835 ms interval. ..........................................112 Figure 48 Switching behaviour of MCH/dsDNA-CY3 (+400 mV → -400 mV, Ebase = +400mV, 5×5 images). Image (A) shows the MIN of the stack, image (B) shows the MAX, ce, and (C) shows MAX minus MIN, ce, with an ROI separating the bright areas (R1) from the dark (R2). The absolute change in fluorescence intensity (Δ cps) is given for R1 (red), R1 (blue), and the whole image (ALL, green). Below the potential step sequence and the measured capacitances are given, scaled with time. Images are not background-subtracted. ...........................................................................................113 Figure 49 Left: a round bead in the holder show and a flat-bottomed bead in the same holder. Right: Seven gold beads immobilized in an epoxy resin. The beads were oriented standing on a surface of modelling clay (plasticine) and the area enclosed by a 2.5 cm long section of plastic tubing, 2.5 cm in diameter and filled with epoxy resin. The resin was allowed to set for 2 weeks, and the modelling clay was removed using soap and a toothbrush followed by chipping off the plastic tubing. ..............................................................................................................................................120 Figure 50 (A). FRET experiment where dsDNA containing a donor (D) dye and an acceptor (A) dye is immobilized on the substrate, the FRET fluorescence measured, and the duplex denatured. (B). FRET experiment where a ssDNA probe with donor (D) is immobilized, fluorescence measured, and then the surface is hybridized in situ or ex situ, and the FRET signal measured. (C). A diagram showing the overlap integral J(λ) for the spectral overlap between the donor fluorescence and the acceptor adsorption, on which the efficiency of FRET depends. ....................................................122 Figure 51 Bottom: A schematic depiction of the correlation between distance and fluorescence intensity for fluorophores near bulk metal surfaces. Quenching close to the metal surface shows a d-3 dependence. Top: Depictions of DNA of various lengths labelled with a fluorescent dye (F). Several monolayer assemblies could be constructed, each with a different DNA chain length in  xiv  order to construct a quenching-distance curve. Desorption of the DNA would provide a means to find the maximum fluorescence, providing a reference point to measure against bulk solution measurements for the same fluorophore. .........................................................................................125 Figure 52 Potential step sequence applied to ssDNA-CY3/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 250 ms exp, 583 ms interval..........................................................................................................................137 Figure 53 Potential step sequence applied to dsDNA-CY3/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (+400 mV → -11e00 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 200 ms exp, 1000 ms interval............................................................................................138 Figure 54 Potential step sequence applied to dsDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 750 ms exp, 1095 ms interval............................................................................................139 Figure 55 Potential step sequence applied to MCH/ssDNA-CY3. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 2500 ms exp, 2836 ms interval........................................................................................................................140 Figure 56 Potential step sequence applied to ssDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 1500 ms exp, 1835 ms interval........................................................................................................................141 Figure 57 Illustration of the timing events in during the collection of a series of images. A is the interval time or period (the inverse of frequency). A is the sum of two components: B, the exposure time, and C, the wait time. The exposure time, B, is the length of time during which light is collected for a given image. Each image is followed by a wait time, C, during which the image is saved and light is not collected. C has a constant sub-component, D, the image write time, which is the time required for the computer to write the file to disk. This has been measured to be slightly less than 350 ms, as is shown in the illustration. The wait time, C, cannot be less than the image time, D. In the generic example shown, the exposure time, B, is 500 ms, while the wait time, C, is 1000 ms. This gives an interval time, A, of 1500 ms, which is a frequency of 40 images per minute. ...............................142 Figure 58 Exposure time vs. measured time interval. Graph used to determine the image "write time". Image sequences were collected using interval times set to zero, with exposure times ranging from 50 ms to 3000 ms. The measured interval time is the sum of the exposure time and the write time. This method gave a write time of 337.8 ms. ...........................................................................................144 Figure 59 Measured interval time for a constant exposure with varying wait time. Graph used to determine the image "write time". Image sequences were collected using set interval times ranging from 0 to 3000 ms, with a constant exposure time of 500 ms. The measured interval time is the sum of the exposure time and the write time, hence only where set interval times are greater than this sum will the measured interval time be equal to the input interval time; those with lesser interval times show a constant measured interval time of exposure time (500 ms) plus the write time (336.6 ms). ....145 Figure 60 Cyclic voltammogram of five polycrystalline gold beads in 100 mM KClO4 at room temperature;  xv  20 mV/s. ...........................................................................................................................................148 Figure 61 CV of the double layer region for five polycrystalline gold beads in 100 mM KClO4 at room temperature; 20 mV/s. .....................................................................................................................148 Figure 62 Differential capacitance curves for polycrystalline gold beads; capacity (μF) values plotted against potential (mV). In 100 mM KClO4 at room temperature; 5 mV/s. ................................................149 Figure 63 Differential capacitance curves for the five polycrystalline normalized to unity at -800 mV. Curves in the negative region show considerable consistency.....................................................................149 Figure 64 Digital microscope image of a gold bead with 0.5 mm gold wire. The lighter circle in the background is etched metal on a slide used for reference, and has a diameter of 2500 μm which was used to calibrate each measurement..........................................................................................150 Figure 65 Correlation between surface areas measured for polycrystalline gold beads using electrochemical and optical (geometric) methods. The two area measurements were shown to be related by Area(electro.) = 1.1582*Area(optic.)+0.4981 with a correlation coefficient of 0.9978..................150 Figure 66 Grafting density of dsDNA as a function of the NaCl concentration in the immobilization buffer (10 mM Tris, pH ~ 7.5). Electrochemically measured (red, squares, long dashes) and estimated using UV-Visible absorbance at 260 nm (green circles, short dashes) of the following adsorption of dsDNA on the bead surface. Curves shown are merely visual guides.............................................153 Figure 67 Electrochemical and UV-Vis absorbance methods compared using the data from Figure 66; electrochemical measurements given on the y-axis and UV-Vis results on the x-axis. Excluding the results at 1000 mM NaCl (after iteratively re-weighting using inverse residuals), the results appear to show an strongly linear relationship, with a least squares fit of y = 0.25700x + 8.2457×1011 with a high coefficient of determination (R2 = 0.9997). .........................................................................154 Figure 68 A comparison of capacitance curves for several samples with inset expansions of the switching region before desorption. Graphs on the left side show samples where DNA was deposited first; those to the right show samples where MCH was deposited first. (A) DNA only; (B) ssDNA-CY3/ MCH, ΓDNA ~ 9.2×1012; (C) ssDNA-CY3/MCH, ΓDNA ~ 1.0×1013; (D) dsDNA-CY5/MCH, ΓdsDNA ~ 3.1×1012; (E) dsDNA-CY5/MCH, ΓdsDNA ~ 9.5×1011; (F) MCH only; (G) MCH/ssDNA-CY3, ΓDNA ~ 1.2×1010; (H) MCH/ssDNA-CY3, ΓDNA ~ 2.6×1010; (I) MCH/dsDNA-CY3, ΓdsDNA ~ 3.7×1010; (J) MCH/dsDNA-CY3, ΓdsDNA ~ 2.4×1011. ........156 Figure 69 Curves from imaging done concurrently with CV measurements. Shown are smoothed , scaled fluorescence intensity curves for one cycle (positive → negative → positive; left to right each graph). (A) ssDNA-CY5 only, 5 mV/s, 6.7 mV/image, Δcps = 235; (B) ssDNA-CY5/MCH, 20 mV/s, 11.7 mV/image, Δcps = 980; (C) MCH/ssDNA-CY5/MCH, 20 mV/s, 36.7 mV/image, Δcps = 400; (D) MCH/dsDNA-CY3/MCH, 20 mV/s, 26.7 mV/image, Δcps = 4800; Curves obtained at 20 mV/s were observed to have the same shape at 5 mV/s. ...........................................................157 Figure 70 Comparison of switching observed for ssDNA-CY3/MCH, MCH/ssDNA-CY3, dsDNACY3/MCH, and MCH/dsDNA-CY3. For details, see respective figures. dsDNA-CY3/MCH is shown without any adjustment of Ebase intensity. Changes in fluorescence intensities are scaled for comparison..................................................................................................................................158 Figure 71 Fluorescence intensity (given in arbitrary units) as a function of time. After approximately 100 seconds, MCH was injected into the cell at OCP. Changes in fluorescence intensity are complicated  xvi  by the photobleaching or reaction of the CY5 dye. Images (A-G) are contrast enhanced; the blurriness of B, C, and D is due to desorption. ...............................................................................159  xvii  List of Symbols and Abbreviations Latin Symbols a a A A Abead AE Ao Aox Ared Aw AC Ar Au(111) AVG b b bET BA BP bp c C C Cd CD Cdl CH Cseries co ce CCD CE Comp-LN1 Comp-LN5 cps CV CY3 CY5 d  unit cell length semiaxis a adenine, one of the bases in DNA area surface area of a gold bead capacitance determined area optically determined area integrated area of oxidation peak (C) integrated are a of reduction peak (C) cross-sectional area of wire alternating current argon gas the (111) crystallographic face of gold image given by average stack intensity linear spacing between charges on a polyion semiaxis b Förster energy transfer barrier filter band pass filter base pair semiaxis c cytosine, one of the bases in DNA capacitance monolayer capacitance diffuse layer capacitance double layer capacitance Helmholtz capacitance capacitance of a series of capacitors. concentration contrast enhanced charge-coupled device counter electrode complementary nucleotide sequence to LN1 complementary nucleotide sequence to LN5 counts per second cyclic voltammetry / cyclic voltammogram cyanine dye with 3 methine linkers cyanine dye with 5 methine linkers separation distance in a parallel plate capacitor xviii  d  d  dfl dr DC DM DNA ds dsDNA dsDNA/MCH dspk E Ebase Einc Ef Estep eV eo efr F F FRET G gb GCS h Hz H2 i ia ic iim ire I Io If(t) IB1 IB5 IHP IPE J(λ) k K kf  metal-fluorophore separation distance dipole moment operator dipole moment diffraction limited resolution differential capacitance dichroic mirror deoxyribonucleic acid double stranded double stranded DNA dsDNA adsorption followed by MCH passivation despeckle potential base potential potential increment final step potential step potential electron volt elementary charge (1.602176487×10−19 C) electron surface roughness factor farad Faraday constant (96485.33 C/mol) Förster (fluorescence) resonance energy transfer guanine, one of the bases in DNA Gaussian blur Gouy-Chapman-Stern height Hertz hydrogen gas current (amps) anodic current cathodic current imaginary current real current charging current initial intensity fluorescence intensity at time t immobilization buffer 1 immobilization buffer 5 inner Helmholtz plane ideally polarizable electrode spectral overlap integral Boltzmann constant Kelvin, unit of temperature fluorescence rate constant xix  ki LN1 LN5 m M MAX MCH MCH/dsDNA MCH/ssDNA MED MIN mol MUDH mV n n N N NA NA nm Nsteps N2 O2 OD OCP OHP p PCR pH pKa Pt px Px,y,z pzc q Q qM qS r r r R R R2  a deactivation pathway rate constant 30mer nucleotide sequence with 5'-thiol and 3'-CY3 30mer nucleotide sequence with 5'-thiol and 3'-CY5 number of nucleotides per strand number of images taken at base potential each time image given by the maximum intensity of a stack 6-mercapto-1-hexanol MCH passivation followed by dsDNA adsorption MCH passivation followed by ssDNA adsorption image given by median intensity of a stack image given by minimum intensity of a stack mole (amount of substance) 11-mercapto-1-undecanol millivolt number of carbons in a linear n-alkane chain number of images in a stack number of deactivation pathways number of images at step potential each time numerical aperture Avogadro's number nanometre total number of potential steps nitrogen gas oxygen gas optical density open circuit potential outer Helmholtz Plane a geometric constant = 1.6 polymerase chain reaction acidity; negative log of the hydrogen ion activity negative log of the acid dissociation constant platinum pixel intensity of a pixel at coordinates (x,y,z) in a stack potential of zero charge charge total charge charge on the metal surface charge in solution radius electronic coordinates electronic coordinate operator nuclear coordinates rotated correlation coefficient xx  RE rms RNA ROI rpm RSH RSSR' Ru S SAM SAMs SCE SNPs ss ssDNA ssDNA/MCH STD t T T Tm Tris TCEP UV V Vac W WE x x1 x2 Xe Z z  reference electrode root mean square ribonucleic acid region of interest revolutions per minute (alkyl) thiol (dialkyl) disulphide ruthenium, in reference to [Ru(NH3)6]3+ surface area of an ellipsoid self assembled monolayer self assembled monolayers saturated calomel electrode single nucleotide polymorphisms single stranded single stranded DNA ssDNA adsorption followed by MCH passivation image given by standard deviation of a stack time temperature (Kelvin) thymine, one of the bases in DNA melting temperature of DNA tris(hydroxymethyl)aminomethane, a buffer tris(2-carboxylethyl)phosphine hydrochloride ultraviolet (light) volt AC voltage amplitude (in mV rms) watt working electrode position along an axis normal to metal surface distance from metal surface to IHP distance from metal surface to OHP xenon number of formula units per unit cell electrolyte charge  Greek Symbols β Γ ΓDNA ΓdsDNA ΓssDNA Δcps ΔG ΔS  energy transfer constant surface excess concentration surface concentration (or grafting density) of DNA surface concentration (or grafting density) of dsDNA surface concentration (or grafting density) of ssDNA change in counts per second change in Gibbs free energy change in multiplicity (selection rule) xxi  ε εr εo θ κ κ-1 λ μm μF ξ π π σ σi σd σM σS τ τf τup τdown φf Φ Φo Φ1 Φ2 ΦM ΦS ψof ψoi ψofε ψoiε ψofυ ψoiν ω  dielectric constant relative dielectric constant permittivity of free space or electric constant fraction of a polyion's charge compensated by counterions inverse Debye length Debye length wavelength (nm) micrometre or micron microfarad dimensionless measure of linear charge density electrons in π-bonds pi, the mathematical constant, 3.14159... surface charge density charge density at the inner Helmholtz plane charge density in the diffuse layer excess charge density on the metal excess charge density in solution observed lifetime of a fluorescence decay natural lifetime of a fluorophore time constant for DNA “standing up” time constant for DNA “lying down” quantum yield potential potential drop across the double layer potential at the inner Helmholtz plane potential at the outer Helmholtz plane potential at the metal surface potential in bulk solution wavefunction of excited (final) state wavefunction of ground (initial) state electronic wavefunction of excited (final) state electronic wavefunction of ground (initial) state vibrational wavefunction of excited (final) state vibrational wavefunction of ground (initial) state frequency  Numerical and Miscellaneous Symbols 3' 5' 3D 32 P ‡ Ǻ  hydroxyl terminus of DNA phosphate terminus of DNA three dimensional a radioactive isotope of phosphorus transition state Angstrom (10-10 m)  xxii  Acknowledgements A number of people whose contributions have assisted me in this project deserve many thanks for the help that they have rendered me. May none be overlooked. First and foremost, I owe a debt of gratitude to my co-supervisors, Dan Bizzotto and Hogan Yu, whom accepted me, a wandering graduate student, into their labs. Their guidance and support enabled the success of this project. Dan, I truly appreciate the guidance and enthusiasm you have provided; you are a paragon of patience. Hogan, you have demonstrated to me how to think critically about a problem as a scientist – and to solve it. My thanks are due to my colleagues in the at UBC and SFU over the past year. In the Bizzotto lab at UBC, these include Dr. Robin Stoodley, Aya Sode, Sabrina Lorenz, Jannu Casanova, Hardy Zietsman, and Olivia Yu. And in the Yu lab at SFU these include Alan Cheng, Dr. Yunchao Li, Hidehiko Asanuma, Jeff Jiang, Dr. Madhvi Ramnial, Dr. Juan Shi, Richard Popoff, Lily Ou, Guoyu Gao, Andrew Chou, Kevin Kam, Manu Pallapa, Marcus Kuikka, Dinah Soolaman, and C. W. Joe Wang. My express thanks to Alan Cheng for his help the initial stages of the project, and for instructing me in the art of DNA manipulation; it was a pleasure to work alongside you. Robin, I both appreciated and enjoyed the thoughtful discussions, not to mention the technical troubleshooting! Also I thank Aya, Jannu, Sabrina, Yunchao, Hideh, and Jeff for the wisdom and encouragement shared along the way. All of you have been an enormous source of support, encouragement, and laughter. I would be remiss not to thank those former colleagues whom, during trying times, gave cheer and support to me: thank you Amy, Guillaume, Insun, Vivian, Tyler, and Jennifer. To all my colleagues, keep pressing on! UBC's chemistry department is has several talented and helpful technical staff whom have aided me, particularly in the mechanical shop. Our glassblower, Brian Ditchburn continues to amaze me with his mastery of the craft and interpretations of my schematics. To my friends beyond the lab, I thank you for the fellowship that we have shared and how you have refreshed my spirit. The thoughts and prayers of those in Campus for Christ at UBC and SFU, along with St. Simon's and Maple Ridge Baptist have sustained me. And to my family, Mom, Dad, Dave, and Mike, and all the rest, your love and kindness have nurtured this young scientist and supported him along the way as he wondered, “Why?”. I would not be here today were it not for you, and I hope that my work honours you. xxiii  Dedication This thesis is dedicated to Ben and Sheron Thompson, whom have treated me as their son. The hospitality, kindness, friendship, and love that they have showed to me these years in Vancouver has been a blessing. For this, they have my everlasting appreciation  But we have this treasure in jars of clay to show that this allsurpassing power is from God and not from us. We are hard pressed on every side, but not crushed; perplexed, but not in despair; persecuted, but not abandoned; struck down, but not destroyed... Therefore we do not lose heart. Though outwardly we are wasting away, yet inwardly we are being renewed day by day. For our light and momentary troubles are achieving for us an eternal glory that far outweighs them all. So we fix our eyes not on what is seen, but on what is unseen, since what is seen is temporary, but what is unseen is eternal. II Corinthians 4:7-9,16-18 (TNIV)  xxiv  1 Introduction This thesis deals with the structure, characteristics, and advancement of self assembled DNA monolayers tethered to a gold substrate via thiolate-gold bonds. Such DNA monolayers form the technological basis of DNA-based sensors, which through processes of molecular recognition are capable of detecting a variety of analytes, foremost among which are specific sequences of DNA. Better understanding of the surface and interfacial regions of these monolayer structures is necessary for the fruition of DNA-based sensors. Hence the motivation underlying this is the desire to practically change lives of persons for the better through improvements in screening and testing methods alongside the knowledge that can be gained through the use of such biosensors. DNA is at the core of our being and forms an element of our individuality. Each individual human being has a unique genome, differing from all others most frequently through single base pair changes in the sequence of each gene (with respect to others in the population), termed single-nucleotide polymorphisms (SNPs) which occur, on average, once ever 1900 base pairs [1], although other alterations are present. SNPs can have effects that are beneficial, deleterious, both, or neither; where these effects are harmful, it is often termed a genetic disease. The ability to detect and diagnose patients with particular SNPs portends a time when doctors routinely test patients for particular SNPs and determine their medical treatment accordingly [2]. Recent studies have highlighted the genetic basis for variation in human response to drugs such as beta blockers having no effect on certain heart failure patients [3, 4] or how certain athletes do not excrete testosterone when doped [5, 6]. Both have been demonstrated to have genetic causes. The consequence is a need for the means to quickly, accurately, and cheaply diagnose an individual's genetic state. DNA-based biosensors offer one solution to this problem of detecting monogenic diseases, and is equally applicable to the detection of more common “multifactorial” genetic diseases [7]. Detection of mutations, SNPs specifically, has already been demonstrated with DNA-based sensors [8, 9]. Still, questions exist regarding the extent to which DNA is influenced by interaction with the environment [10], and here again DNA-based 1  biosensors have utility as a means to study the interaction of DNA with proteins [11], biomolecules [12], and other molecules such as drugs [13-15] and even small molecules [16, 17] to better understand their influence. DNA-based biosensors in the form of self-assembled monolayers (SAMs) of DNA have been developed and extensively studied, providing valuable insights and information with application to genetics, cellular and molecular biology, toxicology, self-assembly, and other fields. Several electrochemical DNA-based biosensors have been reported in the literature, and have, in some cases, been coupled with fluorescence. Our interest lies with thiolate tethered DNA monolayers on gold, as devised and characterized by Tarlov, Herne, and Steel [18-23]. This basic motif has been widely emulated and expanded upon by others, given the relatively facile preparation and quantification of probes, and remains widespread given the ability to electrochemically couple such monolayers for sensor applications and to create miniaturized, patterned substrates for it using gold and other noble metals on a wide variety of supports including glass and semiconductors. There remains however a need for better characterization of DNA SAM interfaces; the sensing performance and hybridization of DNA at SAM DNA surfaces is strongly dependent on the surface structure [23-25]. Studies of this nature have been limited and much remains to be understood regarding the interfacial molecular architectures of DNA SAMs used in sensors. It has been shown through theoretical simulation of the surface structure of double stranded DNA that duplex strands tend to favour interactions with neighbouring strands in a colloid-like behaviour through salt-induced shielding of the phosphate groups [24, 26, 27]. Implied in this is the possibility that DNA aggregates may form before DNA probe attachment to the surface [23, 26]. Furthermore, even electrochemical interrogation has provided suggestive evidence that DNA SAMs are not as homogeneous as previously thought [28]. This may explain part of the variability and generally low yields in studies of hybridization [23, 25, 29-32]. It is important to note that few studies of the DNA SAM electrode surface have provided significant spatial resolution [33].  2  1.1 Objectives This research has been a collaboration of the Bizzotto lab at the University of British Columbia in Vancouver, British Columbia and the Yu lab at Simon Fraser University in nearby Burnaby, British Columbia. The Bizzotto lab's interests lie in the spectro-electrochemical characterization of biological and bioactive compounds deposited onto well-defined electrode surfaces as monolayers using in situ electrofluorescence microscopy, which has been used to understand drug interactions [34, 35], study lipid monolayers [36-38], and characterize the electrochemical desorption of thiolate SAMs on gold [39]. The Yu lab has interests in the design, fabrication, and testing of modified surfaces, with particular inclination toward the creation of biosensing surfaces [15] using patterning [40, 41] of SAMs and immobilized DNA [42, 43]. Together the two labs possess the tools and techniques required for the thorough investigation of DNA SAMs on gold electrodes to better understand the organization at a microscopic level. The intention of this project has been to investigate DNA SAMs prepared on gold substrates in order to investigate what heterogeneity may be present and to develop a means to produce more uniform DNA SAMs. Using electro-fluorescence microscopy, a symbiosis of epifluorescence microscopy with electrochemical methods using a specialized spectroelectrochemical cell [36, 39, 44], gold beads with immobilized, fluorescently labelled, single stranded DNA (ssDNA) and mercaptohexanol (MCH) were imaged. This provides a spatially resolved delineation of the surface distribution of the ssDNA using fluorescence. With the imposition of various potentials, the fluorescence changes in response to the orientation of the DNA due to quenching by the gold surface as investigated by Rant et al., measured using a fibre-optic sensor [45-50]. This information is corroborated and extended using the voltammetric response of electrostatically bound [Ru(NH3)6]3+ during its reduction to [Ru(NH3)6]2+ to determine the surface concentration of DNA on the electrode. With the finding that DNA SAM coated electrodes display a significant level of heterogeneity, we undertook the task of finding an means of eliminating the heterogeneity (e.g. sonication) or designing an alternate means of surface preparation (i.e. ligand exchange). 3  1.2 Rationale The continued difficulty of controlling the adsorption of DNA to metal and semiconductor surfaces in order to prepare SAMs for biosensing applications and the constituent heterogeneity has led to variations in the preparation of DNA SAM surfaces most frequently with entirely different substrates, although these too have their own inherent disadvantages. Non homogeneous surfaces is a probable explanation for the already mentioned issues with DNA hybridization of target strands with immobilized probe sequences. Little conclusive evidence exists, however, for the exact nature of any such structures. The possibility of aggregates [23, 26] or other structures present on the surface raises the possibility of having different fluorescence responses due to movement of the assemblies in response to an electric field and the effect of fluorophore quenching near the surface. The fluorescence electromodulation switching response of single and double stranded, fluorescently labelled DNA has been extensively described by Rant et al. [45-50] and would provide a backdrop against which to compare the behaviour of surface aggregates. Later a fresh approach was implemented to create DNA monolayers from gold electrodes pre-passivated with MCH was devised. It is centred on the principle of thiol/thiolate ligand exchange [18, 51-53] and the association of DNA with the surface [54]. Passivation of the surface with MCH limits the opportunity for DNA to aggregate directly on the gold surface [19, 26], while permitting the specific adsorption of thiols to the protected metal surface [51]. These principles were combined yielding homogeneous DNA SAMs.  1.3 Scope of Thesis The interdisciplinary nature of the research embodied in this thesis requires a correspondingly broad background to the fundamental theories required for a sufficient understanding theories encompassed. These are covered in Chapter 2, and include the nature and structure of DNA and SAMs, in addition to the electrochemistry and spectroscopic imaging used to characterize the surfaces. First, an introduction to fluorescence microscopy is provided 4  including details of fluorescence quenching near the metal surface. Next, a concise overview of pertinent electrochemical concepts including theories for the electrode|solution interface follows. This leads to a discussion of thiolate SAMs, including their preparation and characteristics, as is required for DNA SAMs. However before DNA SAMs can be discussed, a brief background in the structural characteristics of DNA is presented. DNA SAMs are the final subject of review, concentrating on work stemming from the methodology of Tarlov et al. [18, 19, 23] and study of the characteristics of the interfaces formed. Chapter 3 provides an overview of the experimental methodology used throughout the project. Included is a summary of the procedures used for sample preparation as well as the materials, methods, and instrumentation employed. Summaries of both electrochemical methods and image collection used in situ with the spectroelectrochemical cells are described. In Chapter 4, experimental results for the spectroelectrochemical characterization of ssDNA/MCH samples prepared in accordance with the methodology of Tarlov and Herne [18, 55] are analyzed, followed by those of double stranded DNA (dsDNA/MCH). Used here is the reductive desorption of the thiolate monolayers from the surface along with the fluorescence switching that results from electromodulation. This is contrasted with the results for samples prepared using a revamped procedure whereby MCH is used first to passivate the surface, followed by immobilization of single stranded (MCH/ssDNA) or double stranded DNA (MCH/dsDNA). Further experimental findings and results are provided in the Appendices. Chapter 5 makes manifest the conclusions, summarizing them and giving them a proper order and context. First in observation that DNA SAMs as presently prepared are heterogeneous and that with a revised method of preparation that this shortfall may be avoided. To conclude, Chapter 6 provides direction to future work in the study of such interfaces using spectroelectrochemistry.  5  2 Theoretical Background 2.1 Fluorescence Microscopy This section will provide a general description of fluorescence followed by a discussion of fluorescence quenching near metal surfaces. It will conclude with a description of fluorescence microscopy. Bizzotto and Shepherd are the main sources upon which this section is based [56, 57], although other standard references were consulted [58, 59]. 2.1.1 Electronic Transitions and the Franck-Condon Principle Molecular luminescence represents the emission of energy from excited state molecules via a radiative pathway as they return to the ground state. Such electronic transitions occur where the molecular absorption of a photon results in the emission of a photon of longer wavelength – that is, one of lower energy. Light is a form of electromagnetic radiation, of which photons are the quantum unit. The photon interacts with the electronic structure inducing a change in the dipole moment of the molecule which results in an electronic excited state. The probability of this transition is given by 2   oi 〉∣ ∣〈 of ∣d∣  (1)  Where ψof and ψoi represent the wavefunctions of the excited (final) and ground (initial) states of the molecule; d is the dipole moment operator. The transition dipole moment is given by the integral  io 〉 =d fl 〈 of ∣d∣  (2)  which is proportionate to the spectroscopic intensity of the transition. The transition is forbidden when d fl is zero, and allowed when non-zero. Furthermore, the transition must involve excited and ground states of the same multiplicity, hence the selection rule, ΔS = 0; transitions where this selection rule is broken can occur, however the probability is quite low.  6  The Franck-Condon principle is used to estimate the intensity of electronic transitions. The principle is, in short, the approximation that electronic transitions are most likely to occur when the positions of the nuclei do not change significantly; formulated quantum mechanically, the intensity of the transition is proportionate to the square of the overlap integral of vibrational states involved in the transition. This is because electrons can be redistributed more rapidly than nuclei due to the difference in mass. The dipole moment can thus be quantified as  〈 of  r ∣r∣oi  r  〉 〈of   R∣ oi   R〉 =d fl  (3)  where ε and ν identify the electronic and vibrational states respectively. The first part gives the overlap between electronic states, while the second part gives the overlap between the vibrational waveforms of initial and final states. Here, r represents the electronic coordinates, while R represents the nuclear coordinates. 2.1.2 Fate of Excited State Molecule Numerous possible relaxation pathways exist for a given molecule in an electronic excited state to transition to the ground state. For fluorescence to occur, the emission of a photon, with a rate of emission, kf, occurs. As this is only one of the possible pathways for relaxation, only a certain fraction of excited molecules will fluoresce. This fraction is the quantum yield (φf), which describes the probability of the process is given by equation .  f =  kf N  ∑ ki  (4)  1  For such an excited state, there are N deactivation pathways, each with a rate constant, ki. The natural lifetime of a fluorophore (τf) is given by the inverse of kf.; in the absence of competing paths, this lifetime would be observed for the overall decay of the excited state. As the fluorescence is dependent on the number of molecules in the excited state, the decay rate observed for the fluorescence (τ) will be that of the excited state molecules, and the result of all of the N deactivation pathways. Hence quantum yield can be defined as 7  f =   f  (5)  Fluorescence intensity as a function of time is dependent on the mean fluorescence lifetime of a single fluorophore. This is given by −t  I f t=I f 0 e   (6)  where If(t) is the intensity at some time, t, after the fluorophore is excited with a short pulse of light. The decrease in the mean lifetime (τ) results in a decrease in the the intensity of fluorescence. A Jablonski diagram describes possible radiative pathways for relaxation in Figure 1. Ground state (A) and excited state (B) have the same multiplicity, whereas C has a different multiplicity. An incident photon of an appropriate frequency promotes one electron to a vibrational level in the excited electronic state. Intermolecular collisions relax the vibrational energy to the ground vibrational energy level of the excited state. Relaxation of this electron to the ground state with the emission of a photon is fluorescence. Energy lost in the relaxation of the vibrational excited state results in a lower energy emitted photon, which is known as the Stokes shift. In cases where the photon isn't emitted following vibrational relaxation, it may undergo intersystem crossing to a triplet state, which subsequently relaxes to the ground state via the emission of a photon, which for triplet to ground transitions is termed phosphorescence. Typical radiative relaxation pathways such as fluorescence and phosphorescence result in the emission of radiation, however non-radiative pathways also exist. Due to the additional energy present in excited state molecules they are highly reactive relative to the corresponding ground state molecules. Where emission photons is sufficiently slow, chemical reaction rates may compete with fluorescence. This can result in photobleaching or photodecomposition, resulting in the continuous loss of fluorophore and intensity with exposure time.  8  Figure 1 A Jablonski diagram showing the electronic excitation and emission transition processes. Electronic states A and B have the same multiplicity but C does not. Incident radiation excites electrons from the ground state of A to excited states of B, which then relax to the ground state followed by fluorescence emission as they transition to state A, or they may undergo intersystem crossing, to a third state, C, which may undergo relaxation by phosphorescence.  Quenching of the molecule can also limit the fluorescence of a given molecule. It refers to several processes that can include energy transfer, complex formation, excited state reactions, and collisional interactions, hence it frequently shows a strong dependence on factors such as temperature, concentration, impurities, and oxygen concentration. 2.1.3 Fluorescence Near Metal Surfaces Near metal surfaces fluorescent dyes are quenched primarily through an energy transfer process whereby the excited state non-radiatively decays via transfer of energy from the excited 9  state to the metal surface; the process is a form of Förster energy transfer, hence metal and the dye need not be in contact for the process to occur. The process shows an inverse cubic dependence of the decay rate constant with separation distance of the dye from the metal surface, calculated as b ET = d  −3  (7)  This relation holds when energy transfer is to sufficiently thick metals; thinner metals show a quenching dependent on d-4. At distances from the electrode comparable to the wavelength of the emitted light, constructive and destructive interferences resulting from reflection of emitted radiation on the metal surface. As already mentioned, this energy transfer process is closely related to Förster resonance energy transfer (FRET). FRET is a non-radiative relaxation process where energy is transferred from a donor molecule to an acceptor molecule through a dipole-dipole interaction over long ranges. For this to occur, there needs to be a sufficient spectral overlap of the acceptor and donor absorption and emission bands as shown in Figure 1.  Figure 2 Overlap between donor fluorescence emission spectrum and acceptor absorption spectrum, with the overlapping region giving the overlap integral J(λ). 2.1.4 Carbocyanine Dyes Cyanine dyes are widespread in their use in fluorescence applications. They are generally characterized as having large quantum yields, high extinction coefficients, and can be chemically tailored to absorb and emit in a variety of wavelengths. The basic structural motif of cyanine dyes is a polymethine linker in the form of X(-CH=CH)n-CH=Y, which bridges two 10  substituents, X and Y, which are typically nitrogenous aromatic groups. The length of the polymethine linker is the primary factor influencing the absorption wavelength maximum. Two commonly used dyes, known by their trade names, CY3 and CY5, are shown in Figure 3. The two dyes are comparable in most respects, however CY5 is reported to be more sensitive to environmental ozone than CY3 [60].  Figure 3 Molecular structures of CY3 and CY5. CY5 contains 5 methine groups in its linker, while CY3 contains 3. CY3 absorbs maximally at 550 nm, and emits maximally at 570 nm (φf = 0.15). CY5 absorbs maximally at 649 nm, and emits maximally at 670 nm (φf = 0.28).  2.1.5 Fluorescence Microscopy Fluorescence microscopy is a relatively mature technology, having decades of use in the study of biological systems. Its primary use, historically, is determination of the spatial distribution of fluorescent dyes or fluorescently labelled species within a biological matrix such as a cell, a membrane, or tissue sample. Difficulties with the environments presented by such samples have led to an abundance and variety of different fluorophore, however controlled environments where quenchers and oxygen are eliminated can lead to improved imaging subjects. This section will provide a short summary of the principles involved in fluorescence microscopy.  11  2.1.6 Microscope Resolution Microscopes use a collection of lens elements to magnify a specimen of interest. Magnification is achieved through the objective, which creates a magnified image of the specimen in the intermediate image plane. The lenses of the eyepiece bring the magnified image into focus for the viewer in the same manner as a magnifying glass. The total magnification is the product of the objective and eyepiece magnifications. The clarity of the image is dependent on the ability of the optical components to collect light that is refracted, diffracted, or emitted from the specimen [58]. Collection of light is achieved through objectives having a large collection angle, or a large numerical aperture (NA). Large NA objectives enable higher resolution images by limiting diffraction. Diffraction results in the inability to distinguish two adjacent points from which light is emanating. The maximum resolution is defined as the ability to resolve two “Airy discs”. Airy discs are the circular diffraction patterns of light passed through a narrow slit after propagation through the objection. Light passed through the slit no longer appears as a sharp, bright spot, but instead as a slightly blurred spot encircled by diffuse diffraction rings, as shown in Figure 4. As the discs begin to overlap, a limit is reached where two bright with a valley between them is displayed by the profile. The Raleigh criterion calculates the diffraction limited resolution (dr) based the wavelength (λ) of light propagating through the objective as  d r=  1.22  2NA  (8)  This is the maximum resolution achievable by the microscope [59]. 2.1.7 Epi-Fluorescence Microscope Fluorescence microscopy functions on the principle of illuminating a fluorophore with visible or ultraviolet light with simultaneous collection of fluorescence using a detector such as a charge coupled device (CCD) similar to those used in most digital cameras. Given the low quantum yields of many fluorophores, the fluorescence intensity is quite weak relative to the excitation light, and hence it must be separated from the fluorescence wavelength by filtering. 12  Figure 4 An illustrative representation of an airy disc (left) and two overlapping Airy discs. Far apart, Airy discs are easily resolved, however when they begin to overlap closely, a limit, do, is reached, which is known as the diffraction limited resolution. Figure adapted from [59]. This is achieved in the inverted epi-arrangement shown in Figure 5 which uses a “filter cube” containing three light-filtering elements: First, an excitation filter, toward which UV/visible light is directed, transmits a limited band of wavelengths (depicted by the green arrow proceeding from the filter) in the fluorophore's absorption region. Second, the dichroic mirror (or beam splitter) reflects short wavelengths of light (i.e. the excitation wavelength) to the objective, where it is focused on the sample. Fluorescence (depicted by the red arrow) is collected by the objective which directs it through the dichroic mirror which passes longer wavelengths (i.e. the emission wavelength) to the third element. Last is the barrier filter which transmits only in a narrow range in order to filter out any light that has leaked through the dichroic mirror, collecting only the fluorescence. The transmission spectra of each element in a typical configuration (Olympus U-MNG2 filter set [61]) are shown in Figure 6. Following the isolation of fluorescence, it is directed to a coupler that connects the CCD detector to the microscope, focusing and expanding the light to fill the entire CCD array, utilizing all the pixels.  13  Figure 5 A diagram depicting the path taken by light in the epi-fluorescence microscope. The light source provides a broad spectrum of light, from which a particular band is selected as it passes through the excitation filter. The light is reflected by the dichroic filter/mirror and is then directed through the objective. The objective focuses the light on the sample surface containing fluorophores. Both fluorescent and reflected light is collected by the objective. The dichroic filter is a long wavelength pass filter, filtering out some of the shorter excitation light. The barrier filter removes any of the excitation wavelengths remaining. The light is then coupled to a detector, a CCD array camera. 14  Figure 6 Transmission spectra for an Olympus U-MNG2 filter cube comprised of an excitation filter (BP530-550), an emission filter (BA590), and a dichroic mirror (DM570). This is designed for use with the CY3 fluorophore. Data taken from Olympus website [61].  15  2.2 Electrochemical Concepts and the Electric Double Layer 2.2.1 Qualitative Description of the Electric Double Layer Ions in bulk solution exist in a spherically symmetric environment that is electrically neutral within a given region of sufficient size. Near a boundary, such as an electrode surface however this is not the case. Ions near a charged electrode | solution interface experience electrostatic forces which are not present in the bulk of solution. A large excess of ions can accumulate near the electrode surface to counterbalance the surface charge. Dipolar solvent molecules near the electrode interface may organize with a preferred orientation resulting in the induction of a net dipole across the interface. It is these features, charges and oriented dipoles, in the region from the metal to the bulk of solution, that constitute the electrical double layer. 2.2.2 Ideally Polarizable Electrodes Because the interface between solution and a metal electrode is of fundamental importance to electrochemistry, we begin here, with the simplest possible case: Ideally polarizable electrodes are electrochemical systems where an electrode | solution interface has a potential range wherein no charge transfer reactions occur, that is, there are no reduction and no oxidation reactions, due to kinetic or thermodynamic unfavourability; any cations or anions present are inert and unreactive, while no appreciable decomposition of the solvent takes place. For aqueous systems, this is limited by either oxygen or hydrogen evolution at the positive and negative ends of the potential range. Within this potential window, such an electrode system is termed an “ideally polarizable electrode” (IPE). A characteristic example would be solid silver in aqueous potassium fluoride. Some small non-faradaic currents may arise due to structural and material changes which can occur at the surface, such as the adsorption or desorption of molecules from the interface. Within the potential window, the interface between electrode and solution is analogous to a simple capacitor: charges may build up, but no charge is transferred. 2.2.3 Simple Capacitors Capacitors store energy in the form of the electric field between separated bodies of charge. A simple, parallel plate capacitor is shown in Figure 7. Capacitance (C, in farads (F)) is 16  a function of the distance (d) separating two parallel plates and the relative dielectric constant (εr) of the dielectric medium between the plates. An applied voltage (E, in volts (V)) results in the accumulation of equal and opposite charges on the plates (q, in Coulombs (C)) in the form of either an excess of electrons resulting in a negative surface charge, or a deficiency of electrons giving a positive charge. The charge will accumulate according to the equation C=   r o q = d E  (9)  The term εo is the permittivity of free space (8.854×10-23 F·m-1). Variation in the potential will induce a charging current (I, in amps (A)) I=  dq dE =C dt dt  (10)  This is referred to as the capacitive or charging current, and is different from current due to electron transfer. A circuit of n capacitors in series will have a total capacitance given by 1 1 1 1 =  ⋯ C series C 1 C 2 Cn  (11)  These relations are required when describing the electrochemical characteristics of a real electrochemical system within the IPE potential region.  Figure 7 A circuit containing a battery and a parallel plate capacitor. The capacitor consists of two parallel plates of area A, separated by a distance d, filled with a dielectric medium with relative dielectric constant εr; the permittivity of free space, to which it is relative, is εo. When connected to the battery, charge builds up on the plates, producing capacity (equal to εrεoAd-1). 17  2.2.4 Capacitance of the Electrode | Solution Interface At a given potential, a charge may exist at the metal surface (qM). This results in a build-up of ions of opposite charge in solution (qS) which act to counterbalance the charge on the metal surface. The separation of charge results in the double-layer capacitance (Cdl) which is a function of potential. Theories describing the electric double layer have evolved over time, starting Helmholtz, followed by Gouy and Chapman, and presently Gouy, Chapman, and Stern which provide the most complete but simple picture of the double layer as shown in Figure 8. Helmholtz proposed that the interface was analogous to a simple capacitor; the charge on the metal surface was counterbalanced by a thin layer of charge in solution in a manner as that of a parallel plate capacitor. Gouy and Chapman proposed that the charge in solution was distributed diffusely over a region near the surface since the conductivities of metals versus those of solutions are different. To account for this, they introduced a diffuse layer of ionic charge on the solution side of the plane of the Helmholtz model. Stern later refined the model by accounting for the finite size of ions: the distance of the closest approach of ions to the surface would be the ionic radius after the loss of any solvation shell, precluding a layer solvent molecules more strongly adsorbed to the surface. This gives an inner layer between the metal and the diffuse layer proposed by Gouy and Chapman. The solution side is divided into several regions as shown in Figure 8. The inner layer exists close to the electrode surface and also contains neutral solvent molecules and specifically adsorbed ions or molecules. Specific adsorption is when chemical interaction of the adsorbate with the metal surface is stronger than the electrostatic repulsion between the adsorbate and the metal surface. The inner layer is comprised of the inner Helmholtz plane (IHP) and the outer Helmholtz plane (OHP). The IHP lies at position x1 from the metal surface (x = 0) and is defined by the locus of the closest approach of the electrical centres of the specifically adsorbed ions. The potential and charge density at the IHP is given by Φ1 (V) and σi (μC·cm-2) respectively. Solvated ions can approach the metal surface only to a distance x2, defined as the closest approach of the electrical centre of the nearest solvated ion; this may differ depending on the degree of solvation of the surface (cf. [62, 63]). This qualitatively defines the OHP;  18  Figure 8 Model for the double-layer region including specifically adsorbed anions. The Inner Helmholtz Plane (IHP) and the Outer Helmholtz Plane (OHP) are shown in the figure with corresponding positions (x1 and x2) from the metal surface. Potential (Φ), and charge density (σ) are also shown. Below the graph shows variation in potential with distance from the electrode surface. Adapted from Bard [63] and Kolb [62]. 19  quantitatively, the potential at point x2 is given by Φ2. The diffuse layer extends from the OHP to the bulk solution. This layer contains the nonspecifically adsorbed ions in solution and has a charge density σd., from the excess ionic charge up to bulk solution. The sum excess charge density on the solution side of the interface (σS) near a given electrode is  S = i d =− M  (12)  where σM is the charge density on the metal. The double layer capacitance is made up of two capacitors in series (the inner and diffuse layers) as shown by the Gouy-Chapman-Stern (GCS) theory. 2.2.5 The Gouy-Chapman-Stern Theory (Quantitative) Here, a summary of the GCS theory, completely derived in Bard and Faulkner [63], is given. The model treats the electric double layer in the absence of specific adsorption. The solution is divided into thin layers (laminae) of equal thickness starting at the OHP and extending into the bulk electrolyte. The laminae represent the population of ionic species with varying energies (with respect to a reference laminae in the bulk). The total charge per volume of any lamina is calculated using the Poisson-Boltzmann equation describing the gradient of potential from the OHP into the bulk of solution. This is represented as    d dx    8kTn o =− r  o x= x 2  1 2    sinh  z e o 2 2kT    (13)  where k is the Boltzmann constant (1.380×10-23 J·K-1), T is the temperature (K), no is the number concentration of the ions in the reference lamina (m-3), z is the signed charge of the ion, eo is the elementary charge (1.602×10-19 C), and Φ is the electrostatic potential (V). A potential plot in the diffuse layer can be accomplished by integrating Equation 10 between the limits of Φ and Φ2 to obtain     ze  4kT e− x− x = (14) ze 2 tanh 4kT where κ represents the inverse Debye length. The inverse Debye length is dependent on the tanh  2    20  ionic strength of solution, and can be calculated from    2 no z 2eo2 = r o kT    1 2  (15)  The inverse of κ is the Debye length representing the diffuse layer thickness, locating the maximum excess charge density. Using no=coNA where co is the concentration and NA is Avogadro's number (6.022×10-19 mol-1) in water (εr = 78.49) at 298 K, the Debye length of sodium in an aqueous solution of 10 mM NaCl is 3.04 nm; for a 1 M solution it would be 0.304 nm. For more dilute solutions, the Debye length, and hence the diffuse layer thickness increases. In the absence of specific adsorption, only water is present in the inner layer and the charge density at any point from the electrode surface to the OHP is zero. Therefore, the total potential drop from the metal surface (Φo) across the double layer in the absence of specific adsorption is given by      o=2−  d dx  x=x s  x2  (16)  The distribution of potential across the double layer can be visualized in Figure 8. From the metal to the OHP, the potential drops linearly, then from the OHP into bulk solution it decreases exponentially A relation between the charge density on the metal and Φo can be obtained using a Gaussian enclosure about the surface of the metal and Gauss' Law to obtain   M =− S =r o    d dx     (17)    (18)  1 o  x= x2  =8 k T r o n  2 sinh  ze 2 2kt  Substituting Φ2 from Equation 9 provides 1 o 2   M =8 k T r o n  sinh      x ze  o− M 2 2kt r  o  21  The double layer capacitance can be found by differentiating σM with respect to Φo. Inverting and simplifying, which yields  x 1 1 = 2 1 2 2 o C dl o 2 r o z e n 2 ze  o cosh kT 2kT        (19)  This formulation of the double layer capacitance resembles the series capacitance for two capacitors, hence it can be simplified to give  1 1 1 =  C dl C H C D  (20)  The double layer capacitance (Cdl) is a function of the inner layer or Helmholtz capacitance (CH) and the diffuse layer capacitance (CD). The inner layer capacitance is independent of potential and resembles a simple capacitor. The double layer capacitance is dominated by the smallest capacitor in the series. For low electrolyte concentrations, Cdl has a minimum at the potential of zero charge (pzc) because the capacitance of the diffuse part of the double layer is small. The minimum capacitance occurs at the pzc (where σM = 0 and therefore Φ2 = 0), since no excess of counter ions is required in the diffuse double layer to balance the charge on the metal. Measuring the capacitance for low electrolyte concentrations of a nonspecifically adsorbing electrolyte is a method for determining the pzc of a particular metal | electrolyte interface. The measured capacitance can be used to determine the Helmholz capaity, CH, through varying the electrolyte concentration and plotting Cdl-1 against the inverse of the calculated diffuse layer (or Gouy-Chapman) capacitance, CD-1, and finding the intercept, CH-1. When the slope of the Parsons and Zobel plot does not result in a straight line, it is indicative of specific adsorption [64].  22  2.3 Thiolate SAMs and Desorption 2.3.1 Introduction Self-assembled monolayers (SAMs) have wide application in materials chemistry and in nanoscience. As materials, they modify the surfaces and interfaces such that the surface will have different physical or chemical properties. Nanoscience is also the realm of surfaces and interfaces, since such materials inherently have higher surface to volume ratios. Additionally, atoms and molecules in the interfacial environment experience a different environment from the bulk or solution, and thus have different free energies, electronic states, reactivities, mobilities, and structures, and it is these different properties on which nanomaterials depend [51]. Adventitious organic materials will readily adsorb to the bare surfaces of metals, as the adsorbates lower the free energy of the metal|solution (or atmosphere) interface. Such organic layers however are not well defined. SAMs are regular, two-dimensional arrays of molecules formed by the adsorption of molecular constituents from solution (or the gas phase) onto the surface of a solid or liquid (e.g. mercury) substrate covalently. The organization of the adsorbates is spontaneous, forming crystalline or semicrystalline structures. SAM-forming molecules have a chemical functionality, a "ligand" or a "headgroup", that with a (high) specific affinity for the substrate, and can frequently displace any adventitious adsorbed molecules. While many different headgroups for forming SAMs are known, the most extensively studied are thiols which covalently adsorb to noble and coinage metals (e.g. gold, copper, palladium, mercury, etc...) [51]. A wide range of substrate-headgroup combinations are known, however thiols will be our focus. Such interfaces have different interfacial properties than those of the bare metal, which are in turn dependent on the adsorbed molecules. SAMs typically have thicknesses in the range of 1-3 nm, furthermore, phase-separated regions can result in heterogeneity or patterning with dimensions from 10 - 100 nm laterally. SAMs of organic molecules can act as a physical or electrostatic barrier and to decrease the reactivity of surface atoms. Furthermore, SAMs can, 23  depending on their structure and functionalization, couple external environmental conditions to the optical, electronic, and other physical properties of the metal surface. Certain functional groups can enable environmental sensing and molecular recognition, as in the case of DNA SAMs. 2.3.2 Preparation Methods SAMs are most frequently prepared via immersion in an aqueous or liquid solution containing the desired ligand [23, 51]. The substrate is immersed in the solution and the ligand is adsorbed either passively or electrolytically. Other methods commonly employed include Langmuir-Blodgett trough adsorption and vapour phase adsorption including ultra high vacuum (UHV) methods. 2.3.3 Gold Substrates The supporting physical material on which the SAM is formed is known as the "substrate". Gold, in combination with alkane thiols, is the most widely studied and remains the most widely used substrate metal for the formation and study of SAMs [51]. Although single crystal substrates are useful for certain studies, polycrystalline materials are sufficient for many applications and are much more economical. Gold is also a relatively inert metal, not reacting with most chemicals nor oxidizing with atmospheric O2. The high binding affinity of thiols on gold and the absence of side-reactions lead to stable SAMs. This is advantageous for biological studies, as gold-thiol SAMS can last for days to weeks in solutions containing buffers used in biological studies [23]. 2.3.4 Metal-SAM Interface Where an underlying ordered metal surface is present, thiolate monolayers at high coverage will adopt ordered arrays. On Au(111), n-alkane thiols form a (√3×√3)R30° overlayer (R = rotated) structure (Z = 4), while on Au(100) a c(2×2) overlayer is adopted [51, 65]. The unit cell lengths are given with respect to the underlying metal lattice structure (e.g. Au(111), a  24  = 2.88Å). The structure of the SAM evolves slowly during immersion, as the number of pinhole defects decreases and the number of conformational defects in alkane chains decrease, over a period of days. Alkane chains in SAMs adopt tilted conformations that permit high degrees of van der Waals interactions and hydrogen bonding with neighbouring molecules in order to minimize their free energy. The contributions from the interaction of the alkyl chains give a segmental heat of interaction of 1.0 to 1.5 kcal/mol per methylene group [51, 66] . This results in a secondary level of organization within the SAMs. Immersion time for SAM formation in most experiments is 12-24h which can limit order in the monolayer, leaving defects. Defects can result from (1) variations in the surface structure of the substrate (e.g. polycrystalline gold) in the form of grain boundaries, atomic steps, and a variety of crystal faces present; (2) reconstruction of the surface by the adsorption of thiols; (3) impurities in the SAM; (4) desorption of the adsorbate [51]. Formation of an adsorbed thiolate requires the activation of the S-H bond or an S-S bond. The adsorption of dimethyl disulphide on Au(111) from vapour has been found to be reversible, having a desorption energy of ~ 30 kcal/mol (125 kJ/mol) [67], which suggests a significant degree of charge transfer in the formation of the Au-S bonds. In solution, barriers to desorption were lower, at 20-25 kcal/mol (83.7-104.6 kJ/mol) [68]. Bimolecular recombinative desorption of dialkyl disulphides is estimated from experiments to be approximately 15 kcal/mol (62.8 kJ/ mol), however this includes both the heats of dissolution of the adsorbate and the heat of immersion of the substrate in the solvent. Overall, the Au-S bond is relatively strong, with a homolytic cleavage on the order of 50 kcal/mol (210 kJ/mol) [67]. In any case, the recombinative Au-S desorption to produce dialkyl disulphides is not observed at room temperature, however at elevated temperatures, the conversion of surface thiolates to disulphides becomes kinetically feasible. (For alkylthiol-tethered DNA probes, stability is limited above 75°C [23].) Overall, the adsorption and assembly process in solution remains poorly understood due to the complexity of the process. Qualitatively, the adsorption is approximated by a Langmuir  25  adsorption model. For thiols, the hydrogen lost in the adsorption also has an ambiguous fate; it is suspected to undergo either reduction to produce H2 or an oxidative reaction with dissolved O2 yielding water [51]. In the case of hydrogen evolution, adsorption of thiols to a gold surface is may be given as: 2 RSH  Au0 surface → 2 RS −Au surfaceH 2  g   (21)  2.3.5 Mixed Monolayers Several methods exist for the preparation of mixed SAMs; monolayers containing a mixture of two or more molecular species in a well-defined fashion. Examples of such methods include: (1) sequential adsorption where adsorption of the first component only leads to low or partial coverage of the surface, (2) coadsorption from solutions containing a mixture of alkylthiols (RSH + R'SH + ... ), (3) adsorption of asymmetric disulphides (RSSR'), (4) partial or selective electrochemical desorption to create monolayer defects followed by adsorption of another thiol [51]. Such surfaces are desirable for a number of reasons, however in the present application, the primary reason is to have a passivated metal surface with a mixture of probe molecules interspersed. 2.3.6 Displacement of SAMs by Exchange Adsorbed thiolate molecules can exchange gradually over periods of minutes to hours in solutions containing thiols or disulphides. In general, this process does not yield a homogeneous nor uniform SAM, however it can provide a route to generate novel organic surfaces from a preexisting SAM. Rapid replacement occurs at grain boundaries, defects, and regions of disorder. Replacement in dense, crystalline regions is slow; exchange can require several days. Hence SAMs on rough, metallic films containing less order and more defects provide faster exchange rates. The rate of exchange depends on multiple parameters, including chain lengths of the SAM and the adsorbate, the degree of order present in the SAM, and the surface roughness. For nalkane thiols, short chains (n < 12) are displaced more rapidly than longer ones; this enables selective exchange of one component. In studies of nanoparticles functionalized with alkane 26  thiolate ligands, it has been found that the alkane thiolates bound at vertices and edges exchange at a higher rate than those in the denser, better-packed planar faces. Rates of exchange are also dependent on steric bulk and chain length of the initial SAM; greater steric bulk and longer chain lengths lead to lower rates of exchange. Molecules with multiple thiols can chelate the metal surface which limits exchange reactions from taking place [69-73]. 2.3.7 Electrochemical Desorption Upon application of a sufficient negative potential to the supporting metal in an appropriate electrolyte, thiols undergo reductive desorption into solution; the electrochemical half-reaction for alkane thiolates adsorbed on gold metal is: RS −Aue - → RS -  Au0  (22)  Typically, for electrochemical desorption, SAMs are immersed in an aqueous solution at pH ≥ 7. (In this thesis, they are immersed in a buffered solution at pH ~ 7.5.) During the desorption process, the thiol and the bare metal surface become solvated as the thiolate diffuses away into solution. Removal of the applied negative potential can induce oxidative readsorption of the thiolates to the metal surface, hence the process can be reversible. Desorption begins at grain boundaries, then at random nucleation sites within well-organized regions; the highest desorption rates occur at defects and grain edges. This parallels the locations in SAMs with the greatest propensity to exchange. The desorption potential for alkane thiolates is typically around -1000 ± 250 mV (versus Ag|AgCl, KClsat), and is dependent on the degree of ordering, chain length, intermolecular interactions, and crystallinity of the substrate. The variation in desorption potential for different thiolates indicates that selective desorption is possible, and has indeed been used to modify SAMs [74-77].  27  2.4 DNA Essentials 2.4.1 Introduction DNA is a unique substance. It can be formed in a myriad of possible permutations. It can encode information, acting both a repository and a transmitter of genetic information in the biological world. DNA is the only biomolecule able to undergo self-duplication, acting as a template for its own replication, in addition to encoding the information for all other functional elements of a cell. DNA is perhaps the most fundamental constituent of all living organisms. And as such it has become a molecule of great renown. Given the ubiquity of DNA in western culture, a certain step backwards must be taken so as to understand a few essential details of its structure and function, in particular as they relate to this work. On a cultural level, DNA is viewed as a blueprint, since it contains nearly all of the biological information that cumulates in life. DNA is also understood to act as an identifier, being able to distinctly identify specific genes, phenotypes, species, and even individuals. But DNA is rarely viewed as a chemical with distinct structure and physiochemical properties. One author describes DNA as "the chemical of heredity" [78] and in this thesis that is how it must be viewed, as a chemical. 2.4.2 Structure of DNA DNA is an acronym for deoxyribonucleic acid, a nucleic acid polymer composed of repeat units called nucleotides. Each nucleotide consists of a sugar, 2'-deoxyribose, a phosphate group, and a base which forms the monomer unit, as shown in Figure 9. Shorter polymers of nucleic acids are called oligonucleotides while longer sequences are dubbed polynucleotides. DNA is an acid because of the phosphate groups of the nucleotides. Dialkyl phosphates are strong acids and have a pKa ~ 1, hence at physiological, near-neutral pH's, effectively all of the phosphates are deprotonated and ionized (negatively charged) [79], however much of the charge is compensated by ions in solution which are closely associated with the phosphates [80-84]. Successive monomer units in the polymer are connected via phosphodiester linkages. The phosphate group is attached to the hydroxyl group on the 5' carbon of one sugar unit and the 3' hydroxyl of the next, forming a repeating chain alternating sugar, phosphate, sugar, phosphate, 28  and so on. This is referred to as the "backbone" of DNA [79, 85]. While such phosphodiester bonds could theoretically be formed by a dehydration reaction between sugar and phosphate eliminating water, the free energy of such a reaction is quite positive at +25 kJ/mol. Such an equilibrium would lie far to the hydrolyzed side of the phosphodiester bond. However, these linkages are exceedingly unreactive in aqueous solutions, having spontaneous hydrolysis rates at near-neutral pH on the order of ≈ 2 × 10-13 s-1, or a half life around 140 000 years [86]. Therefore, DNA is a metastable compound - thermodynamically unfavoured in formation, but kinetically too slow to react unless catalyzed. Nature and synthetic chemists both use nucleoside triphosphates to couple the formation of DNA with a more thermodynamically favourable reaction, giving an overall negative free energy for the reaction enabling the polymerization of nucleotides to form DNA.  Figure 9 A generic DNA nucleotide structure and the monomer (repeat) unit are shown on the left. To the right are shown the four bases found in DNA, with glycosidic bonds shown which anchor the base through the nitrogen atoms to the deoxyribose sugar. Uracil is also shown for comparison, however it is not present in DNA, but is found in ribonucleic acid (RNA). Hydrogen bonding between the base pairs is shown; guanine forms 3 hydrogen bonds with cytosine, whereas adenine forms 2 hydrogen bonds with thymine. Each nucleotide monomer in the chain carries a heterocyclic base attached at the 1' carbon of the sugar via to a nitrogen in the base via a glycosidic bond. Together, the phosphate connected to the 5' carbon of the deoxyribose sugar via a phosphate ester bond, and the base, 29  connected at the 1' carbon of the sugar, comprise the nucleotide. Four major bases are found in DNA: adenine (A), guanine (G), cytosine (C), and thymine (T). Hence there are four nucleotides present in DNA, listed in Table 1.  Nucleotide  DNA Base  deoxyadenosine monophosphate (dAMP)  adenine (A)  deoxyguanosine monophosphate (dGMP)  guanine (G)  deoxycytidine monophosphate (dCMP)  cytosine (C)  deoxythymidine monophosphate (dTMP)  thymine (T)  Table 1 The bases of DNA.  The bases in DNA are heterocyclic, aromatic compounds capable of hydrogen bonding. Two categories of bases exist: purines and pyrmidines. Adenine and guanine are both purines which have 5 membered imidazole ring fused to a 6 membered pyrimidine ring in a structure, which is a core structure shared with caffeine. Cytosine and thymine are pyrimidines, consisting of a single heterocyclic six-membered aromatic ring, with nitrogen atoms at the 1 and 3 positions. All of the bases are capable of tautomerization between either the keto and enol forms (for guanine and thymine) or their amino and imino forms (for adenine and cytosine). The keto and amino forms are the most stable and hence the prevalent forms and are depicted in Figure 9, however imino and enol forms are present in certain special interactions [79]. The concept of base pairing was not immediately obvious, nor was the idea that DNA could act as a molecular blueprint. Until 1950, DNA was viewed as having a simple homogeneous structure and incapable of carrying encoded information. Chargaff showed that DNA was not a homogeneous substance, but that it exists in a wide variety of chemical configurations [78]; that different organisms contain different proportions of the four bases in their DNA, independent of the tissue from which the DNA was derived: “We arrive at the following conclusions. The deoxypentose nucleic acids from animal and 30  microbial cells contain varying proportions of the same four nitrogenous constituents, namely adenine, guanine, cytosine, thymine. The composition appears to be characteristic of the species, but not of the tissue, from which they are derived. The presumption, therefore, is that there exists an enormous number of structurally different nucleic acids; a number, certainly much larger than the analytical methods available to us at present can reveal…I think there will be no objection to that statement that, as far as chemical possibilities go, they could very well serve as one of the agents, or possibly as the agent, concerned with the transmission of inherited properties.” (Chargaff 1950) [87] Perhaps more importantly, it was found from the various DNA compositions that the proportions of the four nucleotides forming DNA appeared to follow a constant set of rules: 1. The sum of the purines (A and G) equals that of the pyrimidines (C and T) 2. The molar ratio of A to T equals 1 3. The molar ratio of G to C equals 1 4. Consequently the mole fractions and numbers of (A + C) = (T + G) This led to the Watson-Crick base-pairing of DNA which states that adenine pairs with thymine, whereas guanine pairs with cytosine [78, 79]. The chemical reason behind base-pairing lies in hydrogen bonding and steric interactions. Hydrogen bonds will form between complementary base pairs, each pair consisting of one purine and one pyrimidine. Steric constraints of imposed by the structure and dimensions of the helix prevent either two pyrimidines from bonding together or two purines from hydrogen bonding together. As such, two purines would have insufficient space to bond together, while two pyrimidine bases would not be sufficiently proximal to hydrogen bond. Hence, the molecular structure of each of the bases is such that each purine is capable of bonding with only one of the pyrimidines and vice versa given the orientation of attachment within the DNA duplex. This allows the adeninethymine base pair to form 2 intermolecular hydrogen bonds and the guanine-cytosine pair to form 3 such hydrogen bonds [85]. To simplify this, one could imagine each of the DNA bases as keyed blocks of different sizes similar to LEGO®, as shown in Figure 10.  31  Figure 10 The possible permutations of the four base pairs. Note that since AT = TA, there are only six possible. Complimentary pairing is evident in (1) and (2) which represent the standard Watson-Crick base pairing of G-C and A-T. The total widths of the base pairs add up to the same distance and the interfaces are complementary. Mismatches (3) G-T and (4) A-C do not have complementary interfaces showing the inability to hydrogen bond properly. Two pyrimidines (5) are too distant to form hydrogen bonds and are non-complementary while two purines (6) do not have sufficient room to fit together and are non-complementary. 2.4.3 Hybridization and Melting The ability of DNA to replicate, and thus to preserve, replicate, and transmit information, stems from the ability to associate with a strand containing a complementary sequence via hydrogen bonding of the base pairs in a predictable fashion. This also enables sequence-specific recognition of complementary DNA sequences, which has direct application to DNA sensors. In the standard Watson-Crick base pairing of DNA, adenine base-pairs with thymine, whereas guanine base-pairs with cytosine, as already mentioned. There are, in certain DNA structures, other hydrogen bonding interactions possible, but these are not applicable here. These hydrogen bonds found in complementary sequences of base pairs enable association of the strands and allow for formation of double-stranded DNA (dsDNA), as classically depicted by the double helix, if complementary and properly aligned. In dsDNA each of the strands is aligned anti32  parallel to the other as shown in Figure 11. The direction of each strand is recorded with respect to the carbons of the 2'-deoxyribose that bond to the phosphates in the backbone, giving DNA a 5' end and a 3' end. By convention, base sequences are recorded in a 5' to 3' orientation. As such, the 5' end of a polynucleotide terminates in a phosphate group and the 3' end terminates in a hydroxyl group [79]. The sequence of the DNA section shown in Figure 11 is: 5' GACT 3' 3' CTGA 5' Note that even reversed, the sequences are not identical. Furthermore, this sequence could simply be written as GACT. Generally, sequences are non-identical unless the sequences are palindromic, whereby the first base in a given sequence would bond to the last, the second to the penultimate, and so on. However such sequences are capable of forming secondary structures other than the double helix. Once hydrogen-bonded together, two single-stranded DNA sequences form a duplex, which twists to form a double helix. This process is called hybridization. Three type of helices are common: the A-form, the B-form, and the Z-form [79]. Formation of a particular helix is dependent on the sequence and environmental conditions; under physiological conditions the Bform is the most common. The, B-form is a right-handed helix with a diameter of 2.0 nm, a rise per turn of 3.32 nm, and an average of 10 base pairs per turn, is depicted schematically in Figure 12. In the resulting helical structure, the bases project at right angles from each of the strands toward the interior of the helix, while the deoxyribose-phosphate backbone forms the outer edge of the helix. This has the further advantage of separating the negative charges of the phosphate groups lining the backbone to minimize inter-strand repulsion [79]. The base pairs at the centre are completely stacked in the double helix structure, which excludes water and further promotes hydrogen bonding. As the bases are planar, aromatic structures, the packing allows for π stacking, a favourable van der Waals interaction between  33  Figure 11 A small section of a DNA duplex containing two hydrogen bonded strands of DNA is shown. The strand on the left runs (top to bottom) from the 5' to 3' end, GACT - guanine, adenine, cytosine, thymine. The complementary strand is aligned opposite to the first strand, from the 3' to the 5' end, and read in 5' to 3' order, the bases are AGTC.  34  Figure 12 A schematic depiction of the B-form of the DNA double helix. Base pairs are shown stacked along the axis of the duplex. The diameter of the double helix is 2.0 nm, with a rise of 3.32 nm, or 10 base pairs (bp) per turn.  35  the π-orbitals of adjacent bases. This non-covalent overlap of the molecular orbitals above and below the plane of each base is a relatively weak interaction, however the sum of all the π stacking interactions in a double-stranded DNA helix can add a significant amount to the net stabilization energy. However this is secondary to the stabilization energies provided by hydrogen bonding. Hybridization of DNA is a reversible, temperature-dependent process; dsDNA denatures by unwinding and separating into two single strands from heating. The temperature at which the helix denatures is called the melting temperature, Tm, and the process is generally quite abrupt due to a sudden un-zipping of the hydrogen bonded duplex. Specifically, this is the temperature at which half of the base pairs are broken and standard free energy is zero, which correlates to the maximum in the derivative of ultraviolet absorbance with respect to temperature [88]. This functions as a measure of stability related to the binding energy of the duplex. The Tm depends on several factors: sequence length, base composition, and solution composition. DNA rich in GC pairs will have a higher Tm because G-C bases form 3 hydrogen bonds between them, whereas A-T bases form only 2 hydrogen bonds. Because of the extra stability conferred, Tm increases by approximately 4°C for each G-C pair and 2°C for each A-T pair. Solution conditions such as pH, buffer composition, and ionic strength can also influence the melting temperature [23, 78, 85]. High ionic strength stabilizes the duplexes by screening of the electrostatic repulsion of the negatively charged phosphate groups. Other forces counteract the forces bringing single strands of DNA together. The charge repulsion of the anionic backbones of strands contributes significantly to this and enables the reversibility of hybridization which is normally observed. However in cases where the sequences are not perfectly complementary, mismatches, as described in Figure 10, may occur. Through steric interactions and non-complementarity for hydrogen bonding, the helical structure and pi stacking of the dsDNA is disrupted and the duplex is made less stable than perfectly complementary sequences. The position of the mismatch will affect the degree to which a mismatch can affect the stability; like faulty teeth on a zipper, a mismatch near the end of the sequence (< 3 bp from the terminus) will have little effect on the overall stability as they can be  36  left unlocked without affecting the rest of the structure. However, a mismatch near the middle of the sequence (> 3 bp from the terminus) becomes a rupture point during the denaturation of DNA, precipitating its melting at a lower temperature than would otherwise be observed for the complementary sequence. Hence mismatches in DNA show a strong positional dependence with non-terminal mismatches having a greater affect on the Tm of DNA. Annealing DNA to induce hybridization requires control over the "stringency" of the conditions, that is, the careful manipulation of buffer conditions or temperature for the purpose of discriminating between perfectly complementary and mismatched strands. High stringency implies conditions that favour the formation and stabilization of perfect complementary strands, whereas low stringency conditions can permit some base-pair mismatches to be tolerated in a duplex [23, 79]. Temperature control requires slow cooling starting above the Tm inhibits the formation of mismatches, since they are unstable relative to the complementary and properly aligned sequence, so they will tend to dissociate immediately, while the complementary and properly aligned sequence will tend to remain hydrogen bonded. 2.4.4 Information Content Discriminating between sequences with as few as one mismatch is imperative if one is to develop a method of DNA detection and identification or sensing. This is particularly important given the exponential rise in the number of permutations for a polynucleotide of a given length. A single nucleotide has four possibilities: A, G, C, T. A dimer, consisting of a mere 2 nucleotides, can form 16 possible permutations: AA, AG, AC, AT, GA, GG, GC, GT, CA, CG, CC, CT, TA, TG, TC, TT. For a three unit nucleotide there are 64 possibilities. The rise is exponential, increasing as 4n, where n is the number of nucleotides in the chain. For a thirty-mer of DNA, there exists just over 1018 possible combinations, as shown in Table 2. Despite the numerous permutations available, DNA is a perfectly ordered - although appearing random at first glance - heteropolymer of containing four monomer units each with a different aromatic base attached. It is the ordering of these units which encodes data within DNA and that enables differentiation of the sequences for the purpose of molecular recognition for DNA or other analytes [23]. 37  Nucleotides  Possible Combinations  1 2 3 4 5 10 20 30  4 16 64 256 1,024 1,048,576 1,099,511,627,776 1,152,921,504,606,850,000  Table 2 The number of unique sequence combinations possible for polynucleotides of a given length of DNA nucleotides.  In biological systems the information contained in DNA is used primarily for the synthesis of proteins. DNA is read by the cellular machinery much as a computer reads a hard-drive in bits and bytes; here, the DNA sequence is read in three-letter (base) sequences called codons. Each codon codes for a particular amino acid in structure of a protein. Given that there are 64 possible sequences for a trinucleotide grouping, and that there are 20 amino acids in addition to start and stop command sequences, the genetic code is degenerate; more than one sequence can code for a specific protein. The code is redundant, but completely unambiguous as each codon encodes one specific amino acid, and only one in a given organism, however a different codon may also encode for the same amino acid [78, 79, 85]. It is possible then, for different individuals to have different sequences of DNA for a given gene while maintaining identical phenotypes - physical expressions of that gene - as they can encode for the same protein. Such redundancy provides a measure of protection against harmful mutations, but it also results in differences from individual to individual even in those genes which are conserved. Individuals, species, or even genetic diseases can, in some cases, be identified by changes as subtle as a change in one single nucleotide. Such a change is called a single nucleotide polymorphism. The ability to detect such minute changes in the structure of DNA enables its use as an identifier [78] and source of information to humankind.  38  2.5 DNA SAM Sensors 2.5.1 Introduction The ability of DNA to selectively hybridize with a complementary sequence and to bind to certain molecules has long provided a tantalizing paragon of molecular recognition to emulate synthetically and to couple to an electronically measurable signal. The structural and biological properties of DNA and their ability to complex and interact with a variety of analytes, including other biomolecules and proteins have also prompted research into building DNA structures on surfaces. These two projects of science have intersected with the design and creation of DNAbased sensors. The past 15 years have shown an exponential growth in research in biosensor technologies and applications [31]. Advances in the design, creation, and characterization of SAMs have contributed the creation of DNA-based SAM sensors where single nucleotide polymorphisms can be detected, and other real-time kinetic and orientational data on surfacetethered DNA can be acquired, including observation, control, and quantification of hybridization [23, 89]. This section will outline the essential details of DNA SAMs with a brief introduction to their role as biosensors, focusing on their application as genosensors through hybridization, and in aspects of the interfacial molecular structure that control the properties of such DNA-based sensors. While this section will omit much useful data in favour of brevity, several excellent reviews are available [23, 31, 89-93], especially on electrochemical DNAbased sensors [3,4,9-16]. 2.5.2 DNA SAM Sensors Biosensors are devices utilizing biological components and systems such as enzymes, tissues, organelles, whole cells, RNA, DNA, and proteins to detect compounds via a transducible signal [102]. Biosensors generally contain some sensitive biological element which is responsive to a given analyte; for DNA biosensors the biological element is DNA, which has been shown capable of responding to numerous analytes, naturally including DNA itself through hybridization. Genosensors are a subset of DNA biosensors capable of detecting specific DNA nucleotide sequences, however the term is less widely used . Typically, such genosensors consist 39  of an immobilized DNA probe sequence which is able to recognize a complementary nucleic acid sequence, the analyte, or target, through binding via hybridization. This section will focus on DNA-based sensors for use in genetic assays via hybridization, however other configurations are commonly used and in many cases are used in tandem. Three commonly used sensing formats are shown in Figure 13.  Figure 13 Three major interactions used with DNA-based sensors: DNA probe-target hybridization for genetic assays, DNA-protein interactions for protein sensing, and DNA interactions with small molecules (usually intercalants) and ions for sensing of smaller analytes and environmental sensing. DNA-based sensors are closely related to DNA microarrays [103] capable of determination of nucleic acid sequencing however these require highly specialized instrumentation and algorithms for the detection and interpretation of fluorescence data [104, 105]. Instead, we focus on electrochemical biosensors [104] and those where this is coupled with optical methods. An electrochemical DNA sensors are best defined as, “a self-contained integrated device, which is capable of providing specific quantitative or semi-quantitative analytical information using a biological recognition element (biochemical receptor) which is retained in direct spatial contact with an electrochemical transduction element.”[106, 107] Such systems combine sensitivity, selectivity, and low cost for the detection of specific DNA 40  sequences. One drawback of such systems is that PCR (polymerase chain reaction) often required for amplification of the sequence of interest in order to produce a significant signal, however much work as been done advancing lower detection limits, and many envision the possibility of single-molecule detection [23]. Finally, while several sensing and signal transduction schemes have been devised, it is not possible here to give a full description of all such devices and methods although several excellent reviews are available among those already mentioned. 2.5.3 Thiolate DNA/MCH SAMs Given the vast field of data on DNA SAMs and biosensors, the scope of this section of the background and review material must be constrained by two foci: First, as this thesis concerns the assembly of thiolate SAMs of DNA with mercaptohexanol prepared on the basis of the pioneering work of Herne and Tarlov [19, 18, 22, 108], systems of this type are heeded while others, save for where principles still apply, are overlooked. Second, our concern lies largely with the interfacial characteristics of DNA biosensor systems, hence the intricacies of various systems are ignored in favour of physiochemical information on the assembly, homogeneity, organization, and electrochemical responses of such systems. DNA oligomers had previously been synthesized containing a 5' hexylthiol group [109], and other researchers had used thiol-modified ssDNA to study hybridization reactions at solid surfaces [110-113], however it was Herne and Tarlov whom discovered the utility of 6mercapto-1-hexanol (MCH) as a diluent thiol in the preparation of two-component thiolate monolayers containing ssDNA [19, 108]. The thiol spacer molecule was chosen in order to minimize “nonspecific” adsorption of ssDNA to the gold substrate through the nucleotide bases in favour of “specific” adsorption through the gold-thiolate bond. The formation of gold-thiolate bonds by MCH would competitively displace any adsorbed nitrogenous bases; using X-ray photoelectron spectroscopy to compare surfaces with thiolate ssDNA versus unmodified DNA before and after exposure to MCH showed that little DNA remained in the absence of the hexylthiol linker. Hybridization with 32P radiolabelled DNA showed that hybridization occurred in proportion to the surface concentration of ssDNA; surfaces with MCH only showed no 41  measurable hybridization [19]. A schematic representation of the monolayer preparation is shown in Figure 14. An increase in the volume fraction, as measured by neutron reflectance spectroscopy, from ssDNA to ssDNA/MCH, was postulated to represent the desorption of nucleotide bases from the gold surface and concomitant reorientation of the ssDNA strands in support of this model [22]. Later results however suggested that much of the change in the structure of the adsorbed ssDNA was due to desorption of ssDNA through ligand exchange with MCH [18], however fluorescence intensity measurements confirm that reorientation of the layer takes place [114].  Figure 14 A schematic outline of the original preparation of ssDNA/MCH SAMs employed by Herne and Tarlov. First, thiol-derivatized ssDNA (1.0 µM) is adsorbed for a specified period of time (15 s to 22 h) in a phosphate buffer (1.0 M KH2PO4, pH 3.8). Adsorption of MCH (1 mM, 1 h) followed. Finally the monolayer was rinsed with deionized water to remove any remaining nonspecifically adsorbed ssDNA. (ssDNA probes shown are dual-labelled)  Many other systems, in particular those developed for coupling DNA to nonmetallic substrates, have been designed for the purpose of creating DNA SAMs, however thiolate monolayers of ssDNA/MCH on gold remain one of the best, despite certain complications which will later be discussed. Few design attempts have been made into improving the SAM without fundamentally changing its structure. One example employed the creation of holes in an MCH monolayer by selectively desorbing a secondary thiol [115] component to create gaps in the monolayer, followed by refilling with ssDNA [116]. Others have tried coadsorption of ssDNA with MCH [117], however all have resulted in phase-separated domains [115, 118].  42  2.5.4 Hybridization Given the necessity of hybridization to occur in order to obtain sequence-specific information for DNA, its optimization is a logical point of interest. A wide range of hybridization efficiencies have been reported, ranging from near 0% to 100% [23], although a number of studies have reported efficiencies greater than 100% depending on the surface conc of DNA [119]. The efficiency depends on several factors including the ionic strength of the buffer, surface concentration, accessibility of sites, and the DNA sequence. Hybridization has been monitored using several different techniques: surface plasmon resonance [108], electrochemically [20, 120-122], and using fluorescence microscopy [47]. Structurally, hybridization has also been observed to produce a change in the DNA layer thickness as measured by neutron reflectance spectroscopy, believed to result from the mutual repulsion and increased rigidity of dsDNA strands [22], which fluorescence measurements coupled with electromodulation has helped to confirm [45]. Arinaga found that DNA hybridization yield was drastically reduced (< 20%) at higher surface concentrations (ΓssDNA > 2×1012 strands/cm2), however layers diluted (ΓssDNA < 7×1011 strands/cm2) using electrochemical desorption showed yields near 100% as measured with [Ru(NH3)6]3+. High surface concentrations of ssDNA may block target ssDNA from reaching the surface; additionally inhibition of hybridization due to coulombic blockage by charges in the 2D layer may limit hybridization [25]. Additionally, DNA hybridization has been modelled at the solid | solution interface. The model has two routes to hybridization: direct hybridization to the probe from solution and nonspecific reversible adsorption to the surface, followed by surface diffusion and subsequent hybridization. The model suggests that probe surfaces should be designed to promote reversible adsorption and two dimensional diffusion of target strands [54]. The DNA sequence is also of particular importance. Probe sequences containing possible secondary structures with interactions greater than 3 base pairs may completely block target strands [32].Furthermore, kinetic data suggests that secondary structures may result on the DNA SAM surface where by target sequences can bridge two adjacent probes as shown in Figure 15. This may be aggravated where a mismatch occurs near the middle of the target, producing a 43  higher than expected yield [31]. This limits the selectivity of high surface concentration DNA sensors. However some degree of control may be achieved using an applied potential to discriminate between matched and mismatched sequences. Positive potentials tend to favour mismatched pairings, whereas negative potentials can be applied to limit hybridization, although it limits the formation of mismatched duplexes far more than perfectly matched sequences, preferentially denaturing them in a form of “electronic stringency” [123]. It should be noted that this also facilitates the reuse of electrochemical sensors by removing target DNA.  Figure 15 Hybridization of sufficiently dense DNA monolayers may have multiple modes: (A) the idealized 1 probe : 1 target hybridization. (B) Target bridging between two probe strands. (C) Target bridging between two probe strands induced by a mismatch. Adapted from Levicky (2005) [31].  2.5.5 Surface Concentration Control of the surface concentration or grafting density of DNA SAMs has been widely studied. Initially higher surface concentrations were desired given the greater signal intensity [23], however as noted, higher surface concentrations have been determined to have significant drawbacks. Methods applied to control surface concentration of ssDNA include dilution with MCH [18], ionic strength of the immobilization buffer [18, 19, 49, 124], and electrochemical desorption [119]. Current protocols have tended toward immobilization buffers containing lower supporting electrolyte concentrations (~ 50 mM NaCl), however in several instances the DNA is 44  deposited as a droplet on a gold substrate and evaporated to dryness, hence near the end, extremely high salt concentrations may be experienced by the DNA. Nucleotide length also tends to limit grafting density for longer oligomers [18, 124]. Quantification of the surface density of DNA probes is easily achieved electrochemically using chronocoulometric [20, 21] and cyclic voltammetric [42, 43] methods. Electrostatically bound [125] redox-active cations, most commonly [Ru(NH3)6]3+, are intercalated with the DNA, providing quantitative information on the structure; these methods are most widely used quantitative methods, although other methods (e.g. radiolabelling [19]) are available [23]. 2.5.6 Electrochemical Desorption of Thiolate DNA The first report of the electrochemical desorption of thiolate dsDNA (350 bp) was reported by Wang et al. Desorption from gold ultramicroelectrodes at -1300 mV vs. Ag|AgCl (-1345 mV vs. SCE) was found to be complete after 2 to 5 minutes [126]. The desorption of ssDNA (24mer, 3'hexylthiol, 5'CY3) was found by Rant et al. to be dependent on the counterion concentration in solution. Applying successive negative potential steps (0, -200, -400, -600, -800, -1000 mV vs. Ag|AgCl), evolution of the fluorescence signal was monitored with time. Desorption solutions with low ionic strength displayed greater levels of desorption at lower potentials (e.g. -200 mV, -400 mV) whereas those with high ionic strengths showed minimal increase in fluorescence intensity at -800 mV despite no prior desorption. The desorption at multiple potentials at low ionic strength is strongly suggestive of heterogeneity in the binding of DNA to the surface, some having no thiolate tether. The electrochemical desorption of thiolate ssDNA on gold, back-filled with MCH, has been found to occur at -650 mV vs. Ag|AgCl (-695 vs. SCE in 10 mM Tris, 50 mM NaCl, pH 7.3). Square wave pulses to this potential were applied by Arigana et al. to control of the surface concentration of DNA via partial desorption in solutions of MCH [119]. The desorption potential was found to be independent of strand length using 24, 48, 72 and 96mer ssDNA nucleotides with 5'hexylthiol and 3'CY3 labels [119], likely due to counterion condensation [83, 127] limiting the effect of electrostatic forces.  45  2.5.7 Counterion Condensation and DNA Aggregation Manning's counterion condensation theory describes the ionic interactions of long, linear polyions with closely spaced ionic repeating units. Monomeric salts in solution, such as sodium acetate, are completely dissociated in solution, whereas polyacrylate contains a substantial fraction of sodium ions bound to the anionic groups in the polymer. These bound ions are in a condensed state, and compensate the charge of the polyion [82, 83]. The condensation is dependent on the linear charge density, which is inversely dependent on the spacing betwen ionic groups in the polymer (b). For polyelectrolytes of high charge density, counterions will be in a condensed state, closely bound to the polyion, independent of concentration of electrolyte. The fraction of the polyion's charge, θ, compensated by counterions from solution is given by =1−−1  (23)  2  =   z eo  r o k B T b  (24)  where ξ is a dimensionless measure of the linear charge density, z is the electrolyte charge, eo is the elementary charge, εrεo is the dielectric constant of the solvent, kB is the Boltzmann constant, T is the temperature in Kelvin, and b is the linear spacing between charges on the polymer. For an electrolyte with charge z, the formula can be given as  =1−  1 z  (25)  Thus counterions of higher charge condense and reduce the net charge on the DNA by a greater amount. Calculated fractions for ssDNA and dsDNA with various counterions is given in Table 3 [26, 81, 127]. For the double helix, the charge density is equal to the axial charge density of the helix [84]. It needs however to be noted that these values are for long polyions. Oligomeric polyions will behave differently, and certain analyses have not taken this into account [127]. Charges near the ends of the chain will be uncompensated at low salt concentrations and nearing high salt concentrations, a plateau is reached . It can be estimated for ssDNA that the first two phosphate groups at each end will be uncompensated for monovalent concentrations below 500 mM [128]. 46  Z  θ(ssDNA)  θ(dsDNA)  1  0.44  0.76  2  0.72  0.88  3  0.81  0.92  4 0.86 0.94 Table 3 Dependence of fraction of charges compensated (θ) on cation charge for ssDNA (b = 0.40 nm) and dsDNA (b = 0.17 nm). This minimum compensation fraction is independent of concentration in solutions of low ionic strength.  This is important for the immobilization of ssDNA, as it shows a distinct transition in the grafting densities for monovalent salt concentrations between 50 to 200 mM [124]. This correlates well with the critical concentration found for ssDNA (~ 100 mM, κ-1 ≈ 1 nm) above which excessive counterion condensation occurred [127]. It suggests that aggregation of DNA is occurring in solution, and, consequently, at the surface. For dsDNA aggregation begins when approximately 90% of the charges have been compensated [129]. While this does not normally occur for short oligonucleotides, it has recently been found that short (≥ 6 bp) oligonucleotide duplexes can segregate even at small mole percentages from DNA mixtures to form liquid crystals within, but segregated from, the bulk phase [130]. These occur via end-to-end stacking [130], not requiring the usual degree of parallel contact for dsDNA aggregation [129]. 2.5.8 Experimental Evidence of Inhomogeneity It has become apparent through several different studies that ssDNA/MCH surfaces are inhomogeneous at a level beyond the image of a two-component SAM as suggested by Figure 14.Primarily this has been observed through electrochemical techniques. Using ssDNA without a thiol tether, it was found that between 5% and 15% of initially adsorbed DNA was not displaced from the gold surface by MCH [18]. This was coupled with the observation that electrostatic contributions to the removal of DNA from the surface were far less than the effects of MCH passivation which points to the possibility of DNA-DNA interactions at the surface. DNA may be viewed as a chelating, multidentate ligand on the gold surface facing competition 47  from several strongly binding smaller MCH molecules. In this case, the thermodynamic advantages of chelation may inhibit displacement. The adsorption of DNA bases to gold has been studied thermodynamically [131] and has even been used to form self-assembled monolayers of DNA in a controlled fashion using polyadenine nucleotide sequences [132]; such sequences can strongly compete for adsorption even with thiols [133]. A theoretical, all-atom molecular dynamics simulation of surface-immobilized dsDNA SAMs found that DNA duplexes spontaneously tilt toward the nearest neighbouring duplex to adopt a reclining position with an interaxial distance of 2.2 nm [26]. The result is explained in part by Manning's counterion condensation theory, as 76% of the charges on the dsDNA would be neutralized within by sodium counterions located within 0.7 nm of the helix surface [24, 27, 82, 83]. This leads to the formation of colloidal structures; as the electrostatic forces are reduced, attractive colloidal forces become dominant. The inference is that assembly protocols may have to be revised to limit solution aggregate formation [23]. Study of the electrochemical desorption of ssDNA at low electrolyte concentrations exhibited DNA release into solution even at low desorption potentials leading the authors to conclude that some of the DNA was nonspecifically adsorbed [127]. Again, the control of interstrand attractive forces was strongly dependent on the ionic strength of solution pointing to counterion condensation enabling nonspecific adsorption of ssDNA. Characterization of the probe density on DNA SAMs has been enabled by chronocoulometric [20, 21] and cyclic voltammetric [42, 43] studies of electrostatically bound redox cations to DNA. Comparison of the two techniques generally show results in good agreement [134], however others have suggested that it may be possible that some DNA-bound [Ru(NH3)6]3+ is “electroinactive” in measurements made with cyclic voltammetry while being “electroactive” in chronocoulometric measurement, suggesting that some of the DNA is nonspecifically bound further from the electrode than when thiol bonded (e.g. as aggregates), limiting electron transfer and “hopping” processes which enable electrochemical interrogation of DNA-bound ions [28].  48  These observations are of great importance, as it has been widely noted that the structure of DNA sensors strongly affects their performance [23, 135]. 2.5.9 Potential-Controlled Switching of Fluorescently Labelled DNA SAMs Finally, as it is particularly pertinent to the subject and methodology of this thesis, we look at the potential-controlled switching of fluorescently labelled DNA SAMs. This has been extensively investigated by a team in the Walter Shottky Institute at the Technische Universität München. The orientation of ssDNA and dsDNA can be controlled via electrostatic interaction between charging of the underlying substrate – usually a gold electrode – and the charged deoxyribose-phosphate backbone of the adsorbed DNA [136]. This has led to the development of dynamic electrical switching of DNA fluorescence [45] as a means to investigate the conformation and even the hybridization of DNA [46, 47, 50]. The fluorescence of various fluorophores in the presence of metal substrates shows a distance-dependent quenching which acts to suppress fluorescence nearer the metal surface. The response is depicted in Figure 16. Using a fibre-optic sensor, changes in fluorescence can be measured with a high sampling frequency, providing real-time data on DNA orientation which can provide meaningful kinetic data for the angular and rotational dynamics of DNA surface-tethered DNA [45-47, 50]. Single stranded DNA shows a much smaller fluorescence modulation compared to double stranded DNA when the applied electric field is changed. This is attributed to the greater rigidity of the dsDNA and the tortuosity of the ssDNA [45]. Additionally, ssDNA and dsDNA show different kinetic responses as they change orientation, with different rates for standing up and for lying down for both ssDNA and dsDNA as given in Table 4.The ratios of the time constants for standing up (τup) and for lying down (τdown) are near the values predicted from simulations [46, 47], with ssDNA standing up much faster than it goes down, whereas dsDNA shows a slightly slower rate for standing up compared to going down. Single stranded DNA also shows a much smaller fluorescence yield compared to dsDNA; at equivalent surface densities and potential, the fluorescence signal for ssDNA is approximately 1.7 times less than that of dsDNA [47]. This can be understood best in view of the tortuosity of 49  ssDNA even in an electric field resulting in a shorter average distance from the electrode. Overall, fluorescence offers a very useful tool in understanding the physical behaviour of ss and dsDNA at electrode surfaces, offering kinetic and orientational data that other techniques do not provide.  Figure 16 Response of fluorescence to potential induced changes in DNA orientation. DNA is repelled by the negative surface charge resulting from application of a negative potentials to the electrode (left), causing it to stand near-normal to the surface. When a positive potential is applied (right), DNA is attracted to the surface, causing it to tilt toward the surface. As the fluorophore (shown as CY3) is on the terminal end, it will experience the greatest change in proximity to the surface; near the surface, fluorescence is mostly quenched, while further away the fluorescence is considerable.  ssDNA  dsDNA  τ (up)  240 μs  220 μs  τ (down)  115 μs  235 μs  Table 4 Up and down switching time constants for 48-mer ssDNA and dsDNA in a buffered 10 mM Tris buffer (pH = 7.3, ionic strength = 10 mM monovalent electrolyte solution (κ-1 = 3.0 nm) measured upon changing of the electrode polarity, going from +400 mV to -200 mV and back to +400 mV.  50  3 Experimental Methodology This chapter will describe the instrumentation, materials, methods, and programs used in this course of the experiments described in this thesis. A general outline of the preparation of DNA/MCH coated gold bead samples including materials used will ensue, followed by electrochemical materials and methods used. Finally, a description of epi-fluorescence microscopy and the potential stepping programs will conclude.  3.1 Systems Studied This thesis investigates the surface structure of mixed thiolate self-assembled monolayers (SAMs) of fluorescently labelled thiol-modified DNA and mercaptohexanol (MCH), on polycrystalline gold beads. The gold beads functioned as the working electrode in in all experiments. MCH was used to passivate the gold surface, while adsorbed DNA was the primary focus of the work. In particular, it focuses on the differences and interplay between chemisorbed or specifically adsorbed DNA and physisorbed or non-specifically adsorbed DNA. While individual samples differed in their preparation, the general outline that follows lays out the fundamental steps in preparing the samples, including the materials and oligonucleotides used, their purification, and deposition on gold bead electrodes. 3.1.1 Materials and Chemicals Deionized 18.3 MΩ·cm water was produced using a Millipore Millipak ® 40 water purification unit. Purchased chemicals, unless otherwise mentioned, were used as is. Mercaptohexanol, or 6-mercapto-1-hexanol (MCH), HO-(CH2)6-SH, and 11-mercapto-1undecanol (MUDH), HO-(CH2)11-SH, were both obtained from Sigma-Aldrich. Gold wire (99.95%) 0.5 mm diameter was obtained from Goodfellow Cambridge Ltd. Tris(hydroxymethyl)aminomethane (Tris), NH2C(CH2OH)3, sodium chloride, NaCl, and magnesium chloride, MgCl2, for sample preparation were obtained from Fisher Scientific. Tris(2-carboxyethyl)phosphine hydrochloride salt (TCEP), P(CH2CH2COOH)3·HCl was 51  purchased from Sigma-Aldrich, and used as is. MicroSpin G-50 columns and Eppendorf tubes were purchased from Fisher Scientific. Hydrochloric acid, HCl, sodium hydroxide, NaOH, concentrated sulphuric acid, H2SO4, and hydrogen peroxide (30%, aqueous), H2O2, were purchased from Fisher Scientific.  3.1.2 Oligonucleotides Two synthetic, dual-labelled oligonucleotide sequences each 30 nucleotides long, were used in this study: HO-(CH2)6-S-S-(CH2)6-O-5'-CTG TAT TGA GTT GTA TCG TGT GGT GTA TTT-[CY3]-3' (LN1) and HO-(CH2)6-S-S-(CH2)6-O-5'-GTT GTG GCC AAG TAC AAA TTA TGG TAT CTA-[CY5]-3' (LN5), both obtained purified from Sigma-Genosys. Complementary, unlabelled sequences, 5' TAG ATA CCA TAA TTT GTA CTT GGC CAC AAC 3' (Comp-LN1) and 5' AAA TAC ACC ACA CGA TAC AAC TCA ATA CAG 3' (Comp-LN5) respectively, were obtained from University Core DNA Services, University of Calgary [137]. Identifier Sequence  Modifications  LN1  5'-thiol, 3'-CY3  5' CTG TAT TGA GTT GTA TCG TGT GGT GTA TTT 3'  CompLN1 3' GAC ATA ACT CAA CAT AGC ACA CCA CAT AAA 5'  none  LN5  5'-thiol, 3'-CY5  5' GTT GTG GCC AAG TAC AAA TTA TGG TAT CTA 3'  CompLN5 3' CAA CAC CGG TTC ATG TTT AAT ACC ATA GAT 5'  none  Table 5 Oligonucleotides sequences. All sequences used were all 30 units long. Complementary sequences were unlabelled.  The possibility of interactions of the DNA sequences to produce secondary structures including intermolecular self-complementarity and intramolecular hairpin loop formation was estimated using OligioCalc, an online oligonucleotide properties calculator [138, 139] and DINAMelt [140]. None of the sequences showed any intermolecular self-complementarity, however the LN5 sequence does posses the possibility of having an intramolecular hairpin loop form: 5' GTT GTG GCC AAG TAC AAA TTA TGG TAT CTA 3' (interacting bases are  52  underlined). The complementary sequence is relatively short, hence it will not be a strong interaction. In addition, this may be destabilized by surface-adsorption of the ssDNA, given the proximity to the surface-adsorbed 5' end and the destabilization caused by steric interactions. The low ionic strength of in situ measurements will eliminate any of the possible intramolecular loops [140].  3.1.3 DNA Purification and Preparation Thiol modified DNA was supplied as a disulphide (DNA-5'-O-C6H6-S-S-C6H6-O-5'DNA). The disulphide modified DNA was reduced in a solution of TCEP (10mM) in Tris buffer (100mM) at pH 7.4 and incubated for 4 hours at room temperature, resulting in the thiol. The solution was then purified using reverse-phase chromatography using centrifugation in a MicroSpin G-50 column (G-50 Sephadex) to afford single stranded (ss) thiol-terminated DNA (e.g. HS-(CH2)6-O-5'-GTT GTG GCC AAG TAC AAA TTA TGG TAT CTA-[CY3]-3') with any of the MCH byproduct removed. The concentration of the ssDNA solutions were subsequently determined using UV spectroscopy (λ = 260 nm) [79].  3.1.4 Sample Preparation Gold wire was melted using a butane torch to produce gold beads of approximately 1-2.5 mm in diameter. These were then heated in a flame until red hot to further to clean them. Reused beads were cleaned using piranha (a 3:1 mixture of concentrated H2SO4 with 30% H2O2) for 5-10 minutes at 90ºC to remove any organics or sulphides present, rinsed with deionized water, then annealed using a butane torch and rinsed again in deionized water to cool, followed by drying with flowing N2.  The “standard” protocol for preparing DNA coated beads is as follows: Deposition solutions of the DNA were prepared by diluting the purified ssDNA to approximately 1μM in an immobilization buffer containing 10mM Tris and NaCl with or without MgCl2 in variable 53  concentrations depending on the experiment. Where dsDNA is used, the deposition solution is made to the same concentrations, but including an equal concentration of the complementary ssDNA strand. The solution is heated to 70-80ºC, then cooled over 1 hour. The gold beads are then placed in a 20 μL drop of the DNA solution in an Eppendorf tube, sealed, wrapped in aluminum foil to prevent photobleaching and incubated overnight (approx. 24 hr) at room temperature.  Following DNA deposition, the beads are rinsed with the same immobilization buffer solution, followed by deionized water. The surface is then passivated with MCH (1mM) in the immobilization buffer for 1 hour. The bead is then rinsed with deionized water, stored in a solution of the immobilization buffer in the absence of light.  An alternative protocol for the preparation of DNA coated beads follows the reversal of the steps: The surface of the gold bead is passivated using MCH as per the “standard” protocol first, then the MCH-coated bead is placed in the DNA deposition solution (containing either ss or dsDNA) for a period of time from 1 to 48 hr, then rinsed and stored in a solution of the immobilization buffer in the absence of light.  3.2 Electrochemical Methods In addition to studies of the fluorescence of the SAMs, the beads were concurrently characterized using various electrochemical techniques. This section gives a detailed description of the materials and general preparative techniques used in addition to the equipment and experimental methods. 3.2.1 Electrochemical Equipment All electrochemical investigations were conducted in standard three-electrode cells with polycrystalline gold beads, which, in most instances were coated with mixed thiolate  54  monolayers of DNA and MCH, serving as the working electrode (WE). The counter electrode (CE) was a coil of either Pt or Au wire. A saturated calomel electrode (SCE) was used as the reference electrode (RE) unless otherwise noted, which was connected to the working solution through salt bridge. The salt bridges used consisted of either a glass-Teflon (poly(tetrafluoroethene), PTFE) stopcock, or a clean, L-valve ground glass stopcock. Both permit the diffusion of ions while preventing the flow of solution. 3.2.2 Chemicals Deionized 18.3 MΩ·cm water was produced using a Millipore Millipak® 40 water purification unit. Potassium perchlorate (>99%), KClO4, was obtained from Fluka, and slowly recrystallized from Millipore water prior to use. The following were used without further purification: Tris(hydroxymethyl)aminomethane (Tris), NH2C(CH2OH)3, was obtained in an ultrapure bioreagent grade (<1 ppm trace impurities for all metals tested) from J.T. Baker for buffering electrochemical solutions; its pKa is 8.3 at 20°C. Sodium tetrafluoroborate (98%), NaBF4, was obtained from Fluka for electrolyte use [141]. Hexaammineruthenium (III) chloride (99%), [Ru(NH3)6]Cl3, was obtained from Strem Chemicals. Fluoroboric acid (50%, aqueous), HBF4, was produced by Riel de Haën Co. Concentrated nitric acid and concentrated sulphuric acid for cleaning glassware were obtained from Fischer Scientific, as were sodium carbonate, Na2CO3, and sodium hydroxide, NaOH. 3.2.3 Electrochemical Instrumentation Electrochemical measurements in the epi-fluorescence microscopy cell were performed using a potentiostat (FHI) in conjunction with a lock-in amplifier (Princeton Applied Research 5208) and data recorded with a National Instruments (BNC-2090) data acquisition system using LabView. For capacitance measurements, an AC waveform (200 Hz) was superimposed on a cyclic voltammetry voltage ramp (5 mV/s); an AC perturbation of 5.80 mV rms was used. The capacitance was calculated from the out-of-phase and in-phase components of the current assuming a series RC circuit; the area of the electrodes were independently determined (see 55  Appendix) and the capacitance calculated. 3.2.4 Electrochemical Procedures Glass components of the electrochemical cells were cleaned in a heated acid bath for a minimum of 2 hours. The acid bath solution, prepared from roughly equal amounts (by volume) of concentrated sulphuric acid and concentrated nitric acid, was reused repeatedly. Following a brief cooling period, the glassware was rinsed with Millipore water, filled with water and allowed to soak for a minimum of 12 hours. In addition to filling with water, a small amount of either NaOH or Na2CO3 was added to prevent the glass from retaining acid. The cell was rinsed again with Millipore water, then electrolyte was poured into the cell. The electrolyte solution was then purged by bubbling Ar in order to remove any dissolved oxygen; after 15 minutes bubbling was ceased and a constant blanket of Ar was maintained above the solution surface at positive pressure to prevent oxygen from re-entering the cell. Subsequently, the salt bridge was first filled on the supporting electrolyte side, then on the reference side with a saturated KCl solution; the two solutions were separated by a stopcock allowing for ionic contact between the SCE and the working solution while avoiding chloride contamination of the supporting electrolyte. In the optical cells, the reference electrode sits in a solution of supporting electrolyte connected to the cell via a similar salt bridge. The CE was flame annealed prior to introduction to the system. Where the WE was a bare gold bead, it was also slowly flame annealed prior to use. 3.2.5 Electrochemical Techniques 3.2.6 Cyclic Voltammetry Cyclic voltammetry (CV) involves sweeping the potential linearly with time and measuring the resulting current. Each cyclic voltammogram consisted of a potential sweep (e.g. 20 mV/s) to a positive limit, at which point the linear ramp was reversed and the potential swept to a negative limit; the ramp was again reversed and process was repeated. Either a LabView program or the Autolab software package were used to acquire and record the data. In this thesis, cyclic voltammetry was used primarily in three scenarios: (1) for determination of the 56  surface concentration of DNA using the reduction/oxidation of Ru(NH3)63+ / Ru(NH3)62+ as a reporter, (2) for the in situ characterization of the DNA-MCH SAM in the optical cell, and (3) in the determination of the surface area of the bare gold beads. The first and third items will be elaborated upon separately. In situ, CV give information about the quality of the adsorbed monolayer. In the absence of faradaic reactions, the CV gives a rectangular trace; the capacitance in such regions can be calculated using the difference between the anodic (ia) and cathodic (ic) currents as shown in Equation 26 which models the system as a series RC circuit.  i=  i c −i a dE =−C d 2 dt     (26)  CV can also be coupled with fluorescence imaging, with some difficulty, to obtain visual information regarding the effects of sweeping the potential on the surface. This requires synchronizing the acquired electrochemical data with image data, each collected with a different sampling frequency. 3.2.7 Differential Capacitance The capacitance of the interface was measured using differential capacitance (DC) measurements in the double layer region. Capacitance was measured using a sinusoidal voltage perturbation (200 Hz, 5.80 mV rms). This was added to a linear CV sweep (5 mV/s) of the potential, giving the capacitance as a function of potential. The resulting current was analyzed by a lock-in amplifier, and the in-phase and the quadrature (90° out-of-phase) components measured. These are the real and imaginary currents respectively which are used to calculate the differential capacitance. Given the assumption that the interface can be modelled as a capacitor and a resistor in series, Equation 27 can be used to calculate the capacitance (C).  [  ]  i i C= im 1 re V ac  i im  57  2  (27)  C is the calculated interfacial capacitance (μF/cm2), Vac is the root mean square amplitude of the AC potential (mV rms), iim and ire are the imaginary and real (in and out of phase) current densities (μA/cm2), and ω is the frequency (Hz). This data was collected and the capacitance calculated using a LabView program described in a previous thesis [142, 56]. 3.2.8 DNA Surface Concentration CV measurements using [Ru(NH3)6]3+ as a reporter molecule for the surface concentration of DNA were preformed using deaerated solutions of [Ru(NH3)6]Cl3 (5 μM) in a Tris (10 mM) buffer solution (pH ~ 7.5) covered with a blanket of argon gas. DNA/MCH coated beads were rinsed with Millipore water, then immersed in the [Ru(NH3)6]3+ solutions for 15 minutes to equilibrate, and 3 CV scans performed. Redox peaks for DNA-bound ruthenium were integrated to determine the total charge and thus the surface concentration. The beads were immersed as shown in Figure 17.  Figure 17 The gold bead is immersed such that the surface of the water is tangent to the top of the bead. This is checked by looking upward through the water, to see the gold bead tangent to its mirror image.  Ruthenium (III) hexaammine, [Ru(NH3)6]3+, is a widely used transition metal for redox chemistry. It has a one-electron reduction as shown in Equation 4-3, with a standard reduction potential of -180 mV versus SCE [143], however this is shifted negative for DNA-bound ruthenium [43]. The diffusion coefficient of free [Ru(NH3)6]3+ is 9.1 × 10-6 cm2/s [143]. [Ru(NH3)6]3+ + e-  [Ru(NH3)6]2+, E° = -180 mV vs. SCE. 58  (28)  Two configurations were used: The first was a μAutolab II potentiostat/galvanostat (EcoChemie B.V., Utrecht, The Netherlands) set up with a single compartment (1 mL) 3electrode cell. The gold bead (not the stem) was immersed in solution (as observed when the surface of the water was tangent to the top of the bead). Ag|AgCl|3M NaCl was used as the reference electrode and a Pt wire was used as the counter electrode. The second configuration was an Autolab potentiostat/galvanostat set up with a double compartment 3-electrode cell. The compartment (80mL) containing the Pt counter electrode and the gold bead was connected via a salt bridge (stopcock) to a saturated KCl solution containing an SCE reference. The cell is shown schematically in Figure 18. Negative charges in the phosphate backbone of DNA are naturally compensated for by various cation counterions. In the preparation of thiolate monolayers of DNA, Na+ and Mg2+ cations were the primary counterions, and were used as their chloride salts to control the ionic strength when depositing and hybridizing DNA SAMs. When placed in a solution containg a redox cation as the only metal cation present, the compensating counterions will exchange from the DNA and be replaced by the redox cations on a 1:1 charge basis, thus three Na+ would be replaced by one [Ru(NH3)6]3+. The redox peaks for the DNA-bound [Ru(NH3)6]3+ are shifted negatively relative to the solution peaks, as shown in Figure 19; Aqueous ruthenium is more easily reduced than DNAbound ruthenium due to the electrostatic interactions preventing the diffusion of DNA-bound ruthenium directly to the surface, hence an electron “hopping” mechanism is required. Integrating the DNA-bound redox peaks gives the total charge, Q (Coulombs, C), from which the surface concentration of ruthenium can be determined as shown by the relation Q=n F A Ru  (29)  where ΓRu is the surface concentration of [Ru(NH3)6]3+ (mol/cm2), n is the number of electrons in the reaction (n = 1), A is the area of the WE, and F is the Faraday constant (96485.33 C/mol).  59  Figure 18 Typical electrochemical cell with working solution and reference electrode in a separate saturated KCl solution. The cell used for ruthenium-DNA measurements with an electrolyte solution contains approximately 5 μM [Ru(NH3)6]3+ and 10mM TRIS buffer. Also used for surface area measurements with a working solution containing KClO4. WE is the working electrode, a DNA-coated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE) in saturated KCl; CE is the counter electrode, a platinum wire.  60  Figure 19 CV of a gold bead coated with an ssDNA/MCH SAM in a ruthenium solution ([Ru(NH3)6]Cl3, 5 μM in 10 mM Tris buffer). Two ruthenium peaks are observed in each direction (cathodic and anodic). Integrated areas of the oxidation (Aox) and reduction (Ared) are shown by the shaded areas; these give peak areas in units of charge (Coulombs), which can in turn be used to calculate the surface density of DNA. Measurement of the surface area is described in the following section. The system is governed by the binding constants of DNA to the ruthenium complex to DNA. For conditions of saturation, the surface concentration of DNA can be directly determined by Equation 29.   DNA= Ru N A  z m  (30)  Here, ΓDNA is the surface concentration of DNA (strands-or-duplexes/cm2), NA is Avogadro's number, z is the charge on the redox cation, and m is the number of nucleotides per strand. For duplexes to be counted instead of strands, the number of nucleotides per duplex is used; typically this is simply twice the number of nucleotides.  61  The validity of this, and other electrochemical means of the quantification of DNA probes relies on the following assumptions: (1) both initial cations and redox cations are bound to DNA strictly through electrostatic interactions; (2) the number of electrostatically bound redox cations can be accurately determined and can be discriminated from unbound cations of the same type; and (3) redox cations are the sole charge compensating ions for the phosphate groups in DNA. Further details regarding this technique are discussed later in this thesis [42, 43].  3.2.9 Surface Area Measurements The surface area of the gold beads was determined using measurements of beads either before any DNA or MCH was adsorbed to the surface or after all of the DNA/MCH was desorbed from the surface. The surface area was initially measured using the capacity of electrodes measured in 10mM NaBF4 and integration of the current using reduction of a gold oxide monolayer, however this method was later replaced in favour of using the differential capacitance to determine more the surface area more exactly. In both cases, the electrochemically measured surface area includes a small wetting effect for the gold bead, resulting in some of the stem of the bead being included in the measurement, hence this was compared to the geometric surface area. The geometric surface area of each bead was determined using digital microscopy with a 2500 μm diameter metal spot as an internal reference to determine the size by fitting the shape with an ellipse, and measuring the axes which were then used to calculate the surface area. Gold beads formed under heating with a butane torch are generally nearly spherical if small, although larger beads are generally prolate ellipsoids (jellybean shaped) where two semiaxes are equal and the third forms the longest axis (a = b < c), although some are scalene (a < b < c). The surface area of such a bead can be calculated to a good approximation as    p  p  p  p  p  a b a c b c S=4  3  p    1 p  (31)  where S is the surface area of an ellipsoid, a, b, and c, are the semiaxes, and p = 8/5 = 1.6 which 62  is optimal for nearly spherical ellipsoids. Although more accurate formulae are available Equation 4-6 is the simplest and produces a maximum relative error of 1.178% [144-147]. The area of the bead used is smaller by the area of the area of contact between the wire and the bead. This can be approximated as in Equation 28:  Abead ≈S − r  2  (32)  Equation 28 gives the geometric surface area to a good approximation, however this is always less than the measured value due to the surface roughness and wetting of the stem. The surface area, electrochemically determined using differential capacitance, was measured in a degassed 25 mM KClO4 solution, pH ≥ 7. CV scans from +1250 to -800 mV at 20 mV/s were first performed in order to ensure the cleanliness of the system. DC measurements of the double layer region, from +650 to -800 mV at 5mV/s, followed. Using value of 18 μF/cm2 for bare gold, the capacity at -800 mV (where the curve flattens) gives the electrochemical surface area.  63  3.3 Spectroscopic Technique 3.3.1 Epi-Fluorescence Microscopy Fluorescence imaging was performed in concert with electrochemical measurements using two specially designed spectroelectrochemical cells shown in Figures 20 and 21. The cells were mounted on an inverted epi-fluorescence microscope (Olympus IX70) equipped with a 20X objective (numerical aperture = 0.4, working distance = 10 mm) either with or without use of the 1.5X multiplying lens. The microscope objective is below the sample in the inverted configuration so that the objective is "looking up" at the bottom of the sample, through an optical window. The objective is focused on the bottom of the immersed gold bead as shown later in Figure 20. Images were captured with a monochromatic Spot RT digital camera over variable exposure times. White light from a DC-powered 75W xenon short-arc bulb (UXLS75XE) was directed toward a filter cube bearing a excitation filter designed to transmit a band of light specific to each fluorophore's absorption spectrum. Two main filter cube configurations were used, each specific to a certain dye. For CY3, an Olympus (U-MNG2) fitler cube containing an excitation filter (530-550 nm), a dichroic mirror (570 nm), and an emission band pass filter (> 590 nm), was used to excite the sample and collect the fluorescence.  For the fluorescence of CY5, a generic filter cube designed for an Olympus microscope (U-M41008) containing an excitation filter (590-650 nm), a dichroic mirror (660 nm), and an emission filter (662-705 nm) was used. The optical window was made from a glass coverslip, 0.17 mm thick, sealed onto the glass body of the cell, giving excellent image quality.  64  Figure 20 The original design of the spectro-electrochemical cell, used, unless otherwise specified, in the electro-fluorescence experiments. The optical window above the objective is made from a coverglass which is 0.17 mm thick. The 20X objective was used as the default, however a 1.5X internal magnification is possible to make a total 30X magnification if necessary. The filters are mounted in a rotating carousel to allow for interchange of filters during one experiment. Where both fluorescence and electrochemical measurements were preformed, the supporting electrolyte contained 10 mM Tris and 10mM NaBF4. WE is the working electrode, a DNA-coated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE); CE is the counter electrode, a platinum wire. 65  Figure 21 The second generation of the spectro-electrochemical cell is shown schematically. The cell rests on a flange built into the 2.5 cm diameter tube which forms the body of the cell. An aluminum holder is used to station the cell in place above the objective and prevents contact between the glass window and any surfaces. This new cell has allowed for testing the adsorption of 1-mercaptohexanol (MCH) in situ, and may be used to image the hybridization of complementary DNA to ssDNA coated surfaces in the near future; the smaller volume will be a particular asset due to the relatively high cost of DNA, and the concentration required for hybridization. WE is the working electrode, a DNA-coated gold bead; RE is the reference electrode, a saturated calomel electrode (SCE) in saturated KCl; CE is the counter electrode, a platinum wire.  66  3.3.1.1 Camera and Optics The camera contains a Kodak (model KAI-2092) interline charge-coupled device (CCD) which is cooled to 37°C below ambient temperature using a Peltier device in order to reduce dark noise. The quantum efficiency of the CCD is shown in Figure 22. All images were acquired using SPOT software with 12 bit resolution (4096 levels of grey). The CCD sensor is approximately 11.1 x 7.9 mm and contains 1520 x 1180 pixels; each pixel (px) is 7.4 μm x 7.4 μm [148]. A 0.76X coupler (Diagnostic Instruments) focuses the fluorescence image onto the CCD enabling full chip usage. With the 20X objective, the images depict an area approximately 0.52 x 0.73 mm wide. With 2x2 binning, the resulting image is 540 x 760 pixels, with a resolution of 0.961 μm/pixel.  Figure 22 Quantum efficiency of the Kodak KAI-2092 CCD. Graph produced with data taken from SPOT RT specifications [148].  During image acquisition, a minimum write time of 350 ms was found necessary following each image exposure in order for the program to control the time interval between images. When instructed to wait less than 350 ms following an image exposure, the triggering program simply waits until the prior image has been written to the hard disk, which on average requires 337 ms. Further details are provided in the Appendix. 67  3.3.1.2 Original Spectro-electrochemical Cell The spectro-electrochemical cells used for the in situ electro-fluorescence investigations were similar to the standard electrochemical cell. A salt bridge connected directly to the cell permitted movement of the entire cell above the objective. The salt bridge was filled by tilting the cell in order to flow the supporting electrolyte through the PTFE stopcock to a reservoir in which the SCE was placed. The SCE was separated in order to prevent chloride contamination of the cell, however concentrated KCl was not used in the reference side of the salt bridge. The spectro-electrochemical cell was originally designed to include an inset viewing port located at the bottom of the cell for the objective. A glass window in the port was made from a 0.17 mm thick borosilicate glass (Pyrex) coverglass. The thin window limits aberration's and optical distortions, allowing for high quality images. Figure 20 shows the experimental set-up of the original spectro-electrochemical, as set-up for electro-fluorescence measurements.  3.3.1.3 Second Generation Spectro-electrochemical Cell One drawback of the original spectro-electrochemical cell design is the relatively large volume of solution required, which is prohibitive for experiments requiring smaller solution volumes. Having the viewing port inset on the bottom of the cell limited the ability to alter the position of the cell laterally with respect to the objective; the inset port also made repair of a broken window impossible. The second generation cell, instead of having the 0.17 mm glass window inset, it was placed on the very bottom. The cell body was formed from a 2.5 cm diameter tube. A flange built into the walls of of the glass body allows the cell to rest in a circular mount for storage and imaging experiments, while enabling greater manoeuvrability. The design has a relatively small solution volume (5 mL) when in use. The design of the second generation cell is shown in Figure 21.  68  3.3.1.4 Programs Used Only one LabView program was used in potential stepping experiments in this thesis. While it is described in a previous thesis, its operation is briefly described. The program is written to control the potential of the WE and measure capacitance during the acquisition of images in conjunction with an image acquisition program. Images were acquired by the SPOT Advanced program which collected image sequences according to a set time interval (which determines the image-to-image frequency) up to a set maximum. Parameters including binning (2x2), bit-depth (12 bit), exposure time, and gain (4x) were defined using the software. Images were saved automatically to a desired folder; this requires a maximum 350 ms following completion of the exposure. LabView sets the potential to a predefined value; once an image file is saved in the desired directory, LabView steps to another potential and awaits the subsequent file. After the collection of each image, the potential, capacitance, and image number values were written to a list file which is saved at the end of the image/potential sequence. In the LabView program, image_step.vi, four variables addressed the potential: initial potential (Ei), final step potential (Ef), base potential (EBase), and the potential increment (Einc). In addition, the number of images taken at each base potential (M) and at each step potential (N) are defined. The four potentials define the number of steps in the sequence, and in turn, the individual step potentials (EStep) in the series; that, with the number of images taken at each base and step potential, define the total number of images taken. The sequence is shown below: [Start] → EBase → Ei → EBase → Ei+Einc → EBase → Ei+2Einc → EBase → Ei+3Einc → EBase → ... → Ef → EBase → [End] This is shown schematically in Figure 23.The total number of steps is given by:  N steps=  E f − Ei 1 E inc 69  (33)  Figure 23 A schematic description of the potential control during the acquisition of images using the image step program. The values depicted would be representative of a typical experiment with an initial potential (Ei) at -25 mV and a final step potential (Ef) at -400 mV with -25 mV increments (Einc) and a base potential (EBase) at 0 mV; the initial and final step potentials can be changed, as can the increment and the base potential; the base potential is independent of the initial and final potentials. In the sequence shown, 5 images (N) are taken at the step potential, and 5 images (M) are taken at the base potential each time; while 5×5 was the most widely used, other integer values of M×N are possible. Given these settings, 165 images were taken in the experiment. A secondary x-axis is shown for time. This is dependent on the interval time from one image to the next; the present scale is calculated based on a interval time of 2 seconds.  70  3.3.2 Image Analysis Image sequences, referred to as “image stacks”, are ordered by image number which corresponds to a the time domain as shown in Figure 23. With 2x2 binning, each image in the stack is 540 px by 540 px; each pixel is the average of four elements in the CCD sensor array. Image stacks form a 3D array of pixels, each with an intensity Px,y,z, are described graphically in Figure 24.  Figure 24 Diagram describing the coordinate systems used for image stacks. (A) Image stack 540 px by 760 px area and 3 images long. (B) Single pixel, of intensity Px,y,z, with coordinates (x,y,z). (C) The xy-coordinates define the image in two dimensions. (D) The z-coordinate provides the position of the image within the stack and in time, giving a third dimension. Depiction of the 3D array of pixels in an image stack is difficult in two dimensions without the loss of information. Various means of projecting the information from the stack into a 2D image can be achieved through image processing. Image stacks were processed using ImageJ, a public domain Java-based image processing program developed at the National Institutes of Health [149, 150]. This enables calculations to be performed on image sequences to provide data on regional variation, and changes with time in particular areas. Typically “stack calculations” are performed as a function of the z-axis. A stack average (AVG) is obtained from the following function applied to each xy-coordinate pixel to a stack of n images: n  1 1 Pxy = ∑ P xyz =  P xy1 P xy2⋯P xyn  n z =1 n 71  (34)  Similarly, a stack standard deviation (STD) may be obtained for each pixel, going through the stack to determine which areas display the greatest variation in their evolution over time. The calculation, applied to each pixel in the stack, is    n  1  xy= ∑  P xyz − Pxy 2 n z=1  (35)  Additionally, most other standard mathematical operations normally performed on a number series may be performed on an image stack, such as finding the maximum value (MAX), the minimum value (MIN), the median value (MED). Other simple operations used include addition (+), subtraction (-), multiplication (×), division (÷). More complex operations are also possible, including 2-dimensional Gaussian blurring (gb), which can be used to smooth out noise in an image. It applies a Gaussian transformation to each pixel in the image: 2  1 G x , y= e 2  2  2  − x  y  2  2   (36)  Other means to remove noise, or despeckling, such as high and low-pass filters are available and are described in the ImageJ documentation [151]. Other specific techniques such as particle analysis and regional analysis methods have found application in the Bizzotto lab and are described elsewhere [56, 57]. It must be noted that most images presented in this thesis are in false colour, defined via a look up table, as in Figure 25, in order to enable direct visual recognition of regional variance.  Figure 25 Look up tables shown as gradients. Below: A greyscale look up table, where white corresponds to high intensity, and black corresponds to low intensity. Above: The “Royal” look up table defined in ImageJ maps values from greyscale to colour.  72  4 Results and Discussion 4.1 Introduction The objective of our research was to investigate SAMs of adsorbed thiolate ssDNA and to measure the layer's response to changes in potential. The layers were prepared following the approach of Tarlov et al.[18, 19, 22], using 6-mercapto-1-hexanol (MCH) to passivate the gold surface following the initial adsorption of a hexylthiol-labelled ssDNA sequence (ssDNA/MCH). Assembled layers were examined with in situ fluorescence microscopy in conjunction with electrochemical measurements and determination of the ssDNA surface concentration cyclic voltammetry to measure of the reduction and oxidation of DNAintercalated ruthenium in a solution of [Ru(NH3)6]3+ [20, 21, 43, 152]. Analysis of the response of ssDNA/MCH has led us to explore alternative means to prepare DNA/MCH SAMs. Given the better ordering possible with dsDNA, immobilization of dsDNA followed by passivation with MCH (dsDNA/MCH) was studied next. Subsequently, a radically different approach was investigated, where the surface was first passivated with MCH and then either single stranded DNA (MCH/ssDNA) or double stranded DNA (MCH/dsDNA) were incorporated via a ligand exchange process. Such monolayers were found to have superior fluorescence switching properties and fewer “hotspots”, which represents a significant result. Images analysis of the switching and desorption sequences provides a profusion of images, hence only selected images produced during the analysis will be included here.  73  4.2 ss/MCH 4.2.1 ssDNA-CY5/MCH Following the methodology of Tarlov and Herne, adapted to gold beads in accordance with procedures of the Yu lab, ssDNA/MCH coated gold beads were prepared as described in Chapter 4. For ssDNA-CY5/MCH, a deposition solution of ssDNA was prepared (~ 0.5 μM HSC6H12-DNA(LN5)-CY5) in immobilization buffer, IB1, (100 mM NaCl, 100 mM MgCl2, 10 mM Tris, pH 7.5) and the bead immersed for 18 hours, removed and rinsed briefly with deionized water, then placed in a solution of MCH (1 mM in IB1) for one hour, and finally rinsed, then returned to a solution of IB1 awaiting characterization. The surface concentration of DNA was then measured by ruthenium CV ( ΓssDNA = 4.9 × 1012 strands/cm2) and replaced into the IB1 solution in an Eppendorf tube. The sample was sealed and kept in the absence of light. Fluorescence response to electromodulation was then tested in a spectro-electrochemical cell for epi-fluorescence microscopy containing an argon-purged buffer solution (10 mM Tris, pH 7.5). CV and capacitance sweeps were obtained. Electromodulation stepping sequences (+50 mV → -500 mV, 0 mV → +400 mV; Einc = ±50 mV) were performed prior to a final desorption sequence (0 mV → -1100 mV, Einc = -50 mV, Ebase = 0 mV), from which images shown in Figure 26 are derived, with fluorescence intensity and electrochemical measurements shown in Figures 27 and 28. It is apparent that the desorption, as measured by fluorescence, begins at -550 mV, however when ascertained from the capacitance, it appears to begin at -650 mV. This suggests that nonspecifically adsorbed DNA is beginning to desorb from the surface in advance of the reductive desorption of thiolates covalently bonded to gold. Figure 26 shows the heterogeneity in the fluorescence intensity of the surface through a variety of image calculations. The images are mapped from a greyscale image using a colour look-up (26 L) table in order to enable human perception of small differences in intensity. As can be seen from the brightfield image (26 A), there is little correlation between surface irregularities and hotspots on the surface. From the average of the stack (26 B) and its contrast enhanced version (26 C), one observes that the hotspots occur in a variety of shapes and sizes 74  and have no particular pattern of distribution. To find areas with larger changes relative to their initial intensities, the standard deviation of the stack is divided by the median intensity (26 F), showing that the low-intensity areas have relatively large changes with desorption. This can be repeated using images in the pre-desorption region (0 mV → -500 mV), instead using the average intensity; the result shows that hotspots have a smaller modulation relative to their mean intensity. This is reiterated by taking the difference between the maximum and minimum intensities within that region. A regional analysis (Figure 28 and 26M) of the surface shows that all regions of the surface release DNA during desorption; even regions that have little initial fluorescence (relative to other areas) nonetheless show a significant signal during the desorption. Hotspots show an intense initial fluorescence, however their intensity at desorption shows an increase far less than that of dark and moderately intense regions. Combined, these observations indicate that the hotspot regions contain DNA that is likely further from the surface than non-hotspot regions, implying the physisorption of DNA. Additionally, the hotspots appear to have globular shapes, strongly suggestive of deposited aggregates; they range in size from 5 μm to 50 μm across. Minimal fluorescence switching is observed in Figure 27 prior to desorption, although following desorption, diffusive behaviour is observed for the DNA, as is indicated by the time dependence of both the fluorescence intensity and the capacitance. The time dependence of the capacitance actually provides a definitive means of measuring where the base capacitance begins to change, as subsequent measurements over time will yield slightly different values; mapping this change in the form of a standard deviation over each step can indicate when chemisorbed monolayers begin to desorb. Examining the capacitance changes, it appears that electrochemical desorption occurs at -625 mV, whereas fluorescence indicates desorption occurring as early as -500 mV.  75  Figure 26 Images of ss-DNA-CY5/MCH#1 taken during a reductive desorption stepping sequence: 0 mV → -1100 mV, Ebase = 0 mV, 5×5 images, 50 ms exposure, 1000 ms interval time, 445 images. A shows a greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. The average (AVG) intensity of each pixel in the stack is shown in B, and a contrast enhanced version is shown in C. D and E show the standard deviation (STD) and median value (MED) of each pixel in the stack, which are used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. G and H are the maximum (MAX) and minimum (MIN) intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA, hence the difference, I, is taken to find areas where DNA desorbs. Using only those images for potentials in the fluoro-electromodulation region (0 mV → -500 mV) prior to desorption, values can be obtained for which areas are showing the greatest levels of switching, either by taking a MAX-MIN of the stack, K, to see an absolute change, or by taking the STD divided by the AVG to obtain a normalized switching map. K also shows an outlined ROI obtained by thresholding J to select for those areas which showed lower modulation relative to the AVG intensity. M shows a larger version of C, with chosen regions of interest (ROIs) highlighted in white boxes. Nine ROIs, each 48 μm x 48 μm, were chosen to investigate specific features of modulation and desorption: hotspots (3,4,5,7,8, & 9) of varying intensity were chosen, in addition to smooth (2,6) and extremely dark (1) areas for comparison.  76  77  Figure 27 Potential step sequence applied to ssDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1100 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 50 ms exp, 1000 ms interval. Other step sequences are shown in the Appendix.  78  Figure 28 Fluorescence intensities and capacitance for the reductive desorption of ss-DNACY5/MCH (0 mV → -1100 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages over each Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 26. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 50 ms exp, 1000 ms interval.  79  4.2.2 ssDNA-CY3/MCH A second ssDNA/MCH experiment is presented to further develop the observations for ssDNA-CY5/MCH. Here, the alternate ssDNA sequence (HS-C6H12-DNA(LN1)-CY3) and corresponding fluorophore, CY3 are used (ssDNA-CY3/MCH). The preparation followed the same outline as for the previous example, changing only the immobilization buffer, IB5 (100 mM NaCl, 500 mM MgCl2, 10 mM Tris, pH 7.5), to a higher magnesium concentration. Surface concentration of DNA was measured to be ΓssDNA = 7.5 × 1012 strands/cm2. The appearance of the surface shows considerable heterogeneity in fluorescence intensity (see Figure 29). Regional analysis of the image (Figure 30) through the desorption sequence shows that upon desorption, areas with low to moderate initial fluorescence intensity show the greatest relative desorption intensities. Again, although there appears to be some correlation of the fluorescence intensity to the structure of the underlying metal, the correlation does not account for all of the inhomogeneity. In the switching region (0 mV → -500 mV), areas of lower intensity show a greater change in fluorescence relative to the average intensity. The absolute change in intensity however is only slightly lower than for brighter regions. Moreoever, the change from intense to dark regions across the bead surface does not appear to follow a smooth gradient; hotspots have defined, irregular edges. From this one may infer a significant difference in structure of such hotspots compared to other regions. Desorption from the surface can be seen from Figure 30 to occur at slightly two different potentials (compare ROI 1 and 3 or 6 and 7). It should be emphasized that desorption is marked primarily by the initial increase in fluorescence intensity resulting in the diffusion of the DNA from the surface, but also by the point where the DNA shows no remaining structure on the surface: The breaking of bonds tethering the DNA to the surface results in diffusion of the strands both laterally and normal to the surface giving the appearance of a smooth mist. This can obscure observation of the latter, however the shifting of the peak in intensity during desorption may be indicative of it. The intensity during Estep is useful in this determination, though the capacitance at Ebase is also useful: Diffusion of desorbed DNA away from the surface during the  80  Figure 29 Images of ss-DNA-CY3/MCH taken during a reductive desorption stepping sequence: 0 mV → -1200 mV, Ebase = 0 mV, 5×5 images, 250 ms exposure, 583 ms interval time, 485 images. A shows a greyscale bright-field image (10 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. The average (AVG) intensity of each pixel in the stack is shown in B, and a contrast enhanced version is shown in C. D and E show the standard deviation (STD) and median value (MED) of each pixel in the stack, which are used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. G and H are the maximum (MAX) and minimum (MIN) intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA, hence the difference, I, is taken to find areas where DNA desorbs. Using only those images for potentials in the electro-fluoroescence modulation region prior to desorption, values can be obtained for which areas are showing the greatest levels of switching, either by taking a MAX-MIN of the stack, K, to see an absolute change, or by taking the STD divided by the AVG to obtain a normalized switching map. K and E also shows an outlined ROI obtained by thresholding J to select for those areas which showed lower modulation relative to the AVG intensity. M shows a larger version of C, with chosen regions of interest (ROIs) highlighted in white boxes. Nine ROIs, each 48 μm x 48 μm pixels, were chosen to investigate specific features of modulation and desorption: hotspots (3,4,6, & 9) of varying intensity were chosen, in addition to smooth (1,7) and darker (2,5) areas for comparison.  81  82  Figure 30 Fluorescence intensities and capacitance for the reductive desorption of ss-DNACY3/MCH (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 29. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. The Estep capacitance curve is trunctuated due to excessive H2 evolution. Imaging settings: 200 ms exp, 853 ms interval.  83  Estep is, in general, faster than diffusion back to the surface during the step to Ebase due to the charged phosphate groups in DNA and the greater electric field present during each Estep in the desorption region. The result is that the overall fluorescence during the subsequent Ebase does not decrease to a point as low as before (until DNA has completely left the region). These features provide information which can be determined using microscopy which cannot be determined for polycrystalline samples from capacitance measurements. It was found useful to use a reslice function in ImageJ to compare various regions through the use of a cross-section of the image stack. Figure 31 shows the one such region in question as well as a reslice through the desorption sequence. The alternating bands result from the potential step sequence; each pixel in the time axis represents one image, and there are 5 images per potential increment. Regions C and D in the image desorb at -500 mV and -575 mV respectively; a 75 mV difference, however capacitance measurements indicate desorption initiating at -625 mV. As described in the background, DNA adopts a mean orientation in an electric field of a given magnitude – which is dependent on electrode potential. Hence the fluorescence intensity of the DNA-fluorophore complex should vary considerably at a given potential, save for photobleaching, and the time required to reorient, which should be a fraction of a second. Fluorescence switching experiments in regions without faradaic reactions of either DNA or thiolate (+400 to -400 mV) were performed to explore the SAMs' response (Figure 32). Here the sequence was done in two parts; a negative stepping sequence (0 mV → -400 mV, Ebase = 0 mV) and a positive (0 mV → +400 mV, Ebase = 0 mV) with both the initial base potential at 0 mV for both. The maximum change in fluorescence intensity on the negative side was slightly smaller than the change measured in going positive. The fluorescence intensity display a time lag over each three second step, both when going to Estep and to Ebase, suggesting that the reorientation time is on the order of a few hundred milliseconds; the effect is more pronounced in the positive sequence, suggesting a time nearer to ~ 1 second; here, time-dependent changes in the capacitance at Estep are also more pronounced indicating the slowness of the reorientation. Again, this would be consistent with the behaviour of an aggregate. It is interesting to note too that the whole surface appears to be reacting relatively homogeneously to the changes, with the response perhaps being dominated by the most intense signal.  84  Figure 31 Reslice image of the desorption the desorption sequence (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) for ss-DNA-CY3/MCH along line AB which is drawn on the MAX-MIN image on the right. Regions C and D display different desorption potentials, with C desorbing approximately 100 mV before region D.  85  Figure 32 Switching of ssDNA-CY3/MCH. Data from two sequential stepping sequences, first going negative (0 mV → -400 mV, Ebase = 0 mV, 5×5 images, 250 ms exp., 583ms/image), then going positive (0 mV → +400 mV, Ebase = 0 mV, 5×5 images); a dotted line marks the starting points. Change in fluorescence intensity (Δcps) is shown, measured from the minimum value; fluorescence intensity of the first sequence was decreased by 260 cps (average) to remove the offset, likely due to photobleaching. Capacitance shown calculated using an estimated area of 0.1 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right).  86  4.2.3 Conclusions Layers imaged at and near their open circuit potential (OCP) using fluorescence appeared very heterogeneous containing regions where minimal fluorescence was observed and "hotspots"; regions which were strongly fluorescent. Each region shows a different response to potential perturbation or stepping to desorption. Furthermore, removal of the SAM by desorption via electrochemical reduction revealed that the whole surface was covered by ssDNA, although not all regions are similarly fluorescent near OCP. The heterogeneity of the surface has received very modest discussion in the literature, and it appears that a considerable proportion of fluorescence observed is due to the hotspots observed, although all fluorescence studies to date appear to completely lack spatial resolution capable of discriminating between such different contributions. It is widely known that ssDNA tends to have difficulties associated with it, given that longer chains tend to act as form random, possibly interpenetrating, coils, and, during adsorption, do not form well organized surfaces due to the ability to aggregate and adsorb by means other than a thiolate-gold bond. Such "sticky" clusters may well explain the hotspots observed. Hotspots may also be due to the heterogeneity of th e electrode surface, however comparison of bright-field images of the surface with fluorescence images of the same surface shows no such correlation to any significant degree; there is some association, however not all hotspots are located at surface defects. Submicroscopic variation also seems unlikely, as the surface was found to be quite smooth, having a maximum measured roughness factor of 1.158 resulting from the surface tension of the molten gold prior to annealing of the surface (see Appendix). Details regarding the surface area are given in Appendix 2. Other experiments (described later) have also indicated substrate independence; the adsorption of thiolate DNA directly on gold occurs with little regard to which crystal face is present.  87  4.3 ds/MCH Following the results of our initial investigations of ssDNA-CY3/MCH and ssDNA-CY5/ MCH SAM preparations, we decided to investigate the effect of using dsDNA in monolayer assembly. Double stranded DNA lack the tortuosity of single stranded DNA and its base pairs are unavailable for adsorption directly to the gold surface due to hydrogen bonding. Again, with base pairs unavailable, and an external ordering of anionic groups, dsDNA-dsDNA interactions are expected to be less attractive than uncomplimentary ssDNA-ssDNA interactions. 4.3.1 dsDNA-CY3/MCH Preparation of dsDNA/MCH coated gold beads follows closely the procedure employed for ssDNA/MCH coated gold beads. The dsDNA-CY3 is prepared by combining the 5'hexylthiol and 3'fluorophore labelled ssDNA-CY3 (HS-C6H12-DNA(LN1)-CY3) in immobilization buffer (IB5) with the corresponding complementary sequence (DNA(CompLN1)) and heating to 80ºC, followed by slow cooling to anneal the strands. The solution is then diluted to the appropriate concentration (1 μM, 20 μL) in IB5, and the freshly cleaned gold bead immersed for ~ 1 day. The bead is then rinsed briefly with deionized water and submerged in a solution of MCH (1 μM in IB5) for one hour, rinsed again, and kept in an Eppendorf tube containing a 20 μL solution of IB5 in the absence of light. The surface concentration of DNA was then measured by ruthenium CV ( ΓdsDNA = 4.9 × 1012 duplexes/cm2). Imaging of the surface near OCP (Figure 33) showed a decrease in the number of hotspots compared to samples prepared with ssDNA/MCH, however the surface exhibits a greater degree of inhomogeneity with irregularly shaped islands of evenly fluorescent regions, though some of greater intensity, separated by darker breaches of continuity. As with the ssDNA/MCH samples, switching in the pre-desorption region was found to have a greater relative change for the dark regions, but a greater absolute change in the bright regions. The desorption, measured by capacitance begins at -625 mV, although increases in the fluorescence intensity begin between -650 mV and -675 mV (Figure 34). Regional analysis shows that the desorption occurs within a relatively narrow range, however a reslice (Figure 33C,M) shows that there is a slight regional disparity in reduction potential. 88  Figure 33 Images of dsDNA-CY3/MCH taken during a reductive desorption stepping sequence: +400 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 200 ms exposure, 1000 ms interval time, 695 images. (A) Greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. (B) AVG intensity of each pixel in the stack. (C) Reslice along line NO, shown in (M); desorption region shown. (D) STD of stack pixels and (E) MED value of each pixel in the stack; (F) Used to find areas which show high intensity during desorption by dividing STD by MED; certain areas are shown to fluoresce extremely brightly relative to the amount of fluorescence visible prior to desorption. (G) and (H) are the MAX and MIN intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA; (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN using only those images for potentials in the electrofluorescence modulation region (+400 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD divided by the AVG for the same region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 48 μm x 50 μm, were chosen to investigate specific features of modulation and desorption: hotspots (1, 2, 5) of varying intensities, moderately intense (4, 6, 7, 8, 9) and extremely dark (3) areas for comparison.  89  90  Figure 34 Fluorescence intensities and capacitance for the reductive desorption of ds-DNACY3/MCH (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 33.Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 200 ms exp, 1000 ms interval.  91  It is worth noting that the pattern appearing on the surface is purely due to the organization of the dsDNA on the surface, and not the underlying structure of the substrate. In the brightfield image (Figure 33M), a faint line corresponding to the grain boundary can be apprehended. Figure 35 clearly shows the grain boundary in the underlying substrate through the desorption; the reslice line intersecting the boundary shows from the ~ 50 mV advance in desorption inside of the region and the region's sharp, triangular edge is characteristic of a grain crystal edge. The overlying structure of the dsDNA islands transcends these boundaries, showing little discontinuity at the grain edges. Hence we can conclude that the island-like structure is likely the result of inter-duplex interactions, and not primarily of DNA-gold interaction. DNA clearly desorbs from the whole surface, although the quantity from the darker regions is definitely less according to the regional analysis (Figure 34).  Figure 35 Images showing a substrate-dependent shift in the desorption potential. A shows a fluorescence image before desorption (-725 mV) shows that although in the region near the yellow line GH does show an inhomogeneous distribution of fluorescence, the "blotches are continuous, not dependent on the underlying surface, as image B (-775 mV) shows that the area in the red triangle has been fully desorbed while adjacent regions have not been fully desorbed. C and D show contrast enhanced versions of A and B, which clearly identify the angular region which has been completely desorbed. The line GH perpendicularly intersects the edge of the region; a reslice taken along the line is shown in E. The desorption region is expanded in F. It is clear from the intensities that one side immediately desorbs, while the other requires more time.  The potential dependence of the dsDNA switching shows a much clearer on/off switching than ssDNA (Figure 36), with both dark and bright regions contributing to the changes in fluorescence intensity, although not proportionately to their intensity. An apparent decay in the intensity of the fluorescence intensity at the base potential (Ebase = 0 mV) was observed (Figure  92  37) and initially corrected (Figure 36), however further analysis of small, local dark and bright regions (38) revealed that there appears to be a more complicated trend involved. Very bright regions were found to decay linearly, a reasonable approximation for short timescales of an exponential decay, however very dark regions showed a U-shaped parabolic trend in their fluorescence intensity, with the initial (prior to the 400 mV step) and final (after the -400 mV step) displaying the greatest intensity. Areas of intermediate intensity (Figure 37) show an intermediate effect of the two extremes. This effect may be explained by some factor of irreversibility in the effect of potential, whereby the initial positive steps “contract” the surface, while later negative steps “expand”the previously compressed surface. In any case, here, as with ssDNA/MCH, the layer shows a form of hysteresis dampening the potential response.  Figure 36 Switching of dsDNA-CY3/MCH. Data for a sequential stepping sequence (+400 mV → -400 mV, Ebase = 0 mV, 5×5 images). Change in fluorescence intensity (Δcps) is shown, measured from the minimum value (400 mV); fluorescence at Ebase changes with time, likely due to loss of DNA or photobleaching, and has been adjusted here to give a constant fluorescence at Ebase; original data is shown in Figure 37. Capacitance shown calculated using an estimated area of 0.1 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right). 93  Figure 37 Original data for dsDNA-CY3/MCH; Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown in Figure 36.  Figure 38 Analysis of the time dependence of the fluorescence at Ebase; Estep values not shown. Graph shows fluorescence intensity relative to the initial average (first 5 measurements) of fluorescence intensity. Fluorescence in dark areas (A, open diamonds) shows a parabolic change in intensity, with the final intensity greater than the initial. Fitting with a parabolic curve gave y = 6.823×10-7t2 – 2.019×10-4t+0.9955 (R2 = 0.617). Bright areas (B, black circles) show a more linear decrease in intensity with time, fit linearly with y = 5.882×10-5t + 0.9990 (R2 = 0.916). Initial intensities were: Io,A = 14148 cps; Io,B = 163759 cps. For areas of intermediate intensity, the Ebase intensity curve tended to intermediate shapes; more intense regions producing more linear curves. Regions (A) and (B) are shown (right) on a MAX-MIN, ce, image of the surface. 94  4.3.2 dsDNA-CY5/MCH We shall take note of one further sample of dsDNA/MCH. As part of a series investigating the effect of concentration on the surface structure of DNA using 10, 50, 200, 500, or 1000 mM NaCl in the immobilization buffer (10 mM Tris, no MgCl2), dsDNA-CY5/MCH SAMs were prepared. Here we look at the sample prepared in 500 mM NaCl buffer (IB2). The sample was prepared using the procedure outlined for dsDNA-CY3/MCH, substituting IB2 and the sequences (HS-C6H12-DNA(LN5)-CY5 + DNA(Comp-LN5)). The surface concentration was characterized (ΓdsDNA = 3.1 × 1012 duplexes/cm2) using ruthenium/CV. While similar to the previously described sample, this example highlights some further features characteristic of the range of results observed for dsDNA/MCH samples in general. Again, the surface shows a greater homogeneity than ssDNA/MCH samples, with fewer hotspots and regions of even fluorescence intensity bordered by darker regions (Figure 39). Analysis of the switching region shows the greatest relative and absolute changes to be in the hotspots present on the surface; bright regions tend to show a more uniform switching, while dark regions show a lower relative and absolute switching. Intensity at desorption however does not seem to be evenly distributed among the bright regions. The intensity profile for the whole image shows a double peak for the desorption (Figure ). Regional analysis shows that there are areas desorbing at each of the two potentials, -750 mV and -850 mV, with most areas showing desorptions beginning at both potentials. The capacitance shows desorption initiating at -750 mV. Close inspection of the capacitance indicates that there appears to be some reorganization of the surface between -425 mV and -750 mV not resulting in desorption, but instead lowering the capacitance at Ebase by a small, but observable amount.  95  4.3.3 Conclusions Double stranded DNA (dsDNA) was assembled onto the electrode surface to see if similar surface characteristics would be observed compared to those of ssDNA/MCH monolayers. The surface was passivated using MCH in the same manner as before. Double stranded DNA is known to be easier to manipulate, as the double helix forces a more rigid conformation on the two strands, leading to a rod-like structure with anionic phosphates along the outer edges of the helix, resulting in greater inter-strand repulsion. That said, it is possible for dsDNA to closely associate, forming liquid crystals depending on ionic strength and pH when hydrated. It has recently been demonstrated by Clark et al. that short sequences containing 6 to 20 base pairs of dsDNA can form nematic and columnar liquid crystal phases which readily segregate from bulk solution to form liquid crystalline droplets. SAMs of dsDNA/MCH displayed a mixture of hotspots and low fluorescence regions similar to those of ssDNA/MCH, although to a smaller degree and less frequently. Electromodulation of the dsDNA/MCH monolayers showed that the fluorescence from hotspots modulated in a manner similar to ssDNA/MCH examples. The question remained whether these hotspots are non-specifically adsorbed DNA aggregates that are further away from the electrode than chemisorbed duplexes, or if it is merely a local increase in the DNA concentration on the surface giving greater fluorescence in these special regions. Several SAMs of ssDNA and dsDNA were made with very high DNA surface concentrations, close to or greater than the theoretical maximum (5.4×1012 duplexes/cm2 for dsDNA). The dsDNA layers showed strong fluorescence, but maintained significant heterogeneity in the spatial distribution of fluorescence at non-desorbing potentials. Little electro-fluorescence modulation was observed at these high coverages; large changes and increases in fluorescence intensity were observed mainly during the electrodesorption, which indicates that the whole surface is covered with DNA strands that were removed at negative potentials (-600 mV to -900 mV).  96  Figure 39 Images of dsDNA-CY5/MCH taken during a reductive desorption stepping sequence: 400 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 750 ms exposure, 1095 ms interval time, 695 images. (A) Greyscale bright-field image (20 ms exp., contrast enhanced) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in L with the analogous linear black-to-white gradient. (B) AVG intensity of each pixel in the stack and (C) contrast enhanced AVG. (D) STD of stack pixels and (E) MED value of each pixel in the stack; (F) Used to find areas which show high intensity during desorption by dividing STD by MED. (G) and (H) are the MAX and MIN intensities of each pixel in the stack; when taken over the whole stack, the MAX value tends to be primarily from the desorption, while the MIN value is the fluorescence image of the electrode in the absence of DNA; (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN, gaussian blur 2px, ce, using only those images for potentials in the electro-fluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD divided by the AVG for the same region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 50 x 50 pixels, were chosen to investigate specific features of modulation and desorption: hotspots (1,3,5,7,8) of varying intensity, moderately intense (2,4,5,9) and extremely dark (6) areas for comparison.  97  98  Figure 40 Fluorescence intensities and capacitance for the reductive desorption of ds-DNACY5/MCH (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 39. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 750 ms exp, 1095 ms interval.  99  Therefore, at high surface concentration, electromodulation is prevented and so the hotspots that respond to potential are not regions of chemisorbed DNA in locally high concentrations. Hence the hotspots may be regions where dsDNA has non-specifically bound to the electrode surface as an aggregate and hence exist further from the surface than specifically adsorbed dsDNA. This may be possible if large numbers of dsDNA strands are conglomerated together or assembled into a liquid crystal or an insoluble mass. The greater separation from the electrode would lead to an automatic increase in fluorescence intensity. The response to electromodulation may be the result of changes in conformation or orientation of such structures, changes in density where inefficient packing is present, or even from partial diffusion into solution, as would be expected for particles within an electric field. It seems likely that such aggregates would be partly tethered to the surface through chemisorbed dsDNA molecules interacting with or interpenetrating such clusters. It may also be possible that we are observing liquid crystals of dsDNA that are adsorbed to the surface via end-to-end stacking, which has recently been observed [130], or aggregation by some other means.  4.4 MCH/ss 4.4.1 MCH/ssDNA-CY5 Forming SAMs by the adsorption of 5'hexylthiol DNA followed by passivation with MCH resulted in layers containing large numbers of hotspots, particularly for ssDNA. From experiments where ssDNA has been adsorbed, with MCH subsequently adsorbed in situ while imaging the electrode surface (see Appendix), it can be seen that the inhomogeneity is present prior to MCH adsorption and that the passivation of the surface does little to change this; the potential range for desorption of thiolate ssDNA from an unpassivated surface, although more broad, closely resembles the desorption of ssDNA/MCH. And finally, capacitance measurements of ssDNA/MCH coated surfaces give values ~ 50% higher than surfaces coated with MCH only, indicating an that the surface is not fully coated by hexylthiolate groups. These observations seem to collectively indicate that part of the problem of inhomogeneity lies in the initial adsorption of the DNA. DNA bases have a strong affinity for 100  gold and this has been exploited by others to create DNA SAMs grafted to the surface without thiol appendages, using the Au-base interactions to control graft density[132, 133]. DNA is known to be a “sticky” molecule to work with, and could also form aggregates on the surface. Depending on the density or availability of sites near such surfaces, MCH may be blocked from adsorbing and removing the less strongly bound DNA.To test this, a new means of creating monolayers of thiolate ssDNA and MCH was designed, where the surface would first be passivated with MCH to see if thiol-modified DNA would “stick” to the surface to form hotspots. It had been shown before by Tarlov and Herne that unmodified DNA would not adhere to MCH coated surfaces. A gold bead was subjected to MCH in an aqueous solution (24 hours, 1 mM MCH) to create a high-coverage coating shown by capacitance measurements. Immersion of the MCHcoated gold bead in a solution of thiol modified DNA (HS-C6H12-ssDNA(LN5)-CY5, 1 μM, 20 μL) for 1 hour was tested to determine whether having a passivated surface would prevent hotspot formation resulting from nonspecific adsorption directly on the gold surface. The results were intriguing. Hotspot formation was limited, but DNA was able to bind to the surface, likely via an exchange process with MCH. The DNA showed good switching characteristics and had a much lower fluorescence at the base potential, suggesting that the DNA was near the surface and largely quenched. The preparation method merited further investigation. 4.4.2 MCH/ssDNA-CY3 Following the previous example, an MCH-passivated gold bead was immersed in a buffer solution (IB5) containing ssDNA-CY3 (HS-C6H12-ssDNA(LN1)-CY3, 1 μM, 20 μL) for approximately 24 hours. The sample was then removed and rinsed with deionized water, as per the normal procedure, and the surface concentration determined with ruthenium/CV. The surface concentration was much lower (ΓssDNA = 1.1 × 1010 strands/cm2) than samples made by the conventional method (ssDNA/MCH) by a factor of approximately 100. Images of the surface showed a number of what appeared to be hotspots (Figure 41), however these areas appear to mainly be localities of a mixture of regions higher concentration of ssDNA and surface-bound aggregates. Desorption occurs earlier than in other samples, with 101  Figure 41 Images of MCH/ssDNA-CY3 taken during a reductive desorption stepping sequence: 0 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 2500 ms exposure, 2835 ms interval time, 525 images. (A) Greyscale bright-field image (15 ms exp, ce) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in (L) with the analogous linear greyscale gradient, and have been background subtracted to remove light leakage. (B) AVG intensity of each pixel in the stack; (C) shows the same image contrast enhanced with a 2 px gaussian blur (gb) applied to reduce speckling. (D) STD, ce of stack pixels and (E) MED value of each pixel in the stack, smoothed by taking the mean of every pixel in a 2 px radius; (F) Used to find areas which show high intensity during desorption by dividing STD by MED; appears rather noisy but homogeneous. (G) and (H) are the MAX, ce, and MIN, ce, intensities of each pixel in the stack; large MIN values result from background subtraction resulting in greater relative noise in dark regions. (I) MAX minus MIN is taken to find areas where DNA desorbs. (K) MAX minus MIN, ce, using only those images for potentials in the electrofluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; (J) gives the STD ÷ AVG, ce, for the same potential region to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 48μm x 48μm, were chosen to investigate specific features of modulation and desorption: hotspots (5, 6, 7) of varying intensity were chosen, in addition to moderately intense (1, 2, 8, 9) and extremely dark (3, 4) areas for comparison.  102  103  Figure 42 Fluorescence intensities and capacitance for the reductive desorption of ds-DNACY5/MCH (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 41. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 2500 ms exp, 2835 ms interval. An artifact at -275 mV resulted from a momentary write-delay by the computer, resulting in overexposure, hence the data point was interpolated; refer to the Appendix for the complete data set. 104  fluorescence increase beginning at -575 mV, and changes in the base capacitance at -600 mV (Figure 42); the end of the desorption also appears to occur earlier, with most fluorescence gone by -750 mV, unlike earlier experiments, although the capacitance curves appear to be entirely normative . This suggests that the ssDNA may have a different attachment site compared to that found in ssDNA/MCH surfaces. Analysis of the surface shows that the desorption is quite homogeneous with some areas showing a very low average fluorescence intensity at 0 mV, as would be expected for such a low surface concentration of DNA. One interesting feature of the few hotspots present is their desorption behaviour. During desorption, a number of them can be observed to dislocate from the surface and migrate along the electric field lines (Figure 43). Tracking one such particle with a reslice line along the path shows from the fluorescence intensity that it is desorbing and re-adsorbing in step with the potential pulses. The particulate behaviour would be expected only in the case of an adsorbed DNA aggregate, confirming this as the identity of at least one type of hotspot.  Figure 43 Analysis of a desorbed hotspot moving. (A) shows a MAX-MIN of the MCH/ssDNA-CY3 sample; a bright, diagonal line extends from the middle top to the middle left edges of the image. This line shows the path taken by one desorbed hotspot as it moves across the surface during desorption. (B) shows line FG, in red, superimposed along the path of the hotspot shown in (A); this line in the runs antiparallel to the electric field vector, shown in green. (C) shows the stack resliced along line FG; yellow arrows point to the path traced along the line by the desorbed hotspot, moving with time/image number and potential. An expansion of the desorption region is shown in (D), to emphasize the movement that the particle makes with time. E shows the look-up table, "Cool Blue", used to colour the resliced image in order to highlight the particle's path.  105  Finally, the switching behaviour of such surfaces was investigated using potential step sequences in a region without faradaic reactions of either DNA or the thiolate (+400 mV → -400 mV, Ebase = +400 mV). Bright and dark regions surveyed displayed clear, consistent switching behaviour (Figure 44), contributing in rough proportion to their intensity, and no occurrence of hysteresis with stepping potential. A small drift in the baseline is due to photobleaching. There appears to be some small correlation with the surface substructure as noted by how certain regions contain linear and block-like boundaries.  Figure 44 Switching of MCH/ssDNA-CY3. Data for a sequential stepping sequence (+400 mV → -400 mV, Ebase = +400 mV, 5×5 images). Change in fluorescence intensity (Δcps) is shown, measured from the minimum value (400 mV); the eighth data point was removed and interpolated to correct for a large overexposure due to computer error. Capacitance shown calculated using an area of 0.1153 cm2. Green (All), red (R1) and blue (R2) intensity curves correspond to the whole image (All), bright areas (R1), and dark areas (R2) defined by the ROI trace shown on the STD image shown (right). Exposure time = 1500 ms; interval time 1845 ms.  106  4.4.3 Conclusions MCH passivation of gold has been shown by Herne and Tarlov using to prevent adsorption of unmodified DNA using X-ray photoelectron spectroscopy and 32P-radiolabelled DNA experiments [19, 33]; samples coated with MCH showed no DNA association on the surface after immersion in DNA solutions, followed by rinsing. It was hypothesized that competitive adsorption of the thiol would displace any of the more weakly adsorbing DNA base pairs. Further theoretical work has suggested that the aligned dipoles of the hydroxyl groups in MCH monolayers act to repel DNA. This has become a standard procedure for thiolate-bound DNA SAMs on gold and other electrode metals [23]. Thus, in order to test this, 5'-hexylthiol DNA was assembled onto gold electrode surfaces already passivated with MCH. The surface coverage was approximately 1% of the layers prepared by the initial method as measured by ruthenium reduction and the resulting fluorescence was correspondingly much smaller. The pre-passivation method was also applied using dsDNA (discussed later); pre-passivation resulted in a greatly decreased number density and percent coverage of hotspots for both dsDNA and ssDNA. Modulation of the fluorescence signal with changes in potential was much more clearly evident. Although exposure times were much longer, it should be noted that the improvement in modulation signal was not the result of kinetics, as shorter exposure times for intensely fluorescent samples was generally compensated for with longer imaging interval times. The thiol modified ssDNA was able to bond to the gold surface either by a process of ligand exchange, displacing MCH, or at defect sites. The data is consistent with such exchange processes, whereby the most weakly bonded thiolates are exchanged fastest and accordingly are the first to be replaced. It may also be that such thiolates are inherently unstable and hence these are gaps left in the surface of MCH due to the thermal desorption of weakly bound adsorbate molecules, however the former seems more consistent, as MCH/ssDNA remain relatively stable at room temperature over equal periods of time. It should be noted however that other researchers have observed some layer disruption over time [23]. Hence one may imagine at least 3 adsorption sites with differing free energies associated with them. MCH and the 5'hexylthiol 107  chain appended to DNA would have similar free energies of attachment at a given site in the absence of a strong electric field; the equal length of the alkyl side-chains would result in near equal lateral interactions in the SAM near the surface. Assuming some relatively equivalent free energy at the transition state of all three sites (though this is not entirely necessary), one could envisage a comparison of three such sites as illustrated in the free energy diagram in Figure 45. The free energies of specifically adsorbed MCH plus aqueous DNA (Au-MCH + DNA(aq)) and the exchanged state where DNA is specifically adsorbed and MCH is in solution (Au-DNA + MCH(aq)) are variable, being primarily dependent on the grafting density of DNA, the effect of ionic strength on charge compensation (θ), and the net charge density on the surface. It is difficult to take into account every possible permutation of conditions, hence the effect of the DNA on the exchange of otherwise equivalent hexylthiol chains is disregarded.  Figure 45 An illustration of three comparative free energy curves for adsorbed 5'hexylthiol DNA and MCH on gold. The transition state (‡) for the purpose of the experiment represents either the free site with no thiolate and both DNA and MCH in solution, as would be found at desorption, or some transition state through which the thiol(ate)-DNA replaces the thiolate MCH with ligand exchange. Both processes would have similar coordinates, however the magnitude of ΔG between the transition state and the adsorbed states would differ. The values for the associated reduction potentials are merely representative of the change.  108  As would be expected from such a system, DNA-thiols exchanged into a pre-formed monolayer of MCH would be expected to exhibit more positive desorption potentials and little DNA desorption in more negative regions. Furthermore, layers formed where thiolate DNA monolayers were later passivated with MCH would have most of the weakly bonding sites exchanged during the passivation process, leaving only more strongly bonded thiolates unaffected. This appears to be in accord with observation, although more so with dsDNA/MCH, as ssDNA/MCH monolayers tend to have more nonspecifically adsorbed species which are physically adsorbed and desorb earlier. The relatively even distribution of ssDNA raises the question of what sites would be the least stable and well distributed on the surface. This may be explained by the distribution of disordered regions between grain boundaries and at step edges; such areas of greater disorder are dispersed and thiolates bound at them are known to reductively desorb at smaller potential perturbations.  4.5 MCH/ds 4.5.1 MCH/dsDNA-CY3 Completing the comparative quartet, we examine dsDNA monolayers formed by exchange on an MCH passivated gold bead. MCH/ssDNA-CY3 showed that MCH passivated surfaces limit the nonspecific adsorption of DNA and enable the formation of well defined MCH/DNA SAMs through surface ligand exchange. When dsDNA/MCH monolayers were formed, fewer hotspots were present, likely due to the minimization of attractive surface (Au-DNA) and intermolecular (dsDNA-dsDNA) interactions. It is expected, therefore, that MCH/dsDNA-CY3 will exhibit the fewest hotspots. Furthermore, given the rigidity of the duplex, it may prove able to form layers of slightly higher graft density than MCH/ssDNA-CY3. An MCH/dsDNA-CY3 coated bead was prepared by immersion of an MCH coated bead (as described previously) in a solution of dsDNA-CY3 (1 μM, 20 μL, IB5) for 24 hours, rinsed briefly with deionized water, and stored in an Eppendorf tube containing buffer (20 μL, IB5). Characterization of the surface concentration of dsDNA by ruthenium/CV showed a higher 109  (ΓdsDNA = 3.7 × 1010 duplexes/cm2), but still only one hundredth of that for dsDNA/MCH samples. Analysis of background-subtracted images of the samples near OCP showed an absence of hotspots; although there were large regions of more intense fluorescence, these regions were marked by angular edges, indicating that this is a preferential adsorption to grains of some particular crystallographic face (Figure 46). Switching was relatively homogeneous in the predesorption region, although due to the background subtraction, areas with minimal intensity show a higher noise and hence calculations using these regions appear especially noisy. In the present configuration this is necessary due to light “leakage” through the filter used for CY3 and the longer exposure times required to achieve a consistent signal result in the superposition of an image of the electrode surface on the fluorescence image. Correction is achieved by subtracting the final image in the desorption sequence; the same technique is applied to MCH/ssDNA-CY3. Regional analysis of the desorption (Figure 47) shows a uniformity in the desorption potential with nearly all DNA completely desorbed withing a relatively narrow potential region beginning at -625 mV as measured by both fluorescence and capacitance. The fluorescence at 0 mV is considerably higher compared to the MCH/ssDNA-CY3 sample, as anticipated given the known tendency for dsDNA to result in a greater metal-to-fluorophore distance due to the molecule's rigidity. The fluorescence intensity at certain regions is up to 6 times more intense than the average surface intensity, suggesting that given the appropriate crystallographic face, higher concentrations may be obtainable. Switching experiments on the surface from +400 mV to -400 mV using an Ebase = +400 mV display a remarkable signal to noise ratio and unsurpassed switching response to potential (Figure 48). Both bright and dark regions show the same switching behaviour, although to different extents. The switching appears to be in proportion to the underlying image intensity for both areas in roughly the same proportion (1:0.75) indicating that the surfaces are different only in the local surface concentration of DNA. The switching intensity remains unmatched by samples obtained by ssDNA/MCH or dsDNA/MCH preparations.  110  Figure 46 Images of MCH/dsDNA-CY3 taken during a reductive desorption stepping sequence: 0 mV → -1300 mV, Ebase = 0 mV, 5×5 images, 1500 ms exposure, 1835 ms interval time, 525 images. (A) Greyscale bright-field image (15 ms exp, ce) of the electrode surface. All other images are colour-mapped using the "Royal" look-up table shown in (L) with the analogous linear greyscale gradient, and have been background subtracted to remove light leakage. (B) AVG intensity of each pixel in the stack; (C) shows the AVG in the pre-desorption region (0 mV → -500 mV). (D) STD of stack pixels and (E) MED value of the stack. (F) STD ÷ MED, despeckled to reduce noise, ce. (G) and (H) are the MAX and MIN intensities of the stack; a broad range of MIN values result from background subtraction resulting in greater background noise at low intensities. (I) MAX minus MIN, ce, shows areas where DNA desorbs. (K) MAX minus MIN, ce, using only those images for potentials in the electro-fluorescence modulation region (0 mV → -500 mV) prior to desorption; depicts areas with the most switching, as an absolute change; an ROI is superimposed on this. (J) STD ÷ AVG, ce, for the same potential region (0 mV → -500 mV) to show relative changes. (M) shows a larger version of (C), with ROIs highlighted in square boxes. Nine ROIs, each 50 x 50 pixels, were chosen to investigate specific features of modulation and desorption: hotspots (1, 2, 3, 4) of varying intensity, moderately intense (5, 8, 9) and extremely dark (6, 7) areas for comparison. 111  Figure 47 Fluorescence intensities and capacitance for the reductive desorption of MCH/ssDNA-CY3 (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) as a function of potential step (Estep, V vs. SCE). Fluorescence intensities taken as averages at Estep for the whole image, ROI 0, and for selected regions, ROIs 1-9, shown in Figure 46. Capacitance (μF∙cm-2) values given for Estep (●) and Ebase (■); capacitance values measured at Ebase are mapped to the subsequent Estep potential for graphing purposes. Imaging settings: 1500 ms exp, 1835 ms interval.  112  Figure 48 Switching behaviour of MCH/dsDNA-CY3 (+400 mV → -400 mV, Ebase = +400mV, 5×5 images). Image (A) shows the MIN of the stack, image (B) shows the MAX, ce, and (C) shows MAX minus MIN, ce, with an ROI separating the bright areas (R1) from the dark (R2). The absolute change in fluorescence intensity (Δ cps) is given for R1 (red), R1 (blue), and the whole image (ALL, green). Below the potential step sequence and the measured capacitances are given, scaled with time. Images are not background-subtracted.  4.5.2 Conclusions MCH/ssDNA-CY3 monolayers were prepared by ligand exchange of an MCH-passivated gold bead thiol labelled dsDNA-CY3 to produce a mixed SAM. The results showed a high level of structural homogeneity, and the absence of hotspots on the surface. Fluorescence switching of the surface showed that the DNA responded in proportion to the fluorescence intensity of the local DNA surface coverage and without any hysteresis. The layers did show that DNA adsorbs preferentially to certain surfaces, although all regions appeared to have dsDNA coverage. This again supports the multi-site hypothesis of differing energies of adsorption at certain crystal faces and that at regions of higher disorder MCH is preferentially displaced by dsDNA. MCH/dsDNA appears to be an optimal means to produce SAMs of DNA and MCH, displaying excellent switching behaviour and an absence of hotspots. However this 113  methodology of producing MCH/dsDNA SAMs is not suited to all needs. In most instances, SAMs containing ssDNA are desired as it is this which is capable of performing molecular recognition necessary for sensor functionality. Although MCH/ssDNA does show better layer characteristics than ssDNA/MCH, the layers may not be equal in quality to those produced with dsDNA. Creating an immobilized ssDNA SAM from an MCH/dsDNA SAM however is a viable avenue. DNA can be melted, that is, denatured or un-hybridized, my manipulation of the appropriate conditions: temperature, pH, and low ionic strength. Adsorbing duplexes may also be optimized to denature more easily by introduction of specific base mismatches in the design of the complementary sequence, providing a more facile means of deprotection. Additionally, compared to fluorescently labelled thiolate DNA, the unlabelled complementary strand is procured at far less a premium. Further experimental details, descriptions, and figures can be found in the appendices.  114  5 Summary and Conclusions This thesis has examined the organization of single stranded and double stranded DNA at the surface of MCH passivated gold electrodes using in situ fluorescence imaging of the electrode surface in an electrochemical cell, observing the fluoro-electromodulation of DNA and the reductive desorption of DNA, along with measurements of the surface density of DNA. The experimental results presented in this thesis provide a picture of inhomogeneity resulting from nonspecific adsorption of ssDNA/MCH layer prepared according to standard procedures. Using pre-hybridized duplexes to form dsDNA/MCH on the electrode surface afforded similarly heterogeneous surfaces. In order to eliminate the problem of nonspecific adsorption, surfaces passivated with MCH were exposed to thiol derivatized ss and dsDNA, which, through a process of ligand exchange, were adsorbed to the surface with chemical specificity. In the absence of applied potentials, 5'hexylthiol derivatized ssDNA labelled with CY3 or CY5 at the 3' end were adsorbed to thermally annealed gold-beads in aqueous immobilization buffers containing Tris (pH ~ 7.5) and chloride salts to control the ionic strength, and subsequently passivated with MCH (1mM). Characterization of the surface density of ssDNA using ruthenium/CV showed densities greater than 1×1012 strands·cm-2; near the theoretical maximum. Analysis of the surface using in situ epi-fluorescence microscopy at potentials near the OCP showed hotspots – regions of excessive fluorescence intensity. These bright regions showed different potential responses than darker regions. The hotspots were later determined to be clumps of nonspecifically adsorbed ssDNA, as indicated by their intensities at desorption and the inelastic response to potential switching. Unmodified ssDNA can bind strongly to gold surfaces and has been shown to effectively compete with the adsorption of thiols [18, 132, 133]. This suggests that the regions of ssDNA may be forming aggregates in solution or at the surface, which preclude MCH adsorption and DNA displacement at those sites. Imaging of ssDNA electrodes without MCH showed heterogeneity, and in situ imaging of the MCH passivation of one such sample showed that the hotspots did not change significantly during passivation (see Appendix). The desorption peak of ssDNA/MCH treated samples was also particularly broad compared to samples prepared by other means, another indication that not all of the DNA was 115  tethered to the surface through thiolate-gold bonds. The observed heterogeneity for ssDNA/MCH prompted an investigation of dsDNA/MCH SAMs as an alternative means of preparation under otherwise identical conditions. Given the greater charge per complex, repulsive duplex-duplex interactions were expected to be greater in addition to the rigidity of short dsDNA duplexes preventing entanglement. However imaging of the dsDNA/MCH monolayers also showed hotspots and inhomogeneity, albeit to a lesser extent than ssDNA/MCH. The surface density of duplexes was also found to be quite high (>1×1012 duplexes·cm-2). Fluorescence switching in response to potential was minimal at the high surface concentrations, and a strange dependence was observed for the intensity at the base potential (0 mV), which should be invariant other than for a slow decay due to photobleaching, suggestive of some form of hysteresis process with potential stepping. Reductive desorption of the dsDNA/ MCH samples showed multiple adsorption sites as indicated by 2 to 3 desorption peaks, indicating a substrate-dependence on the desorption potential, however the initial intensity and assembly was not dependent on the desorption potential of a given region. It became clear that a new method was required in order to produce DNA SAMs where nonspecific adsorption would be nonexistent. It was found that 5'hexylthiol modified DNA could effectively undergo ligand exchange with MCH to produce mixed MCH/ssDNA or MCH/ dsDNA monolayers under conditions similar to those used in the immobilization of ssDNA. Fluorescence microscopy showed the layers to be much less intense compared to ssDNA/MCH and dsDNA/MCH samples, ranging from 0.1 to 1% of the intensity, while the surface concentration of ssDNA was approximately 1% of that for ssDNA/MCH samples. Fluorescence switching was remarkable in that no hysteresis and only a small decay in the fluorescence intensity at the base potential (+400 mV). Moreover, all regions appeared to switch in proportion to their intensities near the OCP. Desorption of the monolayers showed that the fluorescence intensity at positive potentials was small relative to the background intensity after desorption, indicating that near +400 mV, the fluorescence is almost completely quenched. The desorption peak was singular and narrow, coming off at a less negative potential than other samples do. This corroborates well with the understanding that ligand exchange initially targets  116  the most weakly bound thiolates [51], and the observation that the major effect of MCH is to exchange thiolate-bound DNA [18]. It is likely that in samples where dsDNA is assembled first, subsequent MCH assembly exchanges all of the most weakly bound thiolate DNA, leaving those most strongly bound unaffected. When MCH is assembled first, ds or ssDNA displaces the most weakly bound thiolate MCH first. The result is that the exchanged-in DNA is slightly more weakly bound, desorbing at a less negative potential. A very small amount of heterogeneity was present in the surfaces, as fewer, smaller, and less intense, regions of relative brightness on the surface were observed (though still much darker than the hotspots of ds or ssDNA/MCH). It was especially curious to observe several of these smaller hotspots desorb as a clump, moving along the surface in the direction of the counter electrode, as this supports the hypothesis of ssDNA aggregates forming. Finally, MCH pre-passivation was tested with dsDNA to form MCH/dsDNA SAMs. With this process, no hotspots were observed. The intensity of the fluorescence near OCP was lower, although slightly greater than MCH/ssDNA monolayers; were hotspots to exist, the intensities of them would not enable observation of the surrounding dsDNA without overexposure of the hotspot. The surface concentration was on the order of 1% in comparison to dsDNA/MCH. Fluorescence switching was simply magnificent, showing clear changes in fluorescence intensity; the change from +400 mV to -400 mV was in switching was only slightly less than the change in fluorescence intensity from the brightest pre-desorption intensity to the lowest postdesorption intensity. Desorption occurred with a narrow potential region, slightly earlier than dsDNA, with all regions' intensity peaks in the same potential region. However some regional variations in intensity are present, however the response of each region was in proportion to its intensity, indicating that it is merely a difference in local surface density. The bounding lines of these regions appear to be a result of preferences in adsorption region, although this could mean that the density limit for SAMs prepared via this method could be up to 5x higher depending on the nature of the gold substrate used. This may be an upper limit, imposed upon the surface density achievable via this method, however low-density surfaces may be more appropriate as it prevents cross-hybridization of DNA containing mismatches between neighbouring probe strands. Double stranded DNA adsorbed through ligand exchange of MCH may be the optimal  117  means to produce DNA SAMs. From these, ssDNA SAMs could readily be prepared either through some combination of electronic [123], thermal, pH, and ionic stringency controls or through creating an inherently destabilized dimer containing some combination of mismatches providing a metastable state during adsorption which could be removed with minimal difficulty and effort by denaturation to provide an MCH/ssDNA surface. The present MCH/ssDNA methodology may however already be sufficient with minor modifications. The reworking of the original method for making ssDNA/MCH monolayers has produced mixed SAMs of thiolate DNA and MCH where nonspecifically adsorbed DNA is either insignificant or nonexistent. This is a considerable improvement over ssDNA/MCH SAMs which has been the standard thiol SAM preparation for DNA SAMs. Furthermore, because this methodology employs no new reagents, equipment, nor substrates, it is expected that it will find immediate application in the preparation of ssDNA SAMs with MCH on gold. This revamping of the old method will no doubt change the fabrication procedures for simple ssDNA SAMs upon dissemination. Additional modifications and optimizations will no doubt be found, however the technique developed here represents a step forward to greater control of the interfacial molecular architecture needed to advance the practical application of DNA-based sensors.  118  6 Future Work 6.1 Introduction The initial application of electro-fluorescence spectroscopy to DNA SAMs has provided new information regarding their structure and heterogeneity. Opportunities to answer further questions with this technique and to improve on previous work remain, and some specific areas are described. In particular, the technique potentially offers a better understanding of hybridization at the metal-solution SAM interface in particular regarding why such poor hybridization efficiencies are so frequently recorded.  6.2 Flat Substrates While curved bead surfaces have provided a more than adequate substrate for the investigation of DNA SAM monolayer formation, heterogeneity, and desorption using epifluorescence microscopy, there are advantages to be gained using flat substrate. In particular, any issues related to surface roughness would be minimized, while the entire area presented by the bead would be in the plane of focus, enabling a greater field of view. The simplest option is to refashion beads such that the bottom surface is ground flat as shown in Figure 49 (left); this can be achieved by embedding the beads in an epoxy resin as in Figure 49 (right), then grinding and polishing to give a flat surface comprised of different crystalline domains. Several flatbottomed beads are in preparation. A second option would be to use an atomically flat surface. Single crystal Au(111) substrates while available and more than sufficient are highly expensive and easily damaged, thus would likely be neither sufficiently robust for multiple experiments nor feasible. Another option would be to use an electrode derived from a gold coated mica slide. Gold films grow epitaxially on the (100) mica surface, forming a strongly oriented (111) gold surface, providing a "pseudo-single crystal". Such substrates are readily available, however electrical interconnects and mounting would prove to be difficult for fluorescence measurements and would likely affect 119  Figure 49 Left: a round bead in the holder show and a flat-bottomed bead in the same holder. Right: Seven gold beads immobilized in an epoxy resin. The beads were oriented standing on a surface of modelling clay (plasticine) and the area enclosed by a 2.5 cm long section of plastic tubing, 2.5 cm in diameter and filled with epoxy resin. The resin was allowed to set for 2 weeks, and the modelling clay was removed using soap and a toothbrush followed by chipping off the plastic tubing. the electrochemistry of such systems. Furthermore, flat substrates would have the added advantage of being able to be probed by atomic force microscopy (AFM). The selectively desorbing DNA from the surface would leave behind vacant sites; it would be interesting to see where these sites are located in the monolayer.  6.3 In situ Adsorption/Passivation with MCH The inaugural experiment of adsorbing MCH to a ssDNA, while useful in showing that MCH was not eliminating hotspots, was afflicted by the photobleaching of the CY5 dye. Further experiments using a more photo-stable dye (e.g. CY3) and better characterization of both the surface concentration of DNA and the monolayer quality are necessary. In situ imaging of the passivation of a dsDNA monolayer would also be profitable, given that fewer hotspots are generally present, it may speak to the cause. Better ventilation will be necessary for these 120  experiments given the relatively large amount of aqueous MCH used, its volatility, and noxious odour.  6.4 In situ Hybridization of DNA The second generation spectro-electrochemical cell offers the possibility of in situ detection of hybridization of electrode-immobilized ssDNA probe sequences with target sequences using fluorescence microscopy. This is especially feasible using a labelled probe sequence and an unlabelled target sequence. Unlabelled DNA is far more economical to procure, and given the ~ 5 ml volume of solution required, this is the most sensible, and would provide an initial step toward more limiting analytes. One proposed experiment is to use fluorescence (or, more generally "Förster") resonance energy transfer (FRET) to determine unambiguously those regions of the DNA SAM where hybridization is able to occur and those where it is not. This would give further insight to the nature of the "hotspots" observed in ssDNA and dsDNA SAMs which we have observed. FRET involves a nonradiative energy transfer from the excited state of a "donor" molecule to the excited state of an "acceptor" molecule; the acceptor molecule can then undergo relaxation by fluorescence, while the donor molecule is quenched. This dipole-dipole interaction is strongly dependent on proximity of the two fluorophores, and generally less than 10 to 100 Å. An overlap (J(λ)), shown in Figure 50 (C), of the donor fluorescence spectrum and the acceptor absorbance spectrum is required for two fluorophores to form a suitable donor-acceptor FRET pair. Finally, the transition dipoles of donor and acceptor molecules need to have approximately parallel orientations; orthogonal dipoles will not undergo FRET. Using probe and target sequences labelled for FRET would enable co-localization analysis for hybridization using fluorescence at two wavelengths. In the most likely configuration, the donor-dye labelled probe sequence would have a 5' hexylthiol modification and a 3' donor dye; the complementary target sequence would have a 5' acceptor dye, giving a duplex with both donor and acceptor in close proximity in the duplex as pictorially shown in Figure 50 (A&B). Two particular experiments could be performed with such a system: the first would be to test the 121  Figure 50 (A). FRET experiment where dsDNA containing a donor (D) dye and an acceptor (A) dye is immobilized on the substrate, the FRET fluorescence measured, and the duplex denatured. (B). FRET experiment where a ssDNA probe with donor (D) is immobilized, fluorescence measured, and then the surface is hybridized in situ or ex situ, and the FRET signal measured. (C). A diagram showing the overlap integral J(λ) for the spectral overlap between the donor fluorescence and the acceptor adsorption, on which the efficiency of FRET depends.  122  efficacy of FRET near the metal surface in our system and to observe the denaturing of the duplex. A dsDNA oligonucleotide consisting of both probe and target sequence would be immobilized, using either of the methods outlined in this thesis, and the conditions adjusted so as to induce de-hybridization, as shown in Figure 50 (A). The second experiment, depicted in Figure 50 (B), would be to immobilize ssDNA probes on the surface and, ideally, to hybridize the target sequence carrying the acceptor dye in situ, although ex situ could be performed. Colocalization of the fluorescence from each dye would indicate where hybridization occurs whether or not it is occurring at hotspots, and to what extent.  6.5 Investigations of Nanostructured Au and Pd Surfaces Kelley and coworkers have developed nanostructured gold and palladium surfaces for the immobilization and detection of DNA and other biomolecular markers in solution using ruthenium (III) hexaammine, [Ru(NH3)6]3+, mediated electrocatalytic reduction of ferricyanide, [Fe(CN)6]3- [153-155]. Nanostructures in the form of an array of gold nanowires and Pd or Au "florets" produced in silicon pores have been shown to have superior hybridization efficiencies, as measured by charge transfered during electrocatalytic reduction. It remains unclear if this is the result of greater efficiency relative to the number of DNA strands on the surface or if it is a diffusion effect [156]. Using in situ hybridization with FRET active donor and acceptor probe and target sequences to provide a means of characterizing the hybridization of DNA on these surfaces.  6.6 Surface Concentration of DNA Some researchers remain sceptical regarding the use of ruthenium (III) hexaammine, [Ru(NH3)6]3+, in the determination of DNA surface concentration. While in our hands it appears to be generally applicable and effective, some questions remain: for instance, does it measure nonspecifically adsorbed DNA [28], or only chemisorbed DNA? While further work correlating the remaining DNA in deposition solutions would be productive, the method leaves the possibility that some nonspecifically adsorbed DNA is simply washed off of the bead, producing the a possible systematic overestimate. This could possibly be addressed by including an 123  internal redox indicator with a redox potential that would avoid overlap with ruthenium in CV measurements. Other methods available include use of high concentrations of a passivating thiol to exclude all DNA; the DNA concentration is then measured in solution. This may prove to be a more practical control.  6.7 Creation of Distance-Quenching Curves Another possible avenue of research is in creating distance-quenching curves for fluorescent dyes with DNA. The luminescence of molecules near the metal surface is influenced by the non-radiative transfer of energy to the metal surface; this results in quenching of molecular fluorescence near the metal surface as already mentioned. Quenching-distance profiles have been created for several dyes; Bizzotto and Shepherd have created quenching profiles for fluorescein and cyanine type dyes using gold-coated glass slides covered with steps of SiO2 in various thicknesses [57]. Plotting as a function of separation distance (d) gives a curve similar to what is shown in Figure 51, were fluorescence intensity near bulk metal surfaces varies as d3 before levelling off. This data can be used to provide an estimate of the space between the dye and the metal surface. DNA monolayers provide a potential means to obtain such quenching-distance profiles. Fluorescence intensity, given in counts per pixel or per unit area, from a DNA-coated surface, normalized for DNA surface concentration (ΓDNA), would give data for counts per molecule, assuming a reasonably accurate determination of ΓDNA. Using rigid double stranded DNA sequences of longer or shorter lengths in conjunction with controlling the orientation (perpendicular to the surface) with potential would provide a means of controlling the separation distance without fabricating steps. Finally, reductive desorption of the DNA monolayers would provide an "unquenched" reference point for each measurement. In order for this methodology to succeed, fluorophores less susceptible to photobleaching would be required, and a variety of such dyes are commercially available; these could provide multiple reference points in studying the orientation of adsorbed DNA, since each fluorophore's quenching follows a different curve. 124  Figure 51 Bottom: A schematic depiction of the correlation between distance and fluorescence intensity for fluorophores near bulk metal surfaces. Quenching close to the metal surface shows a d-3 dependence. Top: Depictions of DNA of various lengths labelled with a fluorescent dye (F). Several monolayer assemblies could be constructed, each with a different DNA chain length in order to construct a quenching-distance curve. Desorption of the DNA would provide a means to find the maximum fluorescence, providing a reference point to measure against bulk solution measurements for the same fluorophore.  125  6.8 Fluorescence Lifetime Imaging Microscopy Fluorescence lifetime imaging microscopy (FLIM) is a relatively new technique. Whereas conventional fluorescence microscopy measures the wavelength and intensity of fluorescent dyes, FLIM measures the lifetimes of fluorescent dyes [157]. Emission lifetime can be affected by a variety of environmental factors affecting the fluorophore, including FRET and proximity to metal surfaces [57]. This would be of particular use in studying surface fluorescence inhomogeneity in DNA SAMs, where regions of high anomalously high intensity could arise either from areas where packing of the DNA on the surface leads to a local high ΓDNA or where DNA aggregates have formed, having inherently greater metal-dye separation. A new microscope for FLIM is currently being installed and will be available early in the summer of 2008.  6.9 Particle Tracking The observation of desorbed particles moving across the electrode surface during desorption provides an avenue to characterizing their properties, in particular their size. Stationary particles provide a means of mapping such aggregates in two dimensions, however moving and rotating particles would provide the chance to gain a 3D size range. Furthermore, as the particles are moving in response to the electric field, this may provide a measurement in terms of charge or size, as the particles experience drag, with the measurement of particle velocity. ImageJ [149] provides a number of plugin programs for particle tracking [151], which should enable this to be accomplished.  126  References (1)  The International SNP Map Working Group Nature. 2001, 409, 928-933.  (2)  Chakravarti, A. Nature. 2001, 409, 822-823.  (3)  Liggett, S. B.; Cresci, S.; Kelly, R. J.; Syed, F. M.; Matkovich, S. J.; Hahn, H. S.; Diwan, A.; Martini, J. S.; Sparks, L.; Parekh, R. R.; Spertus, J. A.; Koch, W. J.; Kardia, S. L. R.; Dorn II, G. W. Nat. Med. 2008, 14, 510-517.  (4)  Kolata, G. Genes Explain Race Disparity in Response to a Heart Drug. The New York Times, April 29, 2008, http://www.nytimes.com/2008/04/29/health/research/29heart.html.  (5)  Schulze, J. J.; Lundmark, J.; Garle, M.; Skilving, I.; Ekstrom, L.; Rane, A. J. Clin. Endrocrinol. Metab. 2008, advance online publication, (doi: 10.1210/jc.2008-0218).  (6)  Kolata, G. Some Athletes’ Genes Help Outwit Doping Test. The New York Times, April 30, 2008, http://www.nytimes.com/2008/04/30/sports/30doping.html.  (7)  Schafer, A. J.; Hawkins, J. R. Nat. Biotech. 1998, 16, 33-39.  (8)  Patolsky, F.; Lichtenstein, A.; Willner, I. Nat. Biotech. 2001, 19, 253-257.  (9)  Boon, E. M.; Ceres, D. M.; Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotech. 2000, 18, 1096-1100.  (10) Chakravarti, A.; Little, P. Nature. 2003, 421, 412-414. (11) Boon, E. M.; Salas, J. E.; Barton, J. K. Nat. Biotech. 2002, 20, 282-286. (12) Rodriguez-Mozaz, S.; Marco, M.; Lopez de Alda, M. J.; Barceló, D. Anal. Bioanal. Chem. 2004, 378, 588-598. (13) Mascini, M.; Bagni, G.; Pietro, M.; Ravera, M.; Baracco, S.; Osella, D. BioMetals. 2006, 19, 409-418. (14) Minunni, M.; Tombelli, S.; Mascini, M.; Bilia, A.; Bergonzi, M. C.; Vincieri, F. Talanta. 2005, 65, 578-585. (15) Huang, Y. C.; Ge, B.; Sen, D.; Yu, H. J. Am. Chem. Soc. 2008, 130, advance online publication (doi: 10.1021/ja8011066). (16) Mello, L. D.; Hernandez, S.; Marrazza, G.; Mascini, M.; Kubota, L. T. Biosens. Bioelectron. 2006, 21, 1374-1382.  127  (17) Wang, J.; Rivas, G.; Cai, X.; Palecek, E.; Nielsen, P.; Shiraishi, H.; Dontha, N.; Luo, D.; Parrado, C.; Chicharro, M.; Farias, P. A. M.; Valera, F. S.; Grant, D. H.; Ozsoz, M.; Flair, M. N. Anal. Chim. Acta. 1997, 347, 1-8. (18) Steel, A. B.; Levicky, R. L.; Herne, T. M.; Tarlov, M. J. Biophys. J. 2000, 79, 975-981. (19) Herne, T.; Tarlov, M. J. Am. Chem. Soc. 1997, 119, 8916-8920. (20) Steel, A.; Herne, T.; Tarlov, M. Anal. Chem. 1998, 70, 4670-4677. (21) Steel, A.; Herne, T.; Tarlov, M. Bioconjugate Chem. 1999, 10, 419-423. (22) Levicky, R.; Herne, T.; Tarlov, M.; Satija, S. J. Am. Chem. Soc. 1998, 120, 9787-9792. (23) Tarlov, M.J.; Steel, A.B. In Biomolecular films : design, function, and applications; Rusling, J. F., Ed.; Surficant Science Series; Marcel Dekker: New York, 2003; Vol. 111, p. 367. (24) B. Montgomery Pettitt; Vainrub, A.; Wong, K. In Ionic Soft Matter: Modern Trends in Theory and Applications; Nato Science Series; Springer Netherlands: Dordrecht ;; [London], 2005; Vol. 206, pp. 381-393. (25) Vainrub, A.; Pettitt, B. M. Phys. Rev. E. 2002, 66, 041905. (26) Wong, K.; Pettitt, B. M. Theor. Chem. Acc. 2001, 106, 233-235. (27) Ka-Yiu Wong, B. M. P. Biopolymers. 2004, 73, 570-578. (28) Lao, R.; Song, S.; Wu, H.; Wang, L.; Zhang, Z.; He, L.; Fan, C. Anal. Chem. 2005, 77, 6475-6480. (29) Shen, G.; Tercero, N.; Gaspar, M.; Varughese, B.; Shepard, K.; Levicky, R. J. Am. Chem. Soc. 2006, 128, 8427-8433. (30) Gong, P.; Levicky, R. Proc. Nat. Acad. Sci. U.S.A. 2008, 105, 5301-5306. (31) Levicky, R.; Horgan, A. Trends Biotechnol. 2005, 23, 143-149. (32) Gao, Y.; Wolf, L. K.; Georgiadis, R. M. Nucleic Acids Res. 2006, 34, 3370-3377. (33) Gong, P.; Lee, C.; Gamble, L.; Castner, D.; Grainger, D. Anal. Chem. 2006, 78, 3326-3334. (34) Stoodley, R.; Shepherd, J.; Wasan, K. M.; Bizzotto, D. Biochim. Biophys. Acta Biomembr. 2002, 1564, 289-297. 128  (35) Stauffer, V.; Stoodley, R.; Agak, J. O.; Bizzotto, D. J. Electroanal. Chem. 2001, 516, 73-82. (36) Shepherd, J.; Yang, Y.; Bizzotto, D. J. Electroanal. Chem. 2002, 524-525, 54-61. (37) Agak, J. O.; Stoodley, R.; Retter, U.; Bizzotto, D. J. Electroanal. Chem. 2004, 562, 135-144. (38) Bizzotto, D.; Yang, Y.; Shepherd, J. L.; Stoodley, R.; Agak, J.; Stauffer, V.; Lathuilliere, M.; Akhtar, A. S.; Chung, E. J. Electroanal. Chem. 2004, 574, 167-184. (39) Shepherd, J.; Kell, A.; Chung, E.; Sinclar, C.; Workentin, M.; Bizzotto, D. J. Am. Chem. Soc. 2004, 126, 8329-8335. (40) Cho, H.; Parameswaran, M.; Yu, H. Sensor. Actuator. B Chem. 2007, 123, 749-756. (41) Yu, H. Anal. Chem. 2001, 73, 4743-4747. (42) Su, L.; Sankar, C.; Sen, D.; Yu, H. Anal. Chem. 2004, 76, 5953-5959. (43) Yu, H.; Luo, C.; Sankar, C.; Sen, D. Anal. Chem. 2003, 75, 3902-3907. (44) Bizzotto, D.; Pettinger, B. Langmuir. 1999, 15, 8309-8314. (45) Rant, U.; Arinaga, K.; Fujita, S.; Yokoyama, N.; Abstreiter, G.; Tornow, M. Nano Lett. 2004, 4, 2441-2445. (46) Sendner, C.; Kim, Y. W.; Rant, U.; Arinaga, K.; Tornow, M.; Netz, R. R. Phys. Status Solidi A. 2006, 203, 3476-3491. (47) Rant, U.; Arinaga, K.; Tornow, M.; Kim, Y. W.; Netz, R. R.; Fujita, S.; Yokoyama, N.; Abstreiter, G. Biophys. J. 2006, 90, 3666-3671. (48) Rant, U.; Arinaga, K.; Fujita, S.; Yokoyama, N.; Abstreiter, G.; Tornow, M. Langmuir. 2004, 20, 10086-10092. (49) Rant, U.; Arinaga, K.; Scherer, S.; Pringsheim, E.; Fujita, S.; Yokoyama, N.; Tornow, M.; Abstreiter, G. Proc. Nat. Acad. Sci. U.S.A. 2007, 104, 17364-17369. (50) Tornow, M.; Arinaga, K.; Rant, U. In NanoBioTechnology; Humana Press, 2008; pp. 187-214. (51) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. Rev. 2005, 105, 1103-1170. 129  (52) Hong, R.; Han, G.; Fernandez, J.; Kim, B.; Forbes, N.; Rotello, V. J. Am. Chem. Soc. 2006, 128, 1078-1079. (53) Mouna Zachary, V. C. Angew. Chem. Int. Ed. 2007, 46, 3304-3307. (54) Chan, V.; Graves, D. J.; McKenzie, S. E. Biophys. J. 1995, 69, 2243-2255. (55) Cho, Y.; Kim, S.; Kim, Y. A.; Lim, H. K.; Lee, K.; Yoon, D.; Lim, G.; Pak, Y. E.; Ha, T. H.; Kim, K. J. Colloid Interface Sci. 2004, 278, 44-52. (56) Jeffrey L. Shepherd Ph. D. Dissertation, University of British Columbia, 2005. (57) Bizzotto, D.; Shepherd, J. L. In Advances in Electrochemical Science and Engineering; Advances in Electrochemical Sciences and Engineering Series; Wiley-VCH, 2006; Vol. 9, p. 445. (58) Davidson, M. W.; Abramowitz, M. Optical Microscopy; Florida State University: Tallahassee, Florida, 1999; p. 41. (59) Kapitza, H. G. Microscopy from the very beginning; 2nd ed.; Carl Zeiss Jena Gmbh, 1997. (60) Fare, T.; Coffey, E.; Dai, H.; He, Y.; Kessler, D.; Kilian, K.; Koch, J.; LeProust, E.; Marton, M.; Meyer, M.; Stoughton, R.; Tokiwa, G.; Wang, Y. Anal. Chem. 2003, 75, 4672-4675. (61) Olympus America Inc. Green Excitation Mirror Units : U-MNG2 : UIS2 Series : OLYMPUS. http://www.olympusamerica.com/seg_section/uis2/seg_uis2_mr_umng2.asp. (62) Kolb, D. M. In Spectroelectrochemistry Theory and Practice; Gale, R. J., Ed.; Plenum Press: New York, 1988. (63) Faulkner, L. R.; Bard, A. J. Electrochemical Methods: Fundamentals and Applications; 2nd ed.; Wiley, 2000; p. 856. (64) Schmickler, W. Interfacial Electrochemistry; Oxford University Press: Oxford, 1996. (65) Poirier, G. E.; Tarlov, M. J. Langmuir. 1994, 10, 2853-2856. (66) Wetterer, S.; Lavrich, D.; Cummings, T.; Bernasek, S.; Scoles, G. J. Phys. Chem. B. 1998, 102, 9266-9275. (67) Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733-740.  130  (68) Schlenoff, J. B.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528-12536. (69) Donkers, R.; Song, Y.; Murray, R. Langmuir. 2004, 20, 4703-4707. (70) Hostetler, M.; Templeton, A.; Murray, R. Langmuir. 1999, 15, 3782-3789. (71) Song, Y.; Murray, R. J. Am. Chem. Soc. 2002, 124, 7096-7102. (72) Song, Y.; Huang, T.; Murray, R. J. Am. Chem. Soc. 2003, 125, 11694-11701. (73) Woehrle, G.; Brown, L.; Hutchison, J. J. Am. Chem. Soc. 2005, 127, 2172-2183. (74) Mulder, W.; Calvente, J.; Andreu, R. Langmuir. 2001, 17, 3273-3280. (75) Kawaguchi, T.; Yasuda, H.; Shimazu, K.; Porter, M. Langmuir. 2000, 16, 9830-9840. (76) Gorman, C. B.; Biebuyck, H. A.; Whitesides, G. M. Langmuir. 1995, 11, 2242-2246. (77) Imabayashi, S.; Hobara, D.; Kakiuchi, T.; Knoll, W. Langmuir. 1997, 13, 4502-4504. (78) James Robertson; A. M. Ross; L. Burgoyne DNA In Forensic Science: Theory, Techniques And Applications; Ellis Horwood Series in Forensic Science; CRC Press, 1990; p. 197. (79) Mathews, C. K.; Holde, K. E. V.; Ahern, K. G. Biochemistry; 3rd ed.; Prentice Hall, 1999; p. 1200. (80) Li, A. Z.; Marx, K. A. Biophys. J. 1999, 77, 114-122. (81) Li, A. Z.; Huang, H.; Re, X.; Qi, L. J.; Marx, K. A. Biophys. J. 1998, 74, 964-973. (82) Manning, G. S. Q. Rev. Biophys. 1978, 11, 179-246. (83) Manning, G. S. Acc. Chem. Res. 1979, 12, 443-449. (84) Manning, G. S. Biophys. Chem. 2002, 101-102, 461-73. (85) Calladine, C. R.; Drew, H. R. Understanding DNA: The Molecule and How It Works; 2nd ed.; Academic Press: Cambridge, 1997; p. 283. (86) Schroeder, G. K.; Lad, C.; Wyman, P.; Williams, N. H.; Wolfenden, R. Proc. Nat. Acad. Sci. U.S.A. 2006, 0510879103. (87) Chargaff Cell. Mol. Life Sci. 1950, 6, 201-209.  131  (88) Owczarzy, R. Biophysical Chemistry. 2005, 117, 207-215. (89) Kerman, K.; Kobayashi, M.; Tamiya, E. Meas. Sci. Tech. 2004, 15, R1-R11. (90) Wang, J. Nucleic Acids Res. 2000, 28, 3011-3016. (91) Murphy, L. Curr. Opin. Chem. Biol. 2006, 10, 177-184. (92) Wang, J. Biosens. Bioelectron. 1998, 13, 757-762. (93) Yang, M.; McGovern, M. E.; Thompson, M. Anal. Chim. Acta. 1997, 346, 259-275. (94) Gooding, J. J. Electroanalysis. 2002, 14, 1149-1156. (95) He, P.; Xu, Y.; Fang, Y. Anal. Lett. 2005, 38, 2597-2623. (96) Lucarelli, F.; Tombelli, S.; Minunni, M.; Marrazza, G.; Mascini, M. Anal. Chim. Acta. 2008, 609, 139-159. (97) Odenthal, K. J.; Gooding, J. J. Analyst. 2007, 132, 603-610. (98) Susan R. Mikkelsen Electroanalysis. 1996, 8, 15-19. (99) Wang, J. Anal. Chim. Acta. 2002, 469, 63-71. (100) Zhai, J.; Cui, H.; Yang, R. Biotechnol. Adv. 1997, 15, 43-58. (101) Vercoutere, W.; Akeson, M. Curr. Opin. Chem. Biol. 2002, 6, 816-822. (102) Nagel, B.; Dellweg, H.; Gierasch, L. M. Pure Appl Chem. 1992, 64, 143-168. (103) Wilding, P.; Joos, T.; Kricka, L. J.; Shi, L. Pure Appl Chem. 2006, 78, 667-684. (104) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotech. 2003, 21, 1192-1199. (105) Ramsay, G. Nat. Biotech. 1998, 16, 40-44. (106) Thevenot, D. R.; Toth, K.; Durst, R. A.; Wilson, G. S. Biosens. Bioelectron. 2001, 16, 121-131. (107) Thevenot, D.; Toth, K.; Durst, R.; Wilson, G. Pure Appl Chem. 1999, 71, 2333-2348. (108) Peterlinz, K.; Georgiadis, R.; Herne, T.; Tarlov, M. J. Am. Chem. Soc. 1997, 119, 3401-3402.  132  (109) de la Torre, B. G.; Avino, A. M.; Escarceller, M.; Royo, M.; Albericio, F.; Eritja, R. Nucleos. Nucleot. Nucleic Acids. 1993, 12, 993 - 1005. (110) Goursot, A.; Pedocchi, L.; Coq, B. J. Phys. Chem. 1994, 98, 8747-8751. (111) Lee, G.; Chrisey, L.; Colton, R. Science. 1994, 266, 771-773. (112) Okahata, Y.; Matsunobu, Y.; Ijiro, K.; Mukae, M.; Murakami, A.; Makino, K. J. Am. Chem. Soc. 1992, 114, 8299-8300. (113) Rabke-Clemmer, C. E.; Leavitt, A. J.; Beebe, T. P. Langmuir. 1994, 10, 1796-1800. (114) Arinaga, K.; Rant, U.; Tornow, M.; Fujita, S.; Abstreiter, G.; Yokoyama, N. Langmuir. 2006, 22, 5560-5562. (115) Imabayashi, S.; Hobara, D.; Kakiuchi, T.; Knoll, W. Langmuir. 1997, 13, 4502-4504. (116) Satjapipat, M.; Sanedrin, R.; Zhou, F. Langmuir. 2001, 17, 7637-7644. (117) Erts, D.; Polyakov, B.; Olin, H.; Tuite, E. J. Phys. Chem. B. 2003, 107, 3591-3597. (118) Song, Y.; Liu, Y.; Yang, M.; Zhang, B.; Li, Z. Appl. Surf. Sci. 2006, 252, 5693-5699. (119) Arinaga, K.; Rant, U.; Knezevic, J.; Pringsheim, E.; Tornow, M.; Fujita, S.; Abstreiter, G.; Yokoyama, N. Biosens. Bioelectron. 2007, 23, 326-331. (120) Kuhr, W. G. Nat. Biotech. 2000, 18, 1042-1043. (121) Steichen, M.; Decrem, Y.; Godfroid, E.; Buess-Herman, C. Biosens. Bioelectron. 2007, 22, 2237-2243. (122) Triroj, N.; Lapierre-Devlin, M. A.; Kelley, S. O.; Beresford, R. IEEE Sensor J. 2006, 6, 1395-1402. (123) Heaton, R. J.; Peterson, A. W.; Georgiadis, R. M. Proc. Nat. Acad. Sci. U.S.A. 2001, 98, 3701-3704. (124) Castelino, K.; Kannan, B.; Majumdar, A. Langmuir. 2005, 21, 1956-1961. (125) Ho, P. S.; Frederick, C. A.; Saal, D.; Wang, A. H.; Rich, A. J. Biomol. Struct. Dyn. 1987, 4, 521-34. (126) Wang, J.; Rivas, G.; Jiang, M.; Zhang, X. Langmuir. 1999, 15, 6541-6545. (127) Rant, U.; Arinaga, K.; Fujiwara, T.; Fujita, S.; Tornow, M.; Yokoyama, N.; Abstreiter, G. 133  Biophys. J. 2003, 85, 3858-3864. (128) Manning, G. S.; Mohanty, U. Physica A Stat. Theor. Phys. 1997, 247, 196-204. (129) Bloomfield, V. A. Biopolymers. 1997, 44, 269-82. (130) Nakata, M.; Zanchetta, G.; Chapman, B. D.; Jones, C. D.; Cross, J. O.; Pindak, R.; Bellini, T.; Clark, N. A. Science. 2007, 318, 1276-1279. (131) Demers, L.; Ostblom, M.; Zhang, H.; Jang, N.; Liedberg, B.; Mirkin, C. J. Am. Chem. Soc. 2002, 124, 11248-11249. (132) Opdahl, A.; Petrovykh, D. Y.; Kimura-Suda, H.; Tarlov, M. J.; Whitman, L. J. Proc. Nat. Acad. Sci. U.S.A. 2007, 104, 9-14. (133) Kimura-Suda, H.; Petrovykh, D.; Tarlov, M.; Whitman, L. J. Am. Chem. Soc. 2003, 125, 9014-9015. (134) Ge, B.; Huang, Y.; Sen, D.; Yu, H. Journal of Electroanalytical Chemistry. 2007, 602, 156-162. (135) Southern, E.; Mir, K.; Shchepinov, M. Nat. Genet. 1999, 21, 5-9. (136) Kelley, S.; Barton, J.; Jackson, N.; McPherson, L.; Potter, A.; Spain, E.; Allen, M.; Hill, M. Langmuir. 1998, 14, 6781-6784. (137) Oligonucleotide Synthesis (DNA and RNA), University Core DNA Services, University of Calgary. http://www.ucalgary.ca/dnalab/synthesis. (138) Kibbe, W. A. Nucleic Acids Res. 2007, 35, W43-46. (139) Kibbe, W. A. OligoCalc version 3.22: Oligonucleotide Properties Calculator. Online Calculator. http://www.basic.northwestern.edu/biotools/oligocalc.html. March 12, 2008. (140) Markham, N. R.; Zuker, M. Nucleic Acids Res. 2005, 33, W577-581. (141) Antoinette Hamelin; Zofia Borkowska; Janusz Stafiej J. Electroanal. Chem. 1985, 189, 85-97. (142) Bizzotto, D. Ph.D. Dissertation, University of Guelph, 1996, p. 238. (143) Banks, C. E.; Compton, R. G.; Fisher, A. C.; Henley, I. E. Phys. Chem. Chem. Phys. 2004, 6, 3147-3152. (144) Cantrell, D. W. Three approximations for the surface area of a general ellipsoid will be. 134  sci.math. http://home.att.net/~numericana/answer/ellipsoid.htm. May 15, 2004. (145) Thomsen, K. The following symmetrical formula seems to give the surface area of a general ellipsoid with a relative error < 1.42%. Could it be new? sci.math. http://home.att.net/~numericana/answer/ellipsoid.htm. April 26, 2004. (146) Thomsen, K. On 2004-05-13, Knud Thomsen (Denmark) wrote: sci.math. http://home.att.net/~numericana/answer/ellipsoid.htm. May 13, 2004. (147) Michon, G. P. Surface Area of an Ellipsoid - Scalene Ellipsoid - Numericana. Final Answers. http://home.att.net/~numericana/answer/ellipsoid.htm. April 9, 2007. (148) Bender Associates Inc. SPOT RT Specifications. http://www.benderassoc.com/mic/digital&analog%20imaging/DVC/DiagnosticsCameras/ spotrtsp.htm. 1999. (149) Rasband, W. S. ImageJ: Image Processing and Analysis in Java; National Institutes of Health: Bethesda, Maryland USA. (150) Michael D. Abramoff; Paulo J. Magalhaes; Sunanda J. Ram Biophotonics Int. 2004, 11, 36-43. (151) Rasband, W. S. ImageJ Documentation. version 1.38q. http://rsb.info.nih.gov/ij/docs/index.html. April 27, 2007. (152) Su, L.; Sankar, C.; Sen, D.; Yu, H. Anal. Chem. 2004, 76, 5953-5959. (153) Roberts, M.; Kelley, S. J. Am. Chem. Soc. 2007, 129, 11356-11357. (154) Lapierre-Devlin, M.; Asher, C.; Taft, B.; Gasparac, R.; Roberts, M.; Kelley, S. Nano Lett. 2005, 5, 1051-1055. (155) Gasparac, R.; Taft, B.; Lapierre-Devlin, M.; Lazareck, A.; Xu, J.; Kelley, S. J. Am. Chem. Soc. 2004, 126, 12270-12271. (156) Kelley, S. O. Nanostructured Materials as Ultrasensitive Cellular and Molecular Sensors. Lecture. Lectures in Modern Chemistry. University of British Columbia, Vancouver, BC. April 1, 2008. (157) Bastiaens, P. I. H.; Squire, A. Trends Cell. Biol. 1999, 9, 48-52.  135  Appendix A: Image Step Sequences Image step sequences omitted from the Results and Discussion section are included here. The following are the desorption step sequence used for ssDNA-CY3/MCH (Figure 52), dsDNA-CY3/MCH (Figure 53), dsDNA-CY5/MCH (Figure 54), MCH/ssDNA-CY3 (Figure 55), and MCH/dsDNA-CY3 (Figure 56).  136  Figure 52 Potential step sequence applied to ssDNA-CY3/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1200 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 250 ms exp, 583 ms interval.  137  Figure 53 Potential step sequence applied to dsDNA-CY3/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (+400 mV → -11e00 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 200 ms exp, 1000 ms interval. 138  Figure 54 Potential step sequence applied to dsDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (+400 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 750 ms exp, 1095 ms interval.  139  Figure 55 Potential step sequence applied to MCH/ssDNA-CY3. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 2500 ms exp, 2836 ms interval. 140  Figure 56 Potential step sequence applied to ssDNA-CY5/MCH. Average fluorescence intensity (■, counts per second, cps), capacitance (■, μF∙cm-2), and potential step sequence (0 mV → -1300 mV, Ebase = 0 mV, 5×5 images) shown in top, middle, and bottom panels respectively. Imaging settings: 1500 ms exp, 1835 ms interval. 141  Appendix B: Timing During Image Acquisition In the course of experiments, it was found that a small number of my recorded image exposure times appeared to be incorrect or were suspect given an anomalously high fluorescence intensity, or a lower than expected image collection frequency. A brief investigation of the timing characteristics of the SPOT RT camera and image recording was performed in order to eliminate any lingering doubts. Part of the problem was a flaw in understanding how the SPOT camera and image collection worked. Two variables are used to control the timing of image acquisition: the exposure time and the time interval. The exposure time controls how long the CCD collects incident photons for a given image, while the time interval controls the imaging frequency. The time interval is the sum of the exposure time plus some "wait time" between images, as shown in Figure 57. Both the exposure time and the wait time are variable, however the wait time contains a hidden sub-component, the write time. The write time is the amount of time required by the camera and the computer to write the image file to the computer hard drive, and is found to be between 330 and 350 ms, averaging approximately 338 ms, although this was previously unrealized, as mentioned.  Figure 57 Illustration of the timing events in during the collection of a series of images. A is the interval time or period (the inverse of frequency). A is the sum of two components: B, the exposure time, and C, the wait time. The exposure time, B, is the length of time during which light is collected for a given image. Each image is followed by a wait time, C, during which the image is saved and light is not collected. C has a constant sub-component, D, the image write time, which is the time required for the computer to write the file to disk. This has been measured to be slightly less than 350 ms, as is shown in the illustration. The wait time, C, cannot be less than the image time, D. In the generic example shown, the exposure time, B, is 500 ms, while the wait time, C, is 1000 ms. This gives an interval time, A, of 1500 ms, which is a frequency of 40 images per minute. 142  Input time intervals less than the sum of the exposure time and the write time meant that the computer had to wait for the image data to be written to disk, delaying collection of the subsequent image, and giving the appearance of faulty timing issues. Hence it was desirable to know with certainty the limits imposed on the time interval. The actual time interval was determined using two methods: The first used the *.LST data file generated by the LabView program, which records the time at which each potential measurement is taken; each potential corresponds to an image and hence the sampling frequency of the potential data would be the same given that the LabView potential control takes cue by locating image files on the hard drive. The second method used the image write (modification) times in the metadata for the first and the last images in a given sequence and dividing by the total number of images in the sequence. Each method gave approximately the same answer for imaging frequency within a millisecond or two. Two experiments were performed using standard settings for the LabView VI programs and the SPOT imaging software, writing to a hard drive with ample empty space. The first used varying exposure times, ranging from 50 ms to 3000 ms, with time intervals less than the exposure time. The result gave a linear correlation with an offset equal to the average write time, which was constant throughout this and other experiments as shown in Figure 58. Using this method, the average write time was found to be 337.8 ms. A second experiment used a constant exposure time of 500 ms, varying the interval times which were input. Figure 59 shows that for interval times greater than the sum write time and the exposure time that there was a one to one linear correlation between measured interval times and input interval times, with a near-zero intercept. For images taken with an input interval time less than the sum of the exposure time and the write time, the result shows an effectively constant measured interval time equal 836.6 ms: a 500 ms exposure combined with a 336.6 ms write time. This confirms the result of the prior experiment. These two experiments helped to better control image collection and provided a check-step to evaluate data where the exposure times are in question due to insufficient interval times.  143  Figure 58 Exposure time vs. measured time interval. Graph used to determine the image "write time". Image sequences were collected using interval times set to zero, with exposure times ranging from 50 ms to 3000 ms. The measured interval time is the sum of the exposure time and the write time. This method gave a write time of 337.8 ms.  144  Figure 59 Measured interval time for a constant exposure with varying wait time. Graph used to determine the image "write time". Image sequences were collected using set interval times ranging from 0 to 3000 ms, with a constant exposure time of 500 ms. The measured interval time is the sum of the exposure time and the write time, hence only where set interval times are greater than this sum will the measured interval time be equal to the input interval time; those with lesser interval times show a constant measured interval time of exposure time (500 ms) plus the write time (336.6 ms).  145  Appendix C: Purification Procedure for Reduced DNA The seal is removed from the bottom of a MicroSpin G-50 column, and the column is placed in an Eppendorf tube with the column's cap slightly loose. The buffer solution in the column initially is flushed using centrifugation (1 min, 3000 rpm) and discarded. The column is then flushed by centrifugation (1 min, 3000 rpm) three additional times with the immobilization buffer (400 μL each) to clean it. Finally, the reduced DNA solution is carefully introduced to the top of the column, adsorbing it directly onto the packing material, then centrifuging (3 min, 3000 rpm) into a clean Eppendorf tube. The concentration of the resultant solution is determined using UV-Vis absorbance at 260 nm, compared to a baseline value taken at 320 nm.  146  Appendix D: Determination of the Surface Area Measurement of surface areas for five separate polycrystalline gold beads was carried out using both optical imaging and by electrochemical measurements. CV and DC measurements were performed in 100 mM KClO4 at room temperature. CV sweep rate was 20 mV/s; DC sweep rate was 5 mV/s. First we obtained CVs in the region from -800 mV to +1250 mV, including oxide formation and reduction as shown in Figure 60, for each bead. Subsequently, a CV in the double layer region from +650 mV to -800 mV was obtained for each, shown in Figure 61. Finally, differential capacitance curves were measured for each sample, shown in Figure 62. Comparing data from a single crystalline sample of known area, a capacitance value of 18 μF/cm2 at -800 mV was used to determine the surface area of each bead. The DC curves are shown in Figure 63 normalized to unity at -800 mV using the results. The data was compared to areas calculated for each sample using images collected using a digital microscope. A sample image with internal reference is shown in Figure 64.Ellipse and bead surface areas were calculated as described in the Experimental Methodology. Values obtained are given in Table 6. A graph comparing values obtained from capacitance and those obtained from optical and geometric calculations is given in Figure 65.  Capacity (μF)  μF/cm2  AE (cm2)  AE (mm2)  Ao (mm2)  ACM  2.110  18  0.1172  11.72  9.79  ACN  1.940  18  0.1078  10.78  8.82  ACO  1.480  18  0.0822  8.22  6.56  ACP  3.180  18  0.1767  17.67  14.70  ACQ  2.300  18  0.1278  12.78  10.79  Reference  4.585  18  0.2547  25.47  25.4  Sample  Table 6 Electrochemically and optically determined bead surface areas for five samples (ACMACQ) and a single crystal reference sample of known area.  147  Figure 60 Cyclic voltammogram of five polycrystalline gold beads in 100 mM KClO4 at room temperature; 20 mV/s.  Figure 61 CV of the double layer region for five polycrystalline gold beads in 100 mM KClO4 at room temperature; 20 mV/s.  148  Figure 62 Differential capacitance curves for polycrystalline gold beads; capacity (μF) values plotted against potential (mV). In 100 mM KClO4 at room temperature; 5 mV/s.  Figure  63 Differential capacitance curves for the five polycrystalline normalized to unity at -800 mV. Curves in the negative region show considerable consistency.  149  Figure 64 Digital microscope image of a gold bead with 0.5 mm gold wire. The lighter circle in the background is etched metal on a slide used for reference, and has a diameter of 2500 μm which was used to calibrate each measurement.  Figure 65 Correlation between surface areas measured for polycrystalline gold beads using electrochemical and optical (geometric) methods. The two area measurements were shown to be related by Area(electro.) = 1.1582*Area(optic.)+0.4981 with a correlation coefficient of 0.9978. 150  Optically measured areas were determined by fitting an ellipse to the perimeter of one or more images of each bead and calculating each of the semiaxes. These were then converted to millimetres, and the surface area of a prolate ellipsoid with those two axes calculated. Although a small adjustment can be made for the area that is unavailable due to the attachment to the wire (diameter = 0.5mm; Aw = 0.19635 mm2), such an off-set is equal for all beads, so it does not affect either the slope or the correlation coefficient value in Figure 65. As expected, the actual area is higher than the optical (or geometric) area as shown by the slope and the intercept. It should be noted that the positive intercept is larger than the area covered by the wire, which is likely a consequence of the wetting of the wire at the immersion depth. This suggests that the electrochemical area could be approximated as follows: A E = f r Ao −Aw 2 r h  (37)  where fr is the rugosity, or surface roughness factor, defined by the quotient between the true area and the geometric surface area; r (0.25mm) is the radius of the wire and h is the height of the wire's wetting. This gives  A E = f r Ao −Aw 1.57 h  (38)  Taking the intercept of the graph we can calculate the wetting height to be 0.442 mm.  h=  0.4981 Aw =0.442 mm 1.57  (39)  An average wetting height for the bead stem (above the meniscus) is on the order of 0.5 mm which is very reasonable. The a roughness factor of 1.15 is quite low, which may be due to the flame annealing of the gold electrodes.  151  Appendix E: Auxiliary Experiments Several auxiliary experiments were performed to better understand and characterize the DNA SAMs created and to provide comparisons and controls. This section provides an overview of a number of these as they relate to the development of the concepts and conclusions regarding the DNA monolayers.  E-1 Effects of Ionic Strength As part of an experiment series investigating the effect of the ionic strength of the immobilization buffer on DNA adsorption, a series of dsDNA solutions were prepared, containing 10, 50, 200, 500, and 1000 mM NaCl (with no MgCl2) to cover a broad range of grafting densities using the standard adsorption procedure and a time of 1 day followed by treatment with MCH for 1 hour. The solutions of immobilization buffer containing excess DNA were measured using UV-Vis absorbance spectroscopy (Abs = Abs(260nm) - Abs(320nm)) as diluted (10 μL in 100 μL) solutions. The number of moles remaining was calculated from the absorbance (50 μg/OD, Mw = 19308 g/mol, 0.00259 μg/OD, 20 μL), ranging from 1.53 pm (1000 mM) to 3.21 pm (10 mM), with a pre-adsorption estimate of 3.31 pm. This was used to estimate the number of adsorbed moles. Normalizing for surface area of each bead gave initial graft densities (ΓdsDNA, #duplexes/cm2) ranging from 9.81×1012 to 4.90x1011 duplexes/cm2, or a factor of 20 over the series. The beads were characterized, after passivation with MCH (1.0 mM, 20 μL, in immobilization buffer), using CV measurement of intercalated [Ru(NH3)6]3+ reduction. The results shown in Figure 66 illustrate the differences in the two results, and it may be taken as a before-and-after "snapshot" of the surface prior immediately after DNA immobilization and following MCH passivation and rinsing. Steel et al. have noted large decreases in the amount of immobilized ssDNA within the first hour of passivation with MCH (1.0 mM, deionized water), ranging from an 88% loss for a 48-mer probe, to a 43% loss for a 8-mer probe; the losses are lowered slightly when 1M NaCl is used instead of deionized water. They conclude that much DNA desorption occurs because of thiol-thiol exchange, rather than displacement of nucleotidegold contacts, aided by interstrand electrostatic repulsions [18]. This seems to correspond with 152  our results for dsDNA presented in Figure 67 as evidenced by an estimated 75% loss of probes during MCH passivation. The higher graft densities obtained by electrochemical methods for the samples at 10 and 50 mM suggest an underestimate of the amount of DNA initially adsorbed using the absorbance technique or an error in the accuracy of the CV measurements, although the data does not conclusively point to either possibility.  Figure 66 Grafting density of dsDNA as a function of the NaCl concentration in the immobilization buffer (10 mM Tris, pH ~ 7.5). Electrochemically measured (red, squares, long dashes) and estimated using UV-Visible absorbance at 260 nm (green circles, short dashes) of the following adsorption of dsDNA on the bead surface. Curves shown are merely visual guides.  153  Figure 67 Electrochemical and UV-Vis absorbance methods compared using the data from Figure 66; electrochemical measurements given on the y-axis and UV-Vis results on the x-axis. Excluding the results at 1000 mM NaCl (after iteratively re-weighting using inverse residuals), the results appear to show an strongly linear relationship, with a least squares fit of y = 0.25700x + 8.2457×1011 with a high coefficient of determination (R2 = 0.9997).  154  E-2 Capacitance Measurements Measurement of the capacitance accompanied the image-stepping experiments. While differential capacitance curves were measured for each sample, the information is analogous, and in most cases indistinguishable from the capacitances measured at Estep with each image. The data shows a number of interesting trends, although due to the smaller changes relative to those at desorption, the fine details can be easily unnoticed. Figure 68 shows capacitance curves for ssDNA, ssDNA/MCH, dsDNA/MCH, MCH, MCH/ssDNA, and MCH/dsDNA desorption sequences, with expanded regions inset in each graph. In several instances ssDNA/MCH and dsDNA/MCH samples a decrease in the base potential capacitance was observed below -400 mV with a concurrent change in the step potential capacitance; other samples showed a slight increase in the base potential capacitance prior to desorption, however the results have not been correlated to some specific feature. Samples where the surface was first passivated with MCH showed far greater uniformity. The capacitance at the base potential consistently remained essentially flat approaching the desorption potential in a manner more consistent with pure MCH monolayers. In the macroscopic regions of the capacitance curves, two parts can be observed in the base potential curves for the desorption region, characteristic of the two-step desorption process observed for other alkane thiolate SAMs. There is an initial bump in the curve, consisting of a rise and plateau, followed by a rapid increase in the capacitance up to a final plateau. The initial bump is the result of initial desorption at sites where the hexylthiolate groups are weakly bound: defect sites, steps, and grain boundaries along with sites of random nucleation. The subsequent rise result from the mass desorption of the monolayer.[51] It should be noted that each point represents the average of 5 measurements taken at each step and that measurements at Ebase (0 mV) are mapped to the potential of the subsequent step, Estep. The raw data shows small changes in capacitance due to time-dependent changes in the structure at a given potential, providing further information on the structure.  155  Figure 68 A comparison of capacitance curves for several samples with inset expansions of the switching region before desorption. Graphs on the left side show samples where DNA was deposited first; those to the right show samples where MCH was deposited first. (A) DNA only; (B) ssDNA-CY3/MCH, ΓDNA ~ 9.2×1012; (C) ssDNA-CY3/MCH, ΓDNA ~ 1.0×1013; (D) dsDNACY5/MCH, ΓdsDNA ~ 3.1×1012; (E) dsDNA-CY5/MCH, ΓdsDNA ~ 9.5×1011; (F) MCH only; (G) MCH/ssDNA-CY3, ΓDNA ~ 1.2×1010; (H) MCH/ssDNA-CY3, ΓDNA ~ 2.6×1010; (I) MCH/dsDNA-CY3, ΓdsDNA ~ 3.7×1010; (J) MCH/dsDNA-CY3, ΓdsDNA ~ 2.4×1011.  156  E-3 Concurrent CV / Imaging Curves obtained by collecting images concurrently with CV measurements provide a simple, qualitative look at immobilized DNA, are shown in Figure 69. Without MCH treatment, ssDNA shows requires a greater negative potential than MCH-treated samples to induce the strands to “stand up”, presumably due to interaction of the DNA bases with the gold surface, as shown by the sigmoidal curve. Samples treated with MCH either pre- or post-immobilization show bell-shaped curves. It can be inferred from the curves for MCH/ssDNA-CY3 and MCH/dsDNA-CY3 that the dsDNA has a greater average distance from the surface at 0 mV relative to ssDNA, as the ssDNA is at roughly one third intensity at 0mV, while dsDNA is at roughly two thirds of its maximum intensity at that point. This results from the greater tortuosity of ssDNA and the rigidity of dsDNA which results in a greater effect of regions near the surface on the terminus of each duplex. Such a conclusion is in keeping with the observations of others regarding the conformation of immobilized ds and ssDNA.  Figure 69 Curves from imaging done concurrently with CV measurements. Shown are smoothed , scaled fluorescence intensity curves for one cycle (positive → negative → positive; left to right each graph). (A) ssDNA-CY5 only, 5 mV/s, 6.7 mV/image, Δcps = 235; (B) ssDNA-CY5/MCH, 20 mV/s, 11.7 mV/image, Δcps = 980; (C) MCH/ssDNA-CY5/MCH, 20 mV/s, 36.7 mV/image, Δcps = 400; (D) MCH/dsDNA-CY3/MCH, 20 mV/s, 26.7 mV/image, Δcps = 4800; Curves obtained at 20 mV/s were observed to have the same shape at 5 mV/s.  E-4 Comparison of Switching Characteristics Each of the four methods of preparing passivated DNA SAMs presented here was found to have a characteristic response to potential, as depicted in Figure 70. The curves share characteristic features with the CV/imaging curves in Figure 69. Monolayers prepared in the original ssDNA/MCH manner show potential stepping curves that show a minimum response in 157  the region near OCP (0 mV) which appears to closely correspond to ssDNA only (no MCH) in the response, suggesting that the MCH is not affecting the assembly to the extent desired; the hysteresis and sigmoidal nature contrast sharply with the memoryless return to and from the base and step potentials in MCH/ssDNA prepared samples. For dsDNA/MCH samples, the switching shows a mixture of decay and possibly hysteresis - certainly not solely the domain of photobleaching, whereas MCH/dsDNA shows only a slight effect of photobleaching and no memory effect with the potential switching.  Figure 70 Comparison of switching observed for ssDNA-CY3/MCH, MCH/ssDNA-CY3, dsDNA-CY3/MCH, and MCH/dsDNA-CY3. For details, see respective figures. dsDNACY3/MCH is shown without any adjustment of Ebase intensity. Changes in fluorescence intensities are scaled for comparison. Differences are also observed in the general appearance of the adsorbed monolayers. MCH pre-passivated layers here show a striking preference of adsorption location. 158  E-5 In situ MCH Adsorption In situ adsorption of MCH to a gold bead with ssDNA-CY5 was performed using the second generation spectroelectrochemical cell. The adsorption was initiated by injecting a small amount of concentrated MCH (to produce a total concentration of ~ 1 mM) was injected into the cell containing ~ 4.7 mL of Tris buffer. Shown in Figure 71 is the intensity profile before and during MCH adsorption along with contrast enhanced images corresponding to various times during the experiment. Both preceding and following the injection of MCH there are a significant number of hotspots. This demonstrates that while MCH removes ssDNA from the surface, it is not effecting the removal of hotspots.  Figure 71 Fluorescence intensity (given in arbitrary units) as a function of time. After approximately 100 seconds, MCH was injected into the cell at OCP. Changes in fluorescence intensity are complicated by the photobleaching or reaction of the CY5 dye. Images (A-G) are contrast enhanced; the blurriness of B, C, and D is due to desorption.  159  E-6 Other Experiments Finally, there have been a number of brief experiments to corroborate and confirm certain assumptions and conclusions which will be dealt with briefly. To confirm that the MCH/ssDNA and MCH/dsDNA samples were indeed forming thiolate-bonded SAMs, the procedure was done in parallel with samples pre-coated with mercaptoundecanol (MCUD, HS-C11H22-OH) followed by adsorption of dsDNA and ssDNA bearing 5'hexylthiol moieties. Longer chain alkylthiolate monolayers are known to be more stable relative to shorter alkylthiolate and longer alkylthiols will replace shorter alkylthiolates faster than shorter alkylthiols will replace longer alkylthiolates. Hence if the ds and ssDNA were adsorbing in a purely physical and nonspecific manner, the beads would have an approximately equal surface concentration on the MCUD coated surface; if it were adsorption through ligand exchange, the 5'hexylthiol DNA would replace the 11-carbon chains of MCUD more slowly than the 6-carbon chains of MCH. The lower surface concentration measured by ruthenium-CV and the extremely low fluorescence intensity, both near OCP and during desorption, confirm that less DNA is adsorbed to the longer alkylthiolate SAM, hence the process is one of ligand exchange whereby MCH is replaced by the 5'hexylthiol DNA. Potential is known to control the position and density of states in electronic bands in metals and at metal surfaces. This can result in changes in the reflectivity of of the metal. [62] For those samples where CY3 fluorescence was imaged, light leakage through of reflectance from the gold surface was also imaged, hence it was important to know if the changes in reflectivity affected the measurement of fluorescence switching. Imaging during the desorption of MCH coated gold beads showed a small modulation in the reflected light (less than 0.1 cps) only after the complete desorption of the adsorbed monolayer. Thus it can be concluded that changes in the reflectance of the gold substrate are insubstantial and do not affect the validity of our results in any way.  160  Appendix F: Creative Commons Licencing In order to promote the sharing, dissemination, preservation, and progress of scientific knowledge, this thesis is made available for free redistribution, copying, and sharing under a Creative Commons licence. It is the author's view and hope that this represents the future of scientific publishing in all formats. To that end, this thesis is distributed under the terms of the Creative Commons Attribution Non-Commercial 2.5 Canada License (http://creativecommons.org/licenses/by-nc/2.5/ca/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.  161  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0061760/manifest

Comment

Related Items