% 1H V LL...¥‘.... 63+. 1. Aqum‘P-I ‘ .- 31. ".1. \u. . vhf-.7 53...: .9 . "on? . I.“ . 3 , 9 i? (.32.. “3...“... «ctr-a xvafiuwoufifli: an i .vt‘13f. lent: . '7 it .5). l7: Er: H). 52.0.. 4.. . .uf. .: I; 34-: u.,..x:ivl. , two.- EJIib h- ‘ LY..§..1F ..s.£r : o... z . V , . . , L . fig. . #3:... l .. mm“. “a: . A a \ I . . .. , , . L .. . .. ‘ ,L amnimuu 2...: «y... 1H2. .ué. mun... 7.4%" 4..., .n : . . 3%.". , , , . . , V nil. l. .1 u , A. .. .1. . . .. 13...? .. . . . . ‘ .wm.._...me.r.w..!fiJ w. ’23,. . lblllll This is to certify that the dissertation entitled ELECTROCHEMICAL OXIDATION AND DETECTION OF ALIPHATIC POLYAMINES AT BORON—DOPED DIAMOND THIN-FILM ELECTRODES presented by Malgorzata A. Witek has been accepted towards fulfillment of the requirements for the PhD. degree in Chemistry «424/7 p-qéaab Major Professor’s Signature 13- 3 -— aooa Date MSU is an Affirmative Action/Equal Opportunity Institution . LIBRARY Michigan State University PLACE IN RETURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJCIRC/DatoDuepes-pts ELECTROCHEMICAL OXIDATION AND DETECTION OF ALIPHATIC POLYAMINES AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES By Malgorzata A. Witek A DISSERTATION Submitted to Michigan State University in partial fulfilhnent of the requirements for the degree of DOCTOR OF PHILOSOPHY DEPARTMENT of CHEMISTRY 2002 ABSTRACT ELECTROCHEMICAL OXIDATION AND DETECTION OF ALIPHATIC POLYAMINES AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES By Malgorzata A. Witek The electrochemical responsiveness of boron-doped microcrystalline and nanocrystalline diamond thin films was investigated by cyclic voltammetry. The morphology and microstructure of both types of diamond was studied with atomic force microscopy (AFM), scanning electron microscopy (SEM), transmission electron ' microscopy (T EM), x-ray diffraction (XRD), visible Raman spectroscopy, and secondary ion mass spectrometry (SIMS). Highly boron-doped microcrystalline and nanocrystalline diamond thin-film electrodes exhibited a wide working potential window in aqueous electrolytes (3-3.5 V), a low voltammetric background current, and an excellent responsiveness toward a number of redox systems (i.e., Fe(CN)6'3”‘, Ru(NH3)6*2’*3, IrCl6'2"3, and methyl viologen) without pretreatment. The relatively large heterogeneous electron transfer rate constants (10'2- 10“ cm/s) for all the couples indicated that diamond has a sufficient charge carrier density over the wide potential range to support a rapid electron transfer. The electrochemical response of both types of films for the oxidation of aliphatic polyamines was also studied. Boron-doped diamond is unique in the carbon electrode family in that it provides an active and stable response for this class of analytes. The electrode’s ability to support the amine oxidation, a reaction that involves anodic oxygen transfer, strongly depends on the chemical composition at film surface (i.e., presence of surface boron sites and spa-bonded non-diamond carbon impurities). A model for amine oxidation was proposed in which the non-diamond carbon impurity sites, located primarly in the grain boundaries, generate hydroxyl radical (0H0) at lower overpotential than surrounding diamond matrix. The radical attacks the polyamine molecule coordinated at a surface boron site nearby. The fastest reaction rates for amine oxidation were observed for diamond electrodes at which the formation of OH- had appreciable rate, and that were heavily doped with boron. It was shown that aliphatic polyamines can be quantitatively electrooxidized at both boron-doped microcrystalline and nanocrystalline diamond thin-film electrodes. Flow injection analysis investigations of cadaverine, putrescine, sperrnine and spermidine indicated that these analytes can be detected stably and reproducibly in alkaline media at constant applied potential. A linear dynamic range from 0.1 to 1000 uM and a limit of detection near 0.1 pM, 20 pL inj., (S/N=3) were commonly observed for these analytes. Response variability as low as 2-4 % was achieved. The long-term response stability was excellent with no evidence of permanent electrode fouling by the reaction products. Liquid chromatographic results demonstrated the possibility of separating the amines isocratically in a revered-phase mode, and detecting the polyamines at constant applied potential. Polycrystalline boron-doped diamond thin film electrodes possessed the requisite surface structure and chemical composition to support and sustain the oxidation of aliphatic polyamines via an oxygen transfer reaction. The major advantages of using diamond are that no pre- or post-column derivatization, no mobile phase pH adjustment and no pulsed amperometric waveform are necessary for detection. ACKNOWLEDGMENTS It’s time to say goodbye... My sincere thanks go to my advisor, Dr. Swain who provided academic directions, support, and who was reminding me of keeping the “good attitude”! I thank you for your patience and a good word you always had. I would also like to thank the members of my committee: Dr. Blanchard, Dr. Posey, and Dr. Baker. Thank you for all the comments concerning my research, but also for your kindness. When I was accepted to graduate school at USU, Dr. B. Brown, Dr. S. Bialkowski, Dr. B. Davidson, Dr. M. Koppang, and Prof. Strojek were always very supportive and helpful. I will always be grateful for the encouragement and help they were willing to offer. There is the “significant other” in my life, without whom all of this would have never happen. Thank you Mateusz for your courage in life and love throughout. I would like to thank “The Great People of The Swain Group”: Mike, Chen, Jishou, Jian, Show, Grace, Gloria, Prerna, J.B., Jason S., Jin Woo, Zuzanka, Josef, John, Amy, and Doug. It was a pleasure knowing you all and be a part of the team! Without friends the life would be dreary! Thank you Shannon, Michelle, Sandra, Cindy, Wiesia, and Bogdan! I will always cherish your friendship! Finally and most importantly, I would like to thank my wonderful family: Mom, Dad, and Sebastian. You have been with me every step of the way! Thank you for your love and support! To my Mother and Father in Law, Kasia, Marek and Martyna - Thank you for your constant encouragement! I dedicate this dissertation to my wonderful Grandma, who was supportive throughout my life, and taught me to have dreams. What is the most important she taught me to believe, that they can come true... Maéqorzata iv Table of Contents List of Tables ix List of Figures xii List of Symbols and Notations xxv Chapter 1 1 BACKGROUND AND MO‘HVATIONS 1 1.1. Aliphatic Polyamines ............................................................................................ 1 1.1.1. Biological and Enviromental importance ....................................................... 1 1.1.2. Aliphatic Amine Detection ............................................................................. 3 1.2. Properties and Applications of Conductive Diamond Films .................................. 8 1.3. Diamond Film Deposition and Doping ................................................................ 13 1.4. Outline of the Dissertation .................................................................................. 19 Chapter 2 21 EXPERIMENTAL SECTION 21 2.1. Boron-Doped Diamond Thin Film Growth .......................................................... 21 2.1.1. Microcrystalline Diamond Electrodes Doped Using a Solid State Source. 22 2.1.2. Diamond Electrodes Doped with BZHB ......................................................... 24 2.1.3. Diamond Acid Washing and Hydrogenation Procedure ............................... 26 2.2. Diamond Thin Film Characterization .................................................................. 27 2.2.1. Scanning Electron Microscopy (SEM) ......................................................... 28 2.2.2. Transmission Electron Microscopy (TEM) ................................................... 28 2.2.3. Atomic Force Microscopy (AFM) ................................................................. 29 2.2.4. Raman Spectroscopy .................................................................................. 29 2.2.5. X-Ray Diffraction (XRD) .............................................................................. 29 2.2.6. Boron Nuclear Reaction (BNR) ................................................................... 30 2.2.7. Secondary lon Mass Spectrometry (SIMS) .................................................. 30 2.2.8. Resistivity .................................................................................................... 30 2.3. Electrochemical Measurements ......................................................................... 31 2.3.1. Cyclic Voltammetry Experiments ................................................................. 31 2.3.2. Rotating Disc Electrode (RDE) Experiments ............................................... 32 2.3.3. Double-Layer Capacitance .......................................................................... 33 2.4. Flow Injection Analysis (FIA) .............................................................................. 33 2.5. HPLC System .................................................................................................... 35 2.6. Chemicals and Reagents ................................................................................... 35 Chapter 3 - - 38 THE CHARACTERIZATION AND ELECTROCHEMICAL RESPONSIVENES OF BORON-DOPED MICROCRYSTALLINE DIAMOND THIN-FILM ELECTRODES . 38 3.1. Introduction ........................................................................................................ 38 3.2. Results and Discussion ...................................................................................... 40 3.2.1. Electrochemical Characteristics .................................................................. 40 3.2.2. Material Characterization ............................................................................ 54 3.3. Conclusions ....................................................................................................... 63 Chapter 4 65 THE CHARACTERIZATION AND ELECTROCHEMICAL RESPONSIVENESS OF BORON-DOPED NANOCRYSTALLINE DIAMOND THIN-FILM ELECTRODES... 65 4.1. Introduction ........................................................................................................ 65 4.2. Results and Discussion ...................................................................................... 68 4.2.1. Electrochemical Responsiveness ................................................................ 68 4.2.2. Material Characterization ............................................................................ 78 vi 4.3. Conclusions ....................................................................................................... 86 Chapter 5 87 CYCLIC VOLTAMMETRIC STUDIES OF THE ELECTROOXIDATION OF ALIPHATIC POLYAMINE AT BORON-DOPED DIAMOND THIN FILMS 87 5.1. Introduction ........................................................................................................ 87 5.1.1. Model for Polyamine Oxidation at Polycrystalline Diamond ......................... 91 5.2. Results and Discussion ...................................................................................... 92 5.2.1. Cyclic Voltammetry of Polyamines .............................................................. 92 5.2.2. Effect of the Electrolyte on the Voltammetric Response .............................. 97 5.2.3. Effect of pH on the Voltammetric Response ................................................ 97 5.2.4. Effect of Scan Rate on the Voltammetric Response .................................. 101 5.2.5 Effect of Amine Concentration on the Voltammetric Response ................... 103 5.2.6. Importance of Non-Diamond Carbon Impurities ........................................ 104 5.2.7. Effect of CH4/H2 Ratio on the Voltammetric Response .............................. 106 5.2.8. Effect of Boron on the Voltammetric Response ......................................... 112 5.3. Conclusions ..................................................................................................... 1 30 Chapter 6 134 INVESTIGATIONS OF ALIPHATIC POLYAMINE OXIDATION REACTION MECHANISM AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES .............. 134 6.1. Introduction ...................................................................................................... 134 6.1.1. Non-Active Electrodes ............................................................................... 137 6.1.2. Active Electrodes ...................................................................................... 139 6.2. Results and Discussion .................................................................................... 143 6.2.1. Irreversibility of the Process ...................................................................... 143 6.2.2. Reaction of the Primary and Secondary Amine Groups ............................. 144 vii 6.2.3. Determination of the Ratio Between Number of Protons and Number of Electrons ............................................................................................................. 148 6.2.4. Quantification of Adsorption ...................................................................... 149 6.2.5. Rotating Disk Voltammetric Studies .......................................................... 151 6.3. Conclusions ..................................................................................................... 161 Chapter 7 166 FLOW INJECTION ANALYSIS INVESTIGATIONS OF ALIPHATIC POLYAMINE OXIDAflON AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES ................ 166 7.1 . Introduction ...................................................................................................... 166 7.2. Results and Discussion .................................................................................... 168 7.2.1 . Hydrodynamic Voltammetric i-E Curves .................................................... 168 7.2.2. Calibration Curves ..................................................................................... 181 7.2.3. Response Variability and Stability ............................................................. 188 7.3. Conclusions ..................................................................................................... 196 Chapter 8 200 REVERSED-PHASE HPLC SEPARATION AND AMPEROMETRIC DETECTION OF ALIPHATIC POLYAMINES AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES 200 8.1. Introduction ...................................................................................................... 200 8.2. Results and Discussion .................................................................................... 201 8.2.1. Cyclic Voltammetry and Flow Injection Analysis ........................................ 201 8.2.2. Liquid Chromatography ............................................................................. 210 8.3. Conclusion ....................................................................................................... 217 Chapter 9 218 SUMMARY 218 REFERENCES 218 viii List of Tables Table1.1 Analytical Methods Used for Polyamine Separation and Detection .................. 4 Table 2.1 Microcrystalline Diamond Thin Film Deposition Parameters .......................... 24 Table 2.2 Nanocrystalline Diamond Thin Film Deposition Parameters .......................... 25 Table 2.3 Boron Concentration and Resistivity for a Nanocrystalline Diamond ............. 26 Table 2.4 Microcrystalline Diamond Thin Film Deposition Parameters .......................... 27 Table 2.5 Analytical Techniques Used for Film Characterization ................................... 28 Table 2.6 Molecular Formula and Properties of the Aliphatic Amines ........................... 37 Table 3.1 Summary of Cyclic Voltammetric Data and Apparent Heterogeneous Electron Transfer Rate Constants for Boron-Doped Microcrystalline Diamond Films .......... 54 Table 4.1 Cyclic Voltammetric Data for a Boron-Doped Nanocrystalline Diamond Film 71 Table 4.2 Lattice Spacings and Relative Peak Intensities Obtained from XRD Patterns of a Nanocrystalline Diamond Film and a Cubic Diamond Standard ......................... 83 Table 5.1 Cyclic Voltammetric Data for 0.1 mM Aliphatic Amine Oxidation ................... 94 Table 5.2 Cyclic Voltammetric Charge Data for 1 mM Amines in 0.01 M Borax Buffer/0.1 M NaCIO4 .............................................................................................................. 99 Table 5.3 Cyclic Voltammetric Background Current Data for Boron-Doped Microcrystalline Diamond Films ........................................................................... 109 Table 5.4 Cyclic Voltammetric Peak Current and Potential Data for Amines Oxidation at Microcrystalline Diamond Films Deposited with Different CH4/H2 Ratios. ............ 112 Table 5.5 Cyclic Voltammetric Data for Fe(CN)6'3"‘, Ru(NH3)6”3’*2, IrCIB'm at Boron- Doped Nanocrystalline Diamond Films ............................................................... 118 ix Table 5.6 Cyclic Voltammetric Data for Fe“”"3 and t-Butyl Catechol at Boron-Doped Nanocrystalline Diamond Films ........................................................................... 119 Table 5.7 Summary of Cyclic Voltammetric Peak Current Data for Boron-Doped Nanocrystalline Diamond Electrodes ................................................................... 123 Table 5.8 Summary of Cyclic Voltammetric Peak Potential Data for Boron-Doped Nanocrystalline Diamond Electrode .................................................................... 124 Table 5.9 Summary of Cyclic Voltammetric Data for Boron-Doped Microcrystalline Diamond Electrode ............................................................................................. 127 Table 5.10 Summary of Cyclic Voltammetric Data for Aliphatic Amines at Boron-Doped Microcrystalline Diamond Electrode .................................................................... 129 Table 6.1 Aliphatic EM values as a Function of Solution pH ....................................... 149 Tabl66.2 Kinetic Parameters for Amines Oxidation (pH11) at 20 ppm Microcrystalline Boron-Doped Diamond ....................................................................................... 159 Table 6.3 Kinetic Parameters for Anodic Oxidation of Amines .................................... 160 Table 7.1 FIA-EC Background and Noise Signal for Microcrystalline Diamond Film in Carbonate Buffer pH10 ....................................................................................... 169 Table 7.2 FIA-EC Background and Noise Signal for Nanocrystalline Diamond Film in Borax Buffer pH 11. ............................................................................................ 170 Table 7.3 Amperometric Data Obtained During FIA of Aliphatic Amines at pH 11 ....... 178 Table 7.4 FlA-EC Data for Aliphatic Polyamines Detection at Microcrystalline Diamond Thin-Film Electrode in Borax Buffer, pH 11.2 ...................................................... 183 Table 7.5 FlA-EC Data for Aliphatic Polyamines at a Nanocrystalline Diamond Thin Film Deposited from a 1/94/5 CH4/Ar/H2 source gas ratio with 20 ppm Bsz. ............. 185 Table 7.6 FlA-EC Response Variability Data for Aliphatic Polyamines at a Microcrystalline Diamond Thin Film .............................................................. 190 Table 7.7 FlA-EC Data for Aliphatic Mono- and Diamine ............................................ 198 Table 8.1 Amperometric Detection Figures of Merit for the HPLC Separation of the Aliphatic Polyamines ........................................................................................... 214 Table 8.2 Amperometric Detection Reproducibility and Stability for the HPLC Separation of the Aliphatic Polyamines ................................................................................. 216 xi List of Figures Figure 1.1 Structural models of (A) HOPG, (B) GC, and (C) fcc diamond lattice. ........... 9 Figure 1.2 Plot of boron-doped diamond related papers over the last four decades. Results based on SciFinder® search engine ......................................................... 10 Figure 1.3 Schematic representation of the physical and chemical processes occurring during CVD diamond growth. ................................................................................ 14 Figure 1.4 A diagram for the band structure of boron-doped diamond. ........................ 18 Figure 2.1 A photo of the microwave-assisted CVD reactor: (a) diborane and Ar cylinder cabinet, (b) hydrogen cylinder cabinet, (c) methane cylinder cabinet, (d) microwave generator, (6) reactor, (f) mass flow controller and stop valve, (9) power supply for microwave generator, (h) pressure gauge, (i) water chiller, and (j) rotary pump. 22 Figure 2.2 Design of the electrochemical, three-electrode glass cell. ........................... 31 Figure 2.3 Design of the rotating disk electrode. .......................................................... 32 Figure 2.4 Design of the thin-layer electrochemical flow cell. ....................................... 34 Figure 3.1 Capacitance-potential profiles in 1.0 M KCI for uncoated tungsten and molybdenum, and diamond thin films deposited on W, Mo, and Si substrates. Capacitance values are normalized to the geometric area (0.2 cm2) of the electrodes. Frequency = 40 Hz. AC amplitude = 10 mV. ....................................... 41 Figure 3.2 Background cyclic voltammetric i-E curves in 1.0 M KC! for boron-doped microcrystalline diamond thin films deposited on (A) Si, (B) Mo, and (C) W. Scan rate = 0.1 V/s. Electrode geometric = 0.2 cm2. .................................................. 43 Figure 3.3 Background cyclic voltammetric i-E curves in 1.0 M KCl for boron-doped, microcrystalline diamond films deposited on (A) Si, (B) Mo, and (C) W. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. ........................................................ 44 xii Figure 3.4 Background cyclic voltammetric i-E curves in 1.0 M HN03 for boron-doped, microcrystalline diamond thin films deposited on (A) W and (B) Mo. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. .............................................................. 46 Figure 3.5 A Cyclic votammetric i-E curves for 1.0 mM Fe(CN)6'3’“‘ in 1 M KCI at diamond films deposited on different substrates. Scan rate = 0.1 V/s. Working electrode area = 0.2 cm2. ...................................................................................... 49 Figure 3.5 B Cyclic votammetric i-E curves for 1.0 mM Ru(NH3)6“2’+3 in 1 M KCI at a diamond films deposited on different substrates. Scan rate = 0.1 V/s. Working electrode area = 0.2 cm2. ...................................................................................... 50 Figure 3.5 Cyclic votammetric i-E curves for (C) 0.5 mM IrCl,;’2"3 and (D) 0.1 mM methyl viologen in 1 M KCI, at diamond films deposited on different substrates. Scan rate = 0.1 Vls. Working electrode area = 0.2 cm2. ........................................................... 51 Figure 3.6 Plots of ip°" vs scan rate"2 for (A) Fe(CN)6'3’”‘, (B) Ru(NH3)6*2’*3,(C) IrCls'z‘a, and (D) methyl viologen at diamond films deposited on Si, Mo, and W substrates.53 Figure 3.7 Atomic force microscope images (6 x 6 pm) of boron-doped, micro-crystalline diamond films deposited on (A)silicon, (B)molybdenum, and (C) tungsten. ........... 55 Figure 3.8 SEM image of a boron-doped microcrystalline diamond film deposited on Si. ............................................................................................................................. 56 Figure 3.9 Grazing incidence XRD measurement of a boron-doped microcrystalline diamond film deposited on Si at a 0.5 degrees incidence angle. ........................... 57 Figure 3.10 Macro-Raman spectra for boron-doped microcrystalline diamond films deposited on W, Mo, and Si. Enlargement of the spectra reveals the weak scattering intensity by non-diamond spz-bonded carbon impurity centered at ~1soo cm". ....................................................................................................... 59 Figure 3.11 SIMS data for boron-doped microcrystalline diamond films deposited on (A) Si, (8) Mo, and (C) W. ..................................................................................... 62 xiii Figure 4.1 Background cyclic voltammetric i-E curve in 1.0 M KCI for a boron.doped nanocrystalline diamond film deposited on Si. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. Deposition time = 2h. Source gas mixture = CH4/H2/Ar (1 °/o/5%/94%) (v/v) with 10 ppm 8sz added. Power = 800 W. Pressure = 140 torr. Substrate temperature = ~800 °C .......................................................................... 69 Figure 4.2 Background cyclic voltammetric i-E curve in 1.0 M KCI for a boron-doped nanocrystalline diamond film. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. The deposition conditions are the same as stated in Figure 4.1 .................... 71 Figure 4.3 Cyclic voltammetric i-E curves for (A) Fe(CN)6'3’“‘, (B) Ru(NH3)6*3’*2, (C) erls'Z'a, (D) methyl viologen (MV’2’*"°) in 1 M KCI, and (E) 4-tert-butylcatechol, and (F) Fe”‘°"+3 in 0.1 M HClO. at a boron-doped nanocrystalline diamond thin film. Scan rate = 0.1 We. Electrode geometric area = 0.2 cm2. The diamond deposition conditions are the same as stated in Figure 4.1 .................................................... 72 Figure 4.4 Plot of i,,°" vs scan rate"2for Fe(CN)6'3”, Ru(NH3)6*3’*2, erls‘m, methyl viologen (MV*2’*), 4-tert-butylcatechol, and Fei‘i’+3 at a boron-doped nanocrystalline diamond film. The scan rate was varied from 100 to 500 mV/s. The diamond deposition conditions are the same as stated in Figure 4.1. .................................. 75 Figure 4.6 SEM images of a boron-doped nanocrystalline diamond film deposited on Si in (A) cross-section and (B) top-view image modes. The deposition conditions are the same as stated in Figure 4.1. .......................................................................... 79 Figure 4.7 Atomic force microscope images (10x10um) of a boron-doped, nanocrystalline diamond film deposited Si in the (A) deflection and (B) height image modes. The deposition conditions are the same as stated in Figure 4.1. .............. 80 Figure 4.8 Transmission electron micrograph of a nanocrystalline boron-doped diamond film. The deposition conditions are the same as stated in Figure 4.1 ..................... 81 xiv Figure 4.9 Transmission electron diffraction pattern for a grain in the boron-doped nanocrystalline diamond thin film. The deposition conditions are the same as stated in Figure 4.1 .......................................................................................................... 82 Figure 4.10 XRD pattern for a boron-doped nanocrystalline diamond film deposited on Si. The deposition conditions are the same as stated in Figure 4.1 ....................... 83 Figure 4.11 Visible Raman spectrum for a boron-doped nanocrystalline diamond film deposited on Si. The deposition conditions are the same as stated in Figure 4.1.. 84 Figure 5.1 Cyclic voltammetric i-E curves, background (dashed line) and total current (solid line), for 1.0 mM (A) CAD (B) PUT, (C) SPMD, and (D) SPM in 0.01 M borax buffer/0.1 M NaCI, pH 11. The working electrode was a microcrystalline diamond film deposited from a 0.5% CH4/H2 ratio and 10 ppm BZHS. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. ..................................................................... 93 Figure 5.2 Cyclic voltammetric i-E curves for 1.0 mM CAD in 0.01M borax buffer/0.1 M NaCl, pH 11.2, at a nanocrystalline boron-doped diamond thin film. The nanocrystalline diamond electrode was deposited from a 0.5% CHJHzlAr ratio and 1 ppm BzHe. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. ................. 96 Figure 5.3 Cyclic voltammetric i-E curves for 1 mM cadaverine in (a) 0.01 M phosphate buffer, pH 7.2; (b) 0.01 M carbonate buffer/0.1 M NaClO4, pH 9,; (c) 0.01 M carbonate buffer/0.1 M NaCIO... pH 10, and (d) 0.01 M carbonate buffer/0.1 M NaCIO4, pH 11 at microcrystalline diamond film deposited from a 0.5% CH4/H2 ratio and 1 ppm BzHe. Scan rate = 0.1 V/s. Electrode area = 0.2 cm2 ........................... 98 Figure 5.4 The fractional composition of protonate/unprotonated forms of (A) CAD and (B) SPM as a function of pH ................................................................................ 100 Figure 5.5 Plots of the cyclic voltammetric oxidation peak current, ip°" versus scan rate"2 for (A) 0.1 mM amines at moderately doped microcrystalline film, (B) 1 mM amines at highly doped microcrystalline film in CBpH10.6. (C) Plots of the oxidation peak XV potential, Ep°", versus scan rate for 0.1 mM amines in 0.01 M carbonate buffer, pH10.6. ............................................................................................................... 102 Figure 5.6 Linear sweep voltammetric i-E curves for ethylenediamine at concentrations of (a) 0.1, (b) 0.5, (c) 1, (d) 2, (e) 3, (1)4, and (g) 5 mM at microcrystalline diamond film deposited from a 0.5% CHJH2 ratio. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. ................................................................................................... 103 Figure 5.7 Cyclic voltammetric i-E curves for 1 mM cadaverine in 0.01 M carbonate buffer/0.1 M NaCIO... pH10.6, at a microcrystalline diamond film deposited from a 0.67% CH4/H2 ratio. (A) “as deposited“, and (B) acid washed and rehydrogenated diamond film. Scan rate = 10 mV/s. .................................................................... 105 Figure 5.8 Contact-mode atomic force microscopy images of a microcrystalline diamond films deposited from a (A) 0.33%, (B) 0.67%, and (C) 1% CH4/H2 ratio. All images have the same x, y, z scales. .............................................................................. 106 Figure 5.9 Visible Raman spectra for diamond films deposited using a (a) 0.33%, (b) 0.67%, and (c) 1% CH4/H2 ratio. 514.4 nm laser source and 10 5 integration time. ........................................................................................................................... 108 Figure 5.10 Background cyclic voltammetric i-E curves for a microcrystalline diamond films deposited using a (A) 0.33%, (B) 0.50%, and (C) 1% CH4/H2 ratio. The solution pH was (a) 8.4, (b) 9, (c) 10, (d) 10.6, (e) 11, and (f) 12. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. ................................................................... 110 Figure 5.11 Cyclic voltammetric i-E curves for 1 mM cadaverine in 0.01 M carbonate buffer/0.1 M NaClO4, pH10.6, at diamond films deposited using CH4/H2 source gas ratios of (a) 0.33%, (b) 0.5%, and (c) 1%. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. ................................................................................................... 111 Figure,5.12 Cyclic voltammetric i-E curves for 1 mM CAD, along with the corresponding background current (dashed line), in a 0.01 M borax buffer, pH 11.2, at (A) a boron- xvi doped nanocrystalline diamond thin-film electrode and at (B) a nanocrystalline diamond thin-film electrode deposited without intentionally added boron. Scan rate = 0.1V/s. Electrode geometric area = 0.2 cm2 ..................................................... 113 Figure 5.13 A-E Visible Raman spectra for boron-doped nanocrystalline diamond films deposited from a 1194/5 CHJAr/Hz (v/v) source gas ratio with (A) 0, (B) 1, (C) 10, (D) 20, and (E) 30 ppm of added BgHs. 532 nm laser source and 10 s integration time ..................................................................................................................... 115 Figure 5.14 (A) Cyclic voltammetric i-E curves for boron-doped nanocrystalline diamond films in 1 M KCI over (A) wide potential range showing the working potential window and (B) a narrow potential range. Films deposited using a 1/94/5 CH4/Ar/H2 (v/v) source gas ratio and (a) 1, (b) 10, (c) 20, (d) 30 ppm of added 82H... Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2 ............................................................. 117 Figure 5.15 Cyclic voltammetric i-E curves for boron-doped nanocrystalline diamond films in borax buffer pH 10.6. The films were deposited with (a) 1, (b) 20, (c) 30 ppm BzHe. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. .......................... 121 Figure 5.16 Cyclic voltammetric i-E curves for 1.0 mM cadaverine, putrescine, spermidine, and 0.8 mM spermine in BBpH10.6 at boron-doped nanocrystalline diamond films deposited with (A) 1 ppm (B) 10 ppm , (C) 20 ppm, and (D) 30 ppm of added BzHe. .................................................................................................... 122 Figure 5.17 A-C Raman spectra for three microcrystalline films, deposited from a 0.5 % CHJHZ volumetric ratio with (A) 1, (B) 10, and (C) 20 ppm BzHe added. The doping level was in the range of 1019 to 1020 B/cm3 ........................................................ 126 Figure 5.18 Cyclic voltammetric i-E curves for (A) 1.0 mM PUT, (B) 1.0 mM SPMD, and (C) 0.8 mM SPM, in the BBpH10.6 at three microcrystalline diamond films deposited from a 0.5 % CH4/H2 volumetric ratio with (A) 1, (B) 10, and (C) 20 ppm Bsz added .......................................................................................................... 1 28 xvii Figure 5.19 Model of the competition between the 0;. evolution (OER) and O-transfer reaction (OTR). ................................................................................................... 132 Figure 6.1 (A) Cyclic voltammetric i-E curves for 1.0 mM PUT (dashed line) and 0.8 mM SPM (solid line) in borax buffer pH 11 at a boron-doped nanocrystalline film. Scan rate=0.1 Vls. Electrode area = 0.2 cm2. (B) Subtracted SPM-PUT voltammogram. ........................................................................................................................... 1 45 Figure 6.2 Cyclic voltammetric i-E curves, background (dashed line) and total current (solid line) for (A) 1.0 mM PUT, (B) 1.0 mM EDA, and (C) 1 mM SPM in phosphate buffer, pH 7.2 at a boron-doped nanocrystalline film. Scan rate = 0.1 V/s. Electrode area = 0.2 cm”. ................................................................................................... 146 Figure 6.3 (A) Peak current vs concentration profiles for 1 mM cadaverine at boron- doped microcrystalline diamond electrodes deposited in a 0.5 % CHJHZ ratio with (a) 1 ppm 8sz, and (b) 10 ppm Bsz. (B) Reciprocal plots of peak current versus cadaverine concentration at (a) 1 ppm Bsz, (b) 10 ppm Bsz diamond electrode. ........................................................................................................................... 1 51 Figure 6.4 Voltammetric response for 1 mM Fe(CN)3'3"4 in 1 M KCI at a rotating microcrystalline diamond disk electrode as a function of rotation speed. The electrode was deposited with a 0.5 % CH4/H2 ratio and 10 ppm 82H... Scan rate = 0.1 We. Rotational veiocites (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (6) 261.8, and (f) 314.2 ............................................................................................. 153 Figure 6.5 Voltammetric response for 1 mM Ru(NH3)6"3’+2 in 1 M KCI at rotating microcrystalline diamond disk electrode as a function of rotation speed. The electrode was deposited with 0.5 % CH4/H2 ratio and 10 ppm BgH6.Scan rate = 0.1 Vls. Rotational veiocites (rad/s): (a) 52.2, (b) 104.7, (c) 157.1 , (d) 209.4, (6) 261 .8, and (f) 314.2. ...................................................................................................... 154 xviii Figure 6.6 Voltammetric response for 0.2 mM CAD (forward scan) in borax buffer pH 11 at a boron-doped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5 % CH4/H2 and 1 ppm BzHe. Scan rate = 0.1 Vls. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (C!) 209.4, (e) 261.8, and (f) 314.2 ............................................................................................. 155 Figure 6.7 Voltammetric response for borax buffer, pH 11, at a boron-doped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5% CHJHZ and 20 ppm BZHG. Scan rate = 0.05 V/s. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (e) 261.8, and (r) 314.2. ............................................................................................................. 156 Figure 6.8 Voltammetric response for 0.2 mM SPM (forward scan) in borax buffer, pH 11, at a boron-doped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5% CH4/H2 and 20 ppm Bsz. Scan rate = 0.05 Vls. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (e) 261.8, and (1) 314.2. ...................................................................... 157 Figure 6.9 Koutecky-Levich plots for the oxidation of 0.2 mM amines at a boron-doped microcrystalline diamond disc electrode deposited with 0.5% CHJHz and 20 ppm 82H... Scan rate = 0.05 Vls. Potential: 0.7 V vs Ag/AgCI ...................................... 158 Figure 6.10 Proposed surface SN2-type mechanism of aliphatic amine oxidation at boron-doped diamond electrode ......................................................................... 163 Figure 7.1 FlA-EC hydrodynamic voltammetric background i-E curves for a microcrystalline diamond film deposited using 0.33%, 0.50% and 0.67% CH4/H2 ratio in carbonate buffer, pH 10. Flow rate = 1.0 mL/min. .................................... 168 Figure 7.2 FlA-EC hydrodynamic voltammetric background i-E curves for boron-doped nanocrystalline diamond films deposited with 1 ppm, 10 ppm, and 20 ppm of added BzHa in borax buffer, pH 11. Flow rate = 1.0 mUmin. .......................................... 171 xix Figure 7.3 Background current-time profile in FIA after detector tum-on in borax buffer, pH 11, for a nanocrystalline diamond film deposited from 1/94/5 CHJAr/Hz (v/v) ratio and 1 ppm 82H... Detection potential = +695 mV vs Ag/AgCl. Electrode area = 0.08 cm”. Flow rate =1 mL/min ............................................................................ 172 Figure 7.4 Hydrodynamic voltammetric i-E curves for 20 (LL injections of (A) 1.0 mM SPMD and (B) 1.0 mM CAD. The carrier solution was 0.1M NaCIO4 + 0.01 carbonate buffer, pH 10. Plots of the S/B ratio versus the applied potential are shown for (C) SPMD and (D) CAD. Flow rate :10 mUmin. ................................ 174 Figure 7.5 Hydrodynamic voltammetric i-E curves for 20 (LL injections of 0.1 mM amines. The carrier solution was 0.1M NaCl + 0.01 borax buffer, pH 11. Nanocrystalline diamond films were deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio with (A) 1 ppm, (B) 10 ppm, and (C) 20 ppm of added 8%. Flow rate=1.0 mUmin. ................... 176 Figure 7.5 Plots of the S/B ratio versus the applied potential for 0.1 mM amines. The carrier solution was 0.1M NaCl + 0.01 borax buffer, pH 11. Nanocrystalline diamond film were deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio with (D) 1 ppm, (E) 10 ppm, and (F) 20 ppm of added BzHe. Flow rate = 1.0 mUmin. ..................................... 177 Figure 7.6 Variation of the SPMD/CAD current ratio, as a function of pH of the mobile phase. The carrier solution was 0.01 borax buffer/0.1 M NaCl. Microcrystalline diamond films deposited with 0.5% CHJH; source gas mixture. Flow rate = 1.0 mL/min. ............................................................................................................... 180 Figure 7.7 Calibration curves obtained for (A) SPMD and (B) CAD at boron-doped microcrystalline diamond films deposited withdifferent CH4/H2 source gas ratios. The carrier solution was 0.01M carbonate buffer/0.1 M NaCIO4, pH 10. Injected volume = 20 (LL. Flow rate :10 mL/min. Applied potential = +700 mV ................ 182 XX Figure 7.8 Calibration curves for aliphatic amines at a boron-doped nanocrystalline diamond electrode deposited from a 1/94/5 CHJAr/Hz (v/v) source gas ratio with 20 ppm Bsz. injected volume = 20uL. Flow rate = 1.0 mL/min ................................ 184 Figure 7.9 FIA-EC responses for a series of 10 (IL injections of 0.1uM amine: (a) CAD, (b) EDA, (c) DAP, (d) DA, (9) DEP, (f) SPM, (g) PUT, and (h) SPMD. The carrier solution was 0.01 borax buffer/0.1 M NaCI, pH 11.2. Detection potential=+ 660 mV. ........................................................................................................................... 186 Figure 7.10 Relationship between the FIA peak current and first pK, for (a) MA, (b) EA, (c) PA, (d) CAD, (e) PUT, (f) SPMD, (g) SPM, (h) EDA, (i) HA, (j) DAP, (k) SPMD (2,2), (l) SPMD (3,3). Data collected at two microcrystalline films for 0.1 mM amines (Fig.A and C), and a nanocrystalline film for 0.5 mM amines (Fig. B). ................. 187 Figure 7.11 FIA-EC responses for a multiple injections of 1.0 mM CAD at as “deposited” 0.33% CH4/H2 diamond film. The carrier solution was carbonate buffer pH10. Detection potential=+ 810 mV. Injection volume=20 uL. Flow rate=1.0 mL/min... 189 Figure 7.12 FIA-EC responses for a multiple injections of (A) 1.0 mM CAD and (B) 1.0 mM SPMD at a well 'conditioned" 0.50% C/H film. The carrier solution was CBpH10. A detection potential=855 mV and 3-minute delay between injections was used for CAD. A detection potential=780 mV and a 6-minute delay between injections was used for SPMD. Injection volume = 20 uL. Flow rate = 1.0 mUmin. ........................................................................................................................... 191 Figure 7.13 FIA-EC responses for a multiple injections of CAD and SPMD at a HOPG, GC and a nanocrystalline boron-doped diamond thin-film electrode. Applied potential=655 mV vs Ag/AgCl. Flow rate = 1.0 mL/min ........................................ 193 Figure 7.14 FIA-EC responses for a multiple injections of 0.1 mM CAD, PUT, SPMD and SPM at a boron-doped nanocrystalline film deposited from 1/94/5 CH4/Ar/H2 (v/v) xxi ratio with 1 ppm of added Bsz. The carrier solution was borax buffer pH 11. Applied potential=+665 mV. Injection volume of CAD and PUT = 30uL, and SPMD and SPM = 10111.. Flow rate = 1.0 mUmin. .......................................................... 195 Figure 7.15 FIA-EC series of responses for 0.05 mM DAP at a boron-doped nanocrystalline thin film deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio and 10 ppm BzHe. The carrier solution was borax buffer, pH 11. Applied potential=+665 mV. injection volume = 10 (LL. .................................................................................... 196 Figure 8.1 Background cyclic voltammetric i-E curves in 0.01 M borax buffer, pH 10.6, for a boron-doped microcrystalline diamond film deposited from a 0.4% CH4/H2 ratio. Presented are the 1“, 2"", 25 m, 45 m, 60th scans. Scan rate = 0.1V/s. Electrode area = 0.2 cm2. ................................................................................................... 202 Figure 8.2 Background current-time profile in FIA after detector turn-on for (A) a microcrystalline diamond and (B) glassy carbon.30 The carrier solution for diamond was 7/93 (v/v) acetonitriIe/borax buffer, pH 11.2. The carrier solution for GO was borax buffer, pH 11.2. Detection potential = 665 mV vs Ag/AgCl. Electrode area = 0.08 cm"’. Flow rate = 1 mUmin ........................................................................... 203 Figure 8.3 Cyclic voltammetric i-E curves for 1 mM CAD and 1 mM SPMD in 0.01 M borax buffer, pH 11.2, along with the corresponding background current, for a boron-doped microcrystalline diamond film deposited from a 0.4% CH4/H2 ratio. Scan rate = 0.1 V/s. ............................................................................................ 204 Figure 8.4 Cyclic voltammetric i-E curves in 0.1 mM CAD plus (A) 0.01 M borax buffer, pH 10.6, and (B) 93/7 (v/v) 0.01 M borax buffer, pH 10.6 / CHacN, for a boron- doped microcrystalline diamond film deposited from a 0.4% CH4/H2 ratio. Scan rate =O.1 V/s. ............................................................................................................. 206 xxii Figure 8.5 A Hydrodynamic voltammetric i-E curves for a boron-doped microcrystalline diamond film (0.4% CHJHZ ratio), in 100 (M (a) SPM, (b) SPMD, (c) PUT, and (d) CAD. The carrier solution was 7/93 (v/v) acetonitrile/borax buffer, pH 11.2 (9). Injection volume = 20uL. ..................................................................................... 207 Figure 8.5 B Plots of the S/B ratio versus the applied potential for a boron-doped microcrystalline diamond film (0.4% CH4/H2 ratio), in 100 pM (a) SPM, (b) SPMD, (c) PUT, and (d) CAD. The S/B ratio was calculated as Ito... - Immgmund/ lbackgmnd” 208 Figure 8.6 FIA-EC responses for a boron-doped microcrystalline diamond film (0.4% CH4/H2) during a multiple injections of (A) 5 uM PUT and (B) 2 pM SPMD. The carrier solution was a 7/93 (v/v) acetonitriIe/borax buffer, pH 11.2. Detection potential = 665 mV vs. Ag/AgCI. Injection volume = 20 uL. Flow rate = 1.0 mUmin. ........................................................................................................................... 209 Figure 8.7 FIA-EC calibration responses for a boron-doped microcrystalline diamond film (0.4% CHJHz) during multiple injections of 5 uM SPM. Injected volumes were (a-h) 2 to 20 uL. The carrier solution was a 7/93 (v/v) acetonitrile/borax buffer, pH 11.2. Applied potential = 665 mV vs Ag/AgCl. Flow rate = 1.0 mUmin. ............... 210 Figure 8.8 Reversed-phase liquid chromatographic separation of (A) 47 11M PUT, CAD, SPMD, and SPM, and (B) 0.5 uM PUT, CAD, and SPM on a C13 column (X-Terra, 5 pm particle size, 4.6 x 150 mm). The mobile phase was a 7/93 (v/v) acetonitriIe/borax buffer, pH 11.2. Detection potential = 665 mV vs. Ag/AgCl. Injection volume = 20 uL. Flow rate = 1.0 mL/min. The boron-doped microcrystalline diamond film was deposited from a 0.4% CH4/H2 ratio. ....................................... 212 Figure 8.9 Reversed-phase liquid chromatographic separation of PUT, SPMD, and SPM on a C13 column. The solution mixture contained 20 uM of each amine. The mobile phase was a 16/84 (v/v) acetonitrile/borax buffer, pH 11.2. Detection potential = 665 xxiii mV vs. Ag/AgCI. Injection volume = 20 uL. Flow rate = 1.0 mL/min. The boron- doped microcrystalline diamond film was deposited from a 0.4% CH4/H2 ratio.... 213 xxiv List of Symbols and Notations A .......................................................................................... geometric electrode area [cm2] v ............................................................................................................................ scan rate AEp ............................................................................................... peak potential separation AFM .............................................................................................. atomic force microscopy BB ..................................................................................................................... borax buffer BNR ................................................................................................. boron nuclear reaction CAD .................................................................................................................... cadaverine Cd. ................................................................................................ double-layer capacitance CV .......................................................................................................... cyclic voltammetry CVD ............................................................................................ chemical vapor deposition DA ......................................................................................................................... Dytek‘” A DAP ............................................................................................................ diaminopropane DEP ..................................................................................................................... Dytek" EP E ............................................................................................................................. potential EA ....................................................................................................................... ethylamine EDA ........................................................................................................... ethylenediamine Ep ................................................................................................................... peak potential Em .......................................................................................................... half-peak potential F ......................................................................................... Faraday constant (96486 C/eq) FIA .................................................................................................... flow injection analysis FIA-EC ............................................. flow injection analysis with electrochemical detection FWHM ........................................................................................ full width at half-maximum XXV GC ................................................................................................................. glassy carbon HA ............................................................................................................... hexanediamine HOPG ............................................................................... highly ordered pyrolytic graphite HPLC ................................................................... high performance liquid chromatography i ................................................................................................................................. current ip ....................................................................................................................... peak current k°app ................................................. heterogeneous electron transfer rate constants [cm/s] LC ..................................................................................................... liquid chromatography LOD .......................................................................................................... limits of detection LOQ ..................................................................................................... limits of quantitation MA ................................................................................................................... methyiamine MV .............................................................................................................. methyl viologen n ........................................................................................................... number of electrons NHE .......................................................................................... normal hydrogen electrode OER ............................................................................................ oxygen evolution reaction OTR .............................................................................................. oxygen transfer reaction ox ........................................................................................................................... oxidation PA .................................................................................................................... propylamine PAD .................................................................................... pulsed amperometric detection PUT ..................................................................................................................... putrescine Q ............................................................................................................................... charge R ................................................................................... molar gas constant (8.314 J/molK) RDE .................................................................................................. rotating disc electrode red ......................................................................................................................... reduction RSD ........................................................................................... relative standard deviation xxvi S/B ............................................................................................. signal-to-background ratio SIN ....................................................................................................... signal-to-noise ratio SEM ...................................................................................... scanning electron microscopy SIMS .............................................................................. secondary ion mass spectrometry SPMD 2,2 .................................................................................................... spermidine 2,2 SPM ...................................................................................................................... spermine SPMD 3,3 ..................................................................................................... spermidine 3,3 SPMD ................................................................................................................. spermidine T .................................................................................................................. temperature [K] TEM ................................................................................ transmission electron microscopy XPS ................................................................................. x-ray photoelectron spectroscopy XRD ...................................................................................... x-ray diffraction spectroscopy p ............................................................................................................................ resistivity xxvii Chapter 1 BACKGROUND AND MOTIVATIONS 1.1. Aliphatic Polyamines 1.1.1. Biological and Enviromental Importance The diamine, putrescine, and the polyamines, spermidine and spermine, are ubi- quitous substances found in virtually all cells from higher prokaryotes to eukaryotes."3 As polycations, polyamines are known to interact directly with nucleic acids, proteins, and phospholipids, playing critical roles in the proliferation and differentiation of normal and neoplastic cells.2'9 The general function of the polyamines as growth factors, antioxidants, and metabolic regulators, and as nutrients and second messengers has been proposed.8'11 Cadaverine is a diamine, which is normally produced from lysine by the action of lysine decarboxylase, and under norrnai conditions, mammalian cells do not 12,13 produce this diamine. Thus, the presence of cadaverine in physiological fluids may . . , . _ 1 , indicate the presence of Infection In the host. 12’ 3 and ”memes ”'9'" Elevated levels of intracellular polyamines are a common finding in rapidly proliferating tumors cells, in contrast with lower levels measured in their non-malignant 1,2,4,14,15 counterparts. A number of studies have indicated higher concentrations of putrescine, spermidine, and spermine, or total polyamine contents, in cancer patients, compared to healthy patients.1’2’4 For example, in the biological samples of healthy patients polyamines were present in concentrations ranging from 0.5-6.0 nmole/mg of creatinine for spermine, cadaverine, and spermidine, and to 12-14 nmole/mg creatinine for putrescine. In the same study, the concentration of amines was shown to be three times higher in the urine of colon cancer patients. Chowdhury et al.4 explored the relationship of polyamine presence and concentration to the degree of malignancy in breast cancer patients. The results showed that polyamine levels increased with clinical staging of the disease and showed a direct relationship with the advancement of the disease. 4 Aliphatic amines are of concern, since they occur in various environmental matrices. Aliphatic amines are produced in vast quantities and are, thus, likely to be encountered in the environmenta.16 Amines are of industrial interest as important precursors in the synthesis of dyes, pharmaceuticals, stabilizers, and corrosion 16’” Amines can be formed as secondary pollutants from the biodegradation inhibitors. of nitrogen-containing compounds. They can be a source of formation of carcenogenic N-nitrosoamines.18 Aliphatic polyamines may be present at high concentrations in foods and beverages produced using fermentation processes or exposed to microbial . . . . 19-22 contamination during processmg. They are synthesized by microbial decarboxylation of amino acids. In healthy individuals, dietary polyamines, such as histamine, putrescine, cadaverine, tyramine, tryptamine, phenylethylamine, spermine, and spermidine can be rapidly detoxified by the action of amine oxidases. However, humans with low-amine oxidase activity are at risk of amine toxicity. Aliphatic polyamines have been proposed as potential indicators of the extent of food spoilage.23 The Food and Drug Administration (FDA) presented a report on the relationship between the postmortem formation of biogenic amines in fish tissue and consumer illness.18 The conclusion was that histamine is most closely linked to the development of illness, however, other biogenic aliphatic polyamines, (i.e., putrescine, cadaverine, spermine, and spermidine) could induce a toxic effect, by inhibiting histamine-metabolizing . . . 1e 24 enzymes, such as diamlne oxrdase. ' 1.1.2. Aliphatic Amine Detection There is a significant interest in the development of sensitive, reproducible, and stable detection schemes for aliphatic polyamines due to their biological and toxicological importance. These analytes are not detected directly using optical methods (i.e., UV-VIS, fluorescence) because they are not natively chromophoric or fluorophoric. As a consequence, separation and detection requires pre- or post-column derivatization. The most commonly used derivatization agents for primary and secondary amines are c- phthaldehyde (OPA), fluorescamine, dansyl chloride, benzoyl chloride and 1-naphtyi 25 and re'e'ences there'" Such derivatization procedures increase the complexity isocyanate. of the assay, the analysis time, and the risk for possible indeterminate error. Several separation and detection schemes for aliphatic polyamines have been 16,25,26 and references therein reported. The most often employed detection methods are reversed-phase liquid chromatography (10-100 pmol detection limits) and ion-exchange chromatography (1-200 pmol detection limits), both coupled with absorbance or fluorescence detection.26 Capillary zone electrophoresis, with spectrophotometric detection (1-3 nmol/L sensitivity), has also been reported and shown to be superior to HPLC in terms of sensitivity and sample volume required, but the reproducibility was less than adequate.27 Gas chromatography (1-600 pmol) with electron-capture detector (ECD), nitrogen-phosphorus detection (NPD) or mass spectrometry has been also used for the analysis of polyamines in biological samples)?8 A review of the analytical methods used to assay biogenic aliphatic polyamines was published recently.25 The review examines the prospects and the limitations of polyamines as cancer markers. Table 1.1 gives summary of the separation and detection methods for aliphatic polyamines. Table1.1 Analytical Methods Used for Polyamine Separation and Detection Method Derivatization Sensitivity Paper . . Chromatography thydrm 0.1~0.5 pmol TLC Ninhydrin 5-100 nmol dansyl derivatives 10-100 pmol Overpressure TLC Ninhydrin + Cd+2 2-60 nmol Fluorescamine derivatives 50-200 pmol HPLC Enzymatic+chemiluminescent 5-500 pmol Dansyl derivatives 10-100 pmol Ion-exchange Flehydrin 1.25301) moll chromatography uorescamine 5- pmo O-Phthaldehyde 5-200 pmol GC N-Trifluoroacetyl (N-TFA) 100—600 pmol Electrophoresis Ninhydrin 5-200 pmol MS N-Trifluoroacetyl (N-TFA) 10-100 pmol Dansyl derivatives 10-100 pmol N-Trifluoroacetyl (N-TFA) + GC’MS deuterated internal standards 1-10 meI Enzymatic Assay - 5-10 nmol Data adapted from J. Chromatogr. B 764 (2001) 385-407 Electrochemistry is an analytical technique characterized by instrumental simplicity, moderate cost, and portability. Electrochemical detectors can be readily combined with separation techniques (e.g., lC, CE, HPLC). Several electrochemical detection schemes for aliphatic polyamines have been reported, involving the use of noble metal electrodes, transition metal electrodes and chemically modified electrodes (CME).29'34 . The amperometric detection of aliphatic polyamines at a nickel-modified glassy carbon electrode in flow injection analysis (FIA) produces limits of detection of 4-10 pmol.33’34 Amperometric detection at a cobalt wire electrode was reported, however, the electrode response was unstable and poorly reproducible.35 The potentiometric detection of amines, coupled with ion- chromatography, at a copper electrode was reported with limits of detection near 1uM (50 pmol).36 One of the most successful electrochemical detection schemes for aliphatic amines was developed using a gold electrode. Johnson and coworkers demonstrated the effective separation and detection of several polyamines using ion exchange liquid 29,30 chromatography coupled with pulsed amperometric detection (PAD). PAD was developed to circumvent the problem of electrode fouling during the oxidation of 29'37'40 in this method, the electrode activity is carbohydrates, alcohols, and polyamines. maintained by a series of positive and negative potential pulses (E1, E2, and E3). The pulsed waveform involves a three step procedure: (E1) amperometric detection, (E2) oxidative surface cleaning, and (E3) reductive surface reactivation. This procedure accomplishes (i) oxidation of the amine, (ii) desorption of adsorbed carbonaceous . . . 32-34 specres, and (m) regeneration of a clean and oxrde-free surface. The group demonstrated the effective separation and detection of several polyamines with detection limits of 200 pmole for putrescine.33 Pulsed amperometric detection is successful but somewhat complicated to implement and, even though the detector is commercially available, detection at constant applied potential is the method of choice for most routine analysis. Several years ago, the Johnson group began designing and testing metal alloy composite 41-49 electrodes (e.g., Ag-PbOz, Fe-PbOz, Bi-PbOz, Cu-Mn) , which can be used to stably detect/oxidize various analytes (i.e., carbohydrates, amino acids, phenol, and benzene) 50-55 at constant applied potential. The group demonstrated that Ag-doped PbOz 54,55 For electrodes were useful for the amperometric detection of aliphatic polyamines. example, these electrodes provided a linear dynamic range of three orders of magnitude and limit of detection of 0.3 11M (6 pmol) for diaminopropane. The amine oxidation reaction at these and all electrodes involves anodic oxygen transfer with the oxidant being OHo, produced during the discharge of H20.31'32'56 it was a premise of Johnson and coworkers that OH- is adsorbed at specific surface sites on the electrode surface, and transferred to the reactants via an electrocatalytic oxygen transfer reaction (OTR) mechanism. The authors proposed several properties necessary for an electrode to stably support anodic oxygen transfer reactions: (i) the use of metal oxides with minimal solubility at the pH conditions of the application, (ii) the existence of a low density of surface sites at which the anodic discharge of H20 (evolution of 02) occurs with lower potential than the surrounding matrix, and (iii) the existence of surface sites that are effective for coordinating the analytes. The authors indicated that silver, as a doping agent in the composite electrodes, serves several functions: (i) a low H20 discharge overpotential compared to PbOz, (ii) capability of adsorbing the OH- generated during H20 discharge, and (iii) capability of adsorbing the reactant species (i.e., 55,57,58 amines). The group has discussed the specific properties of these composite electrodes in several seminal publications that render them useful for anodic oxygen transfer reactions.46’54 It was the reports from their group and the identified composite electrode properties that were the impetus for our efforts to fabricate and test boron- doped diamond electrodes, designed with the appropriate properties, for anodic oxygen transfer reactions. Boron-doped diamond thin films can be considered a “composite materials” that possess the necessary properties to support anodic oxygen transfer reaction. First, high- quality diamond films are stable and resistant to corrosion in strongly acidic and alkaline media.59-61 and references therein Therefore, at the anodic potentials used to detect the polyamines, the electrode structure is stable. Second, the films may contain spa-bonded non-diamond carbon impurities, distributed very locally over the surface. These impurities can exist at the grain boundaries or as extended defects within the diamond film. Most importantly, they have a lower overpotential for oxygen evolution than does the surrounding diamond, and they can also be controllably introduced into the films 62'63 This means that reactive through adjustment in the diamond deposition conditions. OH- will be generated locally at these sites at low overpotential, and not to any appreciable extent on the diamond lattice. Third, boron dopant atoms located at the surface can serve as adsorption/coordination sites for the analyte (e.g., polyamine). Boron atoms can insert directly into the growing diamond lattice, but they can also cluster and accumulate in the grain boundaries. The polyamine adsorption/coordination at the boron sites near the grain boundaries is supposed to be important mechanistically, as these sites are very near where OH- is being generated at low overpotential. The goal of the research described herein was to experimentally verify that properly designed boron-doped diamond films possess the required properties to support oxidation of aliphatic polyamines via oxygen transfer reaction. A model for amine oxidation at diamond, similar to one recognized by Johnson and coworkers for Ag-PbOZ, was proposed and evaluated. The mechanistic and kinetic aspects of the polyamine oxidation reaction were also investigated. The research, as well, aimed to development of a sensitive, reproducible and stable electrochemical detection scheme for these biologically— and environmentally- important molecules. The amperometric detection was performed using both flow injection analysis and reversed-phase liquid chromatography. 1.2. Properties and Applications of Conductive Diamond Films Until the mid 1980s, the only allotropic form of carbon used as an electrode material was in an sp2 configuration. Carbon materials, in the forms of highly ordered pyrolytic graphite (HOPG), glassy carbon (GC), carbon fibers, and carbon paste, have . . . . 64-69 and references found a niche in the electrochemical community for several reasons. therein First, the surface chemistry of graphitic materials can be modified and manipulated to control electron transfer kinetics and molecular adsorption. Second, graphitic electrodes exhibit a fairly wide working potential window and generally stable electroanalytical responses, and third, these materials are commercially available and inexpensive. Single crystal graphite and HOPG are benchmark carbon materials that embody the perfect anisotropic characteristics. These well ordered materials have dimensions on the order of a few um compared to tens of angstroms for more disordered materials like GC and carbon black. Highly ordered pyrolytic graphite (HOPG) consists of parallel planes of condensed six membered rings (graphite sheets), with an interplane distance of 3.354 A. The structure of hexagonal graphite is shown in Figure 1.1 A. The structure of glassy carbon (GC) is shown in Figure 1.1 B. GC resembles a twisted, bent, and interlocked mass of graphite ribbons. The interplanar spacing is 3.48 A, slightly larger than that of HOPG. The graphitic ribbons are composed of so-called basic structural units made up of 10-20 rings stacked up more or less in parallel by two to four layers. GC contains graphite-like domains ranging from 25 A to 100 A. GC is one of the most widely used sp2 carbon electrode materials, in particular for eiectroanalytical . . 69 applications. Figure 1.1 Structural models of (A) HOPG, (B) GC, and (C) fcc diamond lattice. Diamond with a tetrahedral, face-centered cubic (fcc) crystal structure, is shown in Figure1.1C. Diamond consists of sp3-bonded carbon and in contrast to the conventional carbon electrodes is isotropic and thus does not exhibit severe differences in surface microstructure. Diamond offers some of the most extreme physical properties of any material. It is the hardest known material. It has the highest thermal conductivity at room temperature (e.g., about 5 times larger than the value for copper). It is the stiffest material, and is inert to most chemical reagents. These attributes of diamond are a consequence of the small interatomic distance of the carbon atoms in the diamond lattice, and the hydrogen termination of the surface. 70'73 With such a wide range of exceptional properties, it is not surprising that diamond has been referred to as 'the ultimate engineering material'. 74 The advent of low pressure diamond synthesis along with the facility of dopant incorporation, has prompted a growing number of investigations into its use as an electrode material. The first reports of electrochemical investigations of diamond thin films appeared in the mid 19803, and since that time, the application of diamond is showing increasing interest. Figure 1.2 presents the plot of the number of published papers on diamond, over the last four decades. 160 § 120 . '5 8 . =1 .5 ' '. g 80 i- TTI— V5: '5 T 4O _ 1 5 :. 0 _. .H L . , i? if 1965 1975 1985 1995 2005 Year Figure 1.2 Plot of boron-doped diamond related papers over the last four decades. Results based on SciFinder® search engine. 10 Several interesting and electroanalytically-important electrochemical properties distinguish boron-doped diamond thin films from conventional carbon electrodes. Boron- doped diamond thin-film electrodes possess some strong attributes that are very important for electroanalysis. They include (i) a large working potential window (3 to 3.5 V) in aqueous media, (ii) a low background current, (iii) morphological and microstructure! stability, (iv) good responsiveness for analytes without pretreatment, and (v) good long-term stability.63 8"“ '°'°'°"°°s "W"- The wide working potential (1 100 11A) of diamond in aqueous media stems from the observed large overpotentials for the hydrogen (HER) and oxygen evolution reactions (OER).75'76'81 The wide potential window and response stability of diamond electrodes permit electrochemical reactions at potentials that otherwise would be difficult or impossible to accomplish. The low background current and noise magnitude on diamond gives improved signal-to—background (S/B) and signal-to-noise (S/N) ratios.82 The double layer capacitance of diamond electrodes is low and is attributed to a couple of factors. First, the relative absence of electroactive carbon-oxygen functionalifies on the hydrogen-terminated diamond surface results in a lower current. The absence of these functionalities can explain some, but not all of the small non-faradaic current and capacitance for diamond.63 The second and most important contributing factor may be a slightly lower density of surface electronic states near the Fermi level, caused by the semiconductor nature of boron doped diamond.63 A lower surface charge carrier density at a given potential would lead to a reduced accumulation of counter-balancing ions and water dipoles on the solution side of the interface, thereby lowering the background current and capacitance. A striking characteristic of diamond electrodes is their stability compared to conventional carbon electrodes. The inertness of diamond electrodes arises from its 11 atomic density and its sp3 bonding, the most stable among all carbon-carbon bonds. Diamond films are stable during anodic polarization in harsh acidic and alkaline media. Only after polarization at high current densities does the O/C ratio increase, suggesting the formation of carbon-oxygen functionalities on the surface. Diamond is very stable in reducing environments. The “as deposited” boron-doped diamond electrode surface is hydrogen-terminated, but over time with use, surface oxygen is incorporated.59'61 As mentioned earlier, the hydrogen, oxygen, and chlorine evolution reactions are kinetically sluggish on diamond, (i.e., require large overpotentials above the equilibrium potential to proceed at appreciable rates). However, simple, one-electron transfer reactions that do not involve a specific surface sites, e.g., Ru(NH3)6“2’*3, Fe(CN)e'3"‘, IrCls‘z‘a, and MV“2"‘"o proceed reversibly to quasi-reversibly on diamond. 77’78'82'83 There are a number of factors that can influence the electrochemical responsiveness of polycrystalline diamond films: (i) non—diamond carbon impurity phases, (ii) the surface termination (H vs. 0), (iii) the dopant type, level, and distribution, (iv) the grain boundaries and other morphological defects, and (v) the primary crystallographic orientation. The complexity of the diamond electrode nature resulting from these factors is well known, but is still not completely understood. While interesting and, in some cases, unique behavior of diamond electrodes has been demonstrated, there are sometimes significant variabilities in the reported response. Several groups have contributed to the development and applications of diamond electrodes (i.e., aspects of growth and nucleation mechanisms of diamond, the factors influencing the electrochemical response, application in electroanalysis and electrocataiysis). Among those are groups, Fujishima et al. (Japan),8“"94 Pleskov et al. (Russia),95'106 107410, 72,73,111-123 124431 Angus et al., Miller et al. (U-S-AJ and Compton et al. (England).132' 1 144-15 43, Gruen et. al. 5 12 No other material shows as much versatility as an electrode as does electrically conducting diamond. The material can be used in electroanalysis to provide sensitive detection of analytes with superb precision and stability; for high current density electrolysis (1-10 A/cmz) in aggressive solution environments without any microstructural or morphological degradation; and as an optically transparent electrodes for spectroelectrochemicai measurements in the UVNis and IR regions of the electromagnetic spectrum.156'160 Electrically conductive diamond thin films have been successfully used in amperometric detection schemes for several analytes (e.g. azides,161 nitrite, hydrazine, 162,163 164-167 phenols, chlorophenols, chlorpromazineea, carbamate pesticides,85 , ,1 . 92 . 91 29 3° 68, cysteine, Diamond has been used also for 9 histamine and serotonine, detection of trace metal ions using anodic stripping voltammetry.16 In general, hydrogen-terminated diamond provides superior detection figures of merit as compared to all other sp2 carbon electrodes, specifically linear dynamic range, limit of detection, . .. .. 141,170-176 response variability and response stability. 1.3. Diamond Film Deposition and Doping Synthetic diamond can be produced by high-pressure high-temperature (HPHT) techniques or by low-pressure chemical vapor deposition (CVD).74 During the HPHT process, solid carbon is heated (2000 K) under extreme pressure (tens of thousands of atmospheres) in the presence of a molten salt catalyst. The diamond crystals produced are used for a wide range of industrial applications, which require thermal conductivity, hardness and wear resistance. A drawback, however, is the cost, and the fact that the method produces diamond only in the form of single crystals.7o’72’177'178 l3 All CVD techniques for producing diamond require activation of gas phase precursor molecules. This activation can be accomplished thermally (e.g., a hot filament), by electric discharge (e.g., DC, RF or microwave), or by a combustion flame 70'72'74'177 Microwave plasma CVD (MWCVD) is the most (e.g., an oxyacetylene torch). widely used method for diamond growth. In this method, microwave power is coupled into the chamber via a dielectric window to create a discharge. The microwaves transfer energy into electrons in the gas phase, which in turn transfer their energy to the gas molecules through collisions. This leads to heating and dissociation of the gas molecules, and the formation of active species, which react on the substrate surface to 70'72'74’177 The complex chemical and physical processes which occur form diamond. during diamond CVD comprise several different features and are illustrated in Figure 1 .3.74 introduced gas mixture H2 CH4 CH4 H2 CH4 H2 H2 CH4 CH4 microwave W power 0 .cu. “2 -H 0H f'iggvctagrc‘l oCH3 oCHa '11 oCHa 0H 'H free radicals 0H ........... . iii substrate Figure 1.3 Schematic representation of the physical and chemical processes occurring during CVD diamond growth. 14 Diamond is deposited from a hydrogen-rich hydrocarbon gas mixture 1 (CH 4/H2)-179- 81 plasma, plays a critical role in diamond growth. Hydrogen atoms react with neutral it is recognized that hydrogen, specifically atomic hydrogen found in the species such as CH4 to create CH3. reactive radicals, which can attach to surface sites. Atomic H is known to etch graphitic spz, the H atoms serve to remove back to the gas phase any graphitic clusters that may form on the surface, while leaving the diamond clusters behind. H atoms are also scavengers of long-chained hydrocarbons. This prevents the build-up of polymers or large ring structures in the gas phase, which might ultimately deposit onto the growing surface and inhibit diamond growth. The bulk of diamond is fully sp3 bonded, but at the surface during the growth a dangling bonds are present, which need to be terminated in order to prevent reconstruction of the surface to graphite. This surface termination is performed by hydrogen, which keeps the sp3 diamond lattice stable. CVD diamond film can be found with two different morphologies: microcrystalline and nanocrystalline.74 The microcrystalline diamond CVD growth uses hydrocarbon- hydrogen (e.g., 1% CH4/99% H2) gas mixtures, while nanocrystalline films are deposited from CHJAr/Hz (0.5-1 %) gas mixtures. Work on the CH4/H2/Ar system has made it clear that growth species change as a function of gas phase composition and chemistry. The nanocrystallinity is a result of a new growth and nucleation mechanism, which involves the insertion of Ca, carbon dimer, into surface carbon-carbon and carbon-hydrogen bonds. Very high rates of heterogeneous nucleation are observed, and the resulting films consist of randomly oriented phase pure diamond grains with well-defined grain . 148,182 boundanes. The difference in the films grown from CH4/Ar plasmas compared with films from CHJHz plasmas is the microstructure. The nanocrystalline diamond consists of randomly oriented 3-10 nm crystallites, compared with the columnar 15 microstructure of the microcrystalline film, which is composed of crystallites several microns in size. The difference in individual crystallite size means that secondary nucleation rates associated with growth from hydrogen-poor plasmas are much higher 148’182 The fraction of atoms in the grain boundaries of than from CH4/H2 plasmas. microcrystalline films relative to the bulk is extremely small. Most of the defects conserve the tetrahedral bonding within the interface layer. Grain boundaries of a nanocrystalline diamond are different from those of microcrystalline. It was found that they are mainly threefold-coordinated and form it bonds. The electronic band structure of grains was characterized by a smaller band gap than the bulk diamond and it was shown that these 148'182 The special nature of n-bonded states can participate in hopping conduction. grain boundaries imparts unique electrical properties on nanocrystalline diamond films and the conducting grain boundaries will rule the electronic properties of undoped film. With decreasing crystallite size their number vastly increases, and the entire film becomes electrically conducting. Changes in the size of the nanocrystallites, therefore, can dramatically alter film properties. 148’182 More detailed discussion of the film morphology and microstructure, along with evaluation of the electrochemical properties of boron-doped microcrystalline and nanocrystalline diamond is presented further in the research chapters. The band gap of diamond is 5.5 eV rendering the material a very good insulator, however, impurity incorporation results in extrinsic semiconducting properties.183 Naturally occurring impurities that have implications for the host diamond consist of N, H, 148,182 and B. The nitrogen incorporation gives diamond n-type electronic character. Substitutional nitrogen is a deep donor with ionization energy of about 1.7 eV. The 16 donors are typically too deep to provide any useful semiconducting properties to the material at room temperature. It is known that the electrical properties of diamond are strongly influenced by the presence of hydrogen. Hydrogenation leads to a decrease in resistivity through increased hole conduction leading to p-type character of diamond surface region. The study by Jackman et al.,184 suggest that hydrogen within diamond give rise to shallow acceptor states or deeper gap states, causing band bending and accumulation of holes. The origin of this effect remains still controversial. Boron is the most often used dopant and this element can be introduced into diamond without significant distortion of the lattice. Boron may be introduced into CVD diamond film from: (i) a gaseous source, such as 3%, (ii) a solid state dopant, such as 8203 and h-BN, and (iii) a boron containing solution, such as boron powder dissolved in organic solvents from which vapors are carried into the reaction chamber by the growth gases.184 Boron is an electron acceptor and provides a p-type electronic character to the material.183 Doping levels ranging from 10‘6 to 1021 B/cm3 can be achieved. As boron is introduced into the lattice, the activation energy of boron acceptor is near 0.37 eV above the top of the valence band. Activation energy of acceptors is observed to decrease with increasing boron concentration. This decrease of the energy can be explained by acceptor band width increase, impurity band formation and conduction as result of it. For highly doped films, activation energies as low as 2 meV have been reported.185 For these highly doped films the impurity band conduction merges with the valence band, and metallic conduction is observed. High abundance of boron during diamond deposition can degrade the quality of the diamond,186 which is manifested by increased point defect densities and in extreme instances by graphitization of the 17 86 diamond film.1 '183 Figure 1.4 shows the bandgap diagram for boron-doped diamond electrode. Conduction Band (Cb) Bandgap 5.4 eV B (Impurity Band) 0.37 eV, Moderately Doped OIIOOOOOO EF v 2 meV, Highly Doped Valence Band (Vb) Figure 1.4 A diagram for the band structure of boron-doped diamond. Boron-doped, polycrystalline diamond electrodes deviate from ideal p-type semiconductor behavior because of a high density of mid-gap electronic states. The mid- gap density of states results from at least four factors: (i) boron-doping level, (ii) lattice hydrogen content, (iii) inherent grain boundaries and other defects in the polycrystalline films, and (iv) non-diamond carbon impurity at the surface.63'163 Of the four factors presented that can influence the diamond conductivity, the surface non-diamond carbon phases exert little influence on the response of doped diamond films. It has been shown by Granger at al.;187 that the removal of non-diamond carbon by acid washing and hydrogen plasma treatment does not inhibit the electrode electronic properties. The role of morphological defects and grain boundaries on the electrode kinetics, specifically the electrical conduction, at polycrystalline films is still not clearly understood. Diamond is host for a variety of extended defects such as stacking 18 faults, microtwins, dislocations, grain boundaries, and mixed habit growth features.63 and ”lemmas "main These defects could serve as discrete sites for electron transfer or could simply affect the electronic properties of the material by increasing the density of states. The mid-gap density of states arising from grain boundaries and defects, though, is relatively low compared to the number of states arising from the boron doping and lattice hydrogen. The electrical conduction could be also inhibited by grain boundaries and defects because of a reduction in the electron and hole mobilities. 1.4. Outline of the Dissertation All the experimental protocols are presented in Chapter 2. In Chapter 3, the electrochemical and physical properties of microcrystalline boron-doped diamond thin- film electrodes are described. The films were deposited on three different substrates (W, Mo, and Si). The purpose for this work was to learn how the substrate material directly or indirectly influences the physicochemical properties and the electrochemical response. In Chapter 4, the electrochemical and physical properties of boron-doped nanocrystalline diamond thin-film electrodes are presented. Chapter 5 contains results of voltammetric studies on electrooxidation of aliphatic polyamines: cadaverine (CAD), putrescine (PUT), spermine (SPM) and spermidine (SPMD) at microcrystalline and nanocrystalline diamond are presented. in this chapter the amine oxidation model is proposed and tested. An evidence for importance of non-diamond carbon impurities and boron surface sites in the reaction model is presented. The mechanistic and kinetic aspects of the amine oxidation reaction have been studied by voltammetric methods on stationary and rotated disc diamond electrodes and are presented in Chapter 6. Amperometric detection of aliphatic amine, using diamond has been coupled with flow injection 19 analysis. Results from these studies are presented in Chapter 7. Finally, in Chapter 8 the results from the reversed-phase liquid chromatography with amperometric detection, using boron-doped diamond thin-film electrodes for the determination of aliphatic polyamines are presented. 20 Chapter 2 EXPERIMENTAL SECTION 2.1. Boron-Doped Diamond Thin Film Growth Boron-doped diamond thin films were deposited using a commercial chemical vapor deposition (CVD) system (1.5 kW ASTeX Corp., Lowell, MA). Microwave energy (2.45 GHz), originating from a magnetron, was directed into a quartz reaction chamber, from above, to generate a plasma. Deposition occurred on the surface of the substrate placed on the stainless steel stage positioned in the middle of the reaction chamber. Boron doping was accomplished using two methods: a solid source of boron (B203 + h- BN) placed inside reaction chamber and gaseous 32H6 was mixed, in a controlled way, with a source gas mixture. Figure 2.1 shows the microwave-assisted CVD reactor. Details of the deposition conditions and substrate preparation procedures are presented in the following paragraphs. 21 Figure 2.1 A photo of the microwave-assisted CVD reactor: (a) diborane and Ar cylinder cabinet, (b) hydrogen cylinder cabinet, (c) methane cylinder cabinet, (d) microwave generator, (e) reactor, (f) mass flow controller and stop valve, (9) power supply for microwave generator, (h) pressure gauge, (i) water chiller, and (j) rotary pump. 2.1.1. Microcrystalline Diamond Electrodes Doped Using a Solid State Source. The films were deposited on boron-doped p-Si (100) (Virginia Semiconductor, lnc., Fredricksburg, VA), tungsten (Goodfellow Metals, Cambridge, England), and molybdenum (Goodfellow Metals). The substrates, ca. 1 cm2, were first rinsed with ultrapure water, toluene, methylene chloride, acetone, propanol, methanol, and ultrapure water. The silicon substrates were subsequently etched in concentrated hydrofluoric acid (HF) for 60 s. After rinsing with ultrapure water and drying, the substrates were mechanically polished by hand for 5 min with a mixture of 2 parts 3203 powder (Aldrich) and 1 part diamond powder (0.1 pm, GE Superabrasives, Worthington, OH). The 22 polishing scratches the substrate and seeds the surface with diamond particles, both of which serve as nucleation centers during film growth. Polishing in the presence of 3203 appears to lead to a higher doping level than is normally achieved in its absence. The polished substrates were ultrasonicated in acetone for 1 min and placed in the CVD reactor on top of a boron diffusion source (GS 126, 8203containing disk, BoronPlus, Technegias, Inc.). The nominal particle coverage of 108 cm'z, after sonication, was determined by AFM. In some cases, a piece of h-BN (Goodfellow Metals) was placed adjacent to the substrates in the reactor. The h-BN, in addition to the embedded 8203 particles, and the boron diffusion source, served as sources for the incorporated boron dopant atoms during deposition. Ultrahigh purity (99.999%) methane and hydrogen were used. The films were deposited from either a 0.33, 0.40, 0.50, 0.67, or 1.0% methane/hydrogen (CH4/H2) source gas mixture, a total gas flow of 200 sccm, a forward power of 1 kW, a deposition pressure of 35 torr, an estimated substrate temperature of 800-850 °C, and a growth time of ca. 15-20 h. The plasma was ignited with all gases flowing into the reactor. After deposition, the CH4 flow was stopped and the films remained exposed to the H2 plasma for an additional 15 min at the growth conditions. After this annealing period, the plasma power and system pressure were gradually reduced, over a 30 min period, to 400 W and 20 torr, in order to cool the samples (<400 °C) in the presence of atomic hydrogen. Post- growth annealing in atomic hydrogen is critical for etching adventitious non-diamond phases, minimizing dangling bonds, and ensuring full hydrogen termination. The plasma was then extinguished and the films further cooled to room temperature under a flow of hydrogen. Table 2.1 lists the typical deposition conditions for microcrystalline diamond thin film growth. 23 Table 2.1 Microcrystalline Diamond Thin Film Deposition Parameters Deposition Parameters Substrate Si, Mo, W, CH4/H2 (v/v) ratio 0.3 - 1 % Plasma forward power 1 kW Deposition pressure 30 - 40 torr Deposition time 15-20 h Substrate temperature 800-850 °C Boron dopant source 8203, h-BN Doping level ~1x1019-1x1020 B/cm3 Resistivity 0.1-0.01 O-cm A doping level of approximately ~1x1019-1x1020 B/cm3 was typical for the films, as determined by boron nuclear reaction analysis (BNR) measurements. The film thickness was nominally 4 to 6 pm and the apparent in-plane resistivity was 0.01 Q-cm or less. 2.1.2. Diamond Electrodes Doped with BzHo 2.1.2.1. N anocrystalline Boron-Doped Diamond Thin Films The boron-doped nanocrystalline diamond thin films were deposited on p-type Si(100) substrates (~10'3 O-cm, Virginia Semiconductor inc., Fredricksburg, VA). The surface of the Si substrate was mechanically scratched on a felt polishing pad with 0.1 pm diameter diamond powder (GE Superabrasives, Worthington, OH). The scratched substrate was then sequentially washed with ultrapure water, isopropyl alcohol (IPA), 24 acetone, IPA, and ultrapure water to remove polishing debris. The scratching treatment enhances the rates of nucleation during the initial growth. The films were deposited from CH4/ Ar/ H2 Typical flow rates were 1, 94 and 5 sccm, respectively, for CH4, Ar, and H2 gas mixtures. The microwave power and deposition pressure were maintained at 800 W and 140 torr, respectively. Table 2.2 lists the typical deposition conditions. The substrate temperature was estimated by an optical pyrometer to be about 800 °C. The deposition time was 2 h, and the resulting films were approximately 4 pm thick, as estimated from the substrate weight change. The plasma was ignited with all gases flowing into the reactor. At the end of the deposition period, the CH4 flow was stopped and the Ar and H2 flows continued. The films remained exposed an H2/Ar plasma for approximately 10 min. The substrate was then cooled in , the presence of atomic hydrogen to an estimated temperature of less than 300 °C by slowly reducing the power and pressure over a 4 min period. This leads to the formation of a thin and continuous nanocrystalline diamond film in a relatively short period of time. Table 2.2 Nanocrystalline Diamond Thin Film Deposition Parameters Deposition Parameters Substrate p-Si (100) or (1 1 1) CH4/Ar/H2 (v/v) ratio “945 Plasma forward power 0.8 kW System pressure 140 torr Substrate temperature .300 °C Deposition time 2 h Boron dopant source BgHe Doping level (B/C in gas phase) 200-10,000 ppm 25 The doping level was determined from boron nuclear reaction analysis measurements. Table 2.3 presents a summary of the doping level, apparent resistivity as a function of the BZH6 concentration in the source gas mixture. Table 2.3 Boron Concentration and Resistivity for a Nanocrystalline Diamond BgHs concentration BIC ratio Boron concentration Resistivity Introduced (ppm) (ppm) (ppm) (fl-cm) 0 mm 0* 50 0.3-0.5 1 ppm 200 500 0.3 10 ppm 2,000 810 0.1 20 ppm 4,000 ----- 0.05 30 ppm 6,000 ----- 0.03 50 ppm 10,000 3300 0.02 Note: *- no intentionally added BgHe 2.1.2.2. Microcrystalline Boron-Doped Diamond Thin Films The substrate preparation protocol for the boron-doped microcrystalline diamond films was the same as that for nanocrystalline films. The growth parameters, however, were different and are presented in Table 2.4. 2.1.3. Diamond Acid Washing and Hydrogenation Procedure Prior to electrochemical testing, the diamond films were often cleaned by a 2- step Chemical treatment. The first step involved immersing the films in warm aqua regia for 30 min, followed by rinsing with ultrapure water. The second step involved exposing the samples to a warm solution of H202 (30%) for 30 min, followed by rinsing with ultrapure water, and drying. The films were then hydrogen plasma treated (microwave- assisted CVD) for 1h to remove the surface oxides formed during the acid washing and 26 to terminate the surface with hydrogen. The hydrogen was introduced into the reactor at a flow rate of 200 sccm. The plasma power, system pressure, and temperature were 1 kW, 35 torr, and 800 °C, respectively. Table 2.4 Microcrystalline Diamond Thin Film Deposition Parameters Deposition Parameters Substrate p-Si (100) or (1 1 1) CH4/H2 (v/v) ratio 0.5 % Microwave power 1 kW Deposition pressure 45 torr Substrate temperature 800 °C Deposition time 10 h Boron dopant source BZH6 Doping level ~1x1019-1x1020 B/cm3 Resistivity 0.1- 0.01 Q-cm 2.2. Diamond Thin Film Characterization The films were comprehensively characterized by scanning electron microscopy (SEM), atomic force and scanning tunneling microscopy (AFM/STM), Raman spectroscopy, powder x-ray diffraction (XRD), boron nuclear reaction analysis, and static secondary ion mass spectrometry (SIMS). At a minimum, every film was examined by AFM and Raman spectroscopy. Table 2.5 presents a summary of the techniques used, along with information each technique provides. 27 Table 2.5 Analytical Techniques Used for Film Characterization Technique information Scanning Electron Microscopy Morphology Transmission Electron Microscopy Morphology Atomic Force Microscopy Morphology Raman Spectroscopy Microstructure X-Ray Diffraction Crystallinity Boron Nuclear Reaction Boron dopant concentration X-Ray Photoelectron Spectroscopy Surface chemical composition Secondary Ion Mass Spectrometry Boron dopant concentration 2.2.1. Scanning Electron Microscopy (SEM) SEM was performed with a JEOL—6400V electron microscope. Images were collected at a 2048x1536 pixel resolution with an ADDA ll digital image acquisition system and AnainiS Pro 3.2 image acquisition software (both Soft Imaging System GmbH, Munster, Germany). Typically, an accelerating voltage of 15 W was employed. The samples were mounted on aluminum stubs with conductive carbon tape. 2.2.2. Transmission Electron Microscopy (TEM) TEM was performed with a JEM-4000EX electron microscope. The samples were prepared by growing a nanocrystalline diamond thin film for 10 min on a Si substrate and then dissolving the substrate using an HF/HN03 solution. The free- standing pieces of diamond were collected from solution and washed on a gold TEM grid. 28 2.2.3. Atomic Force Microscopy (AFM) AFM was performed with a NanoScope lll instrument (Digital Instruments, Santa Barbara, CA) in the contact mode. Si3N4 probe tips, mounted on Au cantilevers (100 pm) with spring constants ranging from 0.54 to 0.06 N/m, were used to acquire topographical images of the films, in air. 2.2.4. Raman Spectroscopy Raman spectra were obtained in a back-scattered collection mode with a Raman 2000 spectrometer equipped with a microprobe attachment (2080) and a direct video module (2090) (Chromex, lnc., Albuquerque, NM). The spectra were obtained using a 1200 grove/mm holographic grating. Excitation was from a frequency-doubled Nd-YAG . (532 nm, 30 mW) laser. The laser beam was focused to a spot size of approximately ~5 um yielding an estimated power density of 150 kW/cmz. Spectra were collected with a 10 s acquisition time. The spectrometer, prior to use, was calibrated with a single crystal (Type lla) diamond standard (100 orientation, Harris Diamond). 2.2.5. X-Ray Diffraction (XRD) 20 x-ray diffraction analysis was performed using a Philips X-Pert MPD diffractometer in the continuous scan mode. X-rays of 1.540 A were generated from a Cu anode. The scan range was from 20 to 125 degrees, using a step size of 0.05 degrees. A count time of 1 s/step and a generator power of 45 W at 40 mA were used. The diffractometer was equipped with a parallel plate collimator, a flat graphite monochromator, and a Xe proportional counter detector. Diffraction profiles were collected, using incident angles of 0.5, 2.0, and 5.0 degrees with respect to the sample surface. The samples were mounted onto a top referencing, zero background sample 29 holder, and supported from the backside with wax. The XRD was performed as a courtesy by Dr. Jennifer Spear at Philips. 2.2.6. Boron Nuclear Reaction (BNR) The boron dopant concentration was deterimined by boron nuclear reaction analysis (Surface Characterization Facility, Case Western Reserve University). Calibration was performed with a piece of high-quality boron nitride. 2.2.7. Secondary Ion Mass Spectrometry (SIMS) The secondary mass ion mass spectrometry (SIMS) was performed using a Cameca/ION-TOF SIMS lV instrument (Muenster, Germany) with a time-of-flight mass . analyzer. Positive-ion spectra were acquired using Ar ions at 11 keV. The data were obtained as a courtesy by Prof. Tom Beebe at the University of Utah. 2.2.8. Resistivity Measurements of the in-plane film resistivity were made using a tungsten four point probe with a 1 mm probe spacing. Currents (pA range) were applied between the outer two probes and the resulting voltage drop between the inner probes was measured. Five measurements were taken and W! or R values were determined by averaging the measurements. No correction factor was used for edge effects or probe placement. The resistivity was calculated using: (1.11.) ( ) where l is the film thickness (mm). 30 2.3. Electrochemical Measurements 2.3.1. Cyclic Voltammetry Experiments The cyclic voltammetric measurements were made with a CYSY-2000 computerized potentiostat (Cypress System inc., Lawrence, KS), using a single- compartment, three-electrode glass cell, presented in Figure 2.2. 63 reference electrode auxiliarv electrode / /_ nitrogen é / purge gas port l“.\ I - ,4 ’59 x, 's o-ring current collector working electrode Figure 2.2 Design of the electrochemical, three-electrode glass cell. The diamond thin-film working electrode was pressed against a Viton® o-ring and clamped to the bottom of the glass cell. Ohmic contact was made by pressing a copper plate against the backside of the scratched and cleaned Si substrate, which contained a bead of Ga/ln alloy or silver paste. A graphite rod was used as the counter electrode, and a commercial Ag/AgCi electrode (3 M NaCI, Cypress Systems, Inc.) served as the 31 2 reference. The geometric area of the working electrode was ca. 0.2 cm. All 0 measurements were performed at room temperature, ~25 C. 2.3.2. Rotating Disc Electrode (RDE) Experiments For the RDE measurements, a diamond electrode disc (0.11 and 0.2 cm2) was perpendicularly attached to the stainless steel rod, using conductive epoxy resin. The electrode was then sealed with an insulating epoxy resin. The stainless steel rod was inserted into the Teflon TM holder and then attached to the rotor. This arrangement allowed the electrode to be rotated at angular velocities (w:21rf). ranging from 500 to 3000 n/min (522-3142 rad/s). The RDE measurements were performed with a Metrohm Model 628-10 rotator (Metrohm, Switzerland) connected to a CYSY-2000 computerized . potentiostat (Cypress System Inc., Lawrence, KS) Rotor TeflonTM holder Epoxy Resin a» . “sass“: §§§ {a Diamond Disk Electrode Figure 2.3 Design of the rotating disk electrode. 32 2.3.3. Double-Layer Capacitance The double-layer capacitance (Cdj) measurements were made using a SR830 DSP lock-in-amplifier (LIA, Stanford Research Systems) connected in series with the Omni—90 analog potentiostat (Cypress System Inc., Lawrence, KS). The potentiostat had a low pass filter built into the design with a 20 us time constant. A 10 mV r.m.s. and 40 Hz sine wave was co-added to constant dc potentials between -500 and +1000 mV (100 mV increments), and the imaginary component of the total impedance was monitored. All values reported are normalized to the electrode geometric area of 0.2 cm2. The capacitance was calculated from the relationship: 1 1 Cdl=——"' where 04. is the double-layer capacitance [uF/cmzj, 23m is the imaginary component of the total impedance [VIA], and 0) is the angular frequency. 2.4. Flow Injection Analysis (FIA) Flow injection analysis studies were performed using a thin-layer, electrochemical flow cell. Figure 2.4 shows a cell diagram. 168The two-piece thin-layer flow cell was constructed with Kel-fTM. The top piece contained the entrance and exit ports for the fluid flow and a place for the Ag/AgCl reference electrode (3 M NaCI, Cypress Systems, Inc.). The exit port was fitted with a short piece of stainless steel tubing (-6 cm in length), which served as the auxiliary electrode. The bottom piece of the cell supported the working electrode. Electrical contact was made by pressing a piece of copper foil against the backside of the conducting substrate. Sometimes the copper was mated to the substrate, using silver paste, or after scratching the backside with a graphite pencil. A 0.1 cm-thick neoprene rubber gasket separated the surface of the 33 working electrode (~0.11 cm?) from the top piece of the cell. A rectangular groove (1.1 x 0.1cm2) was manually cut into the gasket, and this defined the flow channel. Assuming a 25% compression of the gasket with the two pieces of the cell clamped together, the cell volume was estimated to be ~8 uL. reference electrode outlet auxiliary inlet electrode Kel-f body, rubber gasket top piece working t electrode curren collector rubber backing Kel-f body, bottom piece screw clamp Figure 2.4 Design of the thin-layer electrochemical flow cell. Flow injection analysis (FIA) was conducted with components of commercial HPLC system (Shimadzu Scientific Inc.): an autoinjector (SlL-IOADVP), a controller (SCL-10AVP), a gradient pump (LC-10 AD VP), and a degasser (DGU-14 A). All the components were electrically grounded to a common point, and the electrochemical flow cell was housed in Faraday cage to reduce the pick-up of spurious electrical noise. FIA 34 measurements were made with an Omni 90 analog potentiostat (Cypress System inc., Lawrence, KS). Data acquisition and processing were performed with the Shimadzu Class-VP 7 chromatography software package. The carrier solution was deoxygenated on-line, using the degasser, and aspirated through a suction filter. Additionally, the carrier solution was continuously purged with nitrogen. The injection volume was 20 uL and the flow rate was 1 mL/min, unless stated otherwise. 2.5. HPLC System The reversed-phase liquid chromatography was performed with the commercial HPLC system described above. The mobile phase was a mixture of 0.01 M borax buffer (pH 11.2) and acetonitrile. The mobile phase was continuously deoxygenated with' nitrogen (99%) purge gas. The injection volume was 20 uL, and the flow rate was 1 mUmin. unless stated otherwise. The isocratic separation of the aliphatic polyamines was carried out using a guard column (X-Terra, 5 pm particle size, 3.9 x 20 mm) and a C13 reversed-phase column (X-Terra, 5 pm particle size, 4.6 x 150 mm) in series (Waters). The stationary phase was 3-(chlorodimethylsilyl)propyl-N-dodecyl carbamate. This column has a modified silica support in which some of the silicon atoms are replaced with carbon. This substitution reduces the solubility at high pH, yet leaves sufficient silanol activity to enable the use of traditional bonding chemistry to covalently attach the stationary phase. The polyamine solutions were passed through a PTFE syringe filter (0.2 um) before being placed in the glass sample vials of the autoinjector. 2.6. Chemicals and Reagents Reagent grade quality chemicals and ultrapure deionized water (18 MO, Barnstead E—pure) were used to prepare all solutions. Sodium perchlorate hydrate (99%, 35 Aldrich), sodium bicarbonate (EM Science), and sodium hydroxide (Fisher) were used to prepare the 0.01 M carbonate buffers. Sodium chloride (Aldrich), sodium borate (Fisher Scientific), and sodium hydroxide (Fisher Scientific) were used to prepare a series of borax buffers ranging in pH from 8 to 11. The borax buffer was prepared with 0.01 moVL of Na28407 and 0.01-0.1 moVL of NaCl, adjusted to the required pH with the addition of 6 M NaOH. The 0.1 M phosphate buffer, pH 7.2, was prepared with potassium phosphate dibasic (Mallincrodt) and potassium phosphate monobasic (Mallincrodt). The methylamine (Aldrich), ethylamine (Aldrich), propylamine (Aldrich), ethylenediamine (Aldrich), 1,3-propyleneamine (Aldrich), putrescine (1,4- butanediamine, 99%, Sigma), cadaverine (1,5-pentanediamine, 99%, Sigma), 1,6- hexanediamine (Aldrich), spermidine (N-(3-Aminopropyi)-1,4-butanediamine, 99%, - Sigma), spermidine (2,2) (N-(2-AminoethyI)-1,2-ethylenediamine, 99%, Sigma), spermidine (3,3) (N-(3-Aminopropyl)-1,3-propylenediamine, 99%, Sigma), spermine (N, N'- Bis(3—aminopropyl)-1,4-butanediamine, 99%, Sigma), Dytek® A (1 ,5-pentanediamine, 2-methyl, DuPont Nylon), Dytek® EP (1,3-propylenediamine, 1-ethyl, DuPont Nylon), and acetonitrile (HPLC grade, Aldrich) were used without any additional purification. 1.0 mM stock solutions were prepared by dissolving the analyte in the buffer. Table 2.6 shows the molecular formula and properties of the aliphatic amines used. The diamond film electrochemical response was characterized by cyclic voltammetry using the following inorganic and organic systems: potassium ferrocyanide -3/-4 +3/+2 (Fe(CN)5 , Aldrich), hexaamineruthenium (Ill) chloride (Ru(NH3)8 . Aldrich), iridium (IV) hexachloride (erlg'ZI'3 , Aldrich), methyl viologen dichloride hydrate (Mv+2/+1/0 +2l+3 , Aldrich), 4-tert-butylcatechol (4-tBC, Aldrich), and ferric sulfate hydrate (Fe , Aldrich). All reagents were used as received. The supporting electrolyte for +3l+2 +2/+1/0 Fe(CN)6'3/'4, Ru(NH3)6 , "016'“, and methyl viologen (MV ) was 1.0 M KCI. 36 *2’*3 was 0.1 M HCIO4. The supporting electrolyte for 4-tert-butylcatechol, and Fe Potassium chloride (Fisher), nitric acid (Fisher), and perchloric acid (Aldrich) were ultrahigh purity (99.999%). All glassware was cleaned by sequentially rinsing in a KOH/ethanol bath, an alconox solution, and ultrapure water. Table 2.6 Molecular Formula and Properties of the Aliphatic Amines 5,189 Formula Amine Formula pK. values Weight methaignine HzNCH3 10.62 31 g/mol ethylEagine H2N(CH2)2 10,70 45 g/mol Proligignine H2N(CH2)3 10.75 59 g/mol _ ethyleggiismine H2N(CH2)2NH2 6.85, 9.93 60 g/mol diamirggégpane H2N(CH2)3NH2 8.47, 10.46 74 g/mol Pugejfgne H2N(CH2)4NH2 9.35 , 10.6 88 g/mol 03218212309 H2N(CH2)5NH2 9.13 ,10.8 102 g/moi hexarzagi?mine H2N(CH2)6NH2 9.8, 10.9 116 g/mol Speggflrg) 3’4 H 2N(CH2)3NH(CH2) 4NH2 8.2, 9.86, 10.85 145 g/mol spggnwignae’ :13 H2N(CH2)3NH(CH2)3NH2 8.1, 9.70.10.70 131 g/mol srzggnwignzzzf H2N(CH2)2NH(CH2)2NH2 3.0, 9.50, 10.51 103 g/mol Dytek® A (DA) H2NCH2 CH(CH3 )(CH2)3NH2 10-69 ‘16 9""0' DygfifP H2N 0,4“;sz )(CH2)2NH2 10.69 102 g/mol srggmy‘; H2N(CH2)3NH(CH2I4NH(CH2)3NH2 1613433 202 9"“°' 37 Chapter 3 THE CHARACTERIZATION AND ELECTROCHEMICAL RESPONSIVENES OF BORON-DOPED MICROCRYSTALLINE DIAMOND THIN-FILM ELECTRODES 3.1. Introduction The work reported herein was motivated by our interest in understanding the structure-function relationship of boron-doped, polycrystalline diamond thin film electrodes. An issue that has not been addressed yet is the role the substrate might have on the film’s electrochemical properties. Several different substrates have been 86,190 191,79,192 75,193 I 194,195 196-198 used for diamond growth, including Si, Mo W, T Pt and Ni.‘96"97 A substrate material must meet several criteria in order to serve as a substrate for diamond deposition. First, the substrate must have a melting temperature higher than the temperature required for diamond growth (700-850 00). Second, the substrate should have a thermal expansion coefficient comparable with that of diamond. The differences in thermal expansion coefficients between the diamond and the substrate 38 can cause significant film stress, either compressive or tensile, depending on which has the higher expansion coefficient. This can lead to film or substrate cracking and film delamination. Third, a useful substrate material is one capable of forming a carbide interlayer. This interfacial layer aids in the adhesion by (partial) relief of stresses at the interface caused by lattice mismatch or substrate contraction. If one considers the carbon- substrate interactions possible, then metal, alloy, and pure element substrates can be subdivided into three broad classes based upon their reactivity with carbon: (i) little or no solubility or reaction with C, (ii) solubility or reaction with C, and (iii) carbide formation. Si is the most commonly used substrate material due to it’s availability, low cost, and favorable properties. It has a sufficiently high melting point (1683 K), it forms a thin, carbide interfacial layer (a few nanometres thick), and it has a relatively low thermal expansion coefficient. Molybdenum and tungsten display similar qualities, so these are often used as well. The chemical nature, the thermal expansion coefficient, and the lattice parameter of the substrate all influence the nucleation, growth, and adhesion of diamond films. The substrate material could impact the electrochemical response of the films in at least three ways. First, pinholes and small crevices between the grains could provide pathways for solution to permeate through the film and reach the underlying substrate. In such cases, electroactive substrates (e.g., W and Mo) will enhance the current magnitude and the potential dependent features in the background voltammetric response. Second, the plasma processing, in the case of microwave-assisted CVD, could result in the incorporation of impurities. For example, silicon impurities from the quartz reactor chamber in diamond are well known.199 Pt nanoparticles have been detected in diamond films deposited on Pt substrates.198 Tungsten impurities are known 39 to be incorporated during hot filament growth.109 The role these metal impurities might have directly or indirectly on the film structure and electrochemical response has not yet been investigated. Metal inclusions exposed at the surface could catalyze electrochemical reactions. They could also react with the solvent/electrolyte causing an increase in the voltammetric background current. Third, deposition of diamond on non- . . 200,201 diamond substrates could lead to internal stress. The magnitude of the stress will depend, at least in part, on the difference in thermal expansion coefficients. The role stress plays in the electrochemical response is unknown. One possible manifestation of localized stress is less morphological and microstructural stability during exposure to harsh electrochemical conditions (8.9. electrolyte, solution pH, and current density). The objectives of this work were: (i) to determine how different substrates affect- the morphological, microstructural, chemical, and electrochemical properties of microwave-assisted CVD diamond thin-film electrodes, (ii) to determine if metal particle impurities from the substrates are incorporated during deposition, and (iii) to learn how any incorporated metal particles, if present, influence the electrochemical response. 3.2. Results and Discussion 3.2.1. Electrochemical Characteristics Figure 3.1 shows capacitance-potential (Cdi-E) profiles for microcrystalline diamond films deposited on Si, W, and Mo, respectively, in 1.0 M KCI. The profiles for the uncoated metals are also shown, for comparison. There is a general trend of increasing capacitance with more positive potential for all three diamond films. The capacitance is low and ranges from 3 to 7 iiF/cm2 over this potential range, similar to previously reported results.163 40 80 - . - L - . . - ' Q) Mo—diamond { I ' _ . *\ O W-diamond I 60 _ 1. A Si-diamond . fi ‘3 Mo-metai E] W—metal c,ll (pr I cm’) I‘\ 400 ‘ 700 1000 Potential (mV vs Ag/AgCl) Figure 3.1 Capacitance-potential profiles in 1.0 M KCI for uncoated tungsten and molybdenum, and diamond thin films deposited on W, Mo, and Si substrates. Capacitance values are normalized to the geometric area (0.2 cm2) of the electrodes. Frequency = 40 Hz. AC amplitude = 10 mV. These values are over an order of magnitude lower than those for the uncoated metals, except at positive potentials where the formation of an oxide layer passivates the surface. The capacitance of the diamond films deposited on the metal substrates is a little lower at potentials from -500 to 0 mV and a little higher at potentials from 700 to 1000 mV than the capacitance of the film on Si. These data might reflect some minor differences in the electronic properties of the films grown on the metal substrates as compared with those on Si. There is, however, no evidence of any direct solution contact with the metal. The low capacitance of the diamond films, compared to glassy carbon 41 (25-50 uF/cmz), for example, is attributed to (i) the absence of redox-active and ionizable surface carbon-oxygen functional groups, and (ii) differences in the electronic properties of diamond due to a lower charge carrier concentration.63 Figure 3.2 shows cyclic voltammetric i-E curves for microcrystalline diamond films deposited on Si (A), Mo (B), and W (C) in 1.0 M KCI. Background voltammetric i-E curves are extremely useful for evaluating the diamond film quality.202 Both the magnitude of the background current and the working potential window are highly sensitive to the presence of electroactive metal particles and non-diamond spa-bonded carbon impuritiy. Both types of impurity are expected to be concentrated at grain boundaries and defect sites (i.e., dislocation). The i-E curves are all flat and featureless, with no peaks associated with redox-active surface carbon-oxygen functionalities. The. scans were stable with multiple sweeps (not shown here), indicating that there is no penetration of the electrolyte solution through the grain boundaries or defects to the reactive substrate. However, the current magnitude for the film deposited on Si is a factor of 2 lower than the current for the films deposited on the metal substrates. For example, at 250 mV, the current for the film on Si is 0.6 uA, while the current for the films deposited on W and M0 is 1.2 11A and 1.1 11A, respectively. The larger background current is somewhat in contrast to the equal or lower magnitude of the double layer capacitance at this potential seen in Figure 3.1. Therefore, we assume that the slightly larger voltammetric current contains a faradaic component, possibly associated with redox processes on the surface. One possibility is surface redox processes associated with trace levels of surface non-diamond carbon impurity incompletely removed by the acid washing and rehydrogenation pretreatment. 42 12.0 6.0 - A i 0.0 .. —: {—7. e— .6.0 - l l l l I A A .500 .250 0 250 500 750 1000 < 12.0 1 , . r . , . , . 4 . 3 E 6.0. B JJ 5 0.0 _ i 1 P— U -6.0 K 1 . 1 1 1 -500 .250 0 250 500 750 1000 12.0 - . 6.0- C w —> 0.0 r . ’6‘) '- i l 1 1 '1 -500 -250 0 250 500 750 1000 Potential (mV vs Ag/AgCl) Figure 3.2 Background cyclic voltammetric i-E curves in 1.0 M KCI for boron-doped microcrystalline diamond thin films deposited on (A) Si, (B) Mo, and (C) W. Scan rate = 0.1 V/s. Electrode geometric = 0.2 cm2. Figure 3.3 presents background cyclic voltammetric i-E curves in 1.0 M KCI for microcrystalline diamond films deposited on Si, Mo, and W, which reveal the full working potential window. The working potential window, defined as the potential range between current limits of i 100 11A, is 3.25 V, 3.3 V, 3.25 V for diamond deposited on Si, Mo, and W, respectively. The voltammograms all have a similar shape. The background current 43 is low and featureless, with approximately 1.25 V as an anodic potential limit. The current at this potential is due presumably to chlorine evolution.124"7"02 100' ML . 1(23250 -1750 -1250 -750 -250 250 750 1250 100 50: B __) O '- I 1. <———— -50 " f _1 l I 1 J 1 I I l 99250 -1750 -1250 -750 -250 100 r 50-C Current (11A) l l l I I l 250 750 1250 J M -250 250 750 1250 -750 499250 -1750 -1250 Potential (mV vs. Ag/AgCl) Figure 3.3 Background cyclic voltammetric i-E curves in 1.0 M KCI for boron—doped, microcrystalline diamond films deposited on (A) Si, (B) Mo, and (C) W. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. There is some anodic current passed just prior to the onset potential for chlorine evolution for the Mo and W diamond films. This charge is likely associated with oxidation of residual non-diamond carbon impurity on the surface not removed by the acid washing and rehydrogenation treatment. This charge also may be associated with oxidation of the diamond surface involving replacement of hydrogen with different chemisorbed oxygen functionalities (e.g., OH).73’187'203 The reduction of chlorine back to the chloride occurs at different potentials on the three electrodes. A weak peak at ca. 550 mV is seen for this reaction at the Si film. The peak occurs at much more positive values at the Mo and W films, ca. 950 and 1000 mV, respectively. Clearly, the Mo and W films are more active for the reduction of chlorine than is the diamond film deposited on Si.124’202 The current at the cathodic limit (> -1700 mV) is attributed to hydrogen evolution. There is some cathodic current that flows on the negative potential sweep at the Mo and W films starting at ca. -750 mV. The current is likely associated with the reduction of trace levels of dissolved oxygen. The slope of the i-E curves in the chlorine and hydrogen evolution regions is similar for all three films. The approximate 3.3 V window for all the films indicates that the macroscopic microstructure and chemical nature of the surfaces are similar. Most importantly, these data, along with the capacitance profiles shown in Figure 3.1, suggest that influential levels of metal impurities from the substrate 99M exist at the surface. This conclusion is further supported by SIMS data presented below. A low background current and wide working potential window (3 to 4 V) are characteristic features of high quality diamond. By high quality, we mean a clean, hydrogen-terminated film with negligible amounts of non-diamond carbon impurity and a low level of secondary nuclei formed on the primary diamond microcrystallites.63’163 45 Figures 3.4A and B show background cyclic voltammetric i-E curves in 1.0 M HN03 for microcrystalline diamond films deposited on W and Mo. Current (uA) 150 ’A 50— J ——> -50 _ -150 1 l 1 1 1 l . l 1 l 1 -1500 -1000 -500 0 500 1000 1500 2000 100 -100 1 l 1 l 1 l 1 l 1 l 1 -l600 -1100 -600 -100 400 900 1400 1900 l 4 l -200 Potential (mV vs Ag/AgCl) Figure 3.4 Background cyclic voltammetric i-E curves in 1.0 M HN03 for boron-doped, microcrystalline diamond thin films deposited on (A) W and (B) Mo. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm . 2 The working potential windows are 3.1 and 3.3 V, respectively. A window of 3.2 V was observed for the film deposited on Si (figure not shown). The currents are low and featureless throughout most of the potential range, as was the case in KCI. The M0 film shows some excess anodic charge starting near 1500 mV just prior to the onset 46 potential for oxygen evolution. This is likely due to the oxidation of the reactive non- diamond carbon impurity in the grain boundaries. The current at the anodic limit is associated primarily with oxygen evolution and 63 and the current at the cathodic limit is due to combination of nitrate reduction1 hydrogen evolution. There is also current cross-over during the reverse scan for the nitrate reduction/hydrogen evolution current suggesting that an activation overpotential is needed on this surface to induce the reduction reaction. It is also interesting to note that the slope of the i-E curves in this potential region is steeper for the W film than for the Mo film. The steeper slope reflects faster electrochemical reaction kinetics. No change in the shape of the traces was observed during multiple sweeps (>10) suggestive of a very stable morphology and microstructure in this corrosive medium. importantly, long term (12h) potential cycling between these limits revealed no indication of electrolyte permeation through the film to the substrate. Based on this observation, we can rule out the possibility of the substrate material participating directly in the electrochemical reactions. in other words, the diamond film is continuous over the surface, and the morphology and microstructure are stable such that solvent/electrolyte does not permeate the film and attack the electroactive substrate. Four different analytes were used to probe the reactivity of the electrodes: +3l+2 +2l+1/0 Ru(NH3)6 , methyl viologen (MV ), irCi6'2"3, and Fe(CN)6‘3"4. The redox systems were chosen based on their E0 values, each’s mechanism of electron transfer, and their sensitivity towards various surface properties, i.e., surface chemistry. The question regarding the electronic nature of the diamond electrodes was addressed with redox systems that undergo uncomplicated electron exchange. The Ru(NH3)6+3/+2, and +2l-i-1/0 MV , and IrCls'ZI'3 are such redox couples. These couples undergo outer-sphere, 47 e 163 I e e o e e one-electron transfer reaction. Their kinetics are rather insenSitive to the presence of I O I 2 surface ox1des or low levels of contaminants at conventional sp carbon electrodes. Ru(NH3)6*2’*3 and MV+2M have E°’ values in a negative potential range from -0.17 to - 0.68 V vs Ag/AgCl, respectively, and thus can be used to probe electronic properties as a function of applied potential. Fe(CN)5'3/-4, on the other hand, was used to probe the surface chemistry of the electrodes.163 This redox couple undergoes a one-electron transfer reaction with the kinetics being highly sensitive to the chemical nature of the diamond electrode surface.163 The redox reaction proceeds via an inner-sphere route mediated by a specific surface interaction available on the hydrogen-terminated surface. Such surface interactions appear to be blocked on the oxygen-terminated surface. Apparent heterogeneous electron transfer rate constants (koapp) were determined by employing the theory developed by Nicholson and Shain:204 _01. (.20. 2x k0 DR ‘11: 01mm}? /RT) ’4 where ‘P is the kinetic parameter, Do and Dr are the diffusion coefficients for oxidized and reduced form, respectively, a is the transfer coefficient, and v is the sweep rate [V/s]. This theory allows one to calculate k°alpp by determining the variation of the peak potential separation (AEp), with sweep rate. No correction was made for double layer effects, therefore, the rate constants presented are referred to as apparent. Calculations of the rate constants were periorrned assuming a transfer coefficient, 01, of 0.5 and diffusion coefficients of 6.8 x 10'6 cm2/s for erls'ZI'a, 5.0 x 10'6 cm2/s for methyl 48 +2/+3 3 viologen, 5.5 x 10'6 cm2/s for Ru(NH3)6 , 6.2 x 10’6 cm2/s for Fe(CN)5' (ferrocyanide), and 7.6 x 10'6 cm2/s for Fe(CN)6'4 (ferricyanide).204 Figure 3.5 shows cyclic voltammetric i-E curves for four redox systems: (A) Fatwa“. (B) RuiNHa)e*“’*3 '2” , (C) erl5 , and (D) methyl viologen at diamond films deposited on the three different substrates. The voltammograms were obtained on the hydrogen-terminated surface prior to the potential cycling experiments shown in Figure 3.4. ...... Si-diamond 60- _ - _. W-diamond Mo-diamond Current (ILA) Potential (mV vs Ag/AgCl) 3"“ in 1 M KCI at diamond films deposited on different substrates. Scan rate = 0.1 V/s. Working electrode area = 0.2 cm2. Figure 3.5 A Cyclic votammetric i-E curves for 1.0 mM Fe(CN)5' Fe(CN)¢;'3/'4 is a very surface sensitive redox system on diamond.63 AEp for all three films is similar (~70 mV), as are the peak currents. The low AEp suggests the surface is largely oxygen free. The apparent heterogeneous electron transfer rate constant, koapp is ca. 0.02 cm/s. This value is similar to what has been recently reported 49 for diamond,‘733 and is within an order of magnitude of the highest rate constant reported for freshly activated glassy carbon?6 Ru(NH3)6+2/+3 is a surface insensitive redox system on diamond.63 AEp for all three films is similar (~ 77 mV), but the peak current for Si are larger than for the other two substrates. k0,,pp is about a factor of 3 higher at the W film and a factor of 2 higher at the Si film, compared to the Mo film. The approximate value of 0.02 cm/s is similar to what has been previously reported for diamond.63 It is supposed that differences in k°app are due to slight variations in the density of electronic states at these potentials. 60 _ ...... Si-diamond 30 _ .. - _ W-diamond _ Mo-diamond D Current (ILA) G I D ' D 0 O - _ '0 l O l I I 0 I -60 _ -800 -400 0 400 Potential (mV vs Ag/AgCl) Figure 3.5 B Cyclic votammetric i-E curves for 1.0 mM Ru(NH3)5+2/+3 in 1 M KCI at a diamond films deposited on different substrates. Scan rate = 0.1 We. Working electrode area = 0.2 cm2. 50 30 i ...... Si-diamond C 15 _. .. _ W-diamond é Mo-diamond ....... i=1 0 - E1 . U -15 _ -30 1 1 1 1 1 1 400 600 800 1000 Potential (mV vs Ag/AgCI) 6 . - - . 3 _ ...... Si-diamond 5‘ D . .. - .. W-diamond 1. 0 9 Mo-dlamon _ ‘11:. x 7 . g -3 L ....... . 5 . S -6 - , . U . . ‘1/ .9 L— I l l . -1800 -l300 -800 -300 Potential (mV vs Ag/AgCl) Figure 3.5 Cyclic votammetric i-E curves for (C) 0.5 mM erl6'2/'3 and (D) 0.1 mM methyl viologen in 1 M KCI, at diamond films deposited on different substrates. Scan rate = 0.1 V/s. Working electrode area = 0.2 cm2. 51 erls'ZI'3 is also a surface insensitive redox system on diamond.63 AEp ranges from 59 to 67 mV and k°app is ca. 0.1 cm/s. koa,pp for the Si film is the largest and is about two times higher than k0,ilpp for either the Mo or W films. The koa,pp values reported herein for erlgs'ZI'3 are also the highest that have been observed for diamond.63 +2/+1 MV is a surface insensitive redox system on diamond. 63 AEp ranges from 59 to 66 mV and koapp is as high as 0.2 cm/s for the Si film. The koapp value is larger by about a factor of 2 for the Si film, compared to the Mo and W films. The koapp value at the Si film is the highest we have observed for diamond 63 and is very near the highest rate constant seen for freshly activated glassy carbon.66 Typically, only the MV‘LZI+1 couple is used for evaluation of the performance of the electrode, as the MV‘q/o couple has been observed to undergo adsorption.136 At higher concentration, the low solubility of MV0 results in the formation of a precipitate on the electrode surface. Usually voltammetry for MV‘L‘O‘IJ'1 shows two separate anodic stripping responses associated with 136 an amorphous and crystalline deposit. This process is dependent upon couple variables (e.g., scan rate, mass transport rate, and analyte concentration).136 For the +2] +1 couple, cyclic voltammogram for all the deposited lower concentration of the MV films was well resolved with diffusion limited peak current. Figure 3.6 shows plots of ipox vs scan rate” (v19). For all films and all redox systems ipox varies linearly with vii (0.1 -0.5 V/s) (r2=0.996 to 0.999) with a near-zero y- oxfipred ratio near 1.0 was also observed. This response linearity is axis intercept. A of ip indicative of reaction rates (i.e., current) limited by semiinfinite linear diffusion of reactants to the interfacial reaction zone. 52 120- VA. 1 120 - 15 80: 80: 1 40; . 40: 1 0. , 1‘11. 0", ‘4 L. 0‘5‘10‘15 20 25 0 5‘1015‘20 25 Oxidation Peak Current (ILA) 60 - ' - ' ' p . 12 - ' b C 1 ' 1 L D 40 - . 8 - 20 . . 4 - 0 1 l g i 1 4 1 1 1 0 - 1 1 i . i 1 m 1 L 0 5 10 15 20 25 0 5 10 15 20 25 Scan Rate"2 ((mV/sf”) ...... Si-diamond _ . 1. W-diamond Mo-diamond -3/-4 +2/+3 Figure 3.6 Plots of ip°" vs scan rate"2 for (A) Fe(CN)6 2/-3 , (B) Ru(NH3)6 ,(C) erls' , and (D) methyl viologen at diamond films deposited on Si, Mo, and W substrates. A summary of the cyclic voltammetry data and the apparent heterogeneous electron transfer rate constants is presented in Table 3.1. It is important to note that the large rate constants were obtained without any conventional electrode pretreatment (e.g., mechanical polishing, laser activation, etc.). In general, all three diamond films are responsive for these redox systems. All the voltammetric data discussed herein support 53 the notion that the basic electrochemical properties of films deposited on the different substrates are similar. As long as there is complete film coverage, the substrate exerts, the substrate seems to exhibit negligible influence on the response. Table 3.1 Summary of Cyclic Voltammetric Data and Apparent Heterogeneous Electron Transfer Rate Constants for Boron-Doped Microcrystalline Diamond Films Analyte Electrode AEP [Pox ko'” (mV) (M) (""31 Si-diamond 72:3 57:2 0.018:0.008 1.0 mM , .314 W-diamond 71:4 59:1 0.020:0.005 FeICN)6 . Mo—diamond 71:3 59:1 0.027:0.003 Si-diamond 76:8 59:2 0.025:0.008 1.0 mM . +2r+3 W—diamond 74:5 57:1 0.037:0.019 RUINHalo . Mo-diamond 76:4 56:3 0.0350004 Si-diamond 67:1 23:1 0.21:0.08 0.5 mM , .214 W-diamond 68:8 24:2 0.12:0.07 II‘CIs . Mo-diamond 62:4 25:1 0.1 6:0.08 Si-diamond 67:2 53:05 0.22:0.09 0.1 mM , MVIZM W—diamond 66:8 4.8:0.2 0.12:0.01 Mo-diamond 63:6 50:01 0.14:0.10 Note: Cyclic voltammetric AEp and ipox values are averages for a 0.1 V/s scan rate. koapp values are based on measurements performed on three Si-diamond, two Mo-diamond, and two W-diamond films over a narrow scan rate range from 0.1 V/s to 0.5 Vls. All solutions were prepared with 1 M KCI. Measurements were performed with 70 % iR compensation. 3.2.2. Material Characterization Figure 3.7A-C, shows 6 x 6 pm2 AFM images (force mode) of diamond films deposited on Si, Mo, and W, respectively. The images reveal that all the films are well faceted and polycrystalline with many randomly-oriented, octahedral and cubo- octahedral-shaped crystallites. The films are continuous over the substrate with no 54 obvious pin-holes or cracks. There is a sizeable number of secondary growths, mostly in the films deposited on Mo and W, as indicated by the arrows. Figure 3.7 Atomic force microscope images (6 x 6 pm) of boron-doped, micro« crystalline diamond films deposited on (A)silicon, (B)molybdenum, and (C) tungsten. These are the smaller grains that form in the grain boundaries and at defects on the facets of the primary crystallites. The increased fraction of secondary growths on the Mo and W films presumably results from a slightly higher inherent defect density. The higher number of secondary growths causes an increase in the fraction of exposed grain 55 boundary, and it is supposed this is the reason for the increased background voltammetric current. A nominal crystallite diameter of 3 pm is estimated from the image features for all three substrates. The roughness, as estimated from the change in the feature height, is 1.5 : 0.2, 1.5 : 0.3, 2.5 : 0.3 pm for Si, W, and Mo films, respectively. SEM was used also to probe the film morphology. Figure 3.8 shows a cross-section image of a diamond thin film deposited on Si. The image reveals a well faceted, polycrystalline morphology. The diamond film is continuous over the substrate, with a thickness of about 7 pm. Figure 3.8 SEM image of a boron-doped microcrystalline diamond film deposited on Si. Glancing angle XRD measurements (courtesy of Jennifer Spear, Philips Analytical) were made to examine the crystalline phases present near the surface and within the film bulk. Three angles of x-ray beam incidence (with respect to the surface) 56 were used: 0.5, 2.0, and 5.0 degrees. XRD analysis of the diamond film deposited on Si showed the major peaks expected for cubic diamond at 43.9, 75.3, and 91.5 degrees. These are assigned to the (111), (220), and (311) crystallographic orientations, respectively.205 Figure 3.9 shows diffraction patterns for the film deposited on Si, at an incidence angle of 0.5 degrees. The intensity of all the peaks increased with increasing incidence angle (data not shown). 1000 : (111) 600 L Counts (220) (311) 50 60 70 80 90 100 2 Theta (degrees) Figure 3.9 Grazing incidence XRD measurement of a boron-doped microcrystalline diamond film deposited on Si at a 0.5 degrees incidence angle. The XRD analysis of the diamond films deposited on W and Mo showed that, in addition to the expected peaks for cubic diamond, both W and tungsten carbide (WC) phases are present (data not shown). The W reflections were at 40.3, 58.3, 73.2, 87.0, 100.6, and 57 114.9 degrees and are assigned to the (110), (220), (211), (220), (310), and (222) crystallographic orientations, respectively.205 WC reflections were at 35.6, 48.3, 73.1, and 77.1 degrees. Since no electrochemical evidence was found for elevated levels of W metal impurities at the surface, the presence of these phases means that the x-rays were probing the interfacial region of the film, even at this shallow incidence angle. In addition to cubic diamond, only a molybdenum carbide (M020) phase was detected on the Mo substrate. The M020 reflections were at 34.5, 37.9, 52.2, 61.6, 69.6, 74.7, and 75.7 degrees and are assigned to the (100), (002), (102), (110), (103), (112), and (201) crystallographic orientations, respectivelyzos Again, since no electrochemical evidence was found for Mo metal impurities at the surface, the presence of these phases means that the x-rays are probing the interfacial region of the film, even at this shallow angle. For both metal substrates, the diamond peaks are shifted to slightly higher angles from where they are on Si. This is consistent with a contraction of the diamond lattice (i.e., compressive stress), which is likely caused by the higher thermal expansion coefficient of the two metal substrates, as compared to diamond, by a factor of 4 to 5. The FWHM of the (111) diamond line on Si was 0.33 degrees for the 0.5 degree incidence angle. The FWHM observed for diamond line on Mo and W was 0.40 and 0.49, respectively. The FWHM is indicative of the defect density in the films with a higher defect density corresponding to a larger peak width. The larger FWHM on the metal substrate is consistent with the higher fraction of exposed grain boundary due to the secondary growths, as evidenced in the AFM images. Figure 3.10 presents Raman spectra for films deposited on Si, Mo, and W. The spectra consist of one intense band near 1332 cm'1 attributed to the first-order phonon scattering mode for diamond.207 This peak is attenuated and shifted toward higher 58 wavenumbers for the Mo and W films. Peak positions of 1336.8 : 0.6 cm'1 and 1334.04 : 0.3 cm'1 are observed for the Mo and W substrates, respectively. 1 _ W- 0nd‘ I g I 13000 - g Mo-diamond . Q g ' Si-diamond ‘ 8 10000 ' Si'diamond 1 I L Li 1 1 I L i 1250 1370 1490 1610 . .1 7000 _ Raman Shift (cm ) . Mo-diamond ‘ W-diamond 4000 1 A 11; W1 :7 _-;--:_ 1. - 4 " 1280 1330 1380 1430 1480 1530 1580 Raman Shift (cm") Figure 3.10 Macro-Raman spectra for boron-doped microcrystalline diamond films deposited on W, Mo, and Si. Enlargement of the spectra reveals the weak scattering intensity by non-diamond sp2-bonded carbon impurity centered at ~1500 cm'i. The peak shift toward higher wavenumbers is indicative of a compressive stress within the film, due to the mismatch in thermal expansion coefficients. The results are in agreement with the XRD data. The line width (FWHM) of the diamond phonon on M0 is 10.4 : 0.3 cm'1 and on W is 8.9 : 0.5 cm'1. These values are slightly higher than the line width on Si of 75:06 cm". The line width for a peak of HPHT diamond was 2 cm'1, for 59 comparison. The peak width provides information regarding the crystalline quality and, to a first approximation, is inversely related to the phonon lifetime (i.e., defect density).207 in other words, the diamond films deposited on Mo and W possess a higher defect density than do the films on Si. This is in agreement with the AFM, SEM, XRD, and electrochemical data. The increased number of secondary growths in the films grown on Mo and W contribute to a higher defect density. Non-diamond sp2 carbon impurity is likely located at these defects. The expanded spectral region in the insert shows that the films deposited on Mo and W have slightly higher levels of spz-bonded carbon (1490 cm'1). These impurities likely contain a mixture of sp2 and sp3 bonding, however, their concentration is low, considering that the peak intensity is weak and that the scattering cross-section for graphite (a model for spa-bonded carbon) is ca. 50 times larger than that of diamond.207 Raman spectroscopy can provide information regarding the residual stress in the diamond film.208 At room temperature, the total film stress is the sum of two . . , , , 1’ contributions: intrinSic and thermal stress.20 208 The origins of the intrinsic stress are not well understood; however, the structural mismatch between the substrate and the film, or defects within the film, can be responsible. The therrnai stress is caused by the difference in thermal expansion coefficients of the film and substrate, and can be . 200,201 evaluated from the expressron: E 5therm.= T9( (Xd - as ) OT 'V in which Ed is the Young modulus of diamond (Ed=1210 GPa), vd is the Poisson ratio (vd = 0.1), and and as are the thermal expansion coefficients for diamond and the substrate, respectively, ST is the difference between the growth temperature and room 60 200’201 At 25°C, the thermal expansion coefficients for Si. W. MO. and temperature. diamond are 3 X106, 4.5 X106, 5 X106, and 1.1 x10'6 K4, respectively.209 The thermal stress was calculated to be -1.7, -4.1, -3.3 GPa for Si, Mo, and W, respectively. Secondary ion mass spectrometry (SIMS) was used to probe the chemical composition of the films, particularly in the near-surface region. A goal of this work was to determine if the surface contained higher levels of Mo and W, compared to the film bulk. Figure 3.11 presents depth profiles for the (A) Si, (B) Mo, and (C) W films. The ion profiles (raw intensities uncorrected for sputter yield) for C (m/z 11.99), C2 (m/z 23.99), B (m/z 11.00) are shown for all films, as are the Si (m/z 27.97), Mo (m/z 97.90), and W (m/z 183.85) signals, depending on the substrate. The films were not acid washed and rehydrogenated, and were not used in any electrochemical measurements, prior to the SIMS analysis. in other words these were “as deposited”. The profile for the diamond film deposited on Si shows intense signals for Si and C in the near-surface region. The intensities of these signals decrease significantly with depth. In fact, the C signal for films on all three substrates is very intense in the near-surface region and decays with depth into the film. This trend was very unexpected and uncharacteristic of other diamond films examined by SIMS. It is supposed that the intense C and Si signals are due to unintentional contamination layer on the surface. Three possible sources of contamination are (i) pump oil condensation in the instrument, (ii) handling of the samples after deposition and during insertion into the vacuum chamber. Relatively high Si (m/z 27.97) signals were detected at the surface of all the films (profiles not shown in Fig 3.11 B and C). This supports the contamination layer cause. This contamination layer was apparently not present on those samples used in the electrochemical measurements, because its presence would have been 61 indirectly detected in either the background voltammograms, the capacitance profiles, or the voltammetric response for surface-sensitive redox analyte, like Fe(CN)5'3/'4. . T 40 _ ' r 1190 i- 30 _ I ”’1 .1 L i- .3... \C2 A :8. + A 20 1— :!- ,5. 1’ \‘ ‘3: if); .‘I a. A. . : , . '-. 1190 .VW 30 _ (:2 1 Intensity (counts) Apparent Depth (nm) Figure 3.11 SIMS data for boron-doped microcrystalline diamond films deposited on (A) Si, (B) Mo, and (C) w. 62 For the diamond film deposited on Si, the Si, C, and B signals are relatively constant with depth after this suspected contamination layer is sputtered through. For the diamond films deposited on Mo and W, the B, Mo, and W ion intensities are constant with depth and are very near the background levels. Importantly, these SIMS results, even though apparently plagued by a contamination layer, do not reveal the presence of any elevated levels of metal impurities at the surface or in the bulk. Positive ion static SIMS measurements for all the diamond films prior to the depth profiling revealed the presence of a number of hydrocarbons: CzH3, C3H3, 03H4, C4H4, C3H4O, C4H8, C2H60, 05H3, C5H4, Cng, C4H50. The 0 content on the surface was very low, based on ion intensities of the CH0 fragments. 3.3. Conclusions Good quality, boron-doped diamond thin films were deposited on three different substrates, Si, Mo, and W. The resulting films were characterized by electrochemical methods of analysis, atomic force microscopy, x-ray diffraction analysis, Raman spectroscopy, and dynamic secondary ion mass spectrometry. The goal for presented research was to ieam how the substrate directly (e.g., chemical impurities) or indirectly (e.g., stress) affects the morphology, microstructure, chemical, and electrochemical properties. A possible chemical contaminant during growth on metal substrates is metal impurity. Impurity at the surface could directly influence the electrochemical response by direct reaction with the analyte or electrolyte, or indirectly by creating more film defects (i.e., dislocations). It was observed that the electrochemical response of all films was independent of the substrate material. All films exhibited properties of high quality diamond - a wide working potential window; a low voltammetric background current and 63 capacitance; and good responsiveness for Fe(CN)5'3”‘, Ru(NH3)6*2’*3, IrCl6'2"3, and methyl viologen without any pretreatment. koapp values ranged from 10'2 to 10" cm/s for all four redox systems. The films deposited on Mo and W did have some physical properties that were slightly different from those deposited on Si. For example, the Mo and W films have a higher defect density and more thermal stress (compression) than do films on Si. There is, however, no evidence of elevated levels of metal inclusions at the diamond surface or in the film bulk. The higher defect density is caused by a larger fraction of grain boundary resulting from secondary growths. Higher levels of spZ-bonded carbon impurity appear to track the defect density, based on Raman spectroscopic measurements. The higher number of defects and the elevated fraction of non-diamond carbon seemingly lead to only a slight increase in the voltammetric background current. in summary, the substrate appears to have little influence, directly or indirectly, on the electrochemical response of microcrystalline diamond films, at least for these four redox systems. it is possible that the higher internal film stress on the Mo and W substrates could affect their dimensional stability during the imposition of harsh electrochemical conditions. This remains for further investigation. Chapter 4 THE CHARACTERIZATION AND ELECTROCHEMICAL RESPONSIVENESS OF BORON-DOPED NANOCRYSTALLINE DIAMOND THIN-FILM ELECTRODES 4.1. Introduction High quality diamond films can be formed with two different morphologies and microstructures: microcrystalline and nanocrystalline. The distinction between these two structures arises from the nominal grain size, which for microcrystalline films is >1 um and for nanocrystalline films is on the order of 15 nm. Conventional microcrystalline diamond CVD growth uses hydrocarbon-hydrogen (e.g., 1% CH4/99% H2) source gas mixtures, and it is known that under such growth conditions, hydrogen plays a number of .. 111,146,147,210-214 critical roles. Among these are stabilization of the diamond lattice and removal of sp2-bonded carbon impurity, when formed, due to preferential gasification over spa-bonded diamond. Gruen and coworkers discovered that phase-pure nanocrystalline diamond can be deposited from CH4/Ar gas mixtures containing very 146-148,151,182,213,214 little or no added hydrogen. There are two kinds of nanocrystalline 65 diamond films often reported on. The first are films deposited from high CH4/H2 (>5 %) gas mixtures. These films have nanometer-sized features, due to the high rate of nucleation, but are generally of low quality (so-called “dirty” diamond) with significant levels of secondary nucleation and spa-bonded carbon impurity phases. The second are films deposited from CH4/Ar (~1 %) gas mixtures. These films consist of randomly oriented, nanometer-sized grains of phase-pure diamond and are generally of higher quality. The grains are on the order of 15 nm in diameter and the grain boundaries consist of n—bonded carbon atoms (2-4 atoms wide). The 1r-bonded grain boundaries have a profound effect on the mechanical, electrical, and optical properties of the films.148’182 The nanocrystalline diamond films reported on herein are of the second type. The most remarkable difference in films grown using hydrogen-poor Ar gas mixtures, compared with conventional hydrogen-rich mixtures, is the nanocrystallinity of the former compared with the microcrystallinity of the latter. The nanocrystallinity is a result of a new growth and nucleation mechanism that involves the insertion of Cg, carbon dimmer, into carbon-carbon and carbon-hydrogen bonds. The Cg addition is believed to occur by a two-step growth mechanism.213 in computational models, a Cg molecule approaches the unreconstructed monohydride diamond surface and inserts into a C-H bond. The Cg molecule then rotates to insert its other C into a neighboring C-H bond on the surface. A Cg molecule then inserts into an adjacent C-H bond, parallel to the newly inserted Cg dimer. The original state of the surface is recovered by a formation of a bond between carbon atoms in the adjacent surface dimers. Very high rates of heterogeneous nucleation are observed, on the order of 1010 cm'z, and the resulting films consist of 182 randomly oriented, phase-pure diamond grains with well-defined grain boundaries. The smooth nanocrystalline films possess interesting mechanical, tribological, and 66 electrical properties, owing to the small grain size. For example, the films transition from an electrically insulating to an electrically conducting material with a reduction in crystallite size from the micrometer to the nanometer level.151 This is largely due to the fact that the grain boundaries contain rr-bonded carbon atoms (i.e., a high density of electronic states over a wide energy or potential range). The grain boundaries are conducting and, because their numbers vastly increase with decreasing crystallite size, the entire film becomes electrically conducting and functions as a good electrode material.153 Theoretical calculations suggest that localized electronic states are introduced into the band gap due to the presence of sp2-bonded dimers and spa- hybridzed dangling bonds in the grain boundaries.182 There is a lack of spatial connectivity among the spz-bonded sites; therefore, the associated gap states do not form an extended n-system but remain localized. Our group reported on the structural and electrochemical characterization of nitrogen-incorporated nanocrystalline diamond thin film electrodes, deposited from CH4/Ng/Ar gas mixtures.153 The electrical conductivity of these nitrogen-containing films increases with increasing nitrogen content in the source gas mixture, up to about 5 %, and are generally more conductive than the early forms nanocrystalline diamond.15°’153 Recent Hall measurements (mobility and carrier concentration) for films deposited with 10 and 20% added Ng revealed carrier concentrations of 2.0 x 1019 and 1.5 x 1020 cm'a, 1 The room temperature carrier mobilities were 5 and 10 cmzN-s, respectively.15 respectively. A negative Hall coefficient indicated the electrons are the major charge carrier in these films. An explanation for the effect of nitrogen is that the impurity disrupts the microstructure by increasing the extent of spz-bonding. The increase in rc-bonding causes an increase in the density of electronic states. Computational work indicated that 67 the incorporation of nitrogen into the grain boundaries is energetically favored. by 3-5 eV, as compared to substitutional insertion into the grains.182 The electrical properties of the nitrogen-containing nanocrystalline diamond films are largely influenced by the it-bonded carbon atoms in the grain . 1 ,151,1 ,182,215 boundaries. 5° 53 While these electrode materials possess good electrochemical properties, much like those for high quality microcrystalline diamond films, their electrical response is strongly linked to the physicochemical properties of the grain boundaries. Therefore, the electrochemical response can be strongly affected by changes in the ir-bonded grain boundary carbon atoms. This is particularly true during exposure to chemically aggressive solutions conditions that can cause the oxidative etching of the grain boundary carbon atoms. It would be better if the nanocrystalline diamond films exhibited a through-grain conduction mechanism as a result of impurity incorporation, such as boron doping. Such films should exhibit conductivity that scales with the doping level and possess electrochemical properties that are largely unaffected by changes in the physicochemical properties of grain boundaries. The objectives of this work were: (i) to determine the morphological, microstructural, chemical, and electrochemical properties of microwave-assisted CVD nanocrystalline diamond thin-film electrodes, deposited with and without boron dopant, (ii) to determine if the properties of the nanocrystalline film are similar to that of microcrystalline diamond. 4.2. Results and Discussion 4.2.1. Electrochemical Responsiveness Figure 4.1 shows a background cyclic voltammetric i-E curve for a boron-doped nanocrystalline diamond thin film in 1.0 M KCI. The film was deposited from a CH4/Hg/Ar (1%/5°/o/94%) source gas mixture with 10 ppm of added BgHe. The deposition time was 68 2 h. The voltammetric background curve is largely featureless, over the potential range, and stable with cycling. There are no peaks present, associated with redox-active surface carbon-oxygen functionalities,63'163 although there is a small anodic charge passed between 500-800 mV, just prior to the onset of chlorine evolution. 10 Current (ILA) c i .10 L 1 - 1 -600 -250 100 450 800 1150 l A A l A Potential (mV vs Ag/AgCl) Figure 4.1 Background cyclic voltammetric i-E curve in 1.0 M KCI for a boron-doped nanocrystalline diamond film deposited on Si. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. Deposition time = 2h. Source gas mixture = CH4/Hg/Ar (1%/5%/94%) (v/v) with 10 ppm BzHe added. Power = 800 W. Pressure = 140 torr. Substrate temperature = ~800 °C. The curve shape ispsimiiar to that for boron-doped microcrystalline diamond; however, the current magnitude of the former is slightly higher. For instance, at 250 mV, the anodic current for the nanocrystalline film is 0.96 pA (4.8 uA/cmz), while the current for the microcrystalline films is about a factor of 1.5 less, 0.6 pA (3.0 iiA/cmz). The larger current is likely due to the higher fraction of exposed sp2-bonded grain boundary. The 69 spa-bonded carbon in the grain boundaries is a source of charge carriers, along with the substitutional boron dopant atoms; hence, the overall carrier concentration is higher. A higher carrier concentration means greater excess surface charge at a given potential and a larger double layer charging current. The current for both diamond electrodes, however, is significantly less than that for freshly polished glassy carbon, 7-10 11A (35-50 pA/cmz). The low background current is a characteristic feature of diamond electrodes and leads to improved SBR in electroanalytical measurements.63’163 Figure 4.2 shows a background cyclic voltammetric i-E curve for the same boron- doped nanocrystalline diamond in 1.0 M KCI over a much wider potential range. The working potential window is about 3.1 V (: 100 11A or 500 uA/cmz) with a largely featureless response between the potential limits. The anodic current at 1200 mV is due to the oxidation of chloride to chlorine. The cathodic current flowing at -1800 mV is due to the reduction of water (hydrogen ion) to hydrogen. There is a current crossover at about -1700 mV, indicative of an activation overpotential necessary for the initiation of the hydrogen evolution reaction. A low background current and wide working potential window (3 to 3.5 V) are characteristic features of both high quality, boron-doped nanocrystalline, and microcrystalline diamond.63'1‘r’3’163 The electrochemical responsiveness of the boron-doped nanocrystalline diamond thin film toward six redox systems was investigated using cyclic voltammetry. The influence of diamonds physical, chemical, and electronic properties on the electrode reaction kinetics and mechanism for each of these systems was tested. Figure 4.3 shows cyclic voltammetric l-E curves for (A) Fe(CN)6'3"4, (B) Ru(NH3)6+3/+2, '2”, (D) methyl viologen (Mv*2’*"° +2,” in 0.1 M HCIO4. The potential scan rate (v) was 0.1 V/s. A summary of (C) erl6 ) in 1 M KCI, and (E) 4-tert-butyicatechol, and (F) Fe some of the cyclic voltammetric data is provided in Table 4.1. 70 230 . 130 ’. 30’. -70 1 Current (11A) -l70 L -270 ' . 1 1 1 - . - . - 1 1 -2100 -1500 -900 -300 300 900 1500 Potential (mV vs Ag/AgCl) Figure 4.2 Background cyclic voltammetric i-E curve in 1.0 M KCI for a boron-doped nanocrystalline diamond film. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. The deposition conditions are the same as stated in Figure 4.1 Table 4.1 Cyclic Voltammetric Data for a Boron-Doped Nanocrystalline Diamond Film AE E OX. I OX I 0X, Analyte P P P PM (mV) (mV) (114) 11 0.5 mM illlv“"”+1 in 1 M KC! 60 -633 34.5 1.1 1 mM lau(Ni-i;1)11“3’+2 In 1 M KCI 59 -135 62.7 0.99 1 mM Fe(CN)5’3M in 1 M KC! 63 309 69.4 1.02 1 mM t-butylcatechoi in 0.1 M HCIO4 419 655 112.2 1.8 0.5 mM lrc111'2"3 in 1 M KC! 61 823 22.7 0.98 42/43 1 mM Fe In 0.1 M HCIO. 679 926 36.0 0.91 71 80 40. Oxidation current (ILA) 150 100. 50'. 01 -501 .100 ’ -600 A A A l A A A 1 ML 4L AAAAAAAAAAAA A-100‘400‘900‘1400 80 40. 0. (__ -40. - AAAAAA A AAAA 60 30 '1 0 i -60 ’ - - - - - - - - -1300 -900 -500 -100 -30 L .60’..-A-1-. -600-1004009001400 Potential (mV vs Ag/AgCl) Figure 4.3 Cyclic voltammetric i-E curves for (A) Fe(CN)6'3/'4, (B) Ru(NH3)6+3/+2, (C) irc16‘2"3, (0) methyl viologen (MV +2l+3 and (F) Fe +2/+1/0 ) in 1 M KCI, and (E) 4-tert-butylcatechol, in 0.1 M HCIO4 at a boron-doped nanocrystalline diamond thin film. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. The diamond deposition conditions are the same as stated in Figure 4.1 72 The E°’ for these redox systems ranges from approximate! +800 to -1100 mV, if so they are very useful for probing the film’s electronic properties over a wide potential -3/-4 range. A nearly reversible response is observed for Fe(CN)6 with a AEp of 63 mV and an ipoxfipred of 1.0. Fe(CN)(;'3/’4 is a surface-sensitive redox system at both glassy 63’66'216 The electrode reaction carbon and boron-doped microcrystalline diamond. kinetics for this couple are strongly influenced by the amount of exposed edge plane on sp2—bonded carbon, as well as the surface cleanliness.66 Granger et al. showed that surface carbon-oxygen functionalifies on microcrystalline diamond significantly influence AEp, with increasing oxygen content causing an increase in the peak potential 217 separation.187 Such an effect was also observed by Fujishima and coworkers. The effect of the surface oxygen is reversed if the film is rehydrogenated in a hydrogen plasma.187 The small AEp seen for the boron-doped nanocrystalline diamond film is indicative of a high level of surface cleanliness and low surface oxide coverage. In other words, this result suggests the diamond surface is largely hydrogenated.187 +2/+3 A reversible response is seen for Ru(NH3)5 with a AEp of 60 mV and an ipoxflpmd of 0.99. A nearly reversible response is also seen for erIs'ZI'3 with a AEp of 61 ox red ”9 mV and an ip of 0.98. The electrode reaction kinetics for both of these redox systems are relatively insensitive to the physicochemical properties of both spz-bonded carbon and spa-bonded diamond electrodess‘n’m'216 The kinetics are mainly influenced by the density of electronic states at the formal potential for each couple. in other words, the nanocrystalline diamond electrode possesses a sufficient density of electronic states at both -200 mV, a potential that is negative of the apparent flatband potential (ca. 500 mV vs. Ag/AgCI) for hydrogen-terminated, semiconducting (p-type) diamond, and 800 mV, 8 potential that is positive of the flatband potential.218'220 73 A reversible response is seen for MVIZ/+ with a AEp of 60 mV and an ipoxfipm’d of 1.1. Like Ru(NH3)6+2/+i3 and erls'Zl'3, the electrode reaction kinetics for MV‘LZI+ are relatively insensitive to the physicochemical properties of both spa-bonded carbon and 63,136,216 diamond electrodes. The kinetics are mainly influenced by the density of electronic states at the formal potentials. in fact, good responsiveness was also observed for the MVW0 redox couple at an even more negative potential of -950 mV. The cathodic peak at -1025 mV is associated with the reduction of MV” to MVO. The peak shape is consistent with a diffusion-limited reaction. However, the corresponding oxidation peak at -1010 mV for the MV0 to MV" transition does not have the shape expected for a diffusion-limited process but, rather, is sharp and narrow, consistent with an oxidative desorption event. MV0 has limited solubility In aqueous media and, depending on the solution conditions, methyl viologen concentration, and the electrode surface properties, MV0 can adsorb and accumulate on the electrode surface.68 This is the case at present. The sharp oxidation peak results from the oxidative desorption of surface confined MVO. The cyclic voltammetric i-E curves for 4-tert-butylcatechol (t-BC) and Few+3 have much larger AEp’s and more asymmetric peak shapes than do the curves for the other redox systems. AEp's of 419 and 679 mV are observed for t-BC and Fem”, ox red +2/+3 ”9 respectively. The ip ratios are 1.8 and 0.91 for t-BC and Fe . The larger peak separations, as compared to the other four redox analytes, are due to more sluggish 63’216 The catechol redox reaction involves the transfer of electrode reaction kinetics. both electrons and protons, and the electrode kinetics are highly sensitive to the surface microstructure, the presence of surface carbon-oxygen functionalities, and the surface cleanliness of sp2-bonded carbon electrodes.66 Surface adsorption catalyzes the reaction.221 It is supposed that the slow kinetics for diamond result from a lack of 74 adsorption on the spa-bonded, hydrogen-terminated surface. Recent results have demonstrated a clear correlation between the fraction of exposed spz-bonded carbon in microcrystalline and nanocrystalline diamond films, and the surface coverage of adsorbed catecholsfg22 The greater the coverage, the smaller the AEp. The electrode kinetics for Few+2 are strongly influenced by the presence of surface carbon-oxygen functional groups, specifically carbonyl groups, which catalyze the reaction at sp2-bonded electrodes.223 The hydrogen-terminated diamond surface is void of such functionalifies, so this is the postulated reason for the sluggish kinetics. Increasing the surface oxygen content on diamond has been observed to slightly decrease the AE,,.217 Much work remains to understand why the kinetics of this analyte are so sluggish at diamond. 300 A fl tert-butyl catechol 3 250 L Fe(CN).,“”" E ’ 2/+1 O MV+ i 11. 1 5 ’ A Fe+21+3 {3‘ 150 - 0 Ru(NH3)5"3’*2 i: ’ o irc11'2"3 .3 100 1 g 1 N O 50 > 0 - L A - A A A A 0 5 10 15 20 25 Sean 11111111"2 ((mV/s)m) Figure 4.4 Plot of ipox vs scan rate“2 for Fe(CN)6'3/'4, Ru(NH3)6+3/+2, ”Clea/'3, methyl viologen (MV+2/+), 4-tert-butylcatechol, and Few“:3 at a boron-doped nanocrystalline diamond film. The scan rate was varied from 100 to 500 mV/s. The diamond deposition conditions are the same as stated in Figure 4.1. 75 Figure 4.4 presents plots of ipox versus v1,2 for the different redox analytes. It can be seen that the oxidation peak current for each analyte varies linearly with the scan rate"2 with a near-zero y-axis intercept, indicative of reactions limited by semi-infinite linear diffusion of reactant to the electrode surface. The quasi-reversible to reversible voltammetry for all the couples indicates that the boron-doped nanocrystalline diamond has a sufficient density of electronic states over a wide potential range to support a rapid rates of electron transfer. The electrical conductivity of previously reported nanocrystalline diamond thin film electrodes, both as deposited and as deposited with incorporated nitrogen, largely results from the it bonding in the intercrystalline grain boundaries.153 If this it bonding is disrupted, then the localized density of electronic states is reduced, and the electrical conductivity decreases proportionally. Hence the electrochemical responsiveness of these films depends, to a great extent, on the chemical and electronic properties of the grain boundaries. The electrochemical responsiveness of the boron-doped nanocrystalline thin films should be much less influenced by the chemical and electronic properties of the grain boundaries. In other words, the diamond grains themselves should be highly conducting due to the carriers provided by the substitutionally-inserted boron dopant atoms. To test this, an acid washing and hydrogen plasma treatment was 63,187 The applied, which is very effective at oxidatively removing sp2-bonded carbon. first step involved immersing the films in warm aqua regia for 30 min. The films were then rinsed with ultrapure water. The second step involved exposing the samples to a warm solution of HgOg (30%) for 30 min. This was followed by rinsing with ultrapure water and drying. The films were then hydrogen plasma treated (microwave-assisted) for 6 h to remove the surface oxides formed during the acid washing and to terminate the 63,1 87 surface with hydrogen. The hydrogen was introduced into the reactor at a flow rate 76 of 200 sccm. The plasma power, system pressure, and temperature were 1 kW, 35 torr, and 800 °C, respectively. The electrochemical response of two nanocrystalline films was evaluated before and after the acid washing/hydrogen plasma treatment. One was a nanocrystalline film deposited in the presence of the boron dopant source, and the other was a nanocrystalline film deposited without any intentionally added boron. Figure 4.5 shows a ”'3’“ in 1 M KCI at both films. cyclic voltammetric i-E curves for Ru(NH3)5 The response for the boron-doped nanocrystalline film is unaffected by the acid washing and hydrogen plasma treatment. A nearly reversible response is observed for Fe(CN)(5'3/'4 with a AEp of 78 mV before and 72 mV after treatment. A nearly reversible +3/+2’ I rC'6-2l-3, and MV+2I+1 response was also observed for Ru(NH3)6 with AEp’s of 74, 71, and 59 mV, respectively, after treatment. The response of the nanocrystalline film deposited without intentionally added boron was altered by the treatment. The AEp for 314 Ru(NH3)5+3/+2 increases to 165 mV after treatment. The AEp’s for Fe(CN)6 , erl3'2" 3, and Mv+2l+1 all increased marginally after treatment to 166, 83, and 99 mV, respectiVely. The AEp increases because of a decrease in the density of electronic states due to the oxidative etching and disruption of the n-bonded carbon in the grain boundary. The AEp increases for this particular nanocrystalline film were not as dramatic as we have seen in tests of other films. The reason is that this film, even though deposited with no intentionally added boron, was doped with some boron because it was prepared in a reactor that is regularly used for depositing boron-doped films. In fact, recent boron nuclear reaction analysis measurements of an “undoped” film, deposited in the same reactor, revealed a doping level of ca. 50 ppm BIC. Previous tests have shown that the electrochemical response of a truly undoped nanocrystalline film can be almost completely inhibited after the same chemicanlasma treatment. These results suggest 77 that the electronic properties of the boron-doped films are dominated by the carriers (holes) concentration in the diamond lattice, rather than by the physicochemical properties of the grain boundaries. 2 80 80 a L A i b ' B c a 40 . 40 - d E . 8 :1 0 i 0 i s 1 . ,3 -40 - .40 . \ g 1 1 -801111.11M111111-801111111111111 -700 -400 -100 200 -700 -400 -100 200 Potential (mV vs Ag/AgCl) Figure 4.5 Cyclic voltammetric i-E curves for 1 mM Ru(NH3)4""’+2 in 1 M KCI for an (a) untreated CH4/H2/Ar (1 %/5%/94%) with 10 ppm BgHs nanocrystalline diamond film, (b) the same electrode after treatment, (c) untreated CHJHg/Ar (1%/5%/94%) with 0 ppm BzHe nanocrystalline diamond film, and (d) the same electrode after treatment. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. 4.2.2. Material Characterization Figure 4.6 A and B shows a cross-section and top-view SEM image of a boron- doped nanocrystalline diamond thin-film electrode deposited for 2 h. The film was deposited from a CH4/Hg/Ar (1%/5%/94%) source gas mixture with 10 ppm of added Bsz. The smooth and uniformly coated film is composed of nodular features ~100 nm in diameter. No voids were found in the coating, and the thickness appeared uniform over the entire substrate. 78 Still nm Figure 4.6 SEM images of a boron-doped nanocrystalline diamond film deposited on Si in (A) cross-section and (B) top-view image modes. The deposition conditions are the same as stated in Figure 4.1. AFM images were also acquired to probe the film morphology. Figure 4.7 shows an AFM image for nanocrystalline diamond thin-film electrode. The root-mean-square surface roughness was found to be 34 nm over a 100 um2 area and was independent of the growth time (i.e., film thickness). The nodular features are actually clusters of 79 individual diamond grains formed as a result of the high nucleation rate, as evidenced by TEM images presented below. 10 pm 10 pm 14 nm 2 4 10 pm 6 a ‘ 1.. Figure 4.7 Atomic force microscope images (10x10um) of a boron-doped, nanocrystalline diamond film deposited Si in the (A) deflection and (B) height image modes. The deposition conditions are the same as stated in Figure 4.1. 80 The individual grain size within the diamond clusters was revealed by TEM. Figure 4.8 shows a TEM image of the film. This image is representative of several that were collected. It can be seen that the film consists of diamond grains with diameters from 10 to 15 nm. These diamond grains make up the nodular features seen in Figure 4.6. Moreover, each diamond grain is a single crystal and randomly oriented in the film. This conclusion is based on the appearance of the lattice fringe patterns and their orientational differences from grain to grain. Figure 4.8 Transmission electron micrograph of a nanocrystalline boron-doped diamond film. The deposition conditions are the same as stated in Figure 4.1. A rough estimate of the fringe spacing was made, even though this is not a high resolution image. The estimated spacing on at least a couple of the grains is ~ 0.20 nm. This is close to the 0.206 nm interplanar distance for diamond {111} planes. The transmission electron diffraction (TED) pattern from one of the grains, in Figure 4.9, 81 shows an intense ring pattern. This ring pattern also suggests atomically ordered, but randomly oriented grains. The diffraction pattern is indexed to the (111), (022), (113), (004), (133), (224), (115), and (333) planes of cubic diamond (ASTM 6-0675). Figure 4.9 Transmission electron diffraction pattern for a grain in the boron-doped nanocrystalline diamond thin film. The deposition conditions are the same as stated in Figure 4.1 Figure 4.10 shows an XRD spectrum of the film. Three broad reflections are observed at 43.8, 75.5, and 915°. These reflections are assigned to the (111), (220), and (311) planes of the cubic diamond, respectively. The peaks are broader than they are for a microcrystalline film of the same approximate thickness, due to smaller diamond grain size in the nanocrystalline film. Table 4.2 summarizes the calculated lattice spacings and the relative peak intensities. Reference data for cubic diamond powder (ASTM 6-0675) are also presented, for comparison. The diffraction data reveal that the bulk structure of the film is sp3-bonded diamond. 82 L (111) Intensity (a.u.) 30 40 50 60 70 80 90 100 2-Theta (degrees) Figure 4.10 XRD pattern for a boron-doped nanocrystalline diamond film deposited on Si. The deposition conditions are the same as stated in Figure 4.1 Table 4.2 Lattice Spacings and Relative Peak Intensities Obtained from XRD Patterns of a Nanocrystalline Diamond Film and 8 Cubic Diamond Standard Measured ASTM (6-0675) Plane Spacing (A) Intensity Spacing (A) Intensity (111) 2.06 100 2.06 100 (220) 1 .261 23 1 .261 25 (311) 1.0747 13 1.0754 16 ASTM (6-0675) - American Society for Testing Material-reference data for cubic diamond powder. Figure 4.11 shows a visible Raman spectrum of the film. This spectrum is characteristic of high-quality nanocrystalline diamond films. A single, sharp peak at 1332 cm1 in the Raman spectrum is frequently used as a signature of high quality, 83 215,224 single crystal, or large-grained diamond. The spectrum for the nanocrystalline films is quite different from this. L 1333 cm -1 _ _ l 1470 cm _ ; ‘1550 cm :3 .. 3 . 1150 cm -1 l E 8 E. 1 .1.I.L.1;L.1 800 1000 1200 1400 1600 1800 2000 Raman Shift (cm'l) Figure 4.11 Visible Raman spectrum for a boron-doped nanocrystalline diamond film deposited on Si. The deposition conditions are the same as stated in Figure 4.1 Broad peaks are seen at 1150, 1333, 1470, and 1550 cm". The peak at 1333 cm", on top of a large background signal, is associated with the first-order phonon mode of spa- bonded diamond. Due to resonance effects, the Raman cross-section scattering coefficient (visible excitation) for spZ-bonded carbon is larger (50x) than that for the spa- bonded carbon, and the scattering intensity for the former can often greatly exceed that for the latter.225 The peak width (FWHM) is much broader than that for microcrystalline diamond, 140 cm'1 vs 10 cm’1. This is due to the small grain size of the nanocrystalline film. Raman line broadening can result from several mechanisms. One possibility is the 84 well established confinement model.226 This model states that the smaller the domain size, the larger the range of phonons modes (with different q vector and different energy) that are allowed to participate in the Raman process. Hence, the line width results from the spread in phonon energy. Another, and more likely, explanation for the line broadening is phonon scattering at impurities and defects (i.e., grain boundaries).226 The scattering event shortens the lifetime of the phonons and thus broadens the Raman line. The peak at 1150 cm'1 is often used as a signature for high quality nanocrystalline diamond.224 Prawer and coworkers, through the study of clean, nanocrystalline diamond particles (~5 nm diameter), have attributed this peak to a surface phonon mode of diamond.227 On the other hand, Ferrari and Robertson have made arguments for this peak being associated with spz-bonded carbon, specifically transpolyactelylene segments at grain boundaries.228 Their assignment of spz-bonded carbon, rather than spa-bonded carbon as has often been proposed,228 is based on the observations that the peak position changes with excitation energy, the peak intensity decreases with increasing excitation energy, and the peak is always accompanied by another peak at ~1450 cm“, which behaves similarly with excitation energy. We, therefore, tentatively assign the peaks at 1470 and 1550 cm'1 to disordered spz-bonded carbon in the grain boundaries. The 1550 cm'1 peak is down-shifted from the expected 1580 cm'1 position for graphite. This shift results because the spz-bonded carbon is amorphous and is mixed with sp3-bonded carbon. It is important to note that, very likely, the spz-bonded carbon is confined to the grain boundaries of the nanocrystalline film and yields a network of three- and four-fold coordinated carbon atoms.144 Additional research is needed to confirm the origins of the 1150 and 1470 cm'1 bands. The boron concentration in the film deposited with 10 ppm of added Bsz was determined to be 810 ppm (1.43x1020 B/cma), by boron nuclear reaction analysis. One 85 preliminary Hall Effect measurement showed the major carrier to be the hole and the carrier concentration and the conductivity to be 6.4x1017/cm3 and 10 O'1cm'1, respectively. The carrier mobility was found to be 90.4 cmzN-s. More measurements are needed before these values can be established as being representation of the material. 4.3. Conclusions A boron-doped nanocrystalline diamond thin film was deposited by CVD from a CH4/H2/Ar source gas mixture. Boron doping was accomplished using 8sz added to the source gas mixture. The films exhibited a wide working potential window, a low voltammetric background current and good responsiveness for Fe(CN)6‘3"‘, Ru(NH3)6*3’*2, erle‘m, and methyl viologen without any pretreatment. The nearly reversible voltammetry for all four redox systems indicates that the boron—doped nanocrystalline diamond has a sufficient density of electronic states over a wide potential range. More sluggish kinetics were found for 4-tert-butylcatechol and Few“. The sluggish kinetics are attributed to weak surface adsorption for the former, and to the absence of catalyzing surface carbonyl groups for the latter. XRD revealed that the bulk crystal structure of the film was cubic diamond. TEM indicated the film consists of 10-15 nm randomly oriented but atomically ordered diamond grains. SEM showed these grains form aggregates ~100 nm in size. Unlike previously reported nanocrystalline diamond films that have electrical properties dominated by the high fraction of spZ-bonded grain boundaries, the present film is doped with boron and the electrical properties are dominated by the charge carriers in the diamond (i.e., substitutionally insertedboron dopant atoms).. 86 Chapter 5 CY CLIC VOLTAMMETRIC STUDIES OF THE ELECTROOXIDATION OF ALIPHATIC POLYAMINE AT BORON-DOPED DIAMOND THIN FILMS 5.1. Introduction Aliphatic alcohols and amines are usually observed to be electroinactive at most electrodes, at constant applied potential in aqueous media, and have been classified as such historically.229 Despite favorable thermodynamics, the oxidation of aliphatic amines, for example, in aqueous media is kinetically limited because the reactions require transfer of oxygen from H20 to the oxidation product(s), rather than only simple removal of an electron from the “lone pair" on the nitrogen atom. Conventional anode materials (e.g., Pt, Au, and C) lack the ability to support these O-transfer 54,55,230 mechanisms at constant applied potentials. A significant problem is fouling of the electrode by reaction intermediates and products. Attempts have been made to utilize metal oxides (i.e., copper oxide, nickel oxide, gold oxide, platinum oxides) or 35,231 composites of metal oxides as anode materials. For instance, anodized nickel 87 electrodes have been applied for amine oxidation. The anodic response has been proposed to involve a surface-mediated mechanism involving Ni(l|l) formed in the oxide film at positive potentials.231 A disadvantage of using the anodized nickel, however, is the apparent need for the addition of (NiOH")4 in the solution phase to minimize dissolution of the soluble metal oxide.231 Amine oxidation in alkaline solution at glassy carbon has been reported by Masui et. al.232 it was observed, however, that only secondary and tertiary amines could be oxidized stably. Recently, successful anodic oxidation of aliphatic polyamines was demonstrated at glassy carbon electrode modified with nickel particles.34 Silver electrodes have been used for the oxidation of primary and secondary 2 - - - ' ' 33235 in alkaline media, the Silver surface lS amines as reported by Hampson et. al. oxidized to A920. Very soluble complexes between the amine and surface oxides form, causing electrode dissolution and as result, very slow kinetics for amine oxidation. However, at high positive potentials (20.6 V vs. NHE), where formation of A90 occurs, amines are rapidly oxidized in alkaline solution (pH12). Amines can be detected at Au in alkaline media. However, the anodic response is quickly attenuated due to the adsorption of reaction intermediates and products. The electrodes can be restored to full activity by a series of positive and negative potential pulses, so called pulsed amperometric detection (PAD). The procedure accomplishes oxidation of amine, desorption of adsorbed carbonaceous species with regeneration of a 1_ . . e . 3 33 The success of Au for amine ox1dation 1s a result of clean and oxide-free surface. electrocatalytic mechanisms believed to involve the anodic discharge of H20 with generation of adsorbed hydroxyl radicals (OHo) as an intermediate product in the formation of surface oxide (AuO). The catalytically active gold surface is short-lived (ca. 1 s), and therefore, continuous detection is impossible. Only with the application of 88 multistep potential-time waveforms, which alternate the processes of electrocatalytic detection with subsequent cathodic reduction of surface oxide (i.e., regeneration of the gold surface), can the reproducible anodic oxidation can be accomplished.31'32'56 41 ,54,55,236 that 0H0 It was a premise of the research by Johnson and coworkers is generated at the electrode surface through the anodic discharge of water. They proposed that OH- is adsorbed at specific surface sites, and transferred to the reactants via an electrocatalytic O-transfer reaction mechanism. O-transfer reactions can be catalyzed by increasing the surface concentration of adsorption sites, thereby, increasing the concentration of adsorbed OH-. The rate of 02 evolution is also enhanced by increasing the coverage of OH- . It was proposed that there is an optimal low density of surface sites for which the desired O-transfer process can occur at near mass transport-limited rates without an unduly high background signal for 02 evolution.29'30’32’46’47'237 The desire to achieve a significant improvement in electrooxidation of aliphatic amines has stimulated a search for new anode materials that retain their catalytic activity. Electrode characteristics suggested by researchers for these composite anodes include: (i) use of metal oxides with minimal solubility at the pH conditions of the application, (ii) existence of a low density of surface sites at which the anodic discharge of H20 (evolution of 02) occurs with lower potential than the surrounding matrix, and (iii) existence of surface sites that are effective for coordinating the analytes.46’47'54 Adsorption/coordination of the analyte is believed to be important because of (i) a suspected need to “desolvate the electronic orbital”, to which oxygen from adsorbed OH- is transferred, (i.e., oxygen transfer cannot occur by a tunneling mechanism), and (ii) a need for longer residence time of analyte at the electrode surface in order to .47 observe O-transfer. 46 89 Yeo et al., doped PbOz film electrodes with Bi(V) and observed significant 49,238-241 activation for several O-transfer reactions. Johnson et al. proposed silver as a dopant for PbOz films for amine oxidation, as it is known that Ag(l) forms complexes with . . . , .57 amines In aqueous media.54 55 The Ag-doped PbOz composite electrodes showed some activity for amine oxidation, however, at high analyte concentrations, some dissolution of the film occurred due to the high solubility of the amine-Ag(l) complex.55 Similar drawbacks were observed for pure silver electrodes.234’235 The partial success of electrodeposited Ag-PbOz electrodes led to examination of anodized Ag-PbOz alloy films.55 These electrodes exhibited superior response stability for aliphatic amines in comparison to electrodeposited Ag-PbOz films. This was attributed to the presence of only AgO, not A920. Comparable voltammetric responses were observed for primary, secondary, and tertiary aliphatic amines. The catalytic activity of the anodized eutetic- phase Ag-PbOz alloy electrodes for oxidation of amines was observed in alkaline solutions of borate (pH 11) and phosphate buffers (pH 12). No response was observed in neutral phosphate buffer (pH 7), presumably because of protonation of the amine functional group which blocks adsorption at the silver sites in the Ag-Pb oxide surface.55 Johnson et. al. proposed the following reaction model for analyte oxidation at composite electrodes via anodic oxygen transfer reaction:55 8[ ] + H20 —) S[OH] + H+ + e- (OH radical generation at A320) 8’ [ ] + R H S' [R] (coordination of the amine at AgzO surface sites) S [OH] + S’[R] —> S [ ] + S’ [ ] + R0 + H+ + e' (anodic oxygen transfer reaction) The authors indicated that silver, as a doping agent serves the following functions: (i) a low H20 discharge overpotential compared to PbOz. (ii) capability of adsorbing the 90 OH- generated during H20 discharge, and (iii) capability of adsorbing the reactant species.55 The reports from the Johnson group provided the impetus for our efforts to fabricate and test diamond electrodes designed with the appropriate properties to support anodic oxygen-transfer reactions.47’54'55’57'237'242 Applications of such electrocatalytic electrodes for electrosynthesis, as electroanalytical sensors, and in the anodic degradation of toxic chemical waste (“electrochemical incineration”) are possible. 32,33,38,42.50,243-255 Cyclic voltammetric studies of aliphatic polyamine oxidation at microcrystalline and nanocrystalline boron-doped diamond thin-film electrodes are reported herein. The goals of this research were to: (i) determine if the amines can be qualitatively and stably detected at boron-doped microcrystalline or nanocrystalline diamond thin-film electrodes, (ii) determine how the oxidation response varies with concentration, scan rate, and solution pH, (iii) determine the role of spz-bonded non-diamond carbon impurity on the oxidation response, and (iv) verify the role of surface boron sites on the oxidation response. 5.1.1. Model for Polyamine Oxidation at Polycrystalline Diamond The proposed mechanism for polyamine oxidation at diamond is given below and is similar to the model for anodic oxygen-transfer reactions put forth by Johnson and . . . 7.230 coworkers. 46 54 55 5 SNDI ] + HzO —-) SND[OH] -i-H+ +e' (OH radical generation at a sp2 sites) Sn [ ] + R (-—> 83 [R] (coordination of the amine at a boron surface site) Sm) [OH] + 53 [R] —9 SND [ l + $3 [ l + R0 + H+ + e' (oxygen-transfer reaction) 91 The subscripts ND and B refer to non—diamond and boron, respectively. Boron- doped diamond thin film electrodes possess several properties important for anodic oxygen-transfer reactions. First, high-quality diamond films are stable and resistant to corrosion in strongly acidic and alkaline media.“’109’256'258 and mfmmes them" Therefore, at the anodic potentials used to detect the polyamines, the electrode structure is stable. Second, films may contain spz-bonded non-diamond carbon impurity distributed very locally over the surface. These impurities can exist at the grain boundaries or as extended defects within the diamond film. These surface impurities, which exhibit lower overpotential for oxygen evolution than does diamond, can also be intentionally introduced into the films by adjustment of the diamond deposition . . ,1 3,259 conditions.63 6 This means that reactive OH- will be generated locally at these sites at low overpotential, and not to any appreciable extent on the diamond lattice. Third, boron dopant atoms located at the surface can serve as adsorption/coordination sites for the lone pair of electrons on the N atoms of the polyamines. Boron atoms can insert directly into the growing diamond lattice, but they can also cluster and accumulate in the grain boundaries. The distribution of boron atoms in the polycrystalline diamond film is not homogeneous as there is a growth sector dependence. For example, the boron dopant concentration in the (111) sector is a factor of 10 times higher than in (100) sector.26°'263 “d ”mm“ "‘°’°‘" The polyamine adsorption/coordination at the boron sites near the grain boundaries is important mechanistically, as these are sites very near where the OH - is being generated at lower overpotential. 5.2. Results and Discussion 5.2.1. Cyclic Voltammetry of Polyamines Figure 5.1 shows typical cyclic voltammetric i-E curves for 1.0 mM (A) cadaverine (CAD) (B) putrescine (PUT), (C) spermidine (SPMD), and (D) 0.8 mM spermine (SPM), 92 in borax buffer pH 11(BBpH11) at a microcrystalline film deposited from a 0.5% CH4/H2 ratio and 10 ppm Bsz. 200 200 .. A .- 3 150 _ 150 _ V ( l‘ 3:; 100 _ 100 _ t , . a U 50 _ 50 _ f) , 0 . . 0 I" 250 550 850 1150 250 200 200 a 150 - 150 _ E 100 _ 100 _. Q) g . . U 50 - 50 _ ’ l 0 ‘ 0 l ‘ 250 250 550 Potential (mV vs Ag/AgCl) 7 z 6 .’ O l 850 1150 Figure 5.1 Cyclic voltammetric i-E curves, background (dashed line) and total current (solid line), for 1.0 mM (A) CAD (B) PUT, (C) SPMD, and (D) SPM in 0.01 M borax buffer/0.1 M NaCl, pH 11. The working electrode was a microcrystalline diamond film deposited from a 0.5% CH4/H2 ratio and 10 ppm BZHG. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. 93 Similarly shaped curves were observed for several other amines and polyamines (e.g., methylamine, ethylamine, propylamine, ethylenediamine, 1,3-diaminopropane, 1,6- hexamethylenediamine, and N-(3—aminopropyl)-1,3-propanediamine). The oxidation peak current for each of the amines is clearly discernible from the background current. The Ep°" values are 872, 868, 876, and 872 mV, and the EW values are 821, 817, 820, and 815 mV for CAD, PUT, SPMD, SPM, respectively. The ip°" values are 75, 93, 116, and 76 uA for 1 mM CAD, PUT, SPMD, and 0.8 mM SPM, respectively. A summary of the cyclic voltammetric data for several amines and polyamines in BBpH11 are presented in Table 5.1. Table 5.1 Cyclic Voltammetric Data for 0.1 mM Aliphatic Amine Oxidation amine p” (mV) E 1,2(mV) wave shape methylamine 865 801 peak ethylamine 851 798 peak propylamine 849 789 peak ethylenediamine 848 780 peak 1,3-diamlnopropane 860 797 peak putrescine 851 770 peak cadaverine 860 783 peak spermidine 867 762 peak/wave spermine 869 769 peak/wave Note: Voltammetric data are for a 0.1 V/s scan rate. Reported potentials are versus a Ag/AgCl reference electrode. Analytes were prepared in 0.01 M borax buffer/0.1 M NaCI, DH 11. Em values correspond to the potential at half the peak current. 94 Whether a well defined peak shape is observed depends on the physicochemical properties of the electrode. The Ep°" values are nearly the same, independent of the amine molecular structure, indicating the importance of OH- generation. The oxidation reaction all occur at potential near the onset of oxygen evolution. Very interestingly, there is an oxidation peak present during the reverse sweep, which is clearly associated with the oxidation of amine. The current goes through the maximum at about the same potential as on the forward sweep. At this potential, consumption of OH- by the oxygen evolution reaction is minimal, and these reactants are available for the oxygen transfer reaction. The current during reverse sweep is lower than that during the forward scan. This is partially due to the depletion of the amine during forward scan, and very likely at this scan rate there is not enough time for the new amine molecules to diffuse from the bulk solution and coordinate at the boron sites on the electrode surface. Hence, smaller number of molecules are available for oxidation and a lower current results. The lower current during the reverse sweep may also be limited by the availability of surface boron sites for adsorption. Some time is necessary for desorption of the oxidation products to free boron sites for amine coordination. Cyclic voltammograms obtained at lower scan rates (5-10 mV/s) oonfirrn this supposition as it was observed that the current during reverse sweep tracked that obtained during forward sweep. The hysteresis depends on the scan rate with a larger hysteresis observed at higher scan rates. Repetitive potential cycling in the amine solution led to progressive attenuation of the current. The voltammetric response could, however, be regenerated by vigorous mixing and/or inserting a time delay interval from 3 to 5 minutes between the cycles. This behavior indicates there is no permanent fouling of the electrode by the oxidation Products, as in the case for other electrodes. The current recovery is consistent with 95 slow desorption of the product(s) from the surface. The absence of fouling and response recovery is demonstrated in Figure 5.2. 70 60'. so; A a L H l' a , O . ‘6: 30. :1 . U l 20- f 10'. F 700 ‘ 800‘ ‘ A900 A ‘10004‘ 1100 Potential (mV vs Ag/AgCl) Figure 5.2 Cyclic voltammetric i-E curves for 1.0 mM CAD in 0.01M borax buffer/0.1 M NaCl, pH 11.2, at a nanocrystalline boron-doped diamond thin film. The nanocrystalline diamond electrode was deposited from a 0.5% CH4/H2/Ar ratio and 1 ppm BZHG. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. Figure 5.2 shows cyclic voltammetric i-E curves for 1.0 mM CAD in BBpH11 at the 0.5% CH4/H2/Ar boron-doped nanocrystalline diamond film. There is a decrease in the peak current during each of the first six scans. This decrease is due to the formation of a depletion layer at the interface, as the oxidation reaction is irreversible, but also to slow desorption of the product(s) from the electrode surface. If a time delay of 3 min is applied between scans, with the potential poised at 0 V, then the peak current for the subsequent scan (cycle 7) is nearly the same as the initial scan. This indicates that the 96 electrode is not irreversibly fouled by reaction products, as is the case for other metallic electrodes, necessitating the need for pulsed waveforms or electrode pretreatment. 5.2.2. Effect of the Electrolyte on the Voltammetric Response The influence of different supporting electrolytes on the voltammetric response for the polyamines was investigated. CAD was not detected in pH 7.2 phosphate buffer, very likely due to the protonation of the amine group. This probably could inhibits the coordination at surface boron sites, and makes the oxidation reaction more kinetically difficult. A signal was observed for amines with lower dissociation constants (e.g., EDA, SPM, SPMD). At pH 10.6, the voltammetric ressponse varied little with the supporting electrolyte. In general, for a given pH, the Ep°" values for each amine were independent of the supporting electrolyte. The highest oxidation currents were observed for carbonate and borax buffers m supporting electrolytes present. In pH 9 borax buffer, with CuClz as the supporting electrolyte, it was observed that the amine oxidation current was attenuated compared with the response in borax buffer, pH 9 with NaCl as a supporting electrolyte. The amine oxidation current decreased proportionally as the CuClz concentration increased because of Cu+2 and amino group complexation. This complexation prohibits the amine from coordinating at surface boron sites. It is expected that other cations (e.g., Mg*2, Ca“) might also attenuate the oxidation current. This observation provides indirect evidence for the importance of amine adsorption in the oxidation reaction mechanism. 5.2.3. Effect of pH on the Voltammetric Response The effect of pH on the voltammetric response for polyamines was tested in borax, phosphate, and carbonate buffers. Figure 5.3 shows cyclic voltammetric i-E curves for (A) the background response and (B) 1.0 mM CAD as a function of the electrolyte pH. The ionic strength of the electrolytes was maintained constant at 0.4M. A 97 negative shift in the onset potential for the background current (i.e., oxygen evolution) is observed with increasing pH. 100 "f 80 60 L 40 L c 20’. b O 100 400 700 1000 1300 Current (uA) '9‘ OO O v T'ir ff ' ON C p I p 40- h p 20’. Potential (mV vs Ag/AgCl) Figure 5.3 Cyclic voltammetric i-E curves for 1 mM cadaverine in (a) 0.01 M phosphate buffer, pH 7.2; (b) 0.01 M carbonate buffer/0.1 M NaClO4, pH 9,; (c) 0.01 M carbonate buffer/0.1 M NaClO4, pH 10, and (d) 0.01 M carbonate buffer/0.1 M NaClO4, pH 11 at microcrystalline diamond film deposited from a 0.5% CHJHZ ratio and 1 ppm BzHe. Scan rate = 0.1 V/s. Electrode area = 0.2 cm2. 98 The onset potential shifts by ~60 mV/pH unit. The Ep°" and Em values for all the polyamines also shifted negatively with increasing pH by ~60 mV/pH unit. These results confirm that the amine oxidation potential is dependent upon the potential at which OH- is generated. As the pH of the electrolyte increased, ip°" and Opox for all the polyamines increased, as well. Table 5.2 summarizes the 0,,” data obtained at a microcrystalline diamond thin-film electrode deposited from 0.5 % CH4/H2 ratio. The charge was measured after background subtraction. Table 5.2 Cyclic Voltammetric Charge Data for 1 mM Amines in 0.01 M Borax Buffer/0.1 M NaCIO. Charge (11C) for the forward scan 0.01 M 3"“ Bum” CAD PUT SPMD SPM pH 9 120 103 --- 245 pH 10 192 163 273 373 pH 11 226 200 295 470 Note: Voltammetric data are for a 0.1 V/s scan rate. At pH 7.2, there was no well resolved oxidation wave for either CAD and PUT. The lowest pKa values for CAD and PUT are 9.13 and 9.35, respectively, and at pH below these values, most of the molecules are protonated. Therefore, they are not available for coordination and oxidation. The oxidation current for SPMD, SPM, and EDA was 2-3 times higher, compared with CAD or PUT, at pH 8.4. The me for SPMD, SPM, and EDA are 8.0, 7.9, and 6.8, respectively.189 They are lower than the dissociation constant of CAD and PUT, and so the currents observed for SPMD, SPM and EDA are higher. 99 At pH 11, almost 97% of molecules have multiple amine groups unprotonated; therefore, the larger oxidation currents for each of the amine are observed at high pH. Since, the amine oxidation mechanism is thought to involve adsorption at surface boron sites via the non-bonding electron pair on the N atom, the higher oxidation peak current at higher pH is due to the limited protonation. Figure 5.4 shows a diagram of the fractional composition of the protonated and unprotonated forms of (A) CAD and (B) SPM as a function of pH. Fractional Composition ( % ) 100 : C]. /"¢ 80 . A // ’ \. A ’ t D / ,2 T “. ." \\ y} l , \ / > ' \\ l/ 40 I- | IL... . A/ /Q ‘A\ 20 C- 'I/ Q’/ [\\ oi .2213 mm---» 3 8 9 10 l l 12 pH 100 $ 80’. B / [:1 *HaN-R-NH; A HgN-R-NH3" i? HzN-R-NHZ [J *HaN-R-NHz-R-NHz-NHa“ A *l-laN-R-NH-R-NHz-NH; O *HgN-R-NH-R-NH-NHa“ O HaN-R-NH-R-NH-NH3* if HzN-R-NH-R-NH-NHZ Figure 5.4 The fractional composition of protonate/unprotonated forms of (A) CAD and (B) SPM as a function of pH. Polyamines, as polyprotic bases, can accept more than one proton. Depending Upon the pH, amines are present at a different molar ratios of the protonated and 100 unprotonated forms. For pH < pKa1, the fully protonated species is the predominant form (11,/1r". For pKa1< pH < pKaz, the form (H,-,A)+(""> is favored.189 At pH values greater than the pKag, the fully basic, unprotonated form (A) is dominant. The important point is that the pH of the electrolyte solution is'critical for this reaction. The electrode response for the amine oxidation strongly depends on the molar ratio of the protonated and unprotonated forms. The maximum oxidation current is seen for the fully unprotonated form. 5.2.4. Effect of Scan Rate on the Voltammetric Response The effect of scan rate on the cyclic voltammetric response for CAD, PUT, SPMD, and SPM in CBpH10.6, at a 0.5% CH4/H2 microcrystalline diamond film, is shown in Figure 5.5 A and B. The scan rate was varied from 10 to 400 mV/s. Examination of the data, using several heavily doped diamond films, revealed that the i,” varied linearly with scan rate‘”, with non-zero y-intercept. It was observed that the ip°" for moderately doped diamond electrodes was nearly independent of scan rate, particularly, above 0.1V/s. In general, more linear behavior was observed for the lower, ~0.1 mM, rather than the higher, 1.0 mM, amine concentrations. This is illustrated in Figure 5.5 A and B. For the higher amine concentrations, the reaction becomes surface site limited and, as a result, ip°" is nearly independent of scan rate. The unchanging peak current at higher scan rates is also attributable, at least in part, to slow desorption kinetics of the reaction product(s) from the surface sites. Therefore, as the scan rate is increased, the current becomes limited by the availability of open coordination sites. The oxidation peak potential, Ep°", shifts positively with increasing scan rate. This is illustrated in Figure 5.50. Johnson et al. attributed a similar positive shift in E," at 101 composite electrodes to an increased flux of the reactants (i.e., amine).46 In order for the amine molecules to be oxidized via anodic oxygen transfer sites are required for the production of OH-. A higher amine flux will require a larger number of OH-, and higher concentration of these will be available at more positive potentials.46 25 70 ’ . i .21" 50 . A 20; A : B ,3 ...... ’ 50 L i l ~ .5 : 15 . ,,..---;::::::..6' 40 ; I3,121 1:: : 1A""ZZI§"": ---------- g ' » Q 1 .- -: -: 2.... ................ : I ”9.- 1: 10 @639 ,,,,, a ------- 3° : ”19! __9 --------- ‘3 ’ arr-‘3" , .53 ------ U pea-'8’ 20 L Goggs 5 L C 10 C o ' ......................... o I .......................... 0.00 0.05 0.10 0.15 0.20 0.25 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Scan Rate"2 (mV/s)"2 Scan Rate”2 (mV/s)”2 885 . C ”.43 .:.‘::"...A g 875 . Aflyfl/fi‘ 5 9.325 {2 CAD is e ...... a ''''''''' 9 o PUT 1: f; 9,2; SPMD g h 6 91> D SPM 9.. 855 . at; A 345 ..................... 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Scan Rate (mV/s) Figure 5.5 Plots of the cyclic voltammetric oxidation peak current, ip°" versus scan rate"2 for (A) 0.1 mM amines at moderately doped microcrystalline film, (B) 1 mM amines at highly doped microcrystalline film in CBpH10.6. (C) Plots of the oxidation peak potential, ,3", versus scan rate for 0.1 mM amines in 0.01 M carbonate buffer, pH10.6. 102 5.2.5 Effect of Amine Concentration on the Voltammetric Response The linear sweep voltammetric oxidation peak current was measured as a function of the amine concentration. Figure 5.6 shows cyclic voltammetric i-E curves for EDA in 0.01 M carbonate buffer, pH 10. The concentrations ranged from 0.10 to 5 mM. 80; g A 70.~ r <1 : r? 60? 1:1 a I § 50? c 9 L g 40; b 'fi L 8 30': 5‘ L O 20 , a 10; __.._.—.——_———__—— A A n l A L AAAAAAAAAA ‘500 600 700 800 900 1000 1100 1200 Potential (mV vs Ag/AgCl) Figure 5.6 Linear sweep voltammetric i-E curves for ethylenediamine at concentrations of (a) 0.1, (b) 0.5, (c) 1, (d) 2, (e) 3, (f) 4, and (g) 5 mM at microcrystalline diamond film deposited from a 0.5% CHJHZ ratio. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. Increasing the amine concentration caused E,” to shift toward more positive values, similar to the scan rate dependence. This is a consequence of the increased number of amine molecules available for oxidation, requiring a higher concentration of OHo. This is achieved at more positive potentials. A plot of ip°" vs concentration was linear up to a 1.0 mM concentration. The plot had non-zero intercept and different slopes (i.e. 103 sensitivities) for the different amines. Therefore, the analytical utility of the voltammetric and amperometric measurements would appear to be limited to concentrations below 1.0 mM. One possibility for the negative deviation from linearity at concentrations above 1mM could be a limited number of surface boron sites. Studies using films deposited with different boron concentrations support this supposition and will be presented further in this chapter. 5.2.6. Importance of N on-Diamond Carbon Impurities Figure 5.7 shows cyclic voltammetric i-E curves for 1.0 mM CAD in CBpH10 at a 0.67% CHJHz microcrystalline diamond film. Figure 5.7 A shows the curve for the “as- deposited” film in the presence of 1 mM CAD along with the background response (dashed line). The onset potential for oxygen evolution at the "as-deposited” film begins at about 1020 mV. The oxidation response for CAD is well-resolved from the background with an Ep°" of 990 mV, an Em, of 920 mV, and an ip°" of 32 11A for a 10 mV/s scan rate. If the film is acid washed (see Experimental Section) to remove the spz-bonded non-diamond carbon impurity from the surface and then hydrogen plasma hydrogenate the surface, then the voltammetric response in Figure 5.7 B is observed. Clearly, the removal of the non-diamond carbon impurity causes a total attenuation of the response.259 The lack of an oxidation response is due to the removal of the non- diamond carbon impurity at which the OH- forms with lower overpotential than the surrounding diamond. It is important to note that the acid washing and rehydrogenation leads to a higher overpotential for oxygen evolution with an onset potential of about 1230 mV. The 200 mV difference in potential between the “as deposited” and rehydrogenated electrodes is evidence for the successful removal of spz-bonded non-diamond carbon impurity. This result provides compelling evidence for the importance of localized spz- bonded non-diamond carbon impurity sites in the amine oxidation reaction mechanism. 104 50 Oxidation current (11A) 0 400 800 1200 1600 Potential (mV vs Ag/AgCl) F'QIire 5.7 Cyclic voltammetric i-E curves for 1 mM cadaverine in 0.01 M carbonate buffer/0.1 M NaC|04, pH10.6, at a microcrystalline diamond film deposited from a 0~67% CH4/H2 ratio. (A) “as deposited“, and (B) acid washed and rehydrogenated diamond film. Scan rate = 10 mV/s. 105 5.2.7. Effect of CH4/Hz Ratio on the Voltammetric Response Non-diamond spZ-carbon impurity can be introduced into the diamond by adjusting the CH4/H2 ratio used for film deposition. Therefore, it is possible to optimize the amine oxidation response through manipulation of this ratio. In general, as the CH4/H2 ratio is increased, the film quality decreases with the highest quality films being deposited at a 0.33% CH4/H2 ratio. Non-diamond spz-carbon impurity and morphological defects tend to increase with increasing CH4/H2 ratio. Figure 5.8 shows AFM images of diamond films deposited with three different CH4/H2 ratios. Figure 5.8 Contact-mode atomic force microscopy images of a microcrystalline diamond films deposited from a (A) 0.33%, (B) 0.67%, and (C) 1% CH4/H2 ratio. All images have the same x, y, z scales. 106 As the CH4/H2 ratio increases, the nominal crystallite size decreases and the extent of secondary nucleation increases. The decrease in crystallite size leads to an increase in the fraction of exposed grain boundary. It is at the grain boundary where non-diamond spz-carbon deposits form. 63 8"" '°'°'°"°°s "mm The higher CH4/H2 ratio causes a higher rate of nucleation leading to the secondary growths and a reduction in grain size. As shown in Figure 5.8, the film surface for the 0.33% CH4/H2 ratio consists of large 1-5 pm crystals. The 0.67% CHJHZ film has a large fraction of much smaller crystallites (0.1-0.5 um) and few large grains (1-2 pm). The surface of the film deposited from 1% CH4/H2 ratio is not well faceted. Small crystallites <0.1 um, are predominant. The small grain size results from the increased secondary nucleation rates. Figure 5.9 shows corresponding visible Raman spectra for the films. The‘laser spot size was ca. 10 pm, meaning that several grains and grain boundaries were probed in the measurement. An intense first-order diamond phonon line is observed at 1332 cm" for the 0.33% CH4/H2 film. The full width at half-maximum (FWHM) of the band is 8 cm". For comparison, the FWHM for a single-crystal diamond standard is 2-3 cm". The FWHM is a measure of the film quality and, to a first approximation, is inversely related to the phonon lifetime.207 In other words, grain boundary scatter decreases the phonon lifetime and increases the FWHM. There is minimal scattering intensity, observed between 1500 and 1600 cm", associated with spa-bonded carbon impurity. The cross-sectional scattering coefficients (514.4 nm) for diamond and graphite (i.e., non-diamond carbon) are 9 x 10'7 and 500 x 10'7 cm"/sr, respectively.2°7'225’264 Therefore, the technique is quite sensitive to the presence of non-diamond carbon impurity. 107 20000 15000 ._ 100001. Total counts 5000.. 0 1 1 1 l 1 l 1 l 1 l 1 L 1 200 400 600 800 1000 1200 1400 1600 Raman Shift (cm'l) Figure 5.9 Visible Raman spectra for diamond films deposited using a (a) 0.33%, (b) 0.67%, and (c) 1% CH4/H2 ratio. 514.4 nm laser source and 10 s integration time. Importantly, there is little photoluminescence background in the spectra for 0.33% and 0.67% CH4/H2 films, consistent with a low level of spz-bonded carbon impurity.63 and references therein Much larger photoluminescence background is observed in the spectrum for 1% CH4/H2 film. As the CH4/H2 ratio increases, the diamond phonon line intensity decreases, and the FWHM, the scattering intensity between 1500 to 1600 cm'1 and the photoluminescence background all increase, consistent with an increased level of defects, in particular, grain boundaries and non-diamond carbon impurity. The band at 520 cm'1 is attributed to scattering by the underlying Si substrate. The intensity of this band decreases with increasing CH4/H2 ratio because of the increasing film thickness 108 and higher optical density due to non-diamond carbon. The presence of the spZ-bonded non-diamond carbon impurity can also be detected using electrochemical measurements. The background cyclic voltammetric current and potential window in aqueous electrolytes are strongly affected by the presence of spa-bonded carbon.63 Background cyclic voltammetric i-E curves for the three different diamond films are shown in Figure 5.10 A-C. The background for each was recorded at different solution pH. For a given pH, the onset potential for oxygen evolution (i.e., OH- generation) decreases with increasing CHJHZ ratio. For example, at pH 11, the onset potential, measured for a current of 7 11A is 1000, 940, and 870 mV, respectively, for films deposited from the 0.33, 0.50, and 1.0% CH4/H2 ratio. The non-diamond spa-carbon impurity exhibits a lower overpotential for oxygen evolution than does the surrounding diamond. As the concentration of this impurity increased, the onset potential decreases. Presumably, the lower overpotential results from the ability of certain sites on the sp2 impurity to stabilize reaction intermediates. The onset potential also shifts negatively with increasing pH. The anodic current measured at 1050 mV (i.e., in the oxygen evolution regime) increases with increasing CH4/H2 ratio and pH. For instance, the current is 2.9, 7.3, 15.3, and 54.6 11A for the 0.33% CH4/H2 film at pH 9, 10, 11, and 12, respectively. Table 5.3 summarizes the voltammetric background current data. Table 5.3 Cycllc Voltammetric Background Current Data for Boron-Doped Microcrystalline Diamond Fllms Current (11A) at 1050 mV vs AgIAgCI Dlamond Film pH9 pH10 pH11 pH12 0.33% CHJH; 2.9 7.3 15.3 54.6 0.50% CHJH; 10.8 26.3 61 .5 168.4 1.0% CHJH; 22.1 42.4 91.2 188.9 Note: Voltammetric data are for a 0.1 V/s scan rate. 109 300 A 200 o 3 100 400 700 1000 1300 < 3: 200 - e '5'. t: 100 . d 5 c O b 0 L a 100 400 700 1000 1300 300 - C I 200 . e 100 . d c t 0 . a 100 400 700 1000 1300 Potential (mV vs Ag/AgCl) Figure 5.10 Background cyclic voltammetric i-E curves for a microcrystalline diamond films deposited using a (A) 0.33%, (B) 0.50%, and (C) 1% CH4/H2 ratio. The solution pH was (a) 8.4, (b) 9, (c) 10, (d) 10.6, (e) 11, and (f) 12. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2. 110 The amine oxidation current also shows a dependence on the CHJHZ ratio. Cyclic voltammetric i-E curves for 1.0 mM CAD in CBpH10.6, at microcrystalline diamond films deposited using three different CH4/H2 source gas ratios are shown in Figure5.11. 120 2 90 l' I")? :2; b 0"," 5 t ’ \ ,v' 3 60 . . .~/ c I, “ I, I, , g \ ‘ ‘s--o”’ I”, / N a ” '1’ I ‘5 30 - \x" /" / O I, ’I’ :~::" o 2/.' _ ,/ /" ’ \ ' " 0--- /1 600 700 800 900 1000 1100 1200 Potential (mV vs Ag/AgCl) Figure 5.11 Cyclic voltammetric i-E curves for 1 mM cadaverine in 0.01 M carbonate buffer/0.1 M NaCIO4, pH10.6, at diamond films deposited using CHJH; source gas ratios of (a) 0.33%, (b) 0.5%, and (c) 1%. Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. The film deposited from the 0.5% CH4/H2 ratio exhibits the largest and most well-defined oxidation wave, with a peak current of ca. 60 11A. The 0.33% film was next best in performance, in terms of the peak current magnitude, with a value of ca. 20-30 pA. The 1.0% film exhibits a quantifiable response, with a peak current of ca. 45 11A, but it is superimposed on a larger background current. The reproducibility of the response was 111 best for the 0.5% film and worst for the 1% films. There is clearly an optimum electrode surface structure and chemical composition required for the polyamine oxidation reaction. Such a structure is achieved with films deposited from a 0.4-0.5% CHJHz source gas ratio. Table 5.4 shows a summary of some of the cyclic voltammetric data. Table 5.4 Cycllc Voltammetric Peak Current and Potential Data for Amines Oxidation at Microcrystalline Diamond Films Deposited with Different CHJH, Ratios. 1,,“ (11A)! E,“ (mV vs AglAgCI) CHJ H, ratio CAD PUT SPMD SPM 0.3 % 32.1 910 34.3 915 30.6 924 31.7 933 . 0.4 °/e 63.0 937 59.4 947 56.4 951 50.7 956 0.5 °/e 58.8 942 45.2 929 48.0 959 53.2 962 0.67 % 35.9 935 ----- ---- ---- ---- 32.9 969 1 .0 % 44.9 957 44.2 955 43.3 951 36.4 965 Note: Voltammetric data obtained at a scan rate = 0.1 We Electrode area = 0.2 cm2 5.2.8. Effect of Boron on the Voltammetric Response An oxidation mechanism has been proposed whereby the amine functional group coordinates with a surface boron sites. This interaction is thought to take place near the grain boundaries where boron aggregates and spz-bonded carbon impurity is located. OH- is the primary oxidant with the surface boron interaction increasing the residence time of the amine at the electrode surface and weakening the C-N bond. The principal reaction is the transfer of 0 from adsorbed OH- species to the surface coordinated amine. The voltammetric results presented below support this supposition and provide compelling evidence for the importance of surface boron in the oxidation reaction mechanism. 112 Figure 5.12 shows cyclic voltammetric i-E curves for 1 mM CAD in 0.01 M borax buffer, pH 10.6, along with the corresponding background current at (A) a nanocrystalline diamond thin film deposited with intentionally added boron, and (B) a nanocrystalline film deposited without any intentionally added boron. 160 A 120 . 80. 40. 150 500 850 1200 160 120 . Oxidation current (11A) 80. 40. 150 500 850 1200 Potential (mV vs Ag/AgCl) Figure 5.12 Cyclic voltammetric i-E curves for 1 mM CAD, along with the corresponding background current (dashed line), in a 0.01 M borax buffer, pH 11.2, at (A) a boron-doped nanocrystalline diamond thin-film electrode and at (B) a nanocrystalline diamond thin-film electrode deposited without intentionally added boron. Scan rate = 0.1V/s. Electrode geometric area = 0.2 cm2 113 Figure 5.12 A shows a well defined oxidation wave, clearly distinguishable from the water electrolysis background current, for the boron-doped film. Ep°" is 927 mV and i,,°" is 136 11A. There is a second oxidation peak on the reverse sweep at about the same potential as on the forward sweep. This wave shape has been previously discussed and is consistent with a redox reaction that involves a competitive surface interaction.35'54’55 Correspondingly, there is a poorly defined oxidation wave for the unintentionally doped film. During the forward and reverse sweeps, the current is slightly increased above the background, but there is no well-defined oxidation curve. The absence of an oxidation response for CAD is not due to poor film electrical conductivity, as the AEps for Fe(CN)6'3’“‘, Ru(NH3)6*2’*3, and erl6'2"3 were 63, 68, 60 mV, and for methyl viologen, 4- tert-butyl catechol, and Fe""”*3 were 60, 764, 396 mV. 63 This result demonstrates the importance of surface boron in the oxidation reaction mechanism of aliphatic polyamines at diamond. 2" More detailed studies investigating the electrode activity for amine oxidation as a function of boron dopant concentration were performed for both nanocrystalline and microcrystalline films. The characterization of a boron-doped nanocrystalline diamond thin-film electrode, deposited from a 1/94/5 CHJAr/Hz (v/v) ratio with 10 ppm BzHe added for doping, was presented in Chapter 4. a set of nanocrystalline films were deposited with 1, 10, 20, and 30 ppm of added B2H6. This corresponds to a 200, 2000, 4000, 6000, and 10,000 ppm B/C ratio in the source gas mixture. The actual doping level was determined from BNR analysis measurements. Table 2.4 in Chapter 2 presents a summary of the doping levels and apparent resistivities for each film. The ultimate goal for the work was to study the effect of doping level on the polyamine oxidation response, but the first task was to perform a complete characterization of the films. 114 Figure 5.13 shows visible Raman spectra for the nanocrystalline films, deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio with 1, 10, 20, and 30 ppm BzHe added, respectively. Broad peaks are seen near 1150, 1225, 1333, 1470 and 1550 cm". - 1333 cm '1 . 1470 cm“ . 1550 cm"1 —L 1225 pm’ _ 1150 cm '1 Intensity (Arb. Units) I l I l I l L l l L 800 1000 1200 1400 1600 1800 2000 Raman Shift (cm'l) Figure 5.13 A-E Visible Raman spectra for boron-doped nanocrystalline diamond films deposited from a 1/94/5 CH4/Ar/H2 (v/v) source gas ratio with (A) 0, (B) 1, (C) 10, (D) 20, and (E) 30 ppm of added BgHe. 532 nm laser source and 10 3 integration time. 115 The peak near 1333 cm'1 is attributed to the first-order phonon mode of diamond. The FWHM is much broader than that for a microcrystalline diamond film, due to the smaller grain size and higher grain boundary or defect density of the nanocrystalline film. All of the presented modes (i.e., 1150, 1225, 1333, 1470 and 1550 cm"), characteristic of nanocrystalline boron—doped diamond, are present in all the spectra. The band at ~1225 cm" is a signature of boron present in the film. Similarly to microcrystalline diamond, it is attributed to a disordered spa-carbon, due to distortion of diamond lattice by boron atoms. With increasing boron concentration, scattering intensity at this position becomes more pronounced. Another difference in the Raman spectra, especially in spectra D and E, is red-shift of the diamond line position from 1333 cm'1 to ~1325 cm", very likely due to high boron concentration and possible defects developed because of it. The scattering intensity of the entire spectrum decreases with increasing doping level. This results from an increase in opacity (i.e., decreased sampling volume) with doping level. 1 The peak at 1150 cm' is used as a signature for high quality nanocrystalline diamond,224 and is being associated with spZ-bonded carbon, specifically transpolyactelylene segments at grain boundaries.228 Figure 5.14 A shows background cyclic voltammetric i-E curves for the boron- doped nanocrystalline diamond films in 1.0 M KCI, over a wide potential range. The working potential window (1 100 11A or 500 uA/cmz) for each film decreases with increasing doping level. For example, the window is 3.1 V for film A, B, and C, and 3.0 for film D. The response is featureless for all films between the potential limits. The anodic current at positive potentials is due to the onset of chlorine evolution. The reduction of chlorine back to chloride is most pronounced at film D with a peak at 750 mV. The cathodic current at the negative potentials is attributed to hydrogen evolution. Figure 5.14 B shows a background cyclic voltammetric i-E curves for a boron- doped nanocrystalline diamond films in 1.0 M KCI over a more limited potential range. 116 250 el l 150 '. e c 2 b E; a a 50 r O S —’ U . <— -50 . r A -150 1 1 1 1 1 1 -2000 -1000 0 1000 Potential (mV vs. Ag/AgCl) 8 Current (11A) l 1 l -600 -100 400 900 Potential (mV vs. Ag/AgCl) Figure 5.14 (A) Cyclic voltammetric i-E curves for boron-doped nanocrystalline diamond films in 1 M KCI over (A) wide potential range showing the working potential window and (B) a narrow potential range. Films deposited using a 1/94/5 CHJAr/Hz (v/v) source gas ratio and (a) 1, (b) 10, (c) 20, (d) 30 ppm of added BzHe. Scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2 117 The curves are similar in shape and stable with cycling. There are no peaks present in the 0-500 mV range, associated with redox-active surface carbon-oxygen functionalifies. There is anodic charge between 400 and 900 mV that increases with doping level. The cause of this current is unknown but may be related to oxidation of spz-bonded carbon in the grain boundaries and defects, which may increase with doping level. The current magnitude of the nanocrystalline diamond film is slightly higher, than the current for microcrystalline films. For example, the current for microcrystalline films is ca. 0.4-0.6 11A (2.0-3.0 11A/cm2) at 250 mV vs Ag/AgCl in this medium, while the currents for the nanocrystalline films are 0.7, 0.9, 1.0, and 1.111A (3.5-5.5 uA/cmz) for films A, B, C, and D, respectively. The larger current, is likely due to the higher fraction of exposed spa- bonded carbon in grain boundaries. The electrochemical responsiveness of the boron-doped nanocrystalline diamond film toward five redox systems was next investigated using cyclic voltammetry. A summary of the cyclic voltammetric data is given in Tables 5.5 and 5.6. Table 5.5 Cycllc Voltammetric Data for Fa(CN).'°"‘, nu(NH,).*"*’, IrCl.""" at Boron- Doped Nanocrystalline Dlamond Films 1 mM 54cm.“ 1 mM Ru(NH3).*3”" 0.25 mM erlg'm Film (ppm) Epox ipox AEp Epox lpox AEp Epox Ipox AEP (mV) (M) ("M (mV) (PA) (mV) (mV) (M) ("M 1 309 73.1 65 -132 63.2 67 821 13.6 66 10 309 69.4 63 -135 62.7 59 833 215* 63 20 311 74.2 67 -131 64.9 59 824 14.4 64 30 310 70.1 62 -131 67.8 59 830 14.6 64 Note: Voltammetric data obtained at scan rate=0.1V/s, electrode geometric area=0.2 cm”, and * concentration of erl6'2"3 was 0.5 mM. 118 For all of three systems a nearly reversible voltammetric response was observed for all the films. AEps were in the range of 59-67 mV. The small AEps observed for the boron-doped nanocrystalline diamond films are indicative of a high level of surface cleanliness, low surface oxide coverage and high electrical conductivity for all four films. Cyclic voltammetric i-E curves for 4-tert-butylcatechol (t-BC) and Feiz”+3 have much larger AEps and more asymmetric peak shapes. AEp's of 419-487 mV and 632-735 mV +2l+3 are observed for t-BC and Fe , respectively. The larger peak separations, as compared to the other redox analytes, result from more sluggish electrode reaction kinetics.‘7-‘° Table 5.6 Cycllc Voltammetric Data for Fe*”*’ and t-Butyl Catechol at Boron-Doped Nanocrystalline Diamond Films 1 mM Fe’”+3 1 mM t-butyl catechol Film (PPM) E ox E,” (mV) 1,,” (11A) 85, (mV) (":v) 1,,“ (11A) AE, (mV) 1 920 32.3 632 690 1 18.3 439 10 926 36.0 679 655 1 12.2 419 20 960 33.6 735 680 1 17.5 487 30 940 31.7 716 669 115.5 477 Note: Voltammetric data obtained at a scan rate = 0.1 V/s. Electrode geometric area = 0.2 cm2 In general, as discussed in Chapter 3, the electrode kinetics for these two analytes are more sluggish at both microcrystalline and nanocrystalline diamond, as compared to glassy carbon. The AEp for t-butyl catechol shows little variation with doping 119 level and the values are smaller than the 600-800 mV observed for microcrystalline diamond. It is supposed that the slow kinetics for nanocrystalline diamond result from a lack of adsorption on the spa-bonded, hydrogen-terminated surface. The AEp for Fe‘z’+3 increases with increasing doping level. The reason for this is unclear. AEp for this redox system at both microcrystalline and nanocrystalline diamond is larger than that for glassy carbon. This has been attributed at least in part, to the absence of catalyzing surface carbon-oxygen functional groups on the diamond surface, specifically carboxyl groups.63'66 Plots of ip°" versus v"2 for all five redox analytes were linear (r2=0.998), with y- axis intercepts near zero. This trend is indicative of reactions limited by semi-infinite linear diffusion of reactants to the electrode surface. In summary, the nanocrystalline films, with different doping levels, all show a high degree of electrochemical responsiveness for Fe(CN)6'3"‘ , Ru(NH3)6“3’*2, and erl.5'2"3 without any pretreatment. Over this doping range, the density of electronic states in the films is not significantly different. This is an important observation because it proves that the differences in voltammetric response for the polyamines discussed below are not due to variations in the electrical conductivity. Figure 5.15 shows cyclic voltammetric i-E curves for nanocrystalline diamond film electrodes in borax buffer, pH 10.6. At potentials less than 900 mV, the current for all three electrodes have a similar shape. The main difference in the voltammograms is that the onset potential for oxygen evolution, which shifts toward negative potentials with increasing doping level. This trend may indicate that surface boron sites are involved in the oxygen evolution reaction (i.e., stabilization of OH-). 120 200 A150- < e . E 2100' h S o 50. 0 100 300 500 700 900 1100 Potential (mV vs. Ag/AgCl) Figure 5.15 Cyclic voltammetric i-E curves for boron-doped nanocrystalline diamond films in borax buffer pH 10.6. The films were deposited with (a) 1, (b) 20, (c) 30 ppm BzHe. Scan rate = 0.1 We. Electrode geometric area = 0.2 cm2. This trend may also reflect an increased level of spa-bonded carbon present at the surface, caused by the increased doping level. It is known, as shown in Chapter 5, that spz-bonded carbon exhibits a lower overpotential for oxygen evolution than does spa- bonded diamond. Figure 5.16 shows cyclic voltammetric i-E curves for the oxidation of PUT, CAD, SPMD, and SPM at nanocrystalline films with different doping levels. 121 (1V2) .Lfld Current (11A) (IIAIdS WdS 111111 500 800 1100 500 800 1100 500 800 1100 500 800 1100 Potential (mV vs. Ag/AgCl) Figure 5.16 Cyclic voltammetric i-E curves for 1.0 mM cadaverine, putrescine, spermidine, and 0.8 mM spermine in BBpH10.6 at boron-doped nanocrystalline diamond films deposited with (A) 1 ppm (B) 10 ppm , (C) 20 ppm, and (D) 30 ppm of added Bsz. A well-defined voltammetric wave is observed for each of the amines at all films. Most important is the fact that the peak current increases with increasing boron level for all four amines. This is direct evidence for the effect of surface boron on the oxidation 122 reaction. An increasing doping level leads to a higher number of surface boron sites available for the reaction. The cyclic voltammetric peak current and potential data are summarized in Tables 5.7 and 5.8, respectively. It is clear that the oxidation peak current for each amine increases with increasing doping level. It is also clear that the oxidation peak potentials are largely invariant with doping level. At low doping level, the currents for SPMD and SPM are larger than those for PUT and CAD. However, as the doping level increases this trend changes, with the largest currents observed for CAD and SPMD. Table 5.7 Summary of Cyclic Voltammetric Peak Current Data for Boron-Doped Nanocrystalline Diamond Electrodes Oxidation Peak Current l,” (11A) Film 1 mM CAD 1 mM PUT 1 mM SPMD 0.8 mM SPM 0 ppm no oxidation no oxidation no oxidation no oxidation 1 ppm 32.4 i 4.9 32.5 2!: 1.4 51.0 i 2.4 50.1 i: 1.7 10 ppm 124.5 :t 11.5 115.2 i 39.2 153.7 i 2.4 167.1 i 6.9 20 ppm 261.8 i 62.7 220.2 i 38.9 264.6 :1: 41.4 198.6 :1: 26.4 30 ppm 304.5 251 306 225 Note: Voltammetric data obtained at a scan rate = 0.1 We Electrode geometric area = 0.2 cm2. Data are averages for two or three electrodes at each doping level 123 Table 5.8 Summary of Cyclic Voltammetric Peak Potential Data for Boron-Doped Nanocrystalline Diamond Electrode Oxidation Peak Potential E,” (mV) Film 1 mM CAD 1 mM PUT 1 mM SPMD 0.8 mM SPM 0 ppm no oxidation no oxidation no oxidation no oxidation 1ppm 874:19 884:12 896:1 895:2 10 ppm 880:25 882:12 889 :15 896:27 20ppm 880:4 905:5 906:4 910:2 30 ppm 876 900 902 910 Note: Voltammetric data obtained at a scan rate = 0.1 We. Electrode geometric area = 0.2 cm2. Data are averages for two or three electrodes at each doping level Studies investigating the microcrystalline electrode activity for amine oxidation as a function of boron dopant concentration were performed as well. All the microcrystalline films were deposited from 0.5 % CH4/H2 volumetric ratio with 1, 10, and 20 ppm of added Bsz. The concentration of boron in the film is proportional to the boron concentration introduced into CVD reactor. Measured resistivities of the films appear to track the boron concentration. They were 0.38, 0.17, and 0.12 Q-cm, for 1 ppm, 10 ppm, and 20 ppm films, respectively. High quality microcrystalline films, lightly doped with boron possess a Raman spectrum resembling that observed for high quality HPHT synthetic diamond. The only spectral feature is the first- order diamond phonon line at about 1332 cm'1 with a narrow FWHM of about 6 cm". Very heavily doped films (>1020 B/cma) have a more complex 124 spectrum. The diamond phonon line position is usually downshifted from the expected 1332 cm'1 position and possesses a larger FWHM. The increased FWHM is due to the larger number of defects caused by inserted boron dopant atoms.260 The phonon line shape is asymmetric with an upward shift in intensity on the high wavenumber side of the peak. This asymmetric shape is attributed to a Fano-type interference between the scattering by the zone-center optical phonon and the scattering by an electronic continuum caused by the formation of a B impurity band above 1019 B/cm3 whose energy overlaps that of the phonon.260 A band around 1200 cm'1 is also observed and is attributed to the presence of disordered spa— bonded carbon, due to high concentration 260263 The 1200 cm’1 band arises from scattering by of boron atoms in diamond lattice. phonons outside the center of the Brillouin zone, hence a breakdown of the Raman selection rules associated with the loss of long range periodicity in the lattice. This loss can be related to an amorphous state or very small, discrete crystalline regions of diamond. 260263 Figure 5.17 (A-C) shows Raman spectra for three microcrystalline films. The doping levels are in the range of 1019 to 102° B/cma. For 1 ppm, 10 ppm, and 20 ppm films respectively, the FWHM values are 9.7, 11.6, and 13.5 cm". The diamond line intensity also decreases with increasing doping level as the films become more opaque. The first-order diamond phonon line is downshifted about the 2 cm‘1 from the expected 1332 cm“1 position. All the films showed the diamond peak asymmetry and scattering intensity around 1200 cm". The intensity of this mode was the highest for film 20 ppm. 125 k A Huang»: ,. . , s I‘" . w I r-fl 1000 1250 1500 1750 A 5 :3 E 3 4 "’9 >4 3 5* g 2 .5 1 1 1 1 l 1 1 l 1 1 1000 1250 1500 1750 3 C 2 1 - .mt . . . . 1000 1250 1500 1750 Wavenumber (cm'l) Figure 5.17 A-C Raman spectra for three microcrystalline films, deposited from a 0.5 % CHJHZ volumetric ratio with (A) 1, (B) 10, and (C) 20 ppm 8sz added. The doping level was in the range of 10‘9 to 1020 B/cm3. 126 Background voltammetric i-E curves, which are extremely useful for evaluating the diamond film quality, were recorded in 1 M KCI.32 The voltammetric i-E curves were all flat and featureless. The scans were stable with multiple sweeps. The potential windows were 3.1, 3.0, and 3.0 for the 1 ppm, 10 ppm, and 20 ppm films, respectively. The electrochemical responsiveness of a boron-doped diamond thin film toward four redox systems was first investigated using cyclic voltammetry to verify that all three electrodes exhibit a good electrochemical responsiveness. The redox systems were Fe(CN)6'3’“‘, Ru(NH3)6*2’*3, and erl6‘2"3 in 1 M KCI, and Fe*"”+3 in 0.1 M HCIO4 .The influence of diamonds physical, chemical and electronic properties on the electrode reaction kinetics for each of these systems was discussed in Chapter 3. Table 5.9 summarizes some of the cyclic voltammetric data for the four redox systems. All films exhibited a good responsiveness for Fe(CN)6'3"‘, Ru(NH3)5*2’*3, and erl6'2"3 without any kind of pretreatment. Table 5.9 Summary of Cyclic Voltammetric Data for Boron-Doped Microcrystalline Diamond Electrode Fe(CN)e'3"‘ Ru(Nl-i3)5*3’*2 "ch-2’4 F312“3 Fm" '0 As, i,” As, 1,,"x AE, 1,,ox As, 1,,” (mV) (11A) (mV) (11A) “2‘" (11A) (m_V) (M) A 73 62.9 72 57.5 64 20.7 352 33.9 B 82 61.5 68 61.8 63 22.2 782 32.6 C 74 61.2 63 59.9 64 21.7 824 31.7 Note: Voltammetric data obtained at a scan rate = 0.1 V/s, Electrode geometric area = 0.2 cm 127 120 r A 3 9° 5 a 60 1 § . 5 30 - 0 1 L 1 L 1 l 1 400 600 800 1000 1200 120 25. ‘1: 5 U 120 ‘5 Q l: 5 U 400 600 800 1000 1200 Potential (mV vs. Ag/AgCl) Figure 5.18 Cyclic voltammetric i-E curves for (A) 1.0 mM PUT, (B) 1.0 mM SPMD, and (C) 0.8 mM SPM, in the BBpH10.6 at three microcrystalline diamond films deposited from a 0.5 % CH4/H2 volumetric ratio with (A) 1, (B) 10, and (C) 20 ppm B2H5 added. 128 The only redox systems that sluggish kinetics were observed for was Fe*2"‘3 , presumably due to a lack of catalytic surface carbon-oxygen functionalities.63’66 The oxidation peak current for each electrode varied linearly with the scan rate‘” up to 1000 mV/s (r2=0.998). The AEp values are between 60-75 mV for Fe(CN)6‘3’“‘, Ru(NH3)6*2’*3, and "Gig“. Figure 5.18 shows cyclic voltammetric i-E curves for 1.0 mM (A) CAD, (B) SPMD, and (C) 0.8 mM SPM in the BBpH10.6 electrolyte at the three microcrystalline diamond films with different boron doping levels. A well defined oxidation peak current is seen for all three amines at each electrode. However, the highest oxidation currents are observed for the most heavily doped film (i.e., highest surface concentration of boron sites). The ip°" and E,” values are summarized in Table 5.10. Table 5.10 Summary of Cyclic Voltammetric Data for Aliphatic Amines at Boron- Doped Microcrystalline Diamond Electrode 1 mM PUT 1 mM CAD 1 mM SPMD 0.08 mM SPM Fi I m EPOX IPOX EDD! IPOX EPOX 'pOX EDD! IPOX (mV) 01A) (mV) (11A) (mV) (11A) (mV) (1A) A 915 28.3 891 22.1 916 26.3 919 28.8 B 861 63.1 854 58.5 889 73.0 887 59.6 C 896 105.1 872 90.9 879 113.4 907 77.4 Note: Voltammetric data obtained at a scan rate = 0.1 We Electrode area=0.2 cm2 Several observations can be made from the voltammetric data. First, the oxidation potentials show no clear trend with doping level. One might expect there to be a trend of more positive oxidation potential with doping level as the surface coverage of the polyamine should increase with doping level. Second, the oxidation current for each 129 amine increases with increasing doping level. This is consistent with the nanocrystalline film data showing a direct correlation between the surface boron content and the oxidation current. Third, the oxidation current for each amine is nearly the same at the lowest doped film, with value of approximately 2711A, but shows some variation for each amine at the higher doping levels. 5.3. Conclusions The results clearly show that the polyamines can be qualitatively and stably oxidized at both boron-doped microcrystalline and nanocrystalline diamond thin-film electrodes. The oxidation current is highly dependent on the physicochemical properties of the diamond surface, in particular, the spz—bonded non-diamond carbon impurity and surface boron concentration. A CH4/H2 source gas ratio of 0.4-0.5% produces the optimum structure and leads to the largest peak currents. Both boron-doped microcrystalline and nanocrystalline diamond film provide a quantifiable and stable response. The oxidation response also depends on the solution pH. In order to be oxidized, the amines need to be unprotonated. This means that the maximum oxidation currents are observed at pHs above the pK, of the amines. A solution pH of 11 was found to be the optimum. The onset potential for oxygen evolution (OH- formation) shifts toward less positive potentials with increasing pH (~60 mV/pH unit). The E,” values for the amines also shift toward less positive potentials with increasing pH. For a given pH, the oxidation response, both E,” and ip°", were relatively independent of the electrolyte type. At low amine concentrations and low scan rates, the oxidation peak currents vary linearly with the scan rate ”. Increasing the amine concentration caused E,” to shift toward more positive values, similar to the scan rate dependence. This is a 130 consequence of an increased number of amine molecules available for oxidation, requiring a higher concentration of OH-. This is achieved at more positive potentials. Plots of ip°"vs concentration were linear up to a 1.0 mM concentration. The plot had non- zero intercept and different slopes (i.e. sensitivities) for the different amines. Evidence was presented for the importance of (i) spZ-bonded non-diamond carbon impurity and (ii) surface boron sites in the amine oxidation reaction mechanism. The amine oxidation current shows a dependence on the CHJHZ ratio with a maximum signal seen for films deposited from a 0.5 % CHJHZ. If a film is rehydrogenated to remove all the spZ-bonded non-diamond carbon impurity, then a total attenuation of the amine response, prior to oxygen evolution, is observed. The film deposited from the 0.5% CHJHz ratio exhibited the largest and most well-defined waves. The reproducibility of the response was also best for the 0.5% film. It is important to point out again that the prerequisite for the anodic oxygen transfer reaction (OTR) is the anodic discharge of H20 to produce adsorbed OH- as shown in Eq.1. H20 —> S[OOH ads]+ H+ +6 (1) Evolution of 0;. can diminish the current efficiency for OTR, by consumption of adsorbed OHo, as shown in Eq.2. S[OOH] +H20 —-> 51 ] + 02 + 3H+ +33 (2) There is a competition between the Oz evolution (OER) and OTR for the adsorbed OH- and any increase in the rate of OER will result in decrease in the rate of OTR. However, the proper fraction of spz-bonded non-diamond impurity exposed at the surface catalyzing discharge of water at lower overpotential is extremely important for achieving the optimum electrode response for amine oxidation. Such a structure is achieved with 131 films deposited from a 0.4-0.5% CHJHZ source gas ratio. Figure 5.19 shows a simplified model of the competition between the OER and amine oxidation via OTR. Lower density of non-diamond carbon koan < k OTR R+'OHads—> RO+H++e- OH ° / .3 —B‘" HzN-R-NHZ e . 1.9 81’?- ” 0H ' \ . . . OH . ngh den31ty of non-diamond carbon kosa > k on H20 + ' OHads —-> 02 + 3H+ +3E’ Figure 5.19 Model of the competition between the Oz evolution (OER) and O-transfer reaction (OTR). We presented compelling evidence regarding the importance of surface boron sites in the reaction model. At nanocrystalline conductive diamond film deposited without boron, signal for amine oxidation was attenuated. The absence of a well defined oxidation response for cadaverine was not due to poor film electrical conductivity since the high fraction of grain boundary, containing n—bonded atoms, gives film sufficient electrical conductivity, even in the absence of intentional boron-doping. A set of boron-doped nanocrystalline and microcrystalline boron-doped diamond deposited with different dopant level was probed. All microcrystalline films exhibited a good responsiveness and fast kinetics for Fe(CN)6'3"‘, Ru(NH3)6*2’*3, and "Gig“. The AEps of the redox probes studied were independent on doping level. Even the film 132 deposited with the lowest doping level (1 ppm B2H6) had high carrier concentration, adequate to support fast electron transfer. Very fast kinetics for redox systems were observed at nanocrystalline boron- doped diamond electrode, regardless the doping level. A A well-defined voltammetric waves for amine oxidation were observed for both types of films. For both, microcrystalline and nanocrystalline diamond films the anodic signal increased with the increasing dopant level. The differences in anodic oxidation response for amines were not due to poor films electrical conductivity, as it was shown that all the films possessed high density of carriers to support rapid electron transfer for other tested redox systems. This is direct evidence of a surface boron on the oxidation reaction mechanism. Increasing doping level leads to a higher number of surface boron sites available for the reaction. Polycrystalline boron-doped diamond thin film electrodes possessed the requisite surface structure and chemical composition to support and sustain the oxidation of aliphatic amines. Because of the synthetic nature of the film deposition, the appropriate structure and composition can be fabricated into a film. A model is proposed whereby the non-diamond carbon impurity sites, presumably located in the grain boundaries or crystal defects, generate reactive OH- at lower overpotential, than the surrounding diamond matrix. The radicals then attack the polyamine molecules adsorbed/coordinated at surface boron sites near the grain boundafies. 133 Chapter 6 INVESTIGATIONS OF ALIPHATIC POLYAMIN E OXIDATION REACTION MECHANISM AT BORON- DOPED DIAMOND THIN-FILM ELECTRODES 6.1. Introduction Recently a number of reviews have been written about the chemical and 265- 26““ The mechanism and kinetics of the two electrochemical oxidation of amines. types of oxidations have been reported on and, in some cases, studies have shown that the mechanisms are similar for the two routes.269 Evidence for a two-electron transfer reaction has been reported for amine oxidation by H202. PGI'OXY acids, and ozone.”39 and references therein Hydroxylamine was found to be the primary product for these oxidants. Other oxidation reactions, such as the reaction of an aliphatic tertiary amine with chlorine dioxide, produced a secondary amine and an aldehyde.27o It was proposed that the reaction involved abstraction of an electron from the amine in the rate-determining step, followed by rapid loss of a proton and a second electron. The same reaction mechanism was also reported for the oxidation of amines with dibenzoyl peroxide,271 manganese 134 dioxide,272 potassium pehmanganate,273 and N--hromosuooihimide.274 Ag (ll), Ni (in), Co (Ill), and Cu (ll) oxides are all known oxidants,275'276 277,278 with Ag (ll) oxide being particularly efficient for amine oxidation. Heterogeneous reactions with transition-metal oxides are often considered to occur via adsorption of the organic compound on the surface of the oxide, followed by a chemical reaction, commonly hydrogen abstraction. The interest in amine electrooxidation dates back to the mid 1800s, when aniline 266 and references therein oxidation was first studied. Recently, the mechanistic aspects of aromatic and aliphatic amine electrooxidation have received considerable 232265279280 and references them" It was found that the electrooxidation of attention. aromatic amines depends on several factors:266 (i) the presence of hydrogen at the 01- carbon, (ii) the presence of hydrogen on the nitrogen atom, and (iii) the presence and type of substituent groups on the aromatic ring. The oxidation of aromatic aliphatic amines is generally easier than aliphatic ones, due to the strong electron donating 265,266 aromatic ring. This is evidenced electrochemically by more positive aliphatic amine oxidation potentials. The electrooxidation reaction strongly depends on anode material?“282 A survey performed by Caiiizares et al. 281 shows that the type of electrode material used has a great influence on the oxidation reaction mechanism. Two types of anode behavior have been identified. The first type is called active behavior in which the electrode surface undergoes chemical and structural changes (i.e., oxidation) that promote the amine oxidation. The second type is called non-active behavior in which the electrode surface undergoes no change and simply acts as a sink for electrons. Pt, lr02, and stainless steel electrodes are examples of the active type. These electrodes are not fully oxidized and, consequently, are transformed into higher oxides. Fully oxidized metal oxides, such as PbOz, SnOz, or boron-doped diamond are examples of the non-active 135 type. A similar classification of anode materials for the electrooxidation of amines was proposed by Chow et. al.265 The authors assigned Ag, Cu, Co, Ni, Fe, and Au to the active type, and PbOz and glassy carbon to the non-active type. Interestingly, Pt electrode was classified by them as an inert material for amine oxidation. In general, the electrochemical oxidation of amines is considered to involve the removal of an electron from the N atom.265 The formation of a cation radical via a one electron transfer reaction is the first step of oxidation process for both types of electrodes"?65 In simple cases, at inert electrodes, oxidative dealkylation processes dominate because the amine radical cation deprotonates at the rat-carbon, and the or- amino radical formed is oxidized to the immonium salt. The immonium salt then hydrolyzes to the dealkylated amine and a carbonyl compound. The detection of amine oxidation intermediates has received considerable study. In only a few cases, however, has the initially formed cation radical been sufficiently f ' . u a 265 and '9 memes mm" For example, the radical cations long-lived to allow its detection. of conjugated amines are sufficiently stable to be detected. In those cases, the esr data has contributed to an understanding of the electronic structure and spin distribution in non-aromatic aminium radicals, and the electrochemistry yielded information on the ease of formation and stability of these intermediates. To evaluate the ease of formation of aminium radical in solution, the E° value for amine <—> amine-i-o +e- couple is required. Voltammetry is often used to obtain E0 values experimentally. For the reversible process, E° for the system is equal to E1/2. If the electron transfer is irreversible, the oxidation potential is shifted anodically from the “reversible” position of the E°, and the value of the shift depends on the heterogeneous rate constant for the electron transfer. A reliable E° value may be obtained only when the cation radical is sufficiently long lived for a reversible reaction to occur amino» 4» e- <—> amine and oxidation/reduction are 136 observed. Usually, however, the cation radical for simple aliphatic amines is so reactive that only an oxidation wave is observed in voltammetry. In spite of these limitations, it was suggested 265 that for irreversible oxidation of amine to cation radical shift of EPO" from E0 would be similar for all aliphatic amines, so that the differences in Epox for different amines would reflect differences in E°, and the same provide information on ease of formation of cation radical. From these studies it was found that in general, primary amines oxidize much slower than secondary or tertiary amines. The order of reactivity among aliphatic amines is: tertiary>secondary>primary.265 This is as to be expected for electron transfer from nitrogen atom, since the alkyl groups are electron- donating substituents. A short review on inert and active anode materials and mechanistic aspects of amine oxidation on those electrodes is presented below. 6.1.1. Non-Active Electrodes Mann and coworkers investigated the electrochemical oxidation of primary and 269280283284 The rate limiting step was tertiary amines at Pt in alkaline medium. concluded to be loss of the first electron from the N atom, followed by rapid loss of proton from the or-carbon to form radical (ll), as shown in Eqs. 1 and 2. In the next step, the radical (ll) may either lose a second electron to form the iminium ion (III) or may undergo disproportionatlon to form the enamine. It has not been determined which of these intermediates is actually formed. The aldehyde and dealkylated amine were found, however, as the final products. 137 so 9+ RzNCHzR’ Z RzNCHzR’ + e' (1) (I) 0+ coo RZNCHZR’ W RzN -CHR1 '1' HT (2) (II) use + RzN-CHR' W R2N=CHR' -i- 8’ (3) (III) + +H20 RZN=CHR 758T) Rz-NH + R'CHO +H+ (4) Anodic oxidation of amines has been studied by Masui et. al. 232 at glassy carbon (60). The first step in the anodic oxidation of aliphatic tertiary amines at 60' electrode in aqueous solution was considered to be the abstraction of an electron from the lone-pair on the nitrogen. The increasing difficulty of oxidation of amines with increasing electron-withdrawing power of the substituents was observed. It was stated that overall reaction involved transfer of 2 electrons and a proton. The transfer of the first electron was suggested as limiting step, similarly to studies by Mann et. al. at Pt electrode.280 It was proposed that the intermediate aminium cation radical is converted into a quaternary Schiff base, which rapidly hydrolyzes to the amine (ammonia) and the aldehyde. The secondary amine and aldehyde were shown to be produced by anodic oxidation of a tertiary amine. The electrode reaction for amine was deduced to be as in Eq. 5: R2.I 1R2.1¢CH2R' —-I:I—+>R2N-.dI"IR' L R21¢=CHR'} (5) slow fast fast + H20 R2N=CHR' 7;? Rz-NH + R'CHO 138 6.1.2. Active Electrodes Oxidation of primary and secondary amine compounds on Ag electrodes in alkaline medium has been studied by Hampson et al.233'235 The authors reported that the initial step in the reaction sequence corresponds to oxidation of amine to the corresponding imine, as indicated by Eq.6. The imine can undergo hydrolysis to form acetaldehyde and ammonia, as indicated by Eq.7, or it can undergo further oxidation to form acetonitrile as indicated by Eq.8. The ratio of aldehyde to nitrile was determined to be dependent on potential, hydroxide ion concentration and amine concentration. (CH3CH2NH2)ads —) CH3CH=NH+ 2H+ + 26 (6) CHa-CH=NH + H20 —) CHa-CHO -l- NH3 (7)- CHa-CH=N H —) CH3CN + 2H+ + 28' (8) In general, electrochemical studies have shown that primary amines are readily oxidized by A90 in aqueous electrolytes. The oxidation mechanism of secondary amines at A90 was found to be different from that of primary amine. Mechanism according to Hampson et al. is proposed in Eq. 9-13.234 Detected products (II) and (III) of the oxidation were explained by oxidative fission of (I). The product (III) was further responsible for the appearance of alcohols (IV) and olefins (V). It was proposed that Schiff base may also be oxidized at silver electrode to produce aldehydes (VII) and nitriles (VIII). The products of oxidation were the same as products of hydrolyzed Schiff base. It was also proposed that Schiff base RCH2N=CHR produced from secondary amine (quasi M» can be oxidized to primary amine (RCH2NH2) and aldehyde (RCHO) and the product primary amine can be further oxidized to RCN or RCHO. 139 - - 0+ RCHzCHz NHCHzR —e—> RCHzCHz NHCHzR (9) (I) 0 + -e' 0 0 + + RCHzCHz NHCHzR —'_) R-CHz-NH + RCHzCHz (10) (II) (111) + RCHzCHz + OH‘ __) RCH2CH20H -i- RCH2=CH2 (11) (III) (IV) (V) 0 0+ -H+ R-CHz-NH ——> RCH=NH + 011- ——> RCHO (12) (II) (VI) (VII) ’ ’+ 'H+ -2e' _ R-CHz-NH ——> RCH=NH —> Rc=N (13) (11) (v1) ’ ZH‘“ (VIII) ' The kinetics and mechanism of the oxidation of amines were studied at oxide- covered Ni, Cu, Co electrodes by Fleischmann et. al.35 The oxidation potentials for the primary and secondary amines were the same and oxidation occurred at potentials where the oxide transition takes place (i.e., Ni(ll)—>Ni(lll), Co(ll) ->Co(lll)). It was stated that the electrode reaction takes place by a mechanism involving a rate determining chemical reaction between higher oxide and amine intermediate. The nature of the chemical step was investigated, and it was concluded that the ratedetermining step is hydrogen abstraction from or-carbon. The authors indicated that adsorption of the amine is required before the chemical reaction can occur. They based their conclusion on the fact that tertiary amines did not oxidized at metals oxides. The rate of oxidation of secondary amine oxidation was much slower than the rate of oxidation of primary amines. 140 Recently the anodic oxidation of aliphatic amines has been reported at two relatively new electrode materials: nickel-based chemically modified glassy carbon electrode (active) 34 and an anodized eutectic-phase Ag-PbOz (inert).54’5f”57’285 At a nickel-based chemically modified glassy carbon electrode (Ni-CME) the electrocatalytic function of the electrode was attributed to the formation of B-NiOOH.34 Electrode exhibited activity toward mono- and polyamine oxidation and the mechanism was proposed involving adsorption of amine at Ni and a concerted oxidation of the intermediates on the catalytic sites (i.e. B-NiOOH). The radical hydrogen abstraction from the organic compound was considered rate determining step, however, it was pointed out that adsorption played an important role on the overall kinetic mechanism. It was concluded that the oxidation of amino compounds on Ni-CME is under a mixed control of both adsorption and chemical process leading to cleavage of N-H and C-H bonds in a-position. It was stated that the energy released in the formation of adsorption bonds between analytes and catalytic sites partially compensates the activation energy barrier associated with the abstraction of hydrogen atoms and bond cleavage. The primary, secondary, and tertiary aliphatic amines were oxidized at the anodized eutectic-phase Ag-Pb02 in alkaline media.f"i’55’57’285 Acetaldehyde and ammonia were detected as the product of the electrolysis of ethylamine. The overall reaction is presented as in Eq. 14 CHs-CHz-NHz + ZOH‘ —) CHs-CHO + NH3 + H20 + 2E‘ (14) The reaction mechanism was suggested to be electrocatalytic in nature. The authors proposed that hydroxyl radicals can be adsorbed at specific surface sites, and transferred to the reactants via an electrocatalytic oxygen transfer reaction mechanism. 141 Oxygen transfer reactions can be catalyzed by increasing the surface concentration of adsorption sites, thereby, increasing the concentration of adsorbed OH radicals. It was pr0posed that there is an optimal low density of surface sites for which the desired 0- transfer process can occur at near mass transport — limited rates without high background signal for 02 evolution. At Ag-PbOz electrode, silver as doping agent serves the function of H20 discharge and capability of adsorbing the 00H species, and adsorbing the amine molecules via the non-bonded electron pair of the nitrogen atom. The authors presented following mechanistic sequence, describing oxygen transfer at Ag-Pb02 electrode: (i) anodic discharge of water to produce an adsorbed (00H), (ii) adsorption of the amine, (iii) cleavage of the carbon-nitrogen bond with transfer of oxygen to the carbon radical. ln present chapter the mechanism of amine oxidation was investigated in details by use of different electrochemical measurements. First, the number of electrons transferred in the limiting step was determined from the measurements of Tafel slopes, and the ratio between number of protons and electrons taking part in the oxidation process was evaluated from the pH dependence of the Em of amine oxidation peak. Then the adsorption behavior of amines was studied by evaluating how amine concentration and the temperature influence the amine oxidation current. Finally, results of a rotating disc electrode studies were used to determine the heterogeneous electron transfer rate constant and total number of electrons per amine group characteristic for amine oxidation reaction. 142 6.2. Results and Discussion 6.2.1. Irreversibility of the Process Voltammetric investigations of aliphatic amine oxidation at diamond reveal that the overall reaction is irreversible with a single anodic peak observed ~100mV prior to oxygen evolution. The overall irreversibility of the amine oxidation process may be caused by (i) an irreversible electron transfer step, (Eq. 15) A 4"" ' B (15) or (ii) an irreversible electron transfer step followed by a chemical reaction, (Eq.16) ALB—,C (16) Other possible reaction scheme can involve (iii) a reversible electron transfer followed by a fast irreversible chemical step or (iv) a reversible electrochemical step, followed by a fast reversible chemical step. Literature data indicate that electron transfer involved in amine oxidation is irreversible. Cation radicals found from saturated aliphatic amines, are usually short- lived,265 and the equilibrium of the consecutive step is shifted very far toward the chemical step and product formation. Hence, the mechanism (iii) and (iv) with a reversible electron transfer for amine oxidation can be rule out. We have shown that amine did not oxidize at hydrogenated boron-doped diamond electrodes,259 at which simple electron transfer reactions occur (i.e., Fe(CN)6‘3"4, Ru(NH3)6*2’*3,‘63 azides286). Yet, at “as deposited” diamond film we observed amine oxidation reaction. Those results strongly suggest that mechanism of amine oxidation does not involve only electron transfer. This suggests the type of mechanism, where electron transfer is followed by chemical process (EC) as it is shown in Eq.16. The nature of the chemical step then, needs to be considered. 143 Recent studies of the electrooxidation of organic compounds suggest that the oxidation of the amine is assumed to be mediated and performed by hydroxyl . 54,55,230,242,244,281,287-289 radicals. A model for the amine oxidation strongly suggests involvement of those species and their function in chemical part of the mechanism (i.e., oxygen transfer). Hydroxyl radical generation at lower potentials, occurs primarly at spz- bonded non-diamond carbon. lf spZ-bonded non-diamond carbon is removed (i.e., hydrogenation of the. diamond electrode), then the OH- generation occurs at higher overpotential at diamond surface. Amine upon adsorption at electrode surface assumes the same charge as the electrode and as result it may as well rapidly desorb from the electrode surface. This may very likely occur when the following, chemical reaction is very slow. This effect was observed for high quality diamond films where amount of spz-~ bonded non-diamond carbon is low . In extreme case, for hydrogenated boron-doped diamond electrodes,259 the chemical reaction was shut by lack of hydroxyl radicals. These observations suggest that amine oxidation can be indeed classified as EC mechanism, with involvement of hydroxyl radical in chemical step. Fast reaction rates are observed only for diamond electrodes, at which the discharge of water had appreciable rate. 6.2.2. Reaction of the Primary and Secondary Amine Groups In Chapter 5, it was shown that oxidation potentials (Epox') are similar for all tested amines, regardless of their molecular structure, indicating similar oxidation kinetics. Figure 6.1A shows typical cyclic voltammetric i-E curves for 1.0 mM PUT (dashed line) and 0.8 mM SPM (solid line), in borax buffer, pH 11, at a boron-doped nanocrystalline film. Similarly shaped voltammograms are observed for PUT and SPM. However, close inspection of the voltammograms reveals a difference in the shapes of 144 the curves at low potentials (700-750 mV vs Ag/AgCl). In this potential range, the current for SPM is larger than that for PUT. The voltammogram for PUT was subtracted from that for SPM, and the resulting difference voltammogram is presented in 6.1 B. There are two peaks observed at 740 and 870 mV. The difference between these two polyamines is the presence of the two secondary amine groups in SPM. It is expected that the secondary amines will oxidize at lower potential compared to primary amines. To test this supposition the voltammograms were obtained in phosphate buffer pH 7.2. At this pH, all primary amine functional groups are protonated and this reduces the oxidation current response due to the fact that adsorption is inhibited. 120 30 §90 E60 15_ 8 : :1 0w 030 . 0 1 1 -15-111L111l 1.41 1111111 1.. 111 111 400 600 800 1000 1200 400 600 800 1000 1200 Potential (mV vs Ag/AgCl) Figure 6.1 (A) Cyclic voltammetric i-E curves for 1.0 mM PUT (dashed line) and 0.8 mM SPM (solid line) in borax buffer pH 11 at a boron-doped nanocrystalline film. Scan rate=0.1 V/s. Electrode area = 0.2 cm2. (B) Subtracted SPM-PUT voltammogram. Figure 6.2 shows cyclic voltammetric i-E curves for (A)1.0 mM PUT, (B) 1.0 mM EDA and (C) 1.0 mM SPM in phosphate buffer, pH 7.2, at a boron-doped nanocrystalline film. There is no net oxidation current electrode response for the PUT at pH 7.2. There was also no net oxidation current for DAP at this pH. This can be explained by the fact 145 that the lowest dissociation constant value for PUT is 9.3 and DAP is 8.5.189 At the same pH for EDA, however, there is a net oxidation current observed at 1200 mV (p2) with a current of 16 11A (S/B=4). The lowest pKa for EDA is 6.9 which explains observed . . 18 OXIdatlon current. 120 OO C Current (11A) ‘5 9 120 I A ~ B ’. 80 '. L 40 L l . l . r..-..-...... 0r.......-.. 200 600 1000 1400 200 600 1000 1400 Potential (mV vs Ag/AgCl) 120 a 80 ’. 5 40’. g . :1 . U l 0 200 600 1000 1400 Potential (mV vs Ag/AgCl) Figure 6.2 Cyclic voltammetric i-E curves, background (dashed line) and total current (solid line) for (A) 1.0 mM PUT, (B) 1.0 mM EDA, and (C) 1 mM SPM in phosphate buffer, pH 7.2 at a boron-doped nanocrystalline film. Scan rate = 0.1 V/s. Electrode area = 0.2 cm2. 146 In contrast, the voltammogram for SPM shows an oxidation peak p1 at a much lower potential (940 mV) with the current of 11 11A (S/B=5). This oxidation peak is followed by an increased oxidation current at 1200 mV, (p2). This is the same potential at which oxidation of the primary amine group of EDA is observed. The peak p1 is observed at much lower potential than the p2 wave for EDA. The lowest pKa value for SPM is 7.9189 and corresponds to one of the two secondary amine groups. At this pH only this secondary amino group is in partially deprotonated form, and thus available for oxidation. This result indicates that oxidation of secondary amine occurs at lower potential. Interestingly, at the 1200 mV (p2), significant current for oxidation of SPM is also observed. It is twice as large as current for EDA at this potential. This is interesting since the electrooxidation of the secondary amine groups of SPM would lead to formation of a primary amine with the structure resembling that of PUT or DAP. Both PUT and DAP do not oxidize at pH 7.2 due to protonation of the amino group, what makes them unlikely to be adsorbed and oxidized. So why would the primary amino groups, resulted from secondary amine oxidation, undergo oxidation? This observation can be explained by the mechanism in which produced primary amine does not desorb rapidly from the boron surface site on the time scale of experiment, and it is available oxidation. These results suggest that the oxidation potentials for primary and secondary amine vary slightly at diamond. Secondary amine groups oxidize at less positive potentials than primary amine groups, although the rate of the oxidation is much lower than that for the primary amines. This is because at lower potentials, where the p1 for SPM is observed, the rate of hydroxyl radical generation is low. At higher potentials, OHo generation is much faster. SPM molecule may be coordinated at ~940 mV, at the surface, however, the overall process is limited by concentration of 00H present. Since 147 amine oxidation is performed by hydroxyl radicals, the step of their generation is the limiting one. The fact that secondary amine oxidize at diamond at lower potential suggests that the mechanism with rate determining electron-transfer step (i.e., amine adsorption onto electrode) is rather unlikely. 6.2.3. Determination of the Ratio Between Number of Protons and Number of Electrons. In order to determine whether the amine oxidation reaction involves transfer of protons, the effect of pH on the voltammetric response was tested. For a reaction: Red —) Ox + mH++ ne (17) the Nemst equation is presented as follows: 0.059 on||H+|m Epox = 15% T log [TGd] (18) where R is the molar gas constant (8.314 J/mol-K), T is the temperature [K], F is the Faraday constant (96,485 C/eq), n is number of electrons, and m is number of protons. At room temperature (298 K) and knowing that pH=-log[H“] this equation can be rewritten as: Epox : Eo_ m 0359 [pH] (19) Equation 19 indicates a linear relationship between E," and pH with a slope of to -0.059 mm. Such a plot is commonly used to determine the ratio of protons and . . . . 290 . electrons involved in the electrochemical reaction. For instance, for a slope 148 of -59 mV/pH unit, m/n=1, indicating an equal number of protons and electrons involved in the reaction, whereas the slope of -118 mV/pH unit (m/n=2) indicates twice as many protons as electrons. Voltammetric investigations revealed that the Epox and EM values for all the polyamines shifted negatively with increasing pH. This trend indicates that, indeed, electron transfer is accompanied by proton transfer. Plots of Ep/z vs pH revealed slopes of ca. -60 mV/pH unit. This indicates an even number of protons and electrons are transferred during the oxidation reaction. Table 6.1 summarizes the results obtained at different electrodes. Table 6.1 Aliphatic EM values as a Function of Solution pH aEmlapH slope amine 0.3% CH4IH2 0.5% CHngz 1.0% CHJH2 cadaverine -68.4 -56.7 -52.8 spermidine -63.9 -64.7 -59.2 putrescine -75.4 -51 .7 spermine -54.8 -48.7 6.2.4. Quantification of Adsorption Our oxidation model indicates that amine adsorption occurs at surface boron sites as an initial step in the oxidation reaction?”291 Data presented in Chapter 5 demonstrated the importance of surface boron, as without it, even for electrically conducting films, a quantifiable amine oxidation signal is not observed. The simplest 292,293 type of adsorption - the Langmuir-type can be characterized by the relationship between peak current (ip) and concentration (C) is given 149 1" = b + 8C (20) 1 _b_ where a and b are specific constants of the amine oxidation. They are determined experimentally by linear regression of a plot of 1h vs 1/C. a and b are concluded to correspond to fundamental chemical and electrochemical constants according to equations 22 and 23.292 a = 1/nFAkI‘max (22) and b= l/nFAkKrmax (23) where K is the equilibrium constant for adsorption, k is the heterogeneous rate constant (3") for the faradaic reaction, I“max is the maximum surface coverage (mchmz) of adsorbed analyte at high solution concentration, and n, F, and A have their usual electrochemical meanings. Equation 20 has the form of the Langmuir isotherm and applies in cases of relatively low surface coverage where lateral interactions between the adsorbed molecules are negligible and the adsorption sites are fully equivalent. Figure 6.3A shows current-concentration profiles for CAD at a boron-doped microcrystalline diamond film deposited with (a) 1 ppm and (b) 10 ppm BzHe. The current for both films rises with concentration at low values before reaching a plateau. The limiting current is higher for the more heavily doped film. This is consistent with a higher surface boron site density. The plateau is reached at 0.5 mM for the 1ppm film and 1.0 mM for the 10 ppm B2H6 film. Figure 6.3 B presents reciprocal plots of the peak current vs concentration for CAD at the same diamond electrodes grown with (a) 1 ppm and (b) 10 ppm B2H5.The plots are linear over a wide concentration range (0.2 mM to 3.7 mM) with correlation 150 coefficients greater than 0.996. The linear relationship indicates that the amine adsorption at boron sites follows Langmuir-like type behavior. Since there is only a limited number of surface boron sites, a low surface coverage is expected with one amine molecule interacting with one surface boron sites. '1? 0.05 110 - . G) A A B ..... A b < 0 04 _. ''''''' a < 90 - b :1 _ ,,,,,, 3 * a 0 03 (5 """ 1 E 70 i- E h '6 """ E Q) _ Q ...... ‘1: t: 0 02 @9639 ......... b a 50 l- : a ,,,, U . a U ’ ."fi’ 30 L h o 01 w“ 10 r 1 L . L , 4 , 0.00 1 1 ._ 1 . 1 . 1 a 1 0.0 1.0 2.0 3.0 4.0 0 1 2 3 4 5 Concentration (mM) 1/Concentration (l/mM) Figure 6.3 (A) Peak current vs concentration profiles for 1 mM cadaverine at boron- doped microcrystalline diamond electrodes deposited in a 0.5 % CH4/H2 ratio with (a) 1 ppm B2H5, and (b) 10 ppm B2H6. (B) Reciprocal plots of peak current versus cadaverine concentration at (a) 1 ppm BzHa, (b) 10 ppm B2H5 diamond electrode. 6.2.5. Rotating Disk Voltammetric Studies The use of rotating disk voltammetry for the study of electrode kinetics and . . . 294 and references therein reaction mechanisms lS well known. The technique is useful because the rate of mass transport of reactants to electrode surface is controlled by fixing the rotational velocity of the electrode. The number of electrons (n, eq/mol) transferred per reactant molecule can be estimated from the slope of plots of the reciprocal current (1h) versus the reciprocal square root of rotational velocity (1/w‘”). 151 This is known as a Koutecky-Levich plot derived for a one step electron transfer mechanism, according to Eq. 25.294 1 1 1 1 1 l lk ll nappFAkappCo 6.62113ppFKDozs (1)1755“ C: b (01/2 or i. in - a + (25) where ix represents the kinetically-controlled current in the absence of mass transfer, new, is apparent number of electrons transferred, kapp, is the apparent heterogeneous rate constant (cm/s), F is Faraday constant, A is geometric electrode area (cmz), Do is diffusion coefficient (cm2/s), 0.) is rotational velocity (rad/s), o is kinematic viscosity (7- 10x10'3 cm2/s for aqueous solutions), and Co is the bulk concentration of the redox species (moi/cma). The apparent number of electrons transferred during the redox reaction (napp, [qumol]) and the apparent heterogeneous rate constant (kapp) can be estimated from the slope and intercept, respectively, of the plot. The diamond disk electrodes were characterized using Fe(CN)6 '3” +3l+2 and Ru(NH3)6 to verify that the limiting currents were observed and that the limiting current varied with the rotation rate according to theory. Results for rotating rates ranging from 52.2 to 314.2 rad/s, at scan rate of 0.1 V/s, are presented in Figures 6.4 and 6.5, respectively. For both couples, steady state i-E voltammetric curves are observed for each rotation rate. Plots of 1/l vs 1/001/2 were linear (r2=0.999) with slopes -3/-4 +3/+2 of 0.065 and 0.072 for Fe(CN)5 and Ru(NH3)5 , respectively. The diffusion coefficients (D) for each analyte were determined from the slopes, according to Eq.24, where n=1 and electrode area=0.2 cmz. The calculated values were 7.3 (:03) x10'6 152 +3l+2 cm2/sec and 6.8 (10.4) x10'6 cm2/sec for Fe(CN)5 '3” and Ru(NH3)6 , respectively. These values correspond well to the reported values of 7.6 x 10'6 cm2/s for Fe(CN)6'3 and 5.5 x 10'6 cmzls for Ru(NH3)6+3/+2.204 0.012 400 . A I 3 0.008 . L a g T :: [ 0.000.141...1.1. d = 0.000 0.050 0.100 0.150 C E: 200 _ m m 1:. ll (rot. speed) (rad/sec) b O —’ a 100 . 0 —> -100 0 100 200 300 400 500 600 700 Potential (mV vs Ag/AgCl) Figure 6.4 Voltammetric response for 1 mM Fe(CN)5 '3“ in 1 M KCI at a rotating microcrystalline diamond disk electrode as a function of rotation speed. The electrode was deposited with a 0.5 % CH4/H2 ratio and 10 ppm BgHg. Scan rate = 0.1 We. Rotational veiocites (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (9) 261.8, and (f) 314.2. 153 0 . ‘.—_— a -100 . ’ b ‘— 3. c /’ 0.000 g -200 . d 2 1 t T ' ‘// 3-0004 : :1 F"? ' "é ’ U "‘ 1 $0008 '. -300 ,_ v-i : -0,012'..:m......m 0.00 0.05 0.10 0.15 1/ (rot. speed)"2 (rad/sec) ”2 .4” A_ L 4 1 a L #4 M a a l a a 4 l a a a_ a a a l a a a l a a a -600 -500 -400 -300 -200 -100 0 100 200 Potential (mV vs Ag/AgCl) +3/+2 in 1 M KCI at rotating microcrystalline diamond disk electrode as a function of rotation speed. The electrode was deposited with 0.5 % CH4/H2 ratio and 10 ppm B2H6.Scan rate = 0.1 V/s. Rotational veiocites (rad/s): (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (9) 261.8, and (f) 314.2. Figure 6.5 Voltammetric response for 1 mM Ru(NH3)5 Figure 6.6 shows the voltammetric response for 0.2 mM CAD (forward scan) in borax buffer, pH 11, at different rotational rates. The film was deposited with 0.5% CH4/H2 and 1 ppm BgHs. The current is independent of the rotation rate, at least at this scan rate, and a limiting current plateau is not observed as it is expected for mass transport controlled reactions (Figure 6.4 and 6.5). Similarly, the oxidation peak currents for PUT, SPMD, and SPM were independent of rotation velocity. These results indicate that the oxidation reaction, at this scan rate, is not limited by mass transport of the amine to the electrode surface. It is supposed that the amine flux to the surface, under these 154 conditions, is too high compared to the rate of adsorption (coordination of the amine at surface boron sites). It could also be that the rate of adsorption is high but the number of surface sites is limiting the current. Such behavior was reported by Johnson et al. for ethylamine oxidation at Au-Ag composite electrodes.295 30 . 30 25 - 2 - § 20 1 i t: ’7 20 E We) <3 i 10 a 3 r 5 15 - O 1 1 1 1 1 ‘r: L 0 1000 2000 3000 a 10 _ rot. speed (nlmin) l / 5 - L a 0 1.11111111111111111411144 400 500 600 700 800 900 1000 1100 Potential (mV vs AglAgCl) Figure 6.6 Voltammetric response for 0.2 mM CAD (fonlvard scan) in borax buffer pH 11 at a boron-doped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5 % CH4/H2 and 1 ppm Bsz. Scan rate = 0.1 V/s. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (9) 261.8, and (f) 314.2. 155 80 . 3.0 . 1: A 60 - :3 7”” r f 4 p 8 @9'9 """ (9 ---------------- 6) 5' g 1.0 . e E r u d g 40 - 2 00 5 L ' 0.04 0.08 0.12 0.16 c - 1/ (rot. velocity)”2 (rad/sec) "2 I; 20 - 0 . . . 1 . . . 1 . . . 1 300 500 700 900 1100 Potential (mV vs Ag/AgCl) Figure 6.7 Voltammetric response for borax buffer, pH 11, at a boron—doped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5% CHI/H2 and 20 ppm Bsz. Scan rate = 0.05 V/s. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (0) 157.1, (d) 209.4, (6) 261.8, and (f) 314.2. Figure 6.7 shows the voltammetric response for borax buffer, pH 11, at different rotational rates. The film was deposited from 0.5% CHJHZ and 20 ppm B2H6. The background current is independent of rotational velocity. The oxygen evolution current is not limited by mass transport but rather by the generation of OH- at the electrode sunace. Figure 6.8 shows the voltammetric response for 0.2 mM SPM (forward scan) in borax buffer, pH 11, at different rotation rates. The film was deposited from 0.5% CHJHZ and 20 ppm B2H5. 156 120 _ 0.16 f 3:5 , c: 100 - i 0.12 _ figs-«5” v . e 7 P E". 008 i 69’ A 80 _ v (1 <1 5 l a. 1 Q 0.04 i c E 60 T F‘ 0.00 h 1 1 1 1 1 1 1 1 1 1 1 b t _ 0.04 0.08 0.12 0.16 a 5 40 _ 1/ (rot. velocity)"2 (rad/sec) ”2 N G I 7 1_ 0 _ _ _. 1 1 l 1 1 4 I 300 500 700 900 1 100 Potential (mV vs Ag/AgCl) Figure 6.8 Voltammetric response for 0.2 mM SPM (forward scan) in borax buffer, pH 11, at a boronodoped microcrystalline diamond disk electrode as a function of rotational speed. The electrode was deposited with 0.5% CH1/H2 and 20 ppm Bsz. Scan rate = 0.05 V/s. Rotational velocities (rad/s) = (a) 52.2, (b) 104.7, (c) 157.1, (d) 209.4, (9) 261.8, and (f) 314.2. For this higher doped film, the current increases with increasing rotation rate, however a limiting current plateau is not observed. Similar observations were made for CAD, PUT, and SPMD. The fact that the current showed some increase with rotation rate indicates that, at this lower scan rate, the reaction rate is influenced by mass transport of the amine. Koutecky-Levich plots were constructed from the data and are shown in Figure 6.9. The plots show a dependence of reciprocal current on the reciprocal rotation velocity"2 with two regions of linearity. Interestingly, one region with a steep slope is 157 observed at fast rotation rates and a second region with a shallower slope at low rotation rates. Such behavior, although not typical, was previously observed by Treimer et al.48 for oxidation of toluene at Fe(l|I)-doped Pboz electrodes. It was attributed to the shift in the rate-controlling step in anodic oxygen-transfer reactions from that of reactant transport at low rotational velocities, to OH- generation at fast rotation velocities.48 0.08 . D SPM <9 SPMD "‘A 0'06 ” ............... g g. 0 PUT , .............................. I.“ 4: CAD ‘8 ----------- § 0.04 ' 3.4- 9 r. .14 5 * a” ,1 -1.. a = _ _.,:..._;,-:-,::::::‘.m......... 0.02 - 5:: 5“ '94:}, :3” 0.00 1 1 1 1 1 1 1 1 1 1 1 0.04 0.06 0.08 0.10 0.12 0.14 0.16 1/ (rotational velocity)"2 (rad/sec) 112 Figure 6.9 Koutecky-Levich plots for the oxidation of 0.2 mM amines at a boron-doped microcrystalline diamond disc electrode deposited with 0.5% CH1/H2 and 20 ppm Bsz. Scan rate = 0.05 Vls. Potential: 0.7 V vs Ag/AgCl Values for the number of electrons transferred (napp) and the heterogeneous rate transfer (kapp) for the amine oxidation were calculated from the slopes and intercepts, respectively. The kinetic parameters calculated at 0.7 V vs Ag/AgCl are presented in Table 6.2. Results presented in the Table 6.2 were calculated for low rotational velocities (i.e., low flux region). The apparent number of electrons napp transferred for CAD, PUT is 158 1.9, and for SPMD and SPM he,pp is higher, at 3.6 and 5.6, respectively. The napp values are close to 2 eq/mol for CAD and PUT, 4 eq/mol for SPMD, and 6 eq/mol for SPM. The reaction rates were calculated using the napp and they were found to be 6.6 i 0.1x10'3 cm/s for CAD and PUT and 7.4i 0.2x10'3 cm/s for SPMD and SPM. The measured overall rate constants contain the rate constant for the adsorption and desorption steps, and for the OH- generation, and the anodic oxygen transfer electron transfer chemical reaction. Table6.2 Kinetic Parameters for Amines Oxidation (pH11) at 20 ppm Microcrystalline Boron-Doped Dlamond iii: a + b 111““ QZmMamMe BBpH11 b _, (mV‘rad‘V’ 3‘") "GDP [9“ a ("A ) kapp (CW3) CAD 244 1.9 20.3 6.7x10'3 PUT 232 1.8 22.3 6.5x10’3 SPMD 134 3.6 9.6 7.6x10’3 SPM 90 5.6 6.4 7.3x10’3 Note: Electrode geometric area = 0.2 cm2, D=1x10'5 cmzls for CAD and PUT, D=0.8x10'5 cm2/s for SPMD and SPM. C=O.2x10’3 M, v=10x10'3 cm2/s. The kinetic parameters calculated for microcrystalline diamond are compared for other electrodes in Table 6.3. The kapp values for the polyamine oxidation at diamond are approximately an order of magnitude higher than those for the oxidation of EDA at Ni, Cu, Co, and Ag. They are smaller than the values for EA at Ag-PbOz, and similar to the value at Cu-Mn alloy electrodes. 159 Table 6.3 Kinetic Parameters for Anodic Oxidation of Amines Kinetic parameters Amine Electrode n Reference (14473310 “'9" (”V“) EDA (pH10) Cu 5.2 5.6x10-3 ‘3 EDA (pH10) Cu-Mn (95:5) 7.7 6.7x10'3 ‘3 EDA (pH10) Cu-Mn (90:10) 8.6 3.0x1o'3 ‘3 EA (pH12) Ni 4 7.0x10'5 35 diethylamlne . .5 35 EH12) Ni 6 2.4x10 EA (pH12 Cu 4 1.4 x104 35 diethylamlne -5 35 (EH12) Cu 6 5.6 x10 EA (pH12 Co 4 9.0 x10'6 35 diethylamlne -5 35 19*" 2) Co 6 <10 EA (pH12) Ag 4 1.4 x10“t 35 diethylamlne -5 234 (pH1 2) Ag 4 3.6x10 EA (pH10) Ag-PbOz 1.9 1.3x10‘2 “5557 cm (pH11) Diamond 1.8 6.7x10‘3 This work PUT(pi-l11) Diamond 1.8 6.5x1o'3 This work SPMD(pH11) Diamond 3.6 7.6x10‘3 This work SPM (pH11) Diamond 5.6 7.3x1o'3 This work Secondary amine (e.g., diethylamine) oxidation studies at Ag revealed an overall mechanism involving transfer of 4e' and 3H+ to form a nitrile, 29' and 2H+ to form an aldehyde, and 29' and 1 H“ to form an alcohol and alkene (Eqs. 9-13 in the introduction of this chapter).234 Our studies show the involvement of 49' and 4 H”, and 6e' and 6H+ for SPMD and SPM, respectively. Fleischmann et al. proposed transfer of 69' for secondary amine oxndation. 160 The napp values for the polyamines at diamond are smaller when compared with the values for EA at Ni, Cu, Cu-Mn, Co, and Ag. For the transition-metal and Ag-PbOz electrodes, the primary amine oxidation mechanism involves 4e' and 4H” with the nitrile formation, and for secondary amine 29' and 2 H” to form a primary amine. The mechanism for Ag-Pb02 suggests involvement of hydroxyl radicals and overall transfer of 2e" and 2H*. 6.3. Conclusions The investigations presented in this chapter suggest the following: (i) Amine oxidation can be classified as mechanism, where electron transfer is followed by chemical process (EC), with involvement of hydroxyl radical in chemical step. Fast reaction rates (i.e., high currents) are observed only for diamond electrodes at which the discharge of water has appreciable rate. (ii) The oxidation potentials for primary and secondary amines vary at diamond. Secondary amine groups oxidize at less positive potentials than primary amine groups, although the rate of the oxidation is much lower for the former. (iii) A primary amine is likely formed during the oxidation of a secondary amine remaining adsorbed at the surface boron sites, and available for further oxidation at higher potentials. (iv) Ep°x and Ep/2 values for all the polyamines shifted negatively with increasing pH. This observation indicates that electron transfer is accompanied by proton transfer. Plots of Ep/2 vs pH had slopes of ca. -60 mV/pH unit. This indicates that an even number of protons and electrons are involved in oxidation reaction. 161 (v) The adsorption or coordination of the amines follows a Langmuir-type isotherm. (vi) The oxidation peak current for the amines at boron-doped microcrystalline films deposited with 0.5% CH1/H2 and 1 ppm BQHG was independent of the rotation velocity in rotating disc voltammetric measurements at a scan rate of 0.1 V/s. This indicates that either the adsorption kinetics are rate limiting or that the current is limited by the number of available surface boron sites. (vii) The Koutecky-Levich plots constructed for the amine oxidation at boron- doped microcrystalline films deposited with 0.5% CH1/H2 and 20 ppm 8sz show a peak current dependence with the rotational velocity, with two regions of linearity. It was attributed to the shift in the rate-controlling step in 0- transfer reactions from that of reactant transport at low rotational velocities, to OH- generation at fast rotational velocities. (viii) The apparent number of electrons nalpp transferred, at pH 11, during the oxidation of CAD and PUT was 2. For SPMD and SPM, new is 4 and 6, respectively. The reaction rates were calculated using he,pp and were found to be 6.6 i 0.1x10'3 cm/s for CAD and PUT and 7.421: 0.2x10'3 cm/s for SPMD and SPM. Based on these findings we propose the mechanism of amine oxidation at boron- doped diamond electrode (Eq.26-28) CI'h'CHZ'NHZ —) (CH3'CH2'+NH2)ads + e- (26) H20 —) (OOH)ads + e' + H+ (27) (CHa-CH2-+NH2)ads + (OOH)ads —-) CHa-CHO + NH3 + Ht (28) 162 it is our opinion that at highly doped diamond films, chemical process of radical formation and oxygen transfer to the amine molecule is the limiting step. Adsorption process at boron sites itself is not limiting, but the number of those sites can influence the overall kinetics of oxidation. This can be observed for low doped films for which reaction is kinetically limited by surface sites available. The results presented in this chapter and Chapter 5 strongly support it. it is anticipated that the step in Eq.28 proceeds by the surface equivalent of a SN2-type mechanism. The proposed mechanism is presented in Figure 6.10. c/ ‘ / -H+,-2' H H 171‘! ll || ,ov |Ol H H H S H H G) \o/ H \N/—B H9141 H —_ __. "v, ‘2. I \C/ 1 ~— :1 a —> H / R \ -' R l \o n H 9"” 0 ll H \/ H / PH R H” ‘ H? a H’ H H/ B B +OHO -H20 __, 0/ ’ C.” R—c/ || HP/ \ e / H H0 R Figure 6.10 Proposed surface SN2-type mechanism of aliphatic amine oxidation at boron-doped diamond electrode 163 . .57 . 5455 In the mechanism a non- Such mechanism was proposed by Johnson. et al. bonded electron pair on oOHads interacts with the anti-bonding orbital of the (Jr-carbon atom. This is followed by oxidative dissociation of the H-atom from 00H, and cleavage of the C-N bond with formation of NH3 (for primary group) and primary amine (for secondary groups). It was stated 54’55’57 that such surface mechanism requires the proximate location of adsorption sites for 00H and NH2, as well as significant vibrational motions of these adsorbed species along an axis passing through the O and C atoms. Although the amine oxidation reaction products were not examined, it is our supposition that the products of primary amine oxidation are aldehyde and ammonia. Two electron and two proton process is proposed and presented below. ~2e' H2N(CH2 )4 NHz + H20 3:19 H2N(CH2)3CH0 + NH3 RCHzCHO pH 11 —-> No oxidation NH3 Oxidation of polyamines involves secondary amine group. The SPMD oxidation products may be: 4-aminobutanal (II) and diaminopropane (DAP). SPERRLIDINE (1) -2e' H2N(CH2)3NH2 (II) H2N(CH2)3NH(CH2)4NH2 + H20 -—2—H+-> and H2N(CH2)3CHO -2e' H2N(CH2)3NH2 + H20 Y H2N(CH2)2CH0 + NH3 (11) ' i 164 The DAP can be further oxidized at higher potential, resulting 3-aminopropanal and ammonia. It was found during investigations that ammonia and aldehydes were not electroactive in the studied potential range (0.2-1.2 V vs. Ag/AgCI). SPMD oxidation process involves 4 e'. The mechanism of SPM oxidation is similar. Overall process may involve both secondary amino groups and overall transfer of 6 e'. SPERMINE H2N(CH2)3NH2 (I) _28. (II) H2N(CH2)3NH(CH2)4NH(CH2)3NH2 3; and + H20 OHC(CH2)3NH(CH2)3NH2 (III) -2 - H2N(CH2)3NH2 + H20 3;» NH2(CH2)2CHO + NH; (11) ' + -2e' 0HC(CH2)3NH(CH2)3NH2 ——> OHC(CH2)3NH2 + OHC (CH2)2NH2 -2H+ (III) + H20 165 Chapter 7 FLOW INJECTION ANALYSIS INVESTIGATIONS OF ALIPHATIC POLYAMIN E OXIDATION AT BORON- DOPED DIAMOND THIN-FILM ELECTRODES 7.1. Introduction A key finding from the work presented in previous chapters is the fact that aliphatic polyamines can be electrooxidized at boron-doped microcrystalline and nanocrystalline diamond films.169388359391'296 Voltammetry was employed to study the diamond electrode response for amine oxidation. The objective for the work reported on presently was to use diamond films for the oxidative detection of aliphatic amine in a flowing system, (i.e., in flow injection analysis (FIA) and liquid chromatography (LC».291,296 FIA was originally designed for analysis of homogeneous samples and the 297'” The FIA was first developed almost 30 development of single component assays. years ago, and since that time the perception of it has changed. Chromatogaphic techniques are aimed to determine many analytes in a single injected sample, by resolving them through separation. FIA is aimed at detection of a single (or several) 166 analyte(s) within a single dispersion zone.301 Selective chemical or enzymatic conversion is used to minimize matrix effect. In FIA, a sample is injected into a flowing non-segmented stream of the mobile phase. FlA serves as a convenient means for the analysis of small volumes (uL) of sample.301 The injected sample forms a zone that disperses on its way to the detector and the magnitude of the resulting readout peak reflects the concentration of the injected analyte. In FIA or LC with electrochemical detection, the supporting electrolyte of the system acts as a mobile phase. Since the supporting electrolyte is in motion over the electrode surface, mass transfer is assisted by forced convection. Hence, all electrochemical techniques applied for LC or FIA detection are classified as hydrodynamic techniques.303 The most commonly used electrochemical detectors in HA or LC are thin-layer cells, in which the working electrode is positioned in a thin channel through which the mobile phase flows.303 Schematic of the cell used in these studies was presented in Chapter 2. The thin-layer detector in these studies was operated at constant potential (amperometric detection) so—called dc amperometry.303 The current or charge passed was measured as a function of time. Usually, the active volume of the cell is about 0.1-1uL, and typically about 5% of the analyte is electrolyzed as it is swept past the electrode.303 A well-designed FIA system yields the detector readout within 15 s from the moment of sample injection. Approximately the same time is required for the dispersed sample to clear out of the detector, so that the next sample can be injected.299 Considering it, a big number of samples, in a short period of time, can be analyzed and detector tested. Compared with cyclic voltammetric techniques in which the potential is scanned, the amperometry provides a major advantage. By operating at a fixed potential, the charging current is minimized. 167 Flow injection analysis with amperometric detection was used to evaluate the polycrystalline diamond electrode performance for detection of aliphatic amines in the hydrodynamic mode. The electrode performance was tested as a function of the linear dynamic range, limit of quantitation, response variability, and response stability. 7.2. Results and Discussion 7.2.1. Hydrodynamic Voltammetric i-E Curves Hydrodynamic background i-E curves in carbonate buffer pH 10, for microcrystalline boron-doped diamond thin-film electrodes, deposited using 0.33%, 0.50%, and 0.67% CH4/H2 ratios, are shown in Figure 7.1. 0-06 - -O- 0.67 % CHJH; f)- 0.50 % cm, 2 4:} 033 % CH4/Hz 5 0.04 H 5 l- in 5 o 0.02 0.00 . . . . 750 800 850 900 950 Potential (mV vs. Ag/AgCl) Figure 7.1 FIA-EC hydrodynamic voltammetric background i-E curves for a microcrystalline diamond film deposited using 0.33%, 0.50% and 0.67% CH4/H2 ratio in carbonate buffer, pH 10. Flow rate = 1.0 mUmin. 168 The voltammogram was obtained by incrementally changing the detection potential. Once a stable baseline was observed at a given potential, the background and noise signal were measured and higher potential was applied. The measurements were made at a flow rate of 1.0 mUmin. The background current increases with increasing CH4/H2 ratio. For example, the background current at 815 mV was 3, 6, and 9 nA, respectively for the 0.33 %, 0.5 % and 0.67 % CH4/H2 film. Three films for each of the type of electrode were evaluated. The onset potential for oxygen evolution is least positive for the 0.67 % CH4/H2 film. Table 7.1 shows comparison of the background current and noise signal magnitude for three electrodes at a given potential. Table 7.1 FIA-EC Background and Noise Signal for Microcrystalline Dlamond Film In Carbonate Buffer pH10 Applied Potential vs AgIAgCl Diamond fllm deposited from 815 mV 870 mV °"“"* B‘Sfifiziii“ N... N... 0.33% 2-10 nA 230 pA 10-20 nA 200 pA 0.50% 8-12 nA 190 pA 12-25 nA 200 pA 0.67% 10-20 nA 200 pA 30-60 nA 400 pA The higher background current for films deposited at higher CH4/H2 ratio is associated with a higher faction of exposed grain boundary where non-diamond carbon exists. Non- diamond spz-bonded carbon exhibits a lower overpotential for oxygen evolution than does diamond and is more reactive (i.e., susceptible to oxidation) than does diamond. The carbon in the grain boundary is also a source of charge carriers, along with the substitutional boron dopant atoms. This leads to a greater excess surface charge at a 169 given potential and a larger double-layer charging current. The peak to peak noise for all the tested films ranged from ZOO-230 pA at 815 mV. At a more positive potential of 870 mV, the noise increased to ZOO-400 pA. As was the case in cyclic voltammetric measurements, nanocrystalline films possessed slightly higher background current than microcrystalline diamond films. Similar results were observed in FIA, when hydrodynamic current was measured in the same electrolyte at the same potential. Nanocrystalline films were deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio with 1, 10 and 20 ppm of added diborane. The measurements were made at a flow rate of 1.0 mL/min. The background current varied only slightly at these electrodes, and was generally about 10-30 nA at 730 mV. The onset potential for oxygen evolution at pH 11 was similar for all the films. The peakito peak noise ranged from 150-250 pA at 730 mV. The hydrodynamic voltammetric background current for boron-doped nanocrystalline diamond films, in borax buffer, pH 11, are shown in Figure 7.2. Table 7.2 shows comparison of background current and noise signal magnitude for nanocrystalline electrodes at a given potential. Table 7.2 FIA-EC Background and Noise Signal for Nanocrystalline Dlamond Film in Borax Buffer pH 11. Applied Potential vs AgIAgCI Diborane (82m) concentration 675 mV 730 mV 88:33.21?“ N... N... 1 ppm 3-5 nA 120 pA 10-15 nA 150 pA 10 ppm 3-6 M 200 pA 20-30 nA 150 pA 20 ppm 5-7 nA 150 pA 25-30 nA 250 pA 170 30 25- _ O 10 PP!“ BzHo 20 -- D 1 ppm ano h 20 m B H6 15 A PP 2 Background current (nA) 620 640 660 680 700 720 740 Applied Potential (mV vs. AgIAgCI) Figure 7.2 FIA-EC hydrodynamic voltammetric background i-E curves for boron-doped nanocrystalline diamond films deposited with 1 ppm, 10 ppm, and 20 ppm of added 8sz in borax buffer, pH 11. Flow rate = 1.0 mUmin. At more positive potentials (>740 mV) in this medium, the background current becomes unstable due to the oxygen evolution reaction. The background current for “as deposited” nanocrystalline and microcrystalline films stabilizes quickly after detector turn-on, reaching a constant value in about 10 min. The background current for films in use (conditioned), stabilizes in less than 2 minutes, after detector tum-on. A rapid stabilization time and a low and stable background current are characteristic features of 163,188 diamond electrodes. Figure 7.3 presents a background current-time plot for a diamond thin-film electrode in borax buffer, pH 11. 171 120 pA *5 i5 8 c |-—-| g 2 minutes a . .. e0 %5 3 200 nA Time '——.-' 10 minutes Figure 7.3 Background current-time profile in HA after detector turn-on in borax buffer, pH 11, for a nanocrystalline diamond film deposited from 1/94/5 CHJAr/Hz (v/v) ratio and 1 ppm Bsz. Detection potential = +695 mV vs AgIAgCI. Electrode area = 0.08 cmz. Flow rate :1 mL/min. Figure 7.4 A and B shows hydrodynamic voltammetric i-E curves for 20 uL injections of (A) 1.0 mM SPMD and (B) CAD at boron-doped microcrystalline diamond thin-film electrodes, deposited using CH4/H2 gas source ratio of 0.33%, 0.50%, and 0.67%. The hydrodynamic voltammogram was obtained by making repetitive injections of each amine while incrementally increasing the detection potential. Each datum shown represents an average of at least six injections. Error bars are within the size of the marker. Generally, the largest current for CAD and SPMD is observed for the 0.5% CH4/H2 film. Similar results were observed for PUT and SPM, with the largest current obsrved for SPM. The total current response for the polyamines is larger by about a factor of ca. 4 than that for the diamines. For instance, for a 0.5% CH4/H2 film, at 850 172 mV, the current for 1 mM SPMD is 4.5 uA and for 1 mM CAD is 1.0 uA. The larger current for SPMD, compared to CAD, is an observation that has been made previously in cyclic voltammetric studies for highly doped diamond films and reflects a larger number of electrons being transferred per equivalent of analyte. Well-defined, sigmoidally—shaped hydrodynamic i-E profiles are not observed for either analyte. This is because the oxidation reaction is not limited by mass transport but is strongly influenced by the generation and transport of reactive OH- and the adsorption/desorption of the reactant/product. The thin-layer electrochemical flow cell used did give well-defined sigmoidal hydrodynamic voltammograms for other less mechanistically complicated oxidation reactions (e.g., azide oxidation).286 Figure 7.4 C and D show plots of the signal-to-background (SIB) ratio, as a function of the applied potential for SPMD (C) and CAD (D), at the same three diamond films. The S/B ratio was calculated as lm. - lbakgmund / lbackgmnd. Such plots are particularly useful for determining the optimum detection potential in situations where a mass- transport limited process is not observed in the hydrodynamic voltammogram. The largest SIB ratios are seen for the 0.5% CH4/H2 film. The maximum ratio was observed at about 820-840 mV. Similar observations were made for PUT and SPM. The boron doping level in all three films discussedin the previous figure was approximately the same, ~1x10‘9 B/cma. The doping level was estimated in boron nuclear reaction analysis of other films deposited in a similar manner. Before performing FIA of the polyamines, all the electrodes were tested using other redox systems (i.e., Fe(CN)6'3’“‘, Fiu(NH3)5”2’"3) to ensure that the electrode possessed good responsiveness and adequate electrical conductivity. The tested couples showed near-reversible behavior, therefore, one can conclude that the differences in response toward polyamine 173 oxidation are composition and not electronic property variations. 12.0 10.0 . Current (nA) 2.0 _ 00 250 200- 150 SIB ratio 50 8.0 . 6.0 . 4.0 . '750 800 850 900 950 '750 100. 750 800 850 900 950 750 800 850 900 950 due to differences in the electrode surface structure and chemical 259 3.0 2.5 . 2.0 . 1.5 . 1.0 . Potential (mV vs. AgIAgCI) -O- 0.50% cm; £1- 0.67 % CHJH; i} 033 % CHJH2 Figure 7.4 Hydrodynamic voltammetric i-E curves for 20 uL injections of (A) 1.0 mM SPMD and (B) 1.0 mM CAD. The carrier solution was 0.1M NaClO4 + 0.01 carbonate buffer, pH 10. Plots of the SIB ratio versus the applied potential are shown for (C) SPMD and (D) CAD. Flow rate :10 mUmin. 174 It was disscussed in Chapter 5 that the amine oxidation response depends on the sp2-bonded non-diamond carbon impurity and boron concentration. The oxidation current increases with increasing boron concentration in the film. Similar results were observed in FIA-EC. Figure 7.5 shows hydrodynamic voltammetric i-E curves for 20 uL injections of 0.1 mM PUT, CAD, SPMD, and SPM obtained at nanocrystalline diamond thin-film electrodes deposited using a 1/94/5 CHJAr/Hz source gas ratio with (A) 1 ppm Bsz, (B) 10 ppm B2H6, and (C) 20 ppm of added B2H6. Each datum shown represents an average of four injections. Error bars are within the size of the marker. Relative standard deviations of the response were in the range of 0.4-1.0 %. As was the case for microcrystalline films, well-defined, sigmoidal hydrodynamic i-E curves were not observed for any analyte. The largest oxidation currents are seen for the highest doping level. Figure 7.5 D-F show plots of the signal-to-background (SIB) ratio, as a function of the applied potential for 0.1 mM amines at the same electrodes. The oxidation currents and SIB ratios were similar for PUT and CAD, and SPMD and SPM. The SIB ratios for all four amines increased with increasing boron doping level in the film. For instance at 720 mV, the current for 0.1 mM SPMD was 270, 384, and 1678 nA at 1 ppm, 10 ppm, and 20 ppm 82H6 films, respectively. For 0.1 mM PUT, the current was 44, 57, and 458 nA for thet 1 ppm, 10 ppm, and 20 ppm B2H5 films, respectively. The SIB ratio for the polyamines at 720 mV was larger by about a factor of 7 than that for the diamines at all three nanocrystalline films. The optimum detection potential for all polyamines, regardless of the film was determined to be about 700-720 mV vs AgIAgCI. 175 2500 A —A— PUT 2000 A SPMD E 3- g E —<>— CAD 5 1000 f - 0 background 500 r 0 620 650 680 710 740 2500 _ E 2000 L B E» 1500 L f— F 5 l- O 1000 :— 500 L 0 : 620 650 680 710 740 2500 g 2000 *5 1500 2 ‘5 1000 U 500 0 i 620 650 680 710 740 Potential (mV vs AgIAgCI) Figure 7.5 Hydrodynamic voltammetric i-E curves for 20 uL injections of 0.1 mM amines. The carrier solution was 0.1M NaCl + 0.01 borax buffer, pH 11. Nanocrystalline diamond films were deposited from a 1/94/5 CH4/Ar/H2 (v/v) ratio with (A) 1 ppm, (B) 10 ppm, and (C) 20 ppm of added Bsz. Flow rate=1.0 mUmin. 176 200 E D —A— P‘” .3 150 _ _E]._ SPMD _ M E C ‘0— 2:1) 6 100 - + m E 0 620 640 660 680 700 720 740 200 _ : E 150 j c _ 'g _ I- 100 :- 5. : 50 _— fl 0 _ 620 640 660 680 720 740 200 .2 150 g 100 V) 50 0 . . . . 620 640 660 680 700 720 740 Potential (mV vs AgIAgCI) Figure 7.5 Plots of the S/B ratio versus the applied potential for 0.1 mM amines. The carrier solution was 0.1M NaCl + 0.01 borax buffer, pH 11. Nanocrystalline diamond film were deposited from a 1/94/5 CHJAr/Hz (v/v) ratio with (D) 1 ppm, (E) 10 ppm, and (F) 20 ppm of added 3%. Flow rate = 1.0 mUmin. 177 A lower onset potential for SPMD and SPM oxidation was observed which is the most pronounced in Figures 7.5 C and F for the 20 ppm B2H6 film. Table 7.3 presents a summary of numerical values of the current as a function of the potential. It can be seen that the oxidation current for SPMD and SPM is much greater than that for CAD and PUT at lower potentials (550 mV-620 mV). For example at 550 mV, the current for SPMD, SPM, PUT, and CAD is 17, 33, 3 and 3 nA, respectively. Table 7.3 Amperometric Data Obtained During FIA of Aliphatic Amines at pH 11 Current response (nA) Applied Potential (mV vs. AgIAgCI) PUT CAD SPMD SPM 550 3.2 :i: 0.1 3.0 i 0.1 17.3 i- 0.3 32.5 i 0.4 ‘ 570 4.4 i 0.1 4.0 i 0.1 33.8 i: 0.5 57.4 :t 0.8 595 4.5 i 0.1 4.0 :i: 0.1 41.9 i 0.4 71.0 :i: 1.3 605 4.8 i 0.2 4.1 i 0.1 53.5 i 0.9 95.8 i: 1.2 615 5.4 i 0.2 4.3 i 0.1 79.5 i 1.7 133.9 i 2.1 620 6.3 i 0.1 5.9 :t 0.2 145.8 i 2.8 171.4 i 3.6 640 9.4i 0.3 7.3 i 0.1 235.6 :1: 2.5 297.4 :1: 3.8 655 19.1 i 0.3 11.6 i 0.2 375.8 1 3.8 464.3 1 2.2 675 42.1 i 0.8 25.2 i 0.8 604.8 i- 2.4 700.7 i 2.3 700 123.6 i 2.2 81.8 :i: 1.2 1018.2 i 2.5 1084.5 :1: 6.4 710 294.4 i 4.2 240.3 i 2.9 1413.5 1 2.8 1470.7 :1: 5.4 720 457.9 i 3.3 395.3 i 2.5 1678.3 i 3.1 1654.5 i 9.6 730 919.1 1 3.9 906.2 i 4.1 2324.4 i30.3 2192.8 i 2.1 178 Another observation is that the signal for SPM is approximately twice as big as that of SPMD up to 615 mV. The cyclic voltammetric data shown in Chapter 6, for 1 mM SPM at pH 7.2, showed an oxidation wave at 940 mV with a peak current of 11 uA. The lowest pKa value for SPM is 7.9 and corresponds to one of the two secondary amine groups. 189 At this pH only this secondary amino group is in partially deprotonated form, and thus available for oxidation. This result indicates that oxidation of secondary amine occurs at lower potential. It is harder to oxidize primary amine group, hence, the oxidation potential is more positive. This would also suggests, that the higher currents, for SPMD and SPM, at low potentials are due to oxidation of secondary amino groups. Furthermore, the current for SPM is approximately twice as large as that for SPMD, presumably because SPM possesses two secondary amine groups while SPMD has only one. Hydrodynamic voltammogramms for 0.1 mM CAD and 0.1 mM SPMD were obtained as a function of the carrier solution pH. Figure 7.6 shows the SPMD/CAD current ratio at pH 9, 10.6, and 11.4. The first observation is that the optimum oxidation potential shifts toward less positive values as the pH increases. This is due to a negative shift in the onset potential for the background current (i.e., oxygen evolution) observed with increasing pH. The second observation is that the magnitude of the SPMD/CAD current ratio (as well the SPM/PUT ratio) increases with increasing pH. For example, at pH 9 and 10 the current for SPMD is about 4 times larger than that for CAD. However, at pH 10.6 and 11.4 the current for SPMD is 8 and 17 times larger than that for CAD. This is associated with higher number of electrons transferred per equivalent for SPMD due to oxidation of both primary (from the SPMD and the reaction products) and secondary (from the SPMD only) amine group, while for CAD only one primary group is involved in reaction. The current for SPMD is becoming much larger than that for CAD or PUT at 179 higher pH because the primary amine groups of polyamines (i.e., SPMD and SPM) have high pK. values, and only at high pH they are available for oxidation (i.e., in a unprotonated form). 20 O '5 215_ ‘E 2 l-r 310 Q i- <5 Q . 25. U) 0 500 Potential (mV vs AgIAgCI) Figure 7.6 Variation of the SPMD/CAD current ratio, as a function of pH of the mobile phase. The carrier solution was 0.01 borax buffer/0.1 M NaCl. Microcrystalline diamond films deposited with 0.5% CH4IH2 source gas mixture. Flow rate = 1.0 mLImin. The effect of flow rate on the response magnitude and stability for CAD, PUT, SPMD and SPM was investigated. In the flow rate range between 0.3 and 1.5 mL/min, the background current was insensitive to the flow rate changes at the detection potential of 665 mV (pH 11). The amperometric response, however, for all the amines 180 was inversely related to the flow rate. The current for 100 pM amine decreased by 30-38 % when increasing the flow rate from 0.5 to 1.5 mLImin. The reproducibility was poorest at the highest and lowest flow rates of 1.5 and 0.3 mUmin (2-4% based on 5 injections). The best reproducibility was obtained at 0.8-1 mUmin (0.5-0.8 % also based on 5 injections). These observations can be explained by slow adsorption of the reactants at the surface or desorption of the reaction product(s). 7.2.2. Calibration Curves The working curves for each amine were obtained in carbonate buffer, pH 10, as a function of the injected analyte concentration. Plots of log peak height versus log concentration for (A) SPMD and (B) CAD are shown in Figure 7.7. The calibration curves are based on 7 concentrations between 0.32 uM and 1000 uM or 1 uM and 1000 uM. The ilnjected volume was 20 uL and flow rate was 1.0 mUmin. The current response at pH 10 increases linearly with the concentration for all four amines. The linear dynamic range for all four polyamines, at all three films, is at least three orders of magnitude ranging from 1.0 to 1000 (M (r2 > 0.99). A slightly wider dynamic range, down to 0.32 uM was seen for SPM at 0.5% CH4/H2. The nominal y-axis intercept of the calibration plots for all four analytes was less than 0.02 pA. The average limits of detection for CAD and PUT are 0.8:l: 0.05 (M and for SPMD and SPM are 0.30 i 0.1 uM. The lowest LODs were obtained for films deposited using 0.5% CH4IH2. The working curves for each amine were also obtained in borax buffer, pH 11, using microcrystalline films deposited from a 0.50% CHJH; ratio with 10 ppm of Bsz. Table 7.4 presents summary for the FIA-EC data. 181 1 .. A 0 - -1 - .— -2 L ti _ Cl 0.33% CM; 1," -3 _ i} 0.67% cm; 5 . O 0.50% CH4IH ‘3: .4 r 1 . l . J . 5 -6.50 -5.50 4.50 -350 -2.50 5 Log Concentration [M] 0: CB O B on o _ O r-J r -1 _ -2 l. Cl 0.33% CPL/Hz -3 _ i} 0.67% cm; - O 0.50% CH4/Hz .4 . 1 I 1 . L . -6.50 -550 4.50 -3.50 -2.50 Log Concentration [M] Figure 7.7 Calibration curves obtained for (A) SPMD and (B) CAD at boron-doped microcrystalline diamond films deposited withdifferent CH4/H2 source gas ratios. The carrier solution was 0.01M carbonate buffer/0.1 M NaCIO4, pH 10. injected volume = 20 uL. Flow rate :10 mUmin. Applied potential = +700 mV 182 Table 7.4 FIA-EC Data for Aliphatic Polyamines Detection at Microcrystalline Dlamond Thin-Film Electrode ln Borax Buffer, pH 11.2 Linear Dynamic Sensitivity Polyamine R an ge LOD at SIN23 (uNmM) Cadaverine 0.5 M-1000 uM - (CAD) ( = 0.9985) 0'4 ”M 0'53 Putrescine 0.5 uM-1000 uM (PUT) (r2: 0.9992) 0'4”” 0'61 Spermidine 0.5 uM-1000 uM (SPMD) (r2: 0.9993) 0'1“” 2'97 Spermlne 0.5 uM-1000 uM NotezThe calibration curves are based on 7 concentrations between 0.5 and1000 uM. injected volume=20 uL.Flow rate=1.0 mLImin. The current response and peak charge both varied linearly with the concentration. The linear dynamic range for all four polyamines ranged from 0.5 to 1000 uM (r2 > 0.99). The limits of detection were lower at pH 11 than at pH10. For example, the average limits of detection for CAD and PUT are about 0.4 uM, and for SPMD and SPM are 0.1 (M. The mass detection limits are 0.8 ng for CAD, 0.7 ng for PUT, 0.3 ng for SPMD, and 0.4 ng for SPM. The calibration curves were also obtained using boron-doped nanocrystalline diamond films, deposited from a 1/94/5 CH4IAr/H2 (v/v) source gas ratio with 20 ppm of added BzHe. Working curves were built at pH 11 for 7 different amines (mono-, di- and polyamines). The results are summarized in Table 7.5, and Figure 7.8 shows the plots of peak heights versus concentration. 183 12000 5 § Current (nA) .. E ‘99 0.0000 0.0002 0.0004fii/ 0.0006 0.0008 0.0010 Concentration (M) Figure 7.8 Calibration curves for aliphatic amines at a boron-doped nanocrystalline diamond electrode deposited from a 1/94/5 CHJAr/Hz (v/v) source gas ratio with 20 ppm B2H6. Injected volume = 20uL. Flow rate = 1.0 mlJmin. The peak current and charge varied linearly with the concentration for all the amines. The LCD for MA is 10uM. The LODs for EDA and DAP are 100 times lower at 0.1uM. The same 0.1 uM LODs are achieved for SPMD and SPM. CAD and PUT would be detectable at 0.3 uM. The mass detection limits are 6 ng for MA, 0.12 ng for EDA, 0.15 ng for DAP, 0.6 ng for CAD, 0.5 ng for PUT, 0.3 ng for SPMD and 0.4 ng for SPM. 184 Table 7.5 FIA-EC Data for Aliphatic Polyamines at a Nanocrystalline Diamond Thin Film Deposited from a 1I9415 CHJAr/Hz source gas ratio with 20 ppm 32H... Linear Dynamic LOD at Mass Sensitivity Polyamine Range SIN23 Detected uAImM methylamine 10 uM-i 000 uM 10 11M (MA) (r2: 0.9997) (300 ppb) 6'2 "9 0'13 ethyleggdAlamlne 0.1( =M01 S090;)sjitM ((3):!) 0.12 n g 6.37 “was?“ “193.283.?” .2232“... °:.:x:.';° 09:09:52)” (3.3.9. my ”"2313“ 0.9:yéigggapm (3035;) 0'53 "9 1'3 $753113? 0'1 3675353?” (3915):?» 0'29 ”9 7'7 3'81»??? 018:“ 0T$31M (30153:) 0'40 "9 10'5 NotezThe calibration curves are based on 7 concentrations between 0.1 and 1000 uM. Injected volume = 20 uL. Flow rate = 1.0 mLImin. The values of sensitivity were analyzed for both types of electrodes: boron-doped nanocrystalline (20 ppm) and microcrystalline (10 ppm). Interestingly, there is a vast difference in the sensitivity factor for each of the polyamines at microcrystalline film. The average sensitivities for CAD and PUT in borax buffer, pH 11, measured at microcrystalline film; are similar, (0.5-0.6 pA/mM see Table 7.4). The sensitivity for SPMD is ~3 uA/mM, and for SPM is 4.7 pA/mM. At pH 11 the response factors at nanocrystalline film deposited with 20 ppm B2H6 are about twice larger than that for the microcrystalline film deposited from gas mixture containing 10 ppm Bsz. They are 1.2, 1.3, 7.7, 10.5 uA/mM for CAD, PUT, SPMD, and SPM, respectively. 185 The response factor for nanocrystalline film at pH 11 is the highest for amines: SPMD (3,4), SPMD (3,3), SPMD (2,2), SPM, DAP, EDA, DEP. The lowest is observed for monoamines: MA, PA, EA. For diamine as CAD, PUT, HA, DA sensitivity factor was somewhat in between the highest and the lowest value. Figure 7.9 shows FIA-EC responses for a multiple injections of 0.1uM amine (a) CAD, (b) EDA, (c) DAP, (d) DA, (e) DEP, (f) SPM, (g) PUT, and (h) SPMD in 0.01 borax buffer/ 0.1 M NaCI, pH11.2. For the same concentration of the amine injected, the diamond responses are dramatically different. The oxidation current obtained at pH11.2 for the different amines, at three different diamond electrodes, was plotted versus the first pK,a value of each amine. ; h 125: f 3 c 100.: e 9 75% fl . E3 1 5 50‘: b 25{ a d g ii-il__. lilLLLit ill. Time — 10 minutes Figure 7.9 FIA-EC responses for a series of 10 uL injections of 0.1uM amine: (a) CAD, (b) EDA, (c) DAP, (cl) DA, (e) DEP, (f) SPM, (g) PUT, and (h) SPMD. The carrier solution was 0.01 borax buffer/0.1 M NaCl, pH 11.2. Detection potential=+ 660 mV. 186 (81 200 - h g A . £81 150 . f i (83 100 . i 50 . A ed L @ [1,1,5 3‘ o 1 4 - . . E. *5 6.5 75 8.5 95 10.5 3 :32 g 600 1(8) 3 B U . h @f g . .9 . 185 *5 400 I 1" 5 : a 200 - e i 89" . 0 l 1 A l u 65 75 8.5 95 105 250 US$135! c 200 .- ' (85 r1399 150 . j 100 . 50 . (2% d i o . 1 - . . . . g} 185 7.5 8.0 8.5 9.0 9.5 10.0 FirstpKavalue Figure 7.10 Relationship between the FIA peak current and first pK,a for (a) MA, (b) EA, (c) PA, (d) CAD, (e) PUT, (f) SPMD, (g) SPM, (h) EDA, (i) HA. (1) DAP, (k) SPMD (2,2), (l) SPMD (3,3). Data collected at two microcrystalline films for 0.1 mM amines (Fig.A and C), and a nanocrystalline film for 0.5 mM amines (Fig. B). 187 Figure 7.10 shows relation between the current and first pKa at (A) and (C) microcrystalline films for 0.1 mM amines, and (B) nanocrystalline film for 0.5 mM amines. A good relationship was observed for all amines, suggesting that for analytes with the lowest dissociation constant, the highest electrode response is observed. It was observed for microcrystalline and nanocrystalline diamond films. The discrepancies in the electrode response may be partially explained based on differences in fraction of amine present in unprotonated form, available for oxidation. For instance, at pH 11, there will be 60, 65, 61, 91, and 92 % of deprotonated MA, PUT, CAD, DAP, and EDA, respectively. For polyamines it will be even more complicated, because both, primary and secondary amines will be available for oxidation. The FIA results show, also that the amine oxidation current decreases With increasing molecular size, within the same classification group. For example, for monoamines (MA, EA, and PA) the amperometric response for a 0.1 mM concentration was 11.5 .t 0.3, 4.7 i 0.3, 4.5 i 0.4 nA, respectively. Even more significant differences were observed for diamines. The signal was 231.3 i 1.2, 26.3 :i: 0.1, and 25.1 i 0.1 nA for EDA, PUT and CAD. Another possible explanation for observed differences in sensitivity factors may be related to steric differences of amine (i.e. amine orientation, nature and number of amino-group). Different energies required for amine-boron complex formation, as well as stability of these complexes. The nature and stability of the boron-amine complexes was not studied, though and will remain for future considerations. 7.2.3. Response Variability and Stability A major weakness of the assay in the preliminary studies we conducted was the unacceptably large response variability of 15-25% at electrodes deposited from a 1% CH.,/H2.188’259 We also observed, that when injections of the aliphatic polyamine were 188 made with high frequency (every 1 minute), the amperometric response decreased by as much as 30—50% during the initial 10 injections. Thereafter, the response stabilized. The magnitude of the attenuation tended to be similar for each of the polyamines. it was generally much less at a 0.50% than at a 1% or 0.33% CHJHZ film. Figure 7.11 shows an extreme example of this at as “deposited" 0.33% CHJHZ film. P 00 1 a: Oxidation Current [11A] .° .5 j 9 N I Q G j— )- P- L. )._ P )— i' Injection number Figure 7.11 FIA-EC responses for a multiple injections of 1.0 mM CAD at as “deposited” 0.33% CH4IH2 diamond film. The carrier solution was carbonate buffer pH10. Detection potential=+ 810 mV. injection volume=20 uL. Flow rate=1.0 mUmin. The response for 1.0 mM CAD is attenuated during the first eight injections by about 60%. The peak height variability for all 46 injections is 18%. However, after the first eight injections the response stabilizes. Thereafter, the peak height variability is 5.6%. 189 The response variability at all types of diamond films was significantly reduced simply by adding a 3-6 min delay period between injections. It is supposed that this time allows for desorption of the reaction product from surface boron-sites, making these sites available for adsorption/coordination of new reactant molecules. Nominally at the films deposited from a 0.5% CHJHa ratio the response variability ranges from 2 to 8% for all injections made with a 3—6 min delay. The largest variability of 8% is seen for PUT. Films deposited from 0.33% and 0.67% CH4/H2 ratios show greater variability, but well- controlled delay times between injections were not used to obtain the presented data. With controlled delay periods of 3-6 minutes the variability was in the range of 7-10 %. Table 7.6 presents a summary of the response variability for aliphatic amines obtained at microcrystalline diamond electrodes. The number of injections used in the statistical analysis is shown in parentheses. Table 7.6 FIA-EC Response Variability Data for Aliphatic Polyamines at a Microcrystalline Dlamond Thin Film Response Variability (% RSD) Polyamine 0.33% CHJHz 0.50% CHJH, 0.67 % Cl-IJH: Cadaverine (CA0) 67% (70) 5.0% (90) 20.8% (131) Putresclne o o (PUT) 15.0 A. (80) 7.5 A. (118) 9.7% (50) S rmidlne o a two) 7.3 /o (50) 5.0% (104) 15.9 /o (39) Spermlne o o --- (SPM) 13.1 /o (90) 3.0 /o (40) -- Figure 7.12 A and B shows a series of responses of 1.0 mM CAD and SPMD, respectively, at an extensively used 0.50% CH4/H2 film. By extensively, we mean use over many days and hundreds of injections. Detection was made at 855 and 780 mV, 190 respectively in carbonate buffer, pH 10. The first important observation is that the response is quite stable with no progressive response attenuation. This is because the decrease only seems to occur during the first few injections at a new electrode, not at a used one. A < :1. "a a Time 3 Minutes <1 :1. 1-1 W . — 6 minutes Time Figure 7.12 FIA-EC responses for a multiple injections of (A) 1.0 mM CAD and (B) 1.0 mM SPMD at a well 'conditioned' 0.50% C/H film. The carrier solution was CBpH10. A detection potential=855 mV and 3-minute delay between injections was used for CAD. A detection potential=780 mV and a 6-minute delay between injections was used for SPMD. Injection volume = 20 uL. Flow rate = 1.0 mUmin. 191 A second observation is that a delay time of 3 min between injections gives an acceptable response variability of 3.2% (n=17 injections) for CAD and a delay time of 6 min gives a slightly improved value of 2.8% (n=16 injections) for SPMD. These variabilities are lower than those reported in Table 7.6, because these electrodes were extensively used and there was no response attenuation during the first few injections. Values of 2-4% were commonly observed for all four polyamines at extensively used electrodes. In general, there was not a significant improvement in the response variability for any of the analytes when increasing the delay time between injections from 3 to 6 min. At best, the variability was improved by 0.5-1.0%. For comparison the amine oxidation response at highly oriented pyrolytic graphite (HOPG) and glassy carbon (60) electrodes was also studied. Figure 7.13 shows the response variabilities for these two electrodes and a diamond electrode at the detection potential of 655 mV. The first observation is that the signal for 1 mM CAD and SPMD at HOPG and GC is much smaller than that for boron-doped nanocrystalline diamond. The currents are about 150 nA and 260 nA at HOPG, and 50 nA and 150 nA at CC for 1 mM CAD and 1 mM SPMD, respectively. The current at diamond is about 30 nA and 50 nA for 0.1 mM CAD and 0.025 mM SPMD, respectively. The differences in current are due to couple of factors. First, the surface boron sites are not present on either HOPG or GC. The absence of these sites is leading to a short residence time for attack by OH-. Second, most of the OH- generated probably reacts with the surface of the electrode (i.e., edge plane sites) or is consumed in oxygen evolution reaction. The increasing background current may very likely suggest that. Even if some interactions of the amine with carbon-oxygen functional groups are possible, the driving force for the OER at CC and oxidation of the GC surface at this potential are much greater than OTR. 192 .i 300 HOPG (edge plane) I) g 1 (a) 1 mM CAD v (b) 1 mM SPNID E 2001 5 a 1001 H i I—-I 10 minutes 1501 Glassy Carbon 3° 1) A (a) 1 mM CAD E (b) 1 mM SPMD *5 100 ' 0 E U 5‘” lililinfilii |——-l 20 minutes 60' a Nanocrystalline Film A (a) 0.025 mM SPMD E 40. (b) 0.1 mM CAD I.“ * b § In: 5 20‘ 1L1. -Lull ll _.____ ._.._ -_.. o 1 Figure 7.13 FIA-EC responses for a multiple injections of CAD and SPMD at a HOPG, GC and a nanocrystalline boron-doped diamond thin-film electrode. Applied potential=655 mV vs AgIAgCI. Flow rate = 1.0 mUmin. 193 The second observation is the attenuation and variability in the signal at CC and HOPG electrodes. Both of these electrode microstructures are unstable in alkaline media at these detection potentials, and are easily oxidazed.62 What is interesting though is the dramatic difference in the oxidation response variability for each electrodes. For HOPG the amine response increases with the injection number at HOPG, but for CO, the response decreases. it was shown that electrochemically induced corrosion occurs at HOPG in acidic medium.62’304 304,305,132 HOPG undergoes oxidation, intercaiation, delamination. and exfoliation. These processes can result in alteration of the surface microstructure and an increase of the active area. GC does not undergo as much microstructural changes as HOPG so the active area remains fairly constant. The response at 60 likely decreases, due to fouling by reaction products. Reaction products likely adsorb on HOPG as well, but since the active area increases with time, the response does not decrease but actually shows an increase. The boron-doped diamond films possessed the proper structure-function relationship for stable and reproducible oxidation of aliphatic amines. The long-term response stability of an extensively used 0.50% CHJHZ film was examined. For this specific test, injections of 1.0 mM CAD were made over a 10 h period of continuous use. The background signal during this period remained low and unchanging even at this positive detection potential. This is due to the excellent morphological and microstructural stability of the material. The peak height changed by less than 8% (n = 100 injections) over the period. Also, there was no progressive loss of signal due to any kind of permanent electrode fouling, as is the case for other electrodes. The reproducibility and stability studies of amine oxidation were also performed for boron-doped nanocrystalline diamond electrodes. Figure 7.14 shows a series of 194 responses of 0.1 mM CAD, PUT, SPMD and SPM at a nanocrystalline film deposited from 1/94/5 CHJAr/Hz (v/v) ratio and 1 ppm B2H6. Mm *° 00 o o 300 099333330080808888888888889” E, i] 0nmupur 'g 2m)- 'fi’0anCAD : C>01mmsmuu 5 <0 01mMsmM um- _unnnnnunnnnunnnnnnnnnnnnununn aceteneefisecnneannaoaccesses 0 1 1 . . 1 . . . L A L 1 0 50 100 150 Time (minutes) Figure 7.14 FIA-EC responses for a multiple injections of 0.1 mM CAD, PUT, SPMD and SPM at a boron-doped nanocrystalline film deposited from 1/94/5 CHJAr/Hz (v/v) ratio with 1 ppm of added BzHe. The carrier solution was borax buffer pH 11. Applied potential=+665 mV. injection volume of CAD and PUT = 30uL, and SPMD and SPM = 10uL. Flow rate = 1.0 mUmin. The response is quite stable with no progressive response attenuation. The decrease only seems to occur during the first few injections for SPM. The response variability of the peak height was 3.5, 3.1, 2.1, and 3.3% for PUT, CAD, SPMD, and SPM respectively, based on 30 injections. Similar results were observed for other nanocrystalline films, deposited from 10 and 20 ppm B2H6. Response varied by 3-5% 195 during 20-30 injections. The response stability was measured for 0.05 mM DAP during a 7-h period of use. The background signal during that period remained low and stable. Figure 7.15 shows a series of 11 responses. The peak variability was 5%. 75‘ : 50‘ r: G) h In :1 U 25' 0- Time 60 . um Figure 7.15 FIA-EC series of responses for 0.05 mM DAP at a boron-doped nanocrystalline thin film deposited from a 1/94/5 CH4IAr/H2 (v/v) ratio and 10 ppm B2H6. The carrier solution was borax buffer, pH 11. Applied potential=+665 mV. injection volume = 10 uL. 7.3. Conclusions It has been shown that aliphatic polyamines can be quantitatively electrooxidized and detected at boron-doped nanocrystalline and microcrystalline diamond thin-film electrodes. FIA-EC results indicate that cadaverine, putrescine, spermidine, and 196 spermine can be detected stably and reproducibly at constant potential in alkaline media. The reproducibility of the assay, in terms of the linear dynamic range, limit of quantitation and response variability, was best for microcrystalline films deposited from 0.50% CH4IH2 ratio (i.e., films with the optimum balance of localized non-diamond sp2- carbon impurities and surface boron sites). There was correlation observed between amine oxidation electrode response (current) and the boron-doping level (i.e., surface boron-sites) in the microcrystalline and nanocrystalline diamond film. The highest signals and sensitivity values in FIA were observed for highly boron doped nanocrystalline diamond films (20 ppm). Therefore, these films or films with even higher boron-doping are the best performing for the aliphatic amine detection assay. The linear dynamic range for most of the amines was 4 orders of magnitude. In general, the limits of detection (LOD) for EDA, DAP, SPMD, and SPM were 0.1)1M. A 0.3 uM LODs were observed for CAD and PUT. Monoamine (methylamine) was detected at 10 1M level, however, an increase in the electrolyte pH would likely increase the sensitivity and improve the LCD for this analyte. The response variability was vastly improved over previous work by introducing a 3 to 6 minute delay period between injections. Typical peak height variabilities were in the 2 to 4 % range for both microcrystalline and nanocrystalline diamond. The long-term stability was excellent over a 7 or 10-h period of continuous use with no evidence for permanent electrode fouling by the reaction product(s) for nanocrystalline and microcrystalline films. The advantage of using diamond electrode is that no derivatization, pH adjustment of the carrier solution or pulsed waveform is necessary for detection. The detection figures of merit at diamond are as good or superb to data for other electrode materials reported in the literature. Table 7.7 summarizes some of the literature FIA-EC and LC-EC results for aliphatic amines. A limited amount of data has been reported for 197 CAD, PUT, SPMD, and SPM. There are examples of these analytes being separated and detected, but to the best of our knowledge, there has been no complete set of FIA- EC or LC-EC data published. Most of the data are for EA and DAP. Table 7.7 FIA-EC Data for Aliphatic Mono- and Diamine Limit of Response Material 053:1? Amine Detection Variability Reference (11M) (RSD %) Anodized amperometric 4.7 % 45 hrs, 54 Ag-PbOz detection DAP 0'3 (6 pm" (540 inj.) Gold PAD DAP 0.1 (4 pmol) not given 3‘ Gold ISWD DAP 0.01 (0.5pmoi) 0.5% (7 inj.) 3‘32 ._ amperometric 4 % (7 hrs) 34 NI GC detection CAD 0.8 (3.8 pmol) 2% (22 inj) - - 3.0 % (5 inj.) Copper anégegztmstnc EA 0-4 (3 pmol) short term 231 I | TN . 3.0 % (5 inj.) Nickel afigigggifim EA 1.0 (20 pmol) short term 231 3 0 '7 (5 ‘) - . o "2]. Platinum anégegztrrifrtnc EA 60 (1.2 nmol) short term 231 I | TI amperometric very short- 231 Cobalt detection EA 70 (1.4 nmol) live d . amperometric CAD 0.3 . diamond detection PUT (6 pmol) ""5 work 3-5% 20 inj. . DAP . amperometric 0.1 . diamond detection SSPPMAD (4 pmol) This work Diamond provides comparable detection figures of merit to those reported for DAP 31-33,54 198 Data for the amperometric detection of EA (pH 11) at diamond provide higher LOD than that for copper and nickel electrodes. They are better than that for platinum and cobalt oxide electrodes. Diamond, however, provides superior stability. The limits of detection for diamond are very much the same, or better than what has been achieved for Ag-PbOz alloy electrode.54 The pulsed amperometric detection (iSWD) 32 mode using gold provides one orders of magnitude lower limits of detection for DAP, however detection of CAD and PUT has been shown to be much lower at diamondzg'33 The key finding from these studies is that aliphatic polyamines can be electrooxidized and detected at boron-doped microcrystalline and nanocrystalline diamond films. Both of these materials provide stable and reproducible response for amine oxidation in alkaline medium. Both of these materials provide good limits of quantitation for diamines and polyamines. Electrochemical detection using diamond provides advantage of direct assay for monitoring of amines. Amine can be detected without need of derivatization or pH adjustment of the mobile phase. Simple amperometric detection mode is a suitable technique for this assay, no pulsed waveform is necessary as is the case of gold electrode. 199 Chapter 8 REVERSED-PHASE HPLC SEPARATION AND AMPEROMETRIC DETECTION OF ALIPHATIC POLYAMINES AT BORON-DOPED DIAMOND THIN-FILM ELECTRODES 8.1. Introduction Few electrochemical assays for aliphatic amine detection have been reported in literature. The most successful ones involved noble and transition metal electrodes, alloy composite electrodes, and chemically modified electrodes}34 The Johnson group demonstrated the effective separation and detection of several polyamines with a detection limit of 200 pmole for putrescine, using an ion exchange-liquid chromatography 29'3“” PAD at Au electrodes. while coupled with pulsed amperometric detection (PAD). highly useful, is somewhat complicated to implement and, even though the detector is commercially available, detection at constant applied potential is rather the method of choice for many routine analyses. Stable amperometric detection of aliphatic polyamines was observed at metal alloy composite electrodes (e.g., Ag-PbOz).54’55 200 The results described herein constitute the last phase of the polyamine assay development at diamond electrodes - liquid chromatographic separation and amperometric detection. The first phase of the work involved a cyclic voltammetric study of the electrooxidation of PUT, CAD, SPMD and SPM at boron-doped diamond electrodes. An oxidation reaction mechanism was proposed and some of the electrode properties influencing the amine oxidation response were highlighted. Clearly, boron- doped, polycrystalline diamond thin-film electrodes possess the requisite properties for stable, sensitive, and reproducible detection of aliphatic polyamines at constant applied potential. The next phase of the work involved a detailed flow-injection analysis investigation of the amperometric detection, at constant applied potential, of the four polyamines. The reversed-phase liquid chromatographic separation of the polyamines coupled with amperometric detection is described herein. The goals were to: (i) optimize the reversed-phase separation (isocratic) of PUT, CAD, SPMD, and SPM, and (ii) complete a comprehensive evaluation of the analytical detection figures of merit for the four polyamines in the amperometric detection mode. 8.2. Results and Discussion 8.2.1. Cyclic Voltammetry and Flow Injection Analysis Figure 8.1 shows cyclic voltammetric i-E curves in borax buffer, pH 10.6 at microcrystalline diamond film deposited from a 0.4% CHJHZ ratio. There is anodic current passed between 550 and 950 mV during the first scan, which is not present in subsequent scans. This current is likely associated with the irreversible oxidation of the diamond surface, involving replacement of some surface hydrogen atoms with 185,306,307 chemisorbed oxygen functionalities (e.g., OH). The current at 1100 mV is due 201 to oxygen evolution. After the initial one, the scans are stable with multiple sweeps between 200 and 1100 mV. The low and steady background current results from the morphological and microstructural stability of diamond.163 Current (ILA) 30 2"“,25‘“, 45th, 60“I sweep 25 20 -i 15 -l First sweep 10 h \ 5 l __—-—> 0 . W1 ____________ 150 35 550 750 950 115 Potential (mV vs A2/A2Cli Figure 8.1 Background cyclic voltammetric i-E curves in 0.01 M borax buffer, pH 10.6, for a boron-doped microcrystalline diamond film deposited from a 0.4% CHJHZ ratio. Presented are the 13‘, 2"“, 25 "‘, 45 m, 60th scans. Scan rate = 0.1V/s. Electrode area = 0.2 cm2. Figure 8.2A presents a background current-time plot for a diamond thin-film electrode in borax buffer, pH 11.2, and acetonitrile (93/7 vIv) at 665 mV. The flow rate was 1 mUmin. The measurements were made in the thin-layer flow cell and the plot reveals the time dependence of the background current after detector tum-on. The background current stabilizes quickly, reaching a constant value in less than 2 min. The mean value of the current after stabilization is about 11 nA and the peak-to-peak 202 variation is less than 120 pA. The background current for diamond is lower than that for glassy carbon, and remained low and stable with time. By way of comparison, the 1 minute Current (nA) M (I... mum-TNT} . #07. l ‘ I ' ' I ' l ‘ ' ' I 5 10 15 20 25 Time (minutes) Figure 8.2 Background current-time profile in FIA after detector tum-on for (A) a microcrystalline diamond and (B) glassy carbon.30 The carrier solution for diamond was 7/93 (v/v) acetonitrile/borax buffer, pH 11.2. The carrier solution for CO was borax buffer, pH 11.2. Detection potential = 665 mV vs AgIAgCI. Electrode area = 0.08 cmz. Flow rate = 1 mUmin. 203 stabilization time for glassy carbon, under similar conditions (in borax buffer, pH 11.2 only), is about 20 min, as seen in Figure 8.2B. The current initially decreases as the double layer charging reaches completion. This is followed, however, by an increase in the current reaching a maximum at about the 5-min mark. The current then slowly decays over a 15 min period before stabilizing at approximately 20 nA after 20 min. The peak-to-peak variation in the signal ranges from a low 200 pA to a high of 800 pA. A rapid stabilization time, and a low and stable background current are characteristic features of diamond electrodes.163 Figure 8.3 shows cyclic voltammetric i-E curves for 1 mM solutions of CAD and SPMD (0.01M borax buffer, pH 11.2) at a boron-doped microcrystalline diamond thin film. 15 .. 1mMSPMD \ 3 10 E 0 t: 5 5 o 0.1 - _____ . 0 20 40 Potential (mV vs AgIAgCl) Figure 8.3 Cyclic voltammetric i-E curves for 1 mM CAD and 1 mM SPMD in 0.01 M borax buffer, pH 11.2, along with the corresponding background current, for a boron- doped microcrystalline diamond film deposited from a 0.4% CH4IH2 ratio. Scan rate = 0.1 V/s. 204 The corresponding background voltammogram is also shown, for comparison. The background current is low at potentials negative of 850 mV (1.6 “A at 700 mV and 3.3 uA at 850 mV). At about 850 mV, the current sharply increases due to the onset of oxygen evolution. The background current at potentials negative of 850 mV is associated with charging of the electric double layer and surface faradaic processes. An irreversible anodic oxidation wave near 800 mV is observed for both CAD and SPMD, as was seen for the other two aliphatic polyamines (data not shown). The well- defined peak is typical for the oxidation of aliphatic polyamines in alkaline media, pH210, for boron-doped diamond thin-film electrodes.188’259 E,“ is 855 and 845 mV, respectively, for CAD and SPMD, with corresponding ip°" values of 45 and 118 pA. The signal-to—background ratio is about 14 for CAD and 36 for SPMD. Most interesting is the fact that ip°" for SPMD is ~2.5 times greater than for CAD. The higher current is presumably due to a larger number of electrons being transferred per molecule. The electrode response for amines was also studied in the presence of acetonitrile as this solvent was used as the organic modifier for the reversed-phase separation discussed below. Cyclic voltammetry was employed to confirm that the organic modifier had no adverse effect on the amine oxidation response. The voltammograms were recorded in a 7% acetonitrile/borax buffer solution (v/v). Figure 8.4 shows cyclic voltammetric i-E curves for 0.1 mM solutions of CAD for a microcrystalline boron-doped diamond film in (A) 0.01 M borax buffer, pH 10.6, and (B) 93” (v/v) 0.01 M borax buffer, pH 10.6 ICHacN. The E,“ values are 960 and 930 mV without and with added CH3CN, respectively. Most importantly, the i,” values, 15-17 M, are unchanged, indicating that the organic modifier does not adversely affect the response. The background current, however, was slightly higher with the added acetonitrile. 205 49- 39'. 29’. 19’. Current (11A) 9 i 0 Potential (mV vs AgIAgCI) Figure 8.4 Cyclic voltammetric i-E curves in 0.1 mM CAD plus (A) 0.01 M borax buffer, pH 10.6, and (B) 93/7 (v/v) 0.01 M borax buffer, pH 10.6 / CHacN, for a boron-doped microcrystalline diamond film deposited from a 0.4% CH4/H2 ratio. Scan rate :01 Vls. Flow injection analysis with electrochemical detection (FIA-EC) was used to initially evaluate the electrode performance in (7/93 v/v) acetonitrile/borax buffer, pH 11.2. The flow rate was 1 mUmin. Figure 8.5 A shows hydrodynamic voltammograms for the four different polyamines. The background current at the different potentials is also shown, for comparison. Each marker represents the average signal for 4 injections, and the standard deviations are within the size of the marker. The response increases for all the polyamines as the potential is made more positive. As discussed previously, the curves do not exhibit a limiting current as the reactions are not purely mass transport limited, but rather are controlled by the adsorption of the polyamine and reaction with electrogenerated OH-. The signal for PUT is 6.1 nA at 565 mV and increases to 80.7 nA at 665 mV. The background current is 2.9 M at 565 mV and increases to only 7 nA at 206 665 mV. The background current at potentials higher than 695 mV was somewhat unstable, due to the water discharge reaction. 400 c .5. ; A a A 300 l . ' b g 250 - ' ‘5 ‘ - ' a 200 . . C ‘5 150 - ° ' U ’ I O 100 : O I a O I t 50 i . : i z o . e .9» _,,:____2é32-_2.__ —— - 0 A A A 560 580 600 620 640 66 680 70 Applied potential (mV vs AgIAgCI) Figure 8.5 A Hydrodynamic voltammetric i-E curves for a boron-doped microcrystalline diamond film (0.4% CH4/H2 ratio), in 100 (M (a) SPM, (b) SPMD, (c) PUT, and (d) CAD. The carrier solution was 7/93 (v/v) acetonitrile/borax buffer, pH 11.2 (e). Injection volume = 20uL. Figure 8.5 B shows plots of the signal-to-background (SIB) ratio as a function of the potential. These plots are useful for determining the optimum detection potential in cases, like this one, where a well-defined sigmoidal hydrodynamic voltammogram is not observed.259'291 The maximum SIB for CAD, PUT, SPMD, and SPM is 9.2, 10.6, 24.8, and 38.7, respectively, all at 665 mV. The SIB ratio for CAD and PUT is approximately the same, ~10, while the value for SPMD is approximately two times (~20), and for SPM 207 is approximately four times greater (~40). This reflects a greater number of electron transferred per molecule for SPMD and SPM. 35. 8’. SIB ratio 15* 10+ J A A 560 580 600 620 640 660 680 Applied potential (mV vs AgIAgCI) Figure 8.5 B Plots of the S/B ratio versus the applied potential for a boron-doped microcrystalline diamond film (0.4% CH4IH2 ratio), in 100 M (a) SPM, (b) SPMD, (c) PUT, and (d) CAD. The S/B ratio was calculated as lm. - lmkgmndl lbackgmnd. Figure 8.6 A and B show the FIA-EC responses for multiple injections of 5 pM PUT and 2 uM SPMD at a detection potential of 665 mV. The carrier solution was (7/93 700 v/v) acetonitrile/borax buffer, pH 11.2, and the flow rate was 1 mLImin. For 40 injections of PUT and 60 injections of SPMD, the response variability was 2.6 and 4.3 %, respectively. A 4 min delay time was used between injections, which resulted in an . . . . 1 improvement in the preClSlon.29 208 151 |.—-| , Sminutes mi 5 ‘ j j , H l j A g 20 fl . U 15' 10' Time (minutes) Figure 8.6 FIA-EC responses for a boron-doped microcrystalline diamond film (0.4% CHJHg) during a multiple injections of (A) 5 uM PUT and (B) 2 (1M SPMD. The carrier solution was a 7/93 (v/v) acetonitrile/borax buffer, pH 11.2. Detection potential = 665 mV vs. AgIAgCI. Injection volume = 20 uL. Flow rate = 1.0 mUmin. Figure 8.7 shows a series of FIA-EC responses for 50 pM SPM at a detection potential of 665 mV. The carrier solution was 7/93 (v/v) acetonitrile/ borax buffer, pH 11.2, and the flow rate was 1 mLImin. The response curve was generated by varying the injected volume from 2 to 20 uL (a-h). Both the peak current and charge change linearly with the injected mass of each analyte (r2 2 0.998). Such was the case for the other three polyamines. Linear regression statistics for a plot of current vs. mass injected (r2 = 0.998) had a slope of 0.18 uA/ng and an intercept of 0.5 nA. 209 h g 20minutes f g e E d 2 ‘5 c O b a MHHL - -- - --- Time Figure 8.7 FIA-EC calibration responses for a boron-doped microcrystalline diamond film (0.4% Gilt/Hz) during multiple injections of 5 uM SPM. injected volumes were (a-h) 2 to 20 uL. The carrier solution was a 7/93 (v/v) acetonitrile/borax buffer, pH 11.2. Applied potential = 665 mV vs AgIAgCI. Flow rate = 1.0 mL/min. in summary, the addition of acetonitrile to the supporting electrolyte had no adverse effect on the voltammetric or FIA-EC response for any of the polyamines. A linear response was observed with injected mass, and good response precision was seen. 8.2.2. Liquid Chromatography The reversed-phase HPLC separation of a mixture of the four aliphatic polyamines, coupled with amperometric detection, was investigated. The maximum oxidation response is obtained in alkaline solution, at pH values above the pKa. therefore 210 the separation was investigated using different mobile phase compositions consisting of borax buffer and acetonitrile at pH values of 10.0, 10.5, and 11.2. For the most part, all of the amines are unprotonated at these pHs. Peak assignments were made based on the retention times of the individually injected analytes using identical separation and detection conditions. Much work was performed in an effort to optimize isocratic separation of the polyamines. Acetonitrile was selected as the organic modifier, and after examing the separation with modifier concentrations (v/v) from 0 to 20%, a 7 % concentration was found to be optimum in terms of the efficiency of the separation (capacity factor, peak width, and resolution). Figure 8.8 shows a typical chromatogram for a mixture of 47 (1M CAD, PUT, SPMD, and SPM. A chromatogram for 0.50 (M PUT, CAD and SPMD is presented in the inset. in both cases, amperometric detection was made at 665 mV vs. AgIAgCI, and the mobile phase composition was (7193 v/v) acetonitrile/borax buffer, pH 11.2. Variations in retention time during a 7-h period of testing were 1.4, 1.2, 2.8, and 2.9% for PUT, CAD, SPMD, and SPM, respectively. The capacity factors, k’, are 0.40, 1.0, 1.5, and 4.2 for PUT, CAD, SPMD, and SPM, respectively. The selectivity coefficients, a, for PUT/CAD, CAD/SPMD, and SPMD/SPM are 2.5, 1.5, and 2.8, respectively. The total number of theoretical plates, N, for the separation is 188, based on retention time and peak width of the SPM peak. All four peaks are asymmetrically shaped with the most significant tailing seen for the late eluting solutes, SPMD and SPM. The peak asymmetry is calculated as the distance from the leading to the tailing edge of the peak, measured at 5% of the peak height, divided by the distance from the leading edge to the peak maximum. The asymmetry values are 3.8, 2.8, 5.1, and 6.9 for PUT, CAD, SPMD, and SPM, 211 respectively. The tailing appears to be mainly caused by interactions of the amine functional group with the stationary phase, not with the electrode surface as no such tailing is seen in the FIA —EC results. The tailing seemed to get worse with column use. This was because of stationary phase dissolution which creates silanol sites. The interaction of the amine functional groups with the silanol sites is believed to be the cause of the tailing. Obviously, the low number of theoretical plates for the separation is due to the excessive peak tailing. SPMD H G d) in: '5 PUT D 5: CAD '5' C3 F" 5? O 25 “A SPM —*¢J A I 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 Time (minutes) Figure 8.8 Reversed-phase liquid chromatographic separation of (A) 47 1M PUT, CAD, SPMD, and SPM, and (B) 0.5 uM PUT, CAD, and SPM on a C18 column (X- Terra, 5 pm particle size, 4.6 x 150 mm). The mobile phase was a 7/93 (v/v) acetonitrile/borax buffer, pH 11.2. Detection potential = 665 mV vs. AgIAgCI. Injection volume = 20 uL. Flow rate = 1.0 mUmin. The boron-doped microcrystalline diamond film was deposited from a 0.4% CH4/H2 ratio. 212 The most essential biogenic aliphatic polyamines, and those found at significant concentrations in biological fluids, are PUT, SPM, and SPMD. Figure 8.9 shows a typical chromatogram for the three amines at 20 (M concentration. The separation was achieved using 16/84 (v/v) acetonitrile/borax buffer, pH 11.2, mobile phase, and the detection was made at 665 mV vs AgIAgCI. SPMD PUT 5 20 nA ’ In In :1 U :1 13 ,5 spam 1'2 x O 4 6 8 I 1 ' 12 i 14 16 Time (minutes) Figure 8.9 Reversed-phase liquid chromatographic separation of PUT, SPMD, and SPM on a C18 column. The solution mixture contained 20 uM of each amine. The mobile phase was a 16/84 (v/v) acetonitrile/borax buffer, pH 11.2. Detection potential = 665 mV vs. AgIAgCI. Injection volume = 20 11L. Flow rate = 1.0 mLImin. The boron- doped microcrystalline diamond film was deposited from a 0.4% CH4/H2 ratio 213 The retention times for all three analytes decreased with the addition of the organic modifier between 7 and 20% (refer to Fig.8.9) with the total separation taking about 8 minutes for the 16/84 composition. The reduced retention times lessen the peak asymmetry, although the tailing for SPM on this column is still significant. The retention times are 2.2, 3.9, and 5.6 minutes for PUT, SPMD, and SPM, respectively. Further additions of acetonitrile to the mobile phase lead to insufficient peak resolution and an increase in the background current. Amperometric detection analytical figures of merit were determined for the HPLC separation and detection of the aliphatic polyamines. The electrode response for PUT, CAD, SPMD, and SPM were examined as a function of the injected concentration. The results are summarized in Table 8.1. Table 8.1 Amperometric Detection Figures of Merit for the HPLC Separation of the Aliphatic Polyamines Concentration Molar Mass “mm“ "DR Loo (sm 23) L00 L00 Putresclne 0.5 -500 11M PUT (r2=0.998) 0.5 uM 10 pmol 0.88 ng Cadaverine 0.5 -500 (M CAD (r2=0.998) 0.5 uM 10 pmol 1.02 ng Spermidine 0.5 -500 (M SPMD (r2=0.998) 0.5 uM 10 pmol 1.45 ng Spermine 5 -500 uM SPM (r2=0.998) 5 (M 100 pmol 31.2 ng Note:The calibration curves were constructed with concentrations (4 injections at each) from 0.5 to 500 uM. Injection volume=20 uL. Flow rate=1 mLImin. Detection potential=665 mV. LOO=limits of quantitation. The calibration curves were constructed from the response for 8 different injected concentrations ranging from 0.5 to 500 (M. A linear relationship between both the peak current and charge, and the concentration was observed for PUT, CAD, and SPMD, with 214 a linear regression correlation coefficient >0.998. A smaller dynamic range was observed for SPM ranging from 5 to 500 uM with a regression coefficient of > 0.998. The nominal y-axis intercept for all the calibration plots was less thah 5 nA for CAD, SPMD, and SPM and about 10 nA for PUT. The limits of quantitation for PUT, CAD, and SPMD were 0.5 (M or 10 pmol injected, while the limit for SPM was 5 uM or 100 pmol injected (S/N23). The mass limits of quantitation were 1.0 ng for CAD, 0.9 ng for PUT, 1.5 ng for SPMD, and 31 ng for SPM. The LOD observed for SPM are the highest. it is because calibration curves were constructed based on peak height and the broadness observed for SPM peak reduces LOD. The theoretical limits of detection, obtained by extrapolating the linear calibration plots to the concentration at which the SIN=3, were 0.3 uM for PUT and SPMD, 0.5 uM for CAD, and 3 (AM for SPM. Table 8.2 presents a summary of the LC-EC amperometric detection reproducibility and stability data. The reproducibility was tested at different concentrations. The highest response variability was observed for the highest concentration of each analyte. For 170 (M PUT, the current variation is about 6%, while for the 20 and 0.5 uM solutions, variations are about 3% for six injections. A similar trend is observed for all four amines. Generally, the reproducibility data based on the peak charge are much better than those based on peak height, due to the broadened and assymetric peaks. The long-term response stability was examined over a 7-h period of continuous use. Twelve injections of all four amines (47 (M ea.) were periodically made. The background signal, during this period, remained low and unchanging. The peak height varies by 3.0% for PUT and CAD, 6.3% for SPMD and 11.0% for SPM. The peak charge varies much less over this period with values of 2.0% for PUT and SPM, 4.2% for CAD, and 3.9% for SPMD. Most importantly, there is no progressive loss of signal due to permanent electrode fouling. 215 Table 8.2 Amperometric Detection Reproducibility and Stability for the HPLC Separation of the Aliphatic Polyamines Stability nso % Reproducibility Amine Concmll'atlon RSD % (47 11M) Current Charge Current 170 6.1 Pmfifi'm 20 3.0 2.1 3.0 0.5 3.1 150 7.0 cadgxg'm 20 nIa 4.2 3.0 0.5 3.6 , 108 6.1 59:3:ch 20 2.0 3.9 6.3 0.5 2.5 s I 100 12.0 "gal“ 20 6.0 2.1 11.0 5 5.8 ‘ Reproducibility conditions: n=15 (170, 150, 108, 100 11M), n=11 (20 11M), n=7 (0.5 uM). Stability conditions: n=16 (7 hrs). Separation and detection conditions are the same as in Table 8.1. Diamond provides comparable or better detection figures of merit to those reported in the literature for Au, anodized Ag-PbOz, Cu, Ni, Pt. The limit of detection at conductive diamond is better than that achieved so far using Au (~200 pmole for PUT) in LC--EC33 and comparable to or better than those in flow injection studies with chemically modified electrodes, e.g., nickel-modified glassy carbon34 or Ag-PbOz (0.1 and 0.3 1.1M amine, respectively).54’58 In terms of response reproducibility and stability, diamond 23 provides superior figures of merit to those for Cu, Ni, Pt 1 and comparable to those of Au. Ag-Ploo2 and nickel-modified glassy carbon.31'34’54 216 8.3. Conclusion Four important aliphatic polyamines were separated isocratically by reversed- phase liquid chromatography and detected amperometrically using boron-doped diamond electrodes. Diamond electrodes are unique within the carbon electrode family for their ability to stably and sensitively detect aliphatic polyamines. Detection was made at a constant potential of 665 mV vs. AgIAgCI using microcrystalline diamond electrodes deposited from a 0.4% CHJHZ ratio. The analytes were isocratically separated in a 7/93 (v/v) acetonitrile/0.01 M borax buffer, pH11.2. mobile phase. Good separation and peak resolution for all four analytes was achieved within 14 minutes. There was extensive peak tailing for the last eluting components, SPMD and SPM. The peak broadening is presumably due to interaction of the amines with the residual silanol groups on the stationary phase. A linear dynamic range from 0.5 to 500 (M was observed for cadaverine (CAD), putrescine (PUT) and spermidine (SPMD), with a theoretical limit of detection (SIN=3) of 0.3 11M for PUT and SPMD and 0.5 uM for CAD. The linear dynamic range for SPM was 5 to 500 uM and limit of detection was 3 uM. The mass detection limits were 0.88 ng for PUT, 1.02 ng for CAD, 1.45 ng for SPMD, and 31.2 119 for SPM. Typical response variabilities (n=6) for a 0.5 uM injected concentration were ~3%. Excellent long-term stability was found with no progressive loss of signal due to permanent electrode fouling. The major advantage of using diamond is that no pre- or post-column analyte derivatization, pH adjustment of the mobile phase, or pulsed detection waveform is necessary for detection. 217 Chapter 9 SUMMARY 1) Good quality, boron-doped diamond thin films were successfully deposited on three different substrates: Si, Mo, and W. The resulting films were characterized by electrochemical methods of analysis, AFM, XRD, Raman spectroscopy, and dynamic SIMS. it was observed that the electrochemical response of the films was independent of the substrate material. All films exhibited properties of high quality diamond - a wide working potential window, a low voltammetric background current and capacitance, and good responsiveness for Fe(CN)6'3"‘, Ru(NH3)6”2’*3, erl6°2"3, and methyl viologen without any pretreatment. k°,,pp values ranged from 10°2 to 10'1 cm/s for all four redox systems. 2) Boron-doped nanocrystalline diamond thin films were deposited by CVD from CH4/H2IAr source gas mixture. XRD revealed that the bulk crystal structure of the films was cubic diamond. TEM indicated the films consist of 10-15 nm randomly oriented but atomically ordered diamond grains. SEM showed these grains form 218 aggregates ~100 nm in size. Electrochemicaily, these films exhibited a wide working potential window, a low voltammetric background current and good responsiveness for Fe(CN)6'3“, Ru(NH3)6"3’*2, erl6‘2"3, and methyl viologen without any pretreatment. The quasi-reversible voltammetry for all the couples indicates that the boron—doped nanocrystalline diamond has a sufficient charge carrier density over the wide potential range to support a rapid electron transfer. More sluggish kinetics were found for 4- tert-butyicatechol and Fem“. The sluggish kinetics are attributed to weak surface adsorption for the former, and to the absence of catalyzing surface carbonyl groups for the latter. 3) Aliphatic amines can be quantitatively oxidized at both boron-doped microcrystalline and nanocrystalline diamond thin-film electrodes. The oxidation response is highly dependent on the physicochemical properties of the diamond surface, in particular, the spz-bonded non-diamond carbon impurity and surface boron concentration. High doping level and moderate amount of spZ-carbon impurities were found to be the optimal parameters for amine detection. The oxidation response depends on the solution pH. in order to be oxidized the amines need to be unprotonated. 4) FIA-EC results indicate that aliphatic amines can be stably and reproducibly detected at constant potential in alkaline media at boron-doped diamond thin films. The highest signals and sensitivity values in FIA were observed for highly boron-doped nanocrystalline diamond films (20 ppm). Therefore, these films or films with even higher boron-doping are the best performing for the aliphatic amine detection assay. The linear dynamic range for most of amines was 4 orders of magnitude. The limits of 219 detection (LOD) achieved for EDA, DAP, SPMD, and SPM were 0.1uM, for CAD and PUT 0.3 11M, and for MA 10 11M. 5) Aliphatic polyamines (cadaverine, spermine, spermidine, and putrescine) can be isocratically separated by reversed-phase liquid chromatography and amperometrically detected using boron-doped diamond electrodes. Detection was made at a constant potential of 665 mV vs AgIAgCI using microcrystalline diamond electrodes deposited from a 0.4% CH4IH2 ratio. The analytes were separated in a 7/93 (v/v) acetonitrile/0.01 M borax buffer, pH11.2. Good separation and peak resolution for all four analytes was achieved within 14 minutes. A linear dynamic range from 0.5 to 500 (M was observed for CAD, PUT and SPMD, with a theoretical limit of detection (SIN=3) of 0.3 11M for PUT and SPMD and 0.5 uM for CAD. The linear dynamic range for SPM was 5 to 500 11M and limit of detection was 3 11M. The mass detection limits were 0.88 ng for PUT, 1.02 ng for CAD, 1.45 ng for SPMD, and 31.2 ng for SPM. Typical response variability (n=6) for a 0.5 uM injected concentration were ~3%. Excellent long-term stability was found, with no progressive loss of signal due to permanent electrode fouling. 220 10. 11. 12. 13. References Pegg, A. E. Cancer Res. 1988, 48, 759-774. Tabor, C. W.; Tabor, H. Annu. Rev. Biochem. 1984, 53, 749-790. Tabor, C. W.; Tabor, H. Microbiol. Rev. 1985, 49, 81-99. Russell, D. H. D., B. G. M. Polyamines as Biochemical Markers of Normal and Malignant Growth; Raven Press, 1978. Aikens, D.; Bunce, S.; Onasch, F.; Parker, R., Ill; Hurwitz, C.; Clemans, S. Biophys. Chem. 1983, 17, 67-74. Grant, G.; Bardocz, S.; Brown, D. 8.; Watt, W. 3.; Stewart, J. C.; Pusztai, A. Biochem. Soc. Trans. 1990, 18, 1 009-101 0. Bardocz, S. Eur. J. Clin. Nutr. 1993, 47', 683-690. Bardocz, S.; Grant, G.; Brown, D. 3.; Ralph, A.; Pusztai, A. J. Nutr. Biochem. 1993, 4, 66-71. Tabor, C. W.; Tabor, H. Faseb J. 1995, 9, A1389-A1389. Bardocz, S. Trends Food Sci. Technol. 1995, 6, 341 -346. Bardocz, S.; Duguid, T. J.; Brown, D. S.; Grant, G.; Pusztai, A.; White, A.; Ralph, A. Br. J. Nutr. 1995, 73, 819-828. Nagy, G.; Xu, C. X.; Cosofret, V. V.; Buck, R. P.; Lindner, E.; Neuman, M. R.; Sprinkle, R. H. Talanta 1998, 47, 367-376. Marzouk, S. A. M.; Xu, C. X.; Cosofret, B. R.; Buck, R. P.; Hassan, S. S. M.; Neuman, M. R.; Sprinkle, R. H. Anal. Chim. Acta1998, 363, 57-65. 221 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Seller, N.; Sarhan, S.; Knodgen, B.; Gerhart, F. J. Cancer Res. Clin. Oncol. 1988, 1 14, 71 -80. Seller, N.; Sarhan, S.; Grauffel, C.; Jones, R.; Knodgen, B.; Mouiinoux, J. P. Cancer Res. 1990, 50, 5077-5083. Simon, P.; Lemacon, C. Anal. Chem. 1987, 59, 480-484. Tabor, C. W.; Tabor, H. Methods Enzymol. 1983, 94, 29-36. FDA, U. 8.; Nutrition, Center for Food Safety and Applied Nutrition, 2001. Roig Sagues, A. X.; Hernandez Herrero, M. M.; Rodriguez Jerez, J. J.; Quinto Fernandez, E. J.; Mora Ventura, M. T. Alimentan'a (Madrid) 1998, 294, 59-66. Petaja, E.; Eerola, S.; Petaja, P. European Food Research and Technology. 2000, 210, 280-285. Eerola, S.; Sagues, A. X. R.; Lilleberg, L.; Aaito, H. Zeitschn'ft Fur Lebensmittel- Untersuchung Und-Forschung a-Food Research and Technology 1997, 205, 351 -355. Yen, G. C. Chung-kuo Nung Yeh Hua Hsueh Hui Chih (Abstract) 1986, 24, 211- 227 Su, S. C.; Chou, S. 8.; Chang, P. C.; Hwang, D. F. J. of Chrom. B 2000, 749, 163-169. Hungerford, J. M.; Arefyev, A. A. Anal. Chim. Acta 1992, 261, 351-359. Khuhawar, M. Y.; Qureshi, G. A. J. of Chrom. B 2001, 764, 385-407. Khuhawar, M. Y.; Memon, A. A.; Bhanger, M. i. Journal of the Chemical Society of Pakistan 2000, 22, 1 15-1 18. Ma, Y., Zhang, R. , Cooper, C. L. J. Chrom. A 1992, 608, 93. 222 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. Muskiet, F. A. J.; Dorhout, B.; Vandenberg, G. A.; Hessels, J. J. of Chrom. B 1995, 667, 189-198. Dobberpuhl, D. A.; Hoekstra, J. C.; Johnson, D. C. Anal. Chim. Acta 1996, 322, 55-62. Dobberpuhl, D. A.; Johnson, D. C. EIectroana/ysis1996, 8, 726-731. Hoekstra, J. C.; Johnson, D. C. Anal. Chim. Acta 1999, 390, 45-54. Hoekstra, J. C.; Johnson, D. C. Anal. Chem. 1998, 70, 83-88. Pineda, R.; Knapp. A. D.; Hoekstra, J. C.; Johnson, D. C. Anal. Chim. Acta 2001, 449, 1 11-1 17. Caselia, l. G.; Rosa, 8.; Desimoni, E. Electroanalysis 1998, 10, 1005-1009.. Fleischmann, M.; Korinek, K.; Pletcher, D. J. Chem. Soc., Perkin Trans. 2 1972, 1 396-1 403. Hansen, E. H.; Ruzicka, J.; Ghose, A. K. Soil Nitrogen Fert. Pollut. Proc. Rep. Res. Coord. Meet. 1980, 77-91. Hughes, 8.; Johnson, D. C. J. Agric. Food Chem. 1982, 30, 712-714. Hughes, 8.; Meschi, P. l..; Johnson, D. C. Anal. Chim. Acta 1981, 132, 1-10. Jensen, M. B.; Johnson, D. C. Anal. Chem. 1997, 69, 1776-1781. Johll, M. E.; Williams, D. G.; Johnson, D. C. Electroanalysis 1997, 9, 1397-1402. Feng, J.; Johnson, D. C. J. Electrochem. Soc. 1990, 137, 507-510. LaCourse, W. R.; Hsiao, Y. L.; Johnson, D. C. J. Electrochem. Soc. 1989, 136, 3714-3719. Mho, S. L; Johnson, D. C. J. Electroanal. Chem. 2001 , 495, 152-159. 223 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. Mho, S. L; Johnson, D. C. J. Electroanal. Chem. 2001, 500, 524-532. Pampiin, K. L.; Johnson, D. C. J. Electrochem. Soc. 1996, 143, 2119-2125. Popovic, N. D.; Cox, J. A.; Johnson, D. C. J. Electroanal. Chem. 1998, 456, 203- 209. Popovic, N. D.; Cox, J. A.; Johnson, D. C. J. Electroanal. Chem. 1998, 455, 1 53- 160. Treimer, S. E.; Feng, J.; Scholten, M. D.; Johnson, D. C.; Davenport, A. J. J. Electrochem. Soc. 2001, 148, E459-E463. Yeo, l. H.; Johnson, D. C. J. Electrochem. Soc. 1987, 134, 1973-1977. Allen, R. E.; Johnson, D. C. Talanta 1973, 20, 799-809. Chang, H.; Johnson, D. C. J. Electrochem. Soc. 1990, 137, 2452-2457. Chang, H.; Johnson, D. C. J. Electrochem. Soc. 1990, 137, 3108-3113. Feng, J.; Johnson, D. C. J. Electrochem. Soc. 1991, 138, 3328-3337. Ge, J.; Johnson, D. C. J. Electrochem. Soc. 1995, 142, 1525-1531 . Ge, J.; Johnson, D. C. J. Electrochem. Soc. 1995, 142, 3420-3423. Dobberpuhl, D. A.; Johnson, D. C. J. Chrom., A1995, 694, 391-398. Ge, J.; Johnson, D. C. J. Electrochem. Soc. 1996, 143, 2543-2548. Ge, J. J., D. C. J. Electrochem. Soc. 1995, 142, 1525. Swain, G. M. J. Electrochem. Soc. 1994, 141, 3382-3393. Swain, G. M. Adv. Mater. 1994, 6, 388-392. 224 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. Swain, G. M.; Anderson, A. B.; Angus, J. C. MRS Bull. 1998, 23, 56-60. Granger, M. C., USU, Logan, 2000. Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang, M. D.; Butler, J. E.; Lucazeau, G.; Merrnoux, M.; Strojek, J. W.; Swain, G. M. Anal. Chem. 2000, 72, 3793-3804. Arihara, K.; Kitamura, F.; Nukanobu, K.; Ohsaka, T.; Tokuda, K. J. Electroanal. Chem. 1999, 473, 138-144. Edstrom, K.; Herranen, M. J. Electrochem. Soc. 2000, 147, 3628-3632. Chen, P.; McCreery, R. L. Anal. Chem. 1996, 68, 3958-3965. Duebgen, R. Can. Chem. News 1987, 39, 13-16. Engelman, E. E.; Evans, D. H. Langmuir 1992, 8, 1637-1644. Kamau, G. N. Anal. Chim. Acta 1988, 207, 1-16. Angus, J. C.; Argoitia, A.; Gat, R.; Li, Z.; Sunkara, M.; Wang, L.; Wang, Y. Philos. Trans. R. Soc. Lond. Ser. A-Math. Phys. Eng. Sci. 1993, 342, 195-208. Angus, J. 0.; Collins, A. T. Nature 1994, 370, 601-601. Angus, J. C.; Gardner, N. 0.; Chem. Eng. Div.,Case West. Reserve Univ. Cleveland,Ohio,USA., 1973, pp 53 pp. Angus, J. C.; Martin, H. B.; Landau, U.; Evstefeeva, Y. E.; Miller, B.; Vinokur, N. New Diamond and Frontier Carbon Technology 1 999, 9, 175-187. May, P. W. Philosophical Transactions of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences 2000, 358, 473-495. Sakharova, A. Y. P., Y. V.;Quarto, F. D.; Piazza, S.; Sunseri, C.; Teremetskaya, i. G.; Varnin, V. P. J. Electrochem. Soc. 1995, 142, 2704. 225 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. Pleskov, Y. V. Uspekhi Khimii 1999, 68, 41 6-429. Pleskov, Y. V. J. Anal. Chem. 2000, 55, 1045-1050. DeClements, R.; Swain, G. M.; Dallas, T.; Holtz, M. W.; Herrick, R. D.; Stickney, J. L. Langmuir 1996, 12, 6578-6586. Ramesham, R.; Rose, M. F. Diamond Relat. Mater. 1997, 6, 17-27. Ramesham, R.; Rose, M. F. Thin Solid Films 1997, 300, 144-153. Ramesham, R.; Roppel, T.; Ellis, 0.; L00, 8. H. J. Electrochem. Soc. 1991, 138, 2981 -2984. Strojek, J. W.; Granger, M. C.; Swain, G. M.; Dallas, T.; Holtz, M. W. Anal. Chem. 1996, 68, 2031-2037. . Chen, 0. Y.; Granger, M. C.; Lister, T. E.; Swain, G. M. J. Electrochem. Soc. 1997, 144. 3806-3812. Terashima, C.; Rao Tata, N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2002, 74, 895-902. Rao Tata, N.; Loo, B. H.; Sarada, B. V.; Terashima, C.; Fujishima, A. Anal. Chem, 2002, 74, 1578-1583. Patel, K.; Hashimoto, K.; Fujishima, A. Denki Kagaku oyobi Kogyo Butsun’ Kagaku 1992, 60, 659-661 . Fujishima, A.; Popa, E.; Wu, Z.; Rao, T. N. Novel Trends in Electroorganic Synthesis, [Papers presented at the International Symposium on Electroorganic Synthesis], 3rd, Kurashiki, Japan, Sept. 24-27, 19971998, 421-424. Ohta, N.; Nakabayashi, 8.; Fujishima, A. Chem. Lett. 1999, 1147-1148. Rao, T. N.; Yagi, l.; Miwa, T.; Tryk, D. A.; Fujishima, A. Anal. Chem. 1999, 71, 2506-251 1. 226 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A. J. Electroanal. Chem. 2000, 491, 175-181. Sarada, B. V.; Rao, T. N.; Tryk, D. A.; Fujishima, A. Proceedings - Electrochemical Society 2000, 99-32, 502-506. Chailapakul, O.; Aksharanandana, P.; Frelink, T.; Einaga, Y.; Fujishima, A. Sens. Actuators B 2001 , 880, 193-201 . Notsu, H.; Fukazawa, T.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. Electrochemical and Solid-State Letters 2001, 4, H1 -H3. Tsunozaki, K.; Einaga, Y.; Rao, T. N.; Fujishima, A. Chem. Lett. 2002, 502-503. Abaturov, M. A.; Eikin, V. V.; Krotova, M. D.; Mishuk, V. Y.; Sakharova, A. Y.; Pleskov, Y. V. Russ. J. Electrochem. 1995, 31, 1 1 19-1 124. Pleskov, Y. V.; Mishuk, V. Y.; Abaturov, M. A.; Eikin, V. V.; Krotova, M. D.; Varnin, V. P.; Teremetskaya, l. G. J. Electroanal. Chem. 1995, 396, 227-232. Pleskov, Y. V.; Eikin, V. V.; Abaturov, M. A.; Krotova, M. D.; Mishuk, V. Y.; Varnun, V. P.; Teremetskaya, l. G. J. Electroanal. Chem. 1 996, 413, 105-1 10. Pleskov, Y. V.; Sakharova, A. Y.; Churikov, A. V.; Varnin, V. P.; Teremetskaya, i. G. Russ. J. Electrochem. 1996, 32, 1075-1079. Modestov, A. D.; Evstefeeva, Y. E.; Pleskov, Y. V.; Mazin, V. M.; Varnin, V. P.; Teremetskaya, I. G. J. Electroanal. Chem. 1997, 431, 211-218. Pleskov, Y. V.; Mishuk, V. Y.; Abaturov, M. A.; Eikin, V. V.; Krotova, M. D.; Varnin, V. P.; Teremetskaya, l. G. Russ. J. Electrochem. 1997, 33, 61-68. Pleskov, Y. V. J. Chem. Vap. Deposition 1997, 5, 126-132. Evstefeeva, Y. E.; Krotova, M. D.; Pleskov, Y. V.; Mazin, V. M.; Eikin, V. V.; Mishuk, V. Y.; Varnin, V. P.; Teremetskaya, l. G. Russ. J. Electrochem. 1998, 34, 1 352-1357. 227 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. Pleskov, Y.; Tameev, A.; Varnin, V.; Teremetskaya, l. J. Solid State Electrochem. 1998, 3, 25-30. Pleskov, Y. V.; Evstefeeva, Y. E.; Krotova, M. D.; Eikin, V. V.; Mazin, V. M.; Mishuk, V. Y.; Varnin, V. P.; Teremetskaya, I. G. J. Electroanal. Chem. 1998, 455, 139-146. Mazin, V. M.; Evstefeeva, Y. E.; Pleskov, Y. V.; Varnin, V. P.; Teremetskaya, I. G.; Laptev, V. A. Russ. J. Electrochem. 2000, 36, 580-585. Pleskov, Y. V.; Krotova, M. D.; Evstefeeva, Y. E.; Ral'chenko, V. G.; Vlasov, II; Khomich, A. V. Russ. J. Electrochem. 2001, 37, 1 123-1 127. Bouamrane, F.; Tadjeddine, A.; Butler, J. E.; Tenne, R.; Levy-Clement, C. J. Electroanal. Chem. 1996, 405, 95-99. Bouamrane, F.; Tadjeddine, A.; Tenne, R.; Butler, J. E.; Kalish, R.; Levy- Clement, C. J. Phys. Chem. B1998, 102, 134-140. Tenne, R.; Levy-Clement, C. Israel Journal of Chemistry 1998, 38, 57-73. Giesener, J. W.; Natishan, P. M.; O'Grady, W. E.; Morrish, A. A.; Stoner, B. R.; Hagans, P. L. In US; (United States Dept. of the Navy, USA). Us, 2001, pp 5 pp. Angus, J. C.; Will, H. A.; Stanko, W. S. J. Appl. Phys. 1968, 39, 2915-2922. Poferi, D. J.; Gardner, N. C.; Angus, J. C. J. Appl. Phys. 1973, 44, 1428-1434. Sunkara, M.; Angus, J. C.; Hayman, C. 0.; Buck, F. A. Carbon 1990, 28, 745- 746. Anderson, A. B.; Mehandru, S. P.; Angus, J. C. Abstr. Pap. Am. Chem. Soc. 1991, 202, 36-FUEL. Lambrecht, W. R. L.; Lee, C. H.; Segall, B.; Angus, J. C.; Li, 2. D.; Sunkara, M. Nature 1993, 364, 607-610. Mearini, G. T.; Krainsky, l. L.; Dayton, J. A.; Wang, Y. X.; Zorman, C. A.; Angus, J. C.; Hoffman, R. W. Appl. Phys. Lett. 1994, 65, 2702-2704. 228 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. Prijaya, N. A.; Angus, J. C.; Bachmann, P. K. Diam. Rel. Mat. 1994, 3, 129-136. Anderson, A. B.; Grantscharova, E. J.; Angus, J. C. Phys. Rev. B1996, 54, 14341-14348. Jungnickel, G.; Porezag, D.; Frauenheim, T.; Heggie, M. l.; Lambrecht, W. R. L.; Segall, B.; Angus, J. C. Phys. Status Solidi A-Appl. Res. 1996, 154, 109-125. Morrison, P. W.; Taweechokesupsin, O.; Kovach, C. 8.; Roozbehani, B.; Angus, J. C. Diam. Rel. Mat. 1996, 5, 242-246. Li, L.-F.; Totir, D. A.; Vinokur, N.; Miller, B.; Chottiner, G.; Evans, E. A.; Angus, J. C.; Scherson, D. Proc. - Electrochem. Soc. 1997, 97-17, 394—409. Martin, H. B.; D'Urso, J. J.; Morrison, P. W., Jr.; Landau, U.; Angus, J. C. Proceedings - Electrochemical Society 2000, 99-32, 491 -501. Albu, T. V.; Anderson, A. B.; Angus, J. C. J. Electrochem. Soc. 2002, 149, E143- E147. Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. J. Electrochem. Soc. 1996, 143, L238-L240. Miller, B.; Vinokur, N.; Avyigal, Y.; Kalish, R. Book of Abstracts, 214th ACS National Meeting, Las Vegas, NV, September 7-11 1997, ANYL-191. Vescan, A.; Daumiller, I.; Gluche, P.; Ebert, W.; Kohn, E. IEEE Electron Device Left. 1997, 18, 556-558. Vinokur, N.; Miller, B. Proc. - Electrochem. Soc. 1997, 97-17, 384-393. Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. Electrochemical and Solid-State Letters 1998, 1, 265-267. Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. J. Electrochem. Soc. 1999, 146, 125-130. Yoo, K.; Miller, B.; Kalish, R.; Shi, X. Proceedings - Electrochemical Society 2000, 99-32, 440-447. 229 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. Sopchak, D.; Miller, B.; Kalish, R.; Avyigal, Y.; Shi, X. Electroanalysis 2002, 14, 473-478. Prado, C.; Flechsig, G.-U.; Grundler, P.; Foord John, 8.; Markenc, F.; Compton Richard, G. Analyst 2002, 127, 329-332. Compton, R. G.; Marken, F.; Goeting, C. H.; McKeown, R. A. J.; Foord, J. 8.; Scarsbrook, G.; Sussmann, R. 8.; Whitehead, A. J. Chemical Communications (Cambridge) 1998, 1 961 -1 962. Goeting, C. H.; Marken, F.; Gutierrez-Sosa, A.; Compton, R. G.; Foord, J. 8. New Diamond and Frontier Carbon Technology 1999, 9, 207-228. Saterlay, A. J.; Foord, J. 8.; Compton, R. G. Analyst 1999, 124, 1791-1796. Qiu, F.; Compton, R. G.; Marken, F, Wilkins, S. J. Goeting, C. H.; Foord, J. 8. Anal. Chem. 2000, 72, 2362-2370. Saterlay, A. J.; Tibbetts, D. F.; Compton, R. G. Anal. Sci. 2000, 16, 1055-1060. Holt, K. B.; Del Campo, J.; Foord, J. 8.; Compton, R. G.; Marken, F. J. Electroanal. Chem. 2001, 513, 94-99. Saterlay, A. J.; Foord, J. 8.; Compton, R. G. Electroanalysis 2001, 13, 1065- 1070. Saterlay, A. J.; Wilkins, S. J.; Holt, K. B.; Foord, J. 8.; Compton, R. G.; Marken, F. J. Electrochem. Soc. 2001, 148, E66-E72. Tsai, Y.-C.; Coles, B. A.; Holt, K.; Foord, J. 8.; Marken, F.; Compton, R. G. Electroanalysis 2001, 13, 831 -835. Hyde, M.; Saterlay, A. J.; Wilkins, S. J.; Foord, J. 8.; Compton, R. G.; Marken, F. J. Solid State Electrochem. 2002, 6, 183-187. Prado, C.; Flechsig, G.-U.; Gruendler, P.; Foord, J. 8.; Marken, F.; Compton, R. G. Analyst 2002, 127, 329-332. 230 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. Gruen, D. M.; Krauss, A. R.; Zuiker, C. D.; Csencsits, R.; Tenninello, L. J.; Carlisle, J. A.; Jimenez, l.; Sutherland, D. G. J.; Shuh, D. K.; et al. Appl. Phys. Lett. 1996, 68, 1640-1642. Qin, L. 0.; Zhou, D.; Krauss, A. R.; Gruen, D. M. Nanostructured Materials 1998, 10, 649-660. Zhou, D.; Gruen, D. M.; Qin, L. C.; McCauley, T. G.; Krauss, A. R. J. Appl. Phys. 1998, 84, 1981 -1 989. Zhou, D.; McCauley, T. G.; Qin, L. C.; Krauss, A. R.; Gruen, D. M. J. Appl. Phys. 1998, 83, 540—543. Gruen, D. M. MRS Bull. 1998, 23, 32-35. Erdemir, A.; Fenske, G. R.; Krauss, A. R.; Gruen, D. M.; McCauley, T.; Csencsits, R. T. Surface and Coatings Technology 1999, 120-121, 565-572. Fausett, B.; Granger, M. C.; Hupert, M. L.; Wang, J.; Swain, G. M.; Gruen, D. M. Electroanalysis 2000, 12, 7-15. Bhattacharyya, 8.; Auciello, O.; Birrell, J.; Carlisle, J. A.; Curtiss, L. A.; Goyette, A. N.; Gruen, D. M.; Krauss, A. R.; Schlueter, J.; Sumant, A.; Zapoi, P. Appl. Phys. Lett. 2001, 79, 1441 -1443. Krauss, A. R.; Auciello, O.; Ding, M. 0.; Gruen, D. M.; Huang, Y.; Zhimov, V. V.; Givargizov, E. l.; Breskin, A.; Chechen, R.; Shefer, E.; Konov, V.; Pimenov, S.; Karabutov, A.; Rakhimov, A.; Suetin, N. J. Appl. Phys. 2001, 89, 2958-2967. Chen, 0.; Gruen, D. M.; Krauss, A. R.; Corrigan, T. D.; Witek, M.; Swain, G. M. J. Electrochem. Soc. 2001, 148, E44-E51. Zapoi, P.; Stemberg, M.; Curtiss, L. A.; Frauenheim, T.; Gruen, D. M. Phys. Rev. B 2002, 6504, art. no.-045403. Krauss, A. R.; Gruen, D. M.; Auciello, O. H.; Carlisle, J. A.; Swain, G. M.; Ding, M. In PCT Int. Appl.; (The University of Chicago, USA). Wo, 2002, pp 44 pp. Martin, H. B.; Morrison, P. W., Jr. Electrochemical and Solid-State Letters 2001, 4, E17-E20. 231 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. Witek, M.; Wang, J.; Stotter, J.; Hupert, M.; Haymond, 8.; Sonthalia, P.; Swain, G. M.; Zak, J. K.; Chen, 0.; Gruen, D. M.; Butler, J. E.; Kobashi, K.; Tachibana, T. Journal of Wide Bandgap Materials 2002, 8, 171-188. Argoitia, A.; Martin, H. B.; Rozak, E. J.; Landau, U.; Angus, J. C. Mater. Res. Soc. Symp. Proc. 1996, 416, 349-354. Stotter, J. M.; Haymond, 8.; Butler, J. E.; Swain, G. M.; Zak, J. K. Abstracts of Papers, 222nd ACS National Meeting, Chicago, IL, United States, August 26-30, 2001 2001, ANYL-1 17. Zak, J. K.; Butler, J. E.; Swain, G. M. Anal. Chem. 2001, 73, 908-914. Xu, J. 2.; Swain, G. M. Anal. Chem. 1998, 70, 1502-1510. Reuben, C.; Galun, E.; Tenne, R.; Kalish, R.; Muraki, Y.; Hashimoto, K.; Fujishima, A.; Butler, J. M.; Levy-Clement, C. NATO ASI Sen, Ser. 31995,, 1, 137-142. Granger, M. C.; Xu, J.; Strojek, J. W.; Swain, G. M. Anal. Chim. Acta1999, 397, 145-161. Gherardini, L.; Comnineilis, C.; Vatistas, N. Annali di Chim. 2001 Mar-Apr, 91, 161-168 Terashima, C.; Rao, T. N.; Sarada, B. V.; Fujishima, A. Chemical Sensors 2001, 17, 124-126. Terashima, C.; Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2002, 74, 895-902. Fujishima, A.; Terashima, 0.; Honda, K.; Sarada, B. V.; Rao, T. N. New Diamond and Frontier Carbon Technology 2002, 12, 73-81. Spataru, N.; Sarada, B. V.; Popa, E.; Tryk, D. A.; Fujishima, A. Anal. Chem. 2001 73, 514-519 Show, Y. W., M. A.; Sonthalia, P.; Swain, G. M. accepted Chem. Mater. 232 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. Saterlay, A. J.; Foord, J. 8.; Compton, R. G. Analyst 1999 , 124, 1791 -1796 Manivannan, A.; Tryk, D. A.; Fujishima, A. Electrochemical and Solid-State Letters 1999, 2, 455-456. Saterlay, A. J.; Agra-Gutierrez, C.; Taylor, M. P.; Marken, F.; Compton, R. G. Electroanalysis 1999, 11, 1083-1088. Madore, C.; Duret, A.; Haenni, W.; Perret, A. Proceedings - Electrochemical Society 2000, 2000-19, 159-168. Saterlay, A. J.; Marken, F.; Foord, J. 8.; Compton, R. G. Talanta 2000, 53, 403- 415. Manivannan, A.; Seehra, M. 8.; Tryk, D. A.; Fujishima, A. Anal. Lett. 2002, 35, 355-368. Prado, C.; Wilkins, S. J.; Marken, F.; Compton, R. G. Electroanalysis 2002, 14, 262-272. Angus, J. C. MRS Bull. 1996, 21, 4-4. Angus, J. C.; Li, Z.; Sunkara, M.; Gat,.R.; Anderson, A. B.; Mehandru, S. P.; Gels, M. W. Proc. - Electrochem. Soc. 1991, 91-8, 125-141. Bachmann, P. K. EMIS Datareviews Series 1994, 9, 354-363. Bachmann, P. K. EMIS Datareviews Series 1994, 9, 349-353. Bachmann, P. K. EMIS Datareviews Series 1994, 9, 364-367. Gruen, D. M. Annual Review of Materials Science 1999, 29, 211-259. Spitsyn, B. V.; Aleksenko, A. E.; Sokolina, G. A.; Laptev, V. A. New Diamond Sci. Technol., Proc. Int. Conf., 2nd 1991 , 909-916. Looi, H. J., Pang, L.Y.S. , Molloy, A.B., Jones, F., Foord, J8, and Jackman, RB. 1998, 7, 550. 233 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. Popa, E., Notsu, H., Miwa, T., Tryk, D. A., Fujishima, A. Electrochem. Solid State Lett. 1999,49. Spear, K. E. a. D., J. P Synthetic Diamond: Emerging CVD Science and Technology; John Wiley 8 Sons, Inc: New York, 1994. Granger, M. C.; Swain, G. M. J. Electrochem. Soc. 1999, 146, 4551-4558. Joiley, 8.; Keppang, M.; Jackson, T.; Swain, G. M. Anal. Chem. 1997, 69, 4099- 4107. Perrin, D. D. Dissociation Constant of Organic Bases in Aqueous Solution,; London, Butterworth, 1965. Swain, G. M. J. Electrochem. Soc. 1994, 141, 3382-3393. Natishan, P. M. M., A. Mater. Lett. 1989, 8, 269. Rosman, N. A., L.; Chabert, J.P.; Verven, G.; Lucazeau, G. J. Appl. Phys. 1995, 78, 519. Martin, H. B.; Argoitia, A.; Angus, J. C.; Landau, U. J. Electrochem. Soc. 1999, 146, 2959-2964. Beck, F.; Krohn, H.; Kaiser, W.; Fryda, M.; Klages, C. P.; Schafer, L. Electrochim. Acta 1998, 44, 525-532. Fisher, V.; Gandini, D.; Laufer, 8.; Blank, E.; Comnineilis, C. Electrochim. Acta 1998, 44, 521-524. Belton, D. N. 8., S. J. J. Appl. Phys. 1989, 66, 4223. Belton, D. N. S., S. J. Thin Solid Films 1992, 212, 68-80. Tachibana, T. Y., Y.; Miyata, K.; Onishi, T.; Kobashi, K.; Tarutani, M.; Takai, Y.; Shimizu, R.; Shintani, Y. Phys. Rev. B1997, 56, 15967. Collins, A. T. Diam. Rel. Mat. 1992, 1, 457. 234 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. Merrnoux, M.; Fayette, L.; Marcus, 8.; Rosman, N.; Abello, L.; Lucazeau, G. Physica Status Solidi A: Applied Research 1 996, 154, 55-68. Michler, J.; Merrnoux, M.; von Kaenel, Y.; Haouni, A.; Lucazeau, 6.; Blank, E. Thin Solid Films 1 999, 357, 189-201. Chen, 0.; Granger, M. C.; Lister, T. E.; Swain, G. M. J. Electrochem. Soc. 1997, 144, 3806-3812. Ushizawa, K.; Watanabe, K.; Ando, T.; Sakaguchi, l.; Nishitani-Gamo, M.; Sato, Y.; Kanda, H. Diam. Rel. Mat. 1998, 7, 1719. Nicholson, R. Analytical Chemistry 1965, 37, 1351 . Natl. Bur. Stand. (U8), 0., ll, 5, 1953. La Placa, S. P. l. o. B., Brooklyn, NY, ICDD Grant-ln-Aid, 1959. Dennison, J. R.; Holtz, M.; Swain, G. Spectroscopy 1996, 11, 38-46. Mermoux, M. F., L.; Marcus, B.; Rosman, N.; Abello, L.; Lucazeau, G. Phys. Status Solidi 1 996, 154, 55. CRC Handbook of Chemistry and Physics, t. E. 1985. Butler, J. E.; Windischmann, H. MRS Bull. 1998,23, 22-27. Frenklach, M. J. Appl. Phys. 1989, 65,5142. Hsu, W. L. J. Vac. Sci. Technol., A1988, 6, 1803-1811. McCauley, T. G.; Gruen, D. M.; Krauss, A. R. Appl. Phys. Lett. 1998, 73, 1646- 1648. Jiao, 8.; Sumant, A.; Kirk, M. A.; Gruen, D. M.; Krauss, A. R.; Auciello, O. J. Appl. Phys. 2001 , 90, 1 18-122. 235 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. Keblinski, P.; Wolf, 0.; Cleri, F.; Phillpot, S. H.; Gleiter, H. MRS Bull. 1998, 23, 36-41. Wang, J.; Swain, G. M.; Merrnoux, M.; Lucazeau, G.; Zak, J.; Strojek, J. W. New Diamond and Frontier Carbon Technology 1999, 9, 317-343. Yagi, l. N., H.; Kondo, T.; Tryk, D. A; Fujishima, A. J. Electroanal. Chem. 1999, 173, 473. Alehashem, 8.; Chambers, F.; Strojek, J. W.; Swain, G. M.; Ramesham, R. Anal. Chem. 1995, 67, 2812-2821. Fujishima, A.; Rao, T. N.; Tryk, D. A. Proceedings - Electrochemical Society 2000, 99-32, 383-388. Pleskov, Y. V; Mazin, V. M.; Evstefeeva, Y. E.; Varnin, V. P.; Teremetskaya, l. G.; Laptev, V. A. Electrochem. Solid State Lett.2000,3,141-143. Vail, S. D.; Yang, H.-H.; McCreery, R. L. Proceedings - Electrochemical Society 1999, 99-5, 33-36. Cvacka, J. a. S., G. M. unpublished results. McDermott, C. A. K., K. R.; McCreery, R. L. J. Electrochem. Soc. 1993, 140, 2593. Nemanich, R. J.; Glass, J. T.; Lucovsky, G.; Shroder, R. E. J. Vac. Sci. Technol., A 1988, 6, 1783-1787. Knight, D. S.; White, W. B. J. Mater. Res. 1989, 4, 385-393. Bergman, L.; Nemanich, R. J. J. Appl. Phys. 1995, 78, 6709-6719. Prawer, 8.; Nugent, K. W.; Jamieson, D. N.; Orwa, J. O.; Bursill, L. A.; Peng, J. L. Chemical Physics Letters 2000, 332, 93-97. Ferrari, A. C.; Robertson, J. Phys. Rev. 82001, 63, 121 236 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. Adams, R. N., 1969, Electrochemistry at Solid Electrodes, Marcel Dekker: New York. He, L.; Anderson, J. R.; Franzen, H. F.; Johnson, D. C. Chem. Mater. 1997, 9, 715-722. Luo, F., Zhang, F., Baldwin, R. P. Anal. Chim. Acta1991, 244, 169. Masui, M.; Sayo, H.; Tsuda, Y. J. Chem. Soc., B 1968, 973-976. Hampson, N. A.; Lee, J. B.; MacDonald, K. l.; Shaw, M. J. J. Chem. Soc. B 1970, 1766-1769. Hampson, N. A.; Lee, J. B.; Morley, J. R.; MacDonald, K. l.; Scanlon, B. F. Tetrahedron 1970, 26, 1109-1 1 14. Hampson, N. A.; Lee, J. B.; Morley, J. R.; Scanlon, B. F. Can. J. Chem. 1969, 47, 3729-3736. Feng, J.; Johnson, D. C.; Lowery, 8. N. J. Electrochem. Soc. 1995, 142, 2618-. 2620. P0povic, N. D.; Johnson, D. C. Anal. Chem. 1998, 70, 468-472. Yeo, l. H.; Johnson, D. C. J. Electroanal. Chem. 2001, 495, 1 10-1 19. Yeo, I. H.; Lee, Y. S.; Mho, S. l.; Johnson, D. C. J. Electrochem. Soc. 1992, 139, L49-L50. Yeo, l. H.; Kim, 8.; Jacobson, R.; Johnson, D. C. J. Electrochem. Soc. 1989, 136, 1395-1401 . Yeo, l. H.; Johnson, D. C. J. Electroanal. Chem. 2000, 484, 157-163. Popovic, N. D.; Johnson, D. C. Anal. Chem. 1998, 70, 468-472. Austin-Harrison, D. S.; Johnson, D. C. Electroanalysis 1989, 1,189-197. 237 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. Feng, J.; Houk, L. L.; Johnson, D. C.; Lowery, S. N.; Carey, J. J. J. Electrochem. Soc. 1995, 142, 3626-3632. Feng, J.; Johnson, D. C.; Lowery, 8. N.; Carey, J. J. J. Electrochem. Soc. 1994, 141, 2708-2711. Hines, T. G.; Johnson, D. C. in US; (Pine Instrument Co., USA). Us, 1975, pp 6 PP- Houk, L. L.; Johnson, 8. K.; Feng, J.; Houk, R. 8.; Johnson, D. C. J. Appl. Electrochem. 1998, 28, 1 1 67-1 177. Johll, M. E.; Wong, 8. 8.; Johnson, D. C. Electroanalysis1999, 11, 1049-1054. Johnson, D. C. Clean and Efficient Processing: Electrochemical Technology for Synthesis, Separations, Recycle, and Environmental Improvement, lntemational Forum, Electrolysis in the C hemical Industry, 12th, Clearwater Beach, Fla., Oct. 1 1- 15,1998, 197-220 (Abstract) Johnson, D. C.; Feng, J.; Houk, L. L. Electrochim. Acta 2000, 46, 323-330. Johnson, D. C.; Houk, L. L.; Feng, J. In US; (United States Dept. of Energy, USA), 2001, PD 15 pp. Johnson, D. C.; Polta, J. A.; Polta, T. 2.; Neuburger, G. G.; Johnson, J.; Tang, A. P. C.; Yeo, l. H.; Baur, J. J. Chem. Soc., Faraday Trans. 1 1986, 82, 1081-1098. Johnson, D. C.; Ryan, M. D.; Wilson, G. 8. Anal. Chem. 1986, 58, 33R-49R. Johnson, 8. K.; Houk, L. L.; Feng, J.; Houk, R. 8.; Johnson, D. C. Env. Sci. Tech. 1999, 33, 2638-2644. Mho, 8.; Johnson, D. C. J. Electroanal. Chem. 2001, 495, 152-159. Xu, J. 8.; Granger, M. C.; Chen, D. Y.; Strojek, J. W.; Lister, T. E.; Swain, G. M. Anal. Chem. 1997, 69, A591-A597. Xu, J., 2000, Logan, UT. 238 258. Swain, G. M. Adv. Mater. (Weinheim, Ger.) 1994, 6, 388-392. 259. Koppang, M. D.; Witek, M.; Blau, J.; Swain, G. M. Anal. Chem. 1999, 71, 1188- 1195. 260. Ager, J. W.; Walukiewicz, W.; McCluskey, M.; Plano, M. A.; Landstrass, M. i. Appl. Phys. Lett. 1995, 66, 616-618. 261. Gonon, P.; Deneuville, A.; Fontaine, F.; Gheeraert, E.; Campargue, A.; Chenevier, M.; Rodolphe, 8. J. Appl. Phys. 1995, 78, 7404-7406. 262. Gonon, P.; Gheeraert, E.; Deneuville, A.; Fontaine, F.; Abello, L.; Lucazeau, G. J. Appl. Phys. 1995, 78, 7059-7062. 263. Ushizawa, K.; Gamo, M. N.; Xiao, C.; Watanabe, K.; Sato, Y.; Ando, T. Proceedings - Electrochemical Society 2000, 99-32, 247-251. 264. Robins, L. H.; Farabaugh, E. N.; Feldman, A. J. Mater. Res. 1990, 5, 2456-2468. 265. Chow, Y. L.; Danen, W. C.; Nelsen, S. F.; Rosenblatt, D. H. Chem. Rev. 1978, 78, 243-274. 266. Fatal, 8. The Chemistry of Amino Groups; John Wiley and Sons, 1982. 267. Baizer, M. M. Organic electrochemistry; an introduction and a guide; M. Dekker: New York, 1973. 268. Parsons, R.; VanderNoot, T. J. Electroanal. Chem. lnterfacial Electrochem. 1988, 257, 9-45. 269. Hull, 1.. A.; Davis, G. T.; Rosenblatt, D. H.; Mann, C. K. J. Phys. Chem. 1969, 73, 2142-2146. 270. Portis, L. C.; Bhat, V. V.; Mann, C. K. J. Org. Chem. 1970, 35, 2175-2178. 271. Buckley, D.; Dunstan, 8.; Henbest, H. B. J. Chem. Soc. 1957, 4901-4905. 272. Curragh, E. F.; Henbest, H. B.; Thomas, A. J. Chem. Soc. 1960, 3559-3563. 239 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. Shechter, H.; Rawalay, S. S. J. Am. Chem. Soc. 1964, 86, 1706-1709. Dunstan, S.; Henbest, H. B. J. Chem. Soc. 1957, 4905-4908. Nakagawa, K.; Onoue, H. Chem. Commun. 1965, 396. Nakagawa, K.; Onoue, H. Tetrahedron Letters 1965, 1433-1436. Clarke, T. G.; Hampson, N. A.; Lee, J. B.; Morley, J. R.; Scanlon, B. Tetrahedron Lett. 1968, 5685-5688. Clarke, T. G.; Hampson, N. A.; Lee, J. B.; Morley, J. R.; Scanlon, B. F. J. Chem. Soc, C1970, 815-817. Rosenblatt, D. H.; Davis, G. T.; Hull, L. A.; Forberg, G. D. J. Org. Chem. 1968, 33, 1649-1650. Mann, C. K. Anal. Chem. 1964, 36, 2424-2426. Canizares, P.; Martinez, F.; Diaz, M.; Garcia-Gomez, J.; Rodrigo, M. A. J. Electrochem. Soc. 2002, 149, D1 18-D124. Foti, G.; Gandini, D.; Comnineilis, C.; Perret, A.; Haenni, W. Electrochemical and Solid-State Letters 1 999, 2, 228-230. Hull, L. A.; Giordano, W. P.; Rosenblatt, D. H.; Davis, G. T.; Mann, C. K.; Milliken, S. B. J. Phys. Chem. 1969, 73, 2147-2152. Hull, L. A.; Davis, G. T.; Rosenblatt, D. H. J. Amer. Chem. Soc. 1969, 91, 6247- 6250. Ge, J.; Johnson, D. C. Book of Abstracts, 210th ACS National Meeting, Chicago, IL, August 20-24 1995, ANYL-065. Xu, J.; Swain, G. M. Anal. Chem. 1998, 70, 1502-1510. Treimer, S. E.; Feng, J. R.; Scholten, M. D.; Johnson, D. C.; Davenport, A. J. J. Electrochem. Soc. 2001, 148, E459-E463. 240 288. 289. 290. 291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. Kawagoe, K. T.; Johnson, D. C. J. Electrochem. Soc. 1994, 141, 3404-3409. Vitt, J. E.; Johnson, D. C. J. Electrochem. Soc. 1992, 139, 774-778. Bard, A. J. F., L. R. Electrochemical Methods; John Wiley & Sons, 1980. Witek, M. A.; Swain, G. M. Anal. Chim. Acta 2001, 440, 119-129. Hughes, 8.; Johnson, D. C. Anal. Chim. Acta1983, 149, 1-10. Caselia, i. G.; Cataldi, T. R. l.; Salvi, A. M.; Desimoni, E. Anal. Chem. 1993, 65, 3143-3150. Treimer, S.; Tang, A.; Johnson, D. C. Electroanalysis 2002, 14, 165-171. Johll, M. E.; Asala, K. 8.; Swarts, M.; Anderegg, J. W.; Johnson, D. C. Electroanalysis 2000, 12, 248-254. Witek, M.; Swain, G. M. submitted to Anal. Chem. 2002. Ruzicka, J.; Hansen, E. H. Chemtech 1979, 9, 756-764. Ruzicka, J.; Hansen, E. H. Anal. Chim. Acta1980, 114, 19-44. Ruzicka, J. Anal. Chem. 1983, 55, 1040A-1042A, 1044A, 1046A-1048A, 1050A, 1 052A-1 053A. Ruzicka, J.; Hansen, E. H. TrAC, Trends in Analytical Chemistry1998, 17, 69-73. Ruzicka, J.; Hansen, E. H. Anal. Chem. 2000, 72, 212A-217A. Ruzicka, J. Anal. Chem. 2001, 73, 385A. LaCourse, W. R. Pulsed Electrochemical Detection in High-Performance Liquid Chromatography, 1 997. 241 304. 305. 306. 307. Goss, C. A.; Brumfield, J. C.; lrene, E. A.; Murray, R. W. Anal. Chem. 1993, 65, 1378-1389. Hathcock, K. W.; Brumfield, J. C.; Goss, C. A.; lrene, E. A.; Murray, R. W. Anal. Chem. 1995, 67, 2201-2206. Notsu, H.; Yagi, l.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. Electrochemical and Solid-State Letters 1999, 2, 522-524. Martin, H. B.; Argoitia, A.; Landau, U.; Anderson, A. B.; Angus, J. C. J. Electrochem. Soc. 1996, 143, L133-L1 36. 242 HILAN STATE UNIVERSITY LIBRARIES 1293 02327 08 57 El'l‘l'l'n'l'n" VJ." Ii i