lHHlliiMl‘HJFl HHIHH ii — m Ml ‘ HM I48 (I)\JCD L.‘ LIBRARy _ Michigan Sm: U . . \\\\\%\ «\Lxxgxmn \\\\\\\\\2 x; W saw 1 \ 1 M OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. ELECTROCHEMICAL CHARACTERIZATION OF QUINONES ADSORBED ONTO EDGE PYROLYTIC GRAPHITE by Daniel P. Deibel A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1979 ABSTRACT ELECTROCHEMICAL CHARACTERIZATION OF QUINONES ADSORBED ONTO EDGE PYROLYTIC GRAPHITE by Daniel P. Deibel The literature indicates recent interest in modifying solid electrodes for catalysis of electrochemical reactions. Covalent bonding and adsorption are techniques for modifying electrodes which have been investigated. These works indicate that electrochemical re- versibility, fast electron transfer and long attachment lifetime of the species are criteria in the preparation of modified electrodes. These electrodes are then used to catalyze specific reactions. An investigation is described which was used to establish which structural characteristics were best suited to meet the criteria for adsorption onto edge pyrolytic graphite electrodes. Cyclic and differential pulse voltammetries were used in these charac- terizations and to note any catalytic effects for specific reactions. Of the three molecules investigated, which included vitamins K1 and K and alpha-tocopherol quinone, vitamin K3 showed the most 3 promise for the above application. The vitamin K3 electrode was electrochemically reversible and exhibited fast electron transfer. It, however, failed to catalyze specific reactions. The vitamin K1 and alpha-tocopherol quinone were not suitable choices due to their irreversible electrochemical behavior. This paper describes the techniques and reactions employed to fully characterize these quinones and their use as electrochemical electrode modifiers. ACKNOWLEDGMENTS I acknowledge Dr. Neal Armstrong for his aid and helpfulness in this project. I thank my wife, Carla, for her support, patience, and help- fulness in the preparation of this thesis. I also thank my parents for their assistance and advice throughout the years. 11 TABLE OF CONTENTS Chapter Page I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . 1 A. ADSORPTION. . . . . . . . . . . . . . . . . . . . . . 2 B. QUINONES. . . . . . . . . . . . . . . . . . . . . . . 8 C. ELECTROCHEMICAL METHODS . . . . . . . . . . . . . . . 14 D. ELECTRODE PREPARATION AND CHARACTERIZATION. . . . . . 18 E. CHEMICALLY MODIFIED ELECTRODES BY COVALENT ATTACHMENT. . . . . . . . . . . . . . . . . . . . . . 22 F. CHEMICALLY MODIFIED ELECTRODES BY ADSORPTION. . . . . 25 G. CONCLUSION. . . . . . . . . . . . . . . . . . . . . . 27 II. EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . . . 29 A. ELECTROCHEMISTRY. . . . . . . . .~. . . . . . . . . . 30 B. ELECTROCHEMICAL CELL. . . . . . . . . . . . . . . . . 30 C. MATERIALS AND REAGENTS. . . . . . . . . . . . . . . . 35 D. AQUEOUS SOLUTIONS . . . . . . . . . . . . . . . . . . 35 E. ADSORPTION METHOD OF QUINONES ONTO EDGE PYROLYTIC GRAPHITE. . . . . . . .’. . . . . . . . . . 36 F. REACTIONS AT THE MODIFIED ELECTRODE . . . . . . . . . 37 III. RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . . . 39 A. ATTACHMENT OF QUINONES TO GRAPHITE ELECTRODE SURFACES BY MEANS OF ADSORPTION . . . . . . . . . . . 40 B. RATE OF LOSS OF ATTACHED REACTANTS FROM THE ELECTRODE SURFACE . . . . . . . . . . . . . . . . . . 53 C. EVALUATION OF SURFACE CONCENTRATIONS. . . . . . . . . 55 D. ELECTROCHEMISTRY IN AN OXYGEN SATURATED SOLUTION. . . . . . . . . . . . . . . . . . . . . . . 65 E. ELECTROCHEMISTRY IN AN ASCORBIC ACID SOLUTION . . . . 68 F. CONCLUSION. . . . . . . . . . . . . . . . . . . . . . 73 IV. LIST OF REFERENCES. . . . . . . . . . . . . . . . . . . . 75 LIST OF TABLES Table I Differential Pulse Voltammogram Areas From Alpha-Tocopherol Quinone Solutions. . . . II Differential Pulse Voltammogram Areas From Vitamin K1 Solutions. iv Page 60 64 LIST OF FIGURES Figure Page 1 Adsorption—desorption reaction scheme. . . . . . . . . . 6 2 Quinone structures. . . . . . . . . . . . . . . . . . . 10 3 Quinone reaction mechanism. . . . . . . . . . . . . . . 12 4 Cyclic voltammetry. . . . . . . . . . . . . . . . . . . l6 5 Potential-time sequence for differential pulse voltammograms . . . . . . . . . . . . . . . . . . . . . 17 6 Sensitivity comparison of cyclic and differen- tial pulse voltammetries. . . . . . . . . . . . . . . . 19 7 Diagram of a five-operational amplifier potentiostat . . . . . . . . . . . . . . . . . . . . . 31 8 The electrolysis cell . . . . . . . . . . . . . . . . . 32 9 The teflon electrode holder design. . . . . . . . . . . 34 10 Cyclic voltammogram of the edge pyrolytic graphite electrode. . . . . . . . . . . . . . . . . . . 41 11 Cyclic voltammetry of p-benzoquinone . . . . . . . . . . 43 12 Linear sweep voltammogram of alpha-tocopherol quinone . . . . . . . . . . . . . . . . . . . . . . . . 45 13 Differential pulse voltammogram of alpha- tocopherol quinone. . . . . . . . . . . . . . . . . . . 46 14 Cyclic voltammogram of alpha-tOCOpherol quinone. . . . . 47 15 Linear sweep voltammogram of vitamin K 48 1. l6 Differential pulse voltammogram of vitamin K 49 1. 17 Cyclic voltammogram of vitamin K1 . . . . . . . . . . . 50 18 Linear sweep voltammogram of vitamin K3. . . . . . . . . 51 19 Cyclic voltammogram of vitamin K 52 3 . . . . 20 Concentration-time plot of vitamin K3. . . . . . . . . . 56 V Figure Page 21 Adsorption isotherm of vitamin K3. . . . . . . . . . 57 22 Concentration-time plot of alpha-tOCOpherol quinone. . . . . . . . . . . . . . . . . . . . . . . 59 23 Concentration-time plot of vitamin K1. . . . . . 63 24 Linear sweep voltammograms of oxygen reductions. . . 66 25 Cyclic voltammograms of ascorbic acid oxidation on vitamin K1 electrode. . . . . . . . . . 7O 26 Cyclic voltammograms of ascorbic acid oxidation on vitamin K3 electrode. . . . . . . . . . . . . . . 71 vi CHAPTER I INTRODUCTION INTRODUCTION A. Adsorption Electrochemical adsorption studies have been undertaken for many years primarily for one important reason: adsorbed ions and molecules can dramatically alter the electrochemical processes which occur at the electrode interface. Even the smallest amounts of an adsorbed substance can sometimes catalyze, inhibit, or even eliminate electrode reactions. There are several versatile tools used for carrying out an electrochemical adsorption study, such as conventional polarography, chronopotentiometry, electrolysis with a constant potential, a.c. polarography, and stationary electrode polarography and voltam— metry. This study employed stationary electrode voltammetry using cyclic and differential pulse techniques. The former was used in characterizing several systems with small surface concentra- tions of attached reactant at a solid electrode and the latter was used in circumstances where its increased sensitivity over cyclic voltammetry (l) was essential for characterization. D.c. and a.c. polarography and linear sweep voltammetry methods were based on simple models of electrode reactions. How- ever, when complications arose, these models needed an extension of the theory. Using conventional polarography, Brdicka (2) first extended this theory in order to explain the experimental "pre- waves" by showing that these were due to reactant adsorption. Later, "post waves" were also observed (3). (Pre- and post waves are polarographic waves which occur before and after, respectively, the main wave which occurs for the electroactive unadsorbed species in solution.) Bond and Hefter (4, 5) explained these phenomena by conducting very detailed experiments in which it was demonstrated that adsorp- tion may be detected by a distortion of the d.c. or a.c. polaro- gram. Similarly by using linear sweep voltammetry, Wopshall and Shain (6) determined diagnostic criteria for both the presence and the type of adsorption. The presence of adsorption may be detected by voltammograms exhibiting either two peaks or a single enhanced peak. Other methods for the detection of adsorption described are: (1) the shape of the current—potential curve, (2) a scan rate dependence method, and (3) a concentration de- pendence method. Criteria for the type of adsorption, e.g. strong or weak, reactant and/or product, were also catagorized. Thus unknown systems could be catagorized by studying the variation of peak shape and peak current as a function of bulk concentration. The theoretical understanding of adsorption from the litera- ture is very diverse, most probably because of the combination of electrode reaction and adsorption processes which lead to a number of possible starting points. Holub's (7, 8) theoretical treatments of linear adsorption used a time stationary state approximation technique in order to describe the time dependent behavior of adsorption reacting systems at expanding electrodes. Sluyters et a1. (9) also gave a simple description of the effect of reactant adsorption in d.c. polarography at the dropping mer- cury electrode. Adsorption is a consequence of the field force at the sur- face of a solid (the adsorbent), which attracts the molecule (the adsorbate). Several different types of forces are responsible for the bonding, but these are mainly of two kinds, physical and chemical. The types of force give rise to physical (van der Waals) adsorption and chemisorption, respectively. The following forces may be identified (10): (1) Dispersion forces. These arise from fluctuations in the electron density clouds of two atoms. These charge fluctuations are sufficient to induce resonance and cause an attraction; the resultant force has a fairly long range. DiSpersion forces are almost always present and they are a major contributor to the total energy of adsorption. (2) The overlap or repulsive forces. They appear when two atoms approach very near to one another and their electron orbitals or eigenfunctions overlap. If the adsorbate and adsorbent are composed of non-polar molecules, then the dispersion and repulsive forces are the only ones which must be taken into account in treating the interaction which takes place. If one or both molecules are polar, then other forces will be effective. (3) Dipole interactions. These forces occur when a polar adsor- bent is adsorbed on a non-polar or a polar adsorbent, or whenever a non-polar adsorbate is adsorbed on a polar adsorbent. (4) Valency forces, like repulsive forces, occur at sufficiently close distances. They are due to electron orbital overlapping and are responsible for chemisorption. (5) Interaction forces between the atoms or molecules of the bound adsorbate themselves. They are considered when the coverage of adsorbate on the adsorbent reaches a stage at which the distance S separating adsorbate molecules is small. In chemisorption, a transfer of electrons between the adsor- bent and the adsorbate occurs. In effect a chemical compound is formed, but it is confined to a single layer of atoms or molecules on the surface of the adsorbent. A scheme which represents the general processes occurring at the electrode surface (9) is given in Figure l. OxA and RedA represent the oxidized and reduced adsorbed species, respectively. The rate constants for the adsorbed and desorbed oxidized species are KAO and KDO: respectively; and similarly for the reduced species, KARed and KDRed- The adsorption process involves a selective uptake of a substance from an external phase into a phase provided by the adsorbent. When an adsorbent is immersed in a solution, adsorption occurs when the solute molecules trans- fer from the solution to the adsorbent leading to a change in the concentration of the solution. A spontaneous occurrence of adsorption results in the system reaching a state of lower energy, or higher stability, as a consequence. In Figure 1, there are shown two oxidation-reduction (redox) reactions: (1) the one between the non-adsorbing species (R1) and (2) the one between the adsorbed species (R2). The adsorption-desorption rates are controlled by the rate constants for their respective species. An adsorption isotherm can also describe the adsorptive- desorptive processes in which oxidation and reduction are involved. Isotherms have been classified by Giles and co-workers (11, 12) into four main types with subdivisions. These are characterized by the curvature of the isotherm near the origin. The "3" type O + ne— Red 1(on KDOx KARed KDRecI Ox + ne- Red sz Figure 1. Adsorption - desorption reaction scheme 7 is convex, "L" or Langmurian type is concave to the concentration axis. The "H" or "high affinity" isotherm is very steep at low concentrations, implying strong preferential adsorption of the solute. And the "C" or constant partition type is straight through— out. Giles discusses the various isotherms in terms of the solid- solvent, solid-solute, and solute-solute interactions, and also by reference to the orientation of the solute molecules on the surface. The Langmuir isotherm concept has the flexibility to cover homogenious and heterogeneous surfaces, chemical and physical forces, and mono- and multi— molecular layers (l3, 14). The isotherm is derived on the basis of treating the adsorption- desorption process as chemical reactions reversible in nature. It assumes that a surface contains a finite number of identical adsorption sites and an adsorption process such that occupation of any site does not affect the characteristics of any other site. For the adsorption of two species, the Langmuir isotherm equation can be written: P =K c r/(1+K c +K c ) (1) 0x Ox 0x S Ox Ox Red Red [Red . KRed cRed [jg/’(l + KOx c0x + KRed cRed ) (2) where r‘and r. are the surface excesses, Cox and c the sur- Red Red face concentrations, '2 the maximum surface excess, and KOx = K /K =1< . AOx DOx ’ KRed ARed/KDRed Standard potentials can then be assigned to the different reactions. E: , pertaining to reaction (1), is defined as the potential of an electrode in equilibrium with equal concentrations 0 of unadsorbed Ox and Red. E2 , pertaining to reaction (2), is defined as the potential of an electrode in equilibrium with equal 8 surface excesses of adsorbed Ox and Red. The Nernst equation may be written as follows: Eeq. = E: + (RT/nF)ln (cox/ CRed) (3) 13er = 133 + (RT/nF)ln (CAOx/ARed) (4) = E‘z’ + (RT/nF)ln (FOX/ fled) (5) With equations (1) and (2) this leads to o 0 E2 = E1 - (RT/nF)ln (KOx/KRed) or (6) O O O 0 E2 = E]. - (AGOx -AGRed)/ nF (7) If the electrode reactions are reversible, the peak poten- tials will be related to the standard free energies of adsorption and therefore the waves will be in a certain order. For the two reactions, (R1) and (R2), from equation (6) it can be summized for the reductions of Ox and AOx that (a) if KOx> KRed’ reaction (R2) produces a postwave (b) if KOx < KRed’ reaction (R2) produces a prewave (c) if KOx = KRed’ E20= El0 and both (R1) and (R2) occur at the same potential. The separation between the peak potentials of the pre- and post peaks and the main peak is related to the standard free energy of adsorption. B. Quinones The quinone family constitutes an interesting system of oxidation-reduction reactions. The family is a typical example of a reversible system which involves rapid transfer of both electrons and protons (15, 17). Such a system can be used to test new electrode materials, new electrochemical techniques, and new solvent systems. Quinone redox reactions are of interest in this study in order to characterize their behavior on edge pyrolytic graphite and also to study reactions with the modified quinone electrode surface. Quinones, (some related to more complicated aromatic systems), have been isolated from biological sources (molds, fungi, higher plants). In many cases, they take part in oxidation-reduction cycles essential to the living organism (18). Anthroquinoid dyes are of enormous technological importance and much work has been done in devising synthesis of large ring systems embodying the quinone structure, e.g. Alizarin, Indanthrene, and Golden Yellow G.K.. The hydro- quinone/quinone couple is used in the photographic industry because of its importance in aiding the conversion of silver ion to free silver. The quinone family has the same basic structural characteristics, i.e. the cyclic diketones (Figure 2). In this figure are the quinones used in this study are shown. Because they are highly conjugated, quinones are colored; p-benzoquinone, for example, is yellow. Also their high conjugation balances the quinones energetically against the corresponding hydroquinones (the phenol containing two -OH groups). Rapid indirect polarographic and visual micromethods have been developed (19) for the determination of the quinone groups by re- duction of titanium (III) at pH 3. Vitamin K3 was first determined at the dropping mercury electrode (DME) by Fieser and Fieser (20). Lingane (21) developed a microcoulometric method for quinone deter- mination which was later updated (22) with electrogenerated cerium (IV) ion. Lindquist and Farroha (23) applied differential pulse polarography (DPP) for the detection of vitamins K1 and K3. Princeton Applied Research (24) also employs a DPP method of detection for these vitamins with E% = -O.54V for vitamin K and -1.05V for vitamin K3 1 10 p-benzoquinone O l l l O vitamin K1 C CH SEE, "'3d”\\~//L\\ 'fl.va/’~‘\I’2i‘. H H vitamin K3 0 illllllfllllllllll’ I I O alpha-tocopherol quinone EH [CH2CH2CH2 Hf3c H3 Figure 2. Quinone structures. 11 (vs. the saturated calomel electrode (SCE)) in acidic medium. Vitamin K3 has been electrochemically characterized by Patriarche and Lingane (21). Buffers of pH values below eight were employed because of vitamin K3's instability in alkaline medium. Using a hanging mercury drop electrode (HMDE), the reduction involved two electrons which was nearly reversible at pH 4.98 with a half wave po- tential of -0.085V vs. Ag/AgCl, KCl (sat.). Also, AEp (the difference between the cathodic and anodic peaks) was only 43mV, which is only 13.8 mV greater than the theorectical value of a two electron reaction. In an ideal system, an electrochemically reversible system is one for whichAEp= 58mV/n, where n equals the number of electrons trans- ferred at 25°C. The quinone/hydroquinone (Q/QHZ) system has been considered the classical organic reversible redox reaction: Q + 2H+ + 2e-‘T—‘3QH2 (8) Several workers (25, 28) have suggested the general scheme for quinone/ hydroquinone redox process of species containing one to five fused ben- zene rings (Figure 3). From a survey of the published material, this scheme provides a proper framework for the mechanism. It does not neces- sarily include all processes or does it give a precise mechanism. Several workers including Peover (29), Wawzonek (30), Elving (31, 32), and Kolthoff (33) have studied the nonaqueous chemistry of the Q/QH2 system. Eggins and Chambers (26) studied the proton effects on the reduc- tion of the Q/QH2 system and characterized the electrochemistry of it in acetonitrile at platinum electrodes. The principle mechanistic con- clusions of their work were that in the presence of strong donors, Q is 12 Q . 3 Q" V A 0‘2 I H H H+ 1+ A A _ on v QH' - QH 2+ 43 + 4;: QH2 ‘ QH2 \ QH2 Reactions across represent redox reactions each involving a single electron transfer Reactions down represent acid-base reactions each involving a single proton transfer. Figure 3. Quinone reaction mechanism 13 reduced in a two electron process, probably via the protonated qui- none, QH+. The protonation equilibrium is fast on the time scale of the electrochemical experiment. After protonation the processes that follow are the transfer of two electrons, then a proton, or symbolized by e-e-H+. This scheme was established by observation of the effect of proton donors on the cyclic voltammograms of the various quinones. Brisset (28) reported electrochemical reduction of alpha-sub- stituted quinones examined polarographically in buffered water-pyri- dine mixtures. Experimental results show a two electron wave in acidic media, and two one electron waves in alkaline media. In the basic medium, it was determined that the first wave was acid independent, and thus the mechanism was e-H+e-H+. Bressard et a1. (25) reported semi-reversible electrochemistry for the quinone couple (AEp = 0.59V) in lO-ZM HClO4 on platinum. They concluded that two mechanisms co-exist after the formation of the single protonated form of the quinone, QH+. The first reaction is H+e-e- and the second co-existing mechanism is H+ e-H+e-. Hale and Parsons (15) described possible pathways for the reduc- tion of the quinone couple in pH 4 buffer at the mercury electrode. They suggested the mechanism is either H+e-e-H+ or e-H+e-H+ and favor the former. Vetter (34) conducted detailed mechanistic studies in acidic and neutral solutions using a platinum electrode. From Tafel plots, he suggested a H+e-H+e- mechanism. In both aqueous and non-aqueous studies, general agreement is found that in acidic medium, protonation is the first step, followed by electron donation. Results are unclear whether it is H+e- or e-H+ for 14 the final two steps. An understanding of the mechanism is essential for the following reasons. An assessment of the advantages and disadvantages of working in protic or aprotic solvents could be obtained. Also characteriza- tion of reactants, intermediates, and products in these mechanistic studies could possibly be utilized in the explanation of the experi- mental data later. Finally, a knowledge of the mechanistic processes of quinones is necessary for the study of the catalytic behavior of reactions with prepared quinone electrode surfaces. Investigations involving the characterization of the Q/QHZ couple at impregnated graphite electrodes buffered protic solvents have been carried out by several workers (35, 39). Current—voltage curves (35) indicated semi-reversible characteristics in buffered solutions at pH 1, 4, and 7 for p-benzoquinone. However, the magnitude of the half peak potential shifts with pH was in good agreement with theory. Data in- dicated an overall two electron change. Reduction in buffers in pH 4.5 - 5.5 (36, 37) yielded what appeared to be well-defined reversible cathodic waves on pretreated electrodes for p-benzoquinone (E8: +0.02V vs. Ag/AgCl (sat.)). Shain and DeMars (38) report that the Q/QH2 system was irreversible at the carbon paste electrode (Nujol) with AEp values as high as 200—300 mV in a pH range from 1 to 9. A later study by Lindquist (39) found three carbon paste electrodes which showed electro- chemical reversibility for the quinone hydroquinone system at pH 1.0 and 4.8. C. Electrochemical Methods Various electrochemical methods were used in this study. Linear sweep, cyclic, and differential pulse voltammetries were employed to 15 evaluate the electrochemical activity of the quinones. Chronoam- perometry was used to determine the exposed surface area of the graphite electrode. Cyclic voltammetry (Figure 4) involves the measurement of current- voltage curves under diffusion-controlled, mass transfer conditions at a stationary electrode. Symmetrical triangular voltage scan rates are utilized which can range in the rate of voltage change from a few to hundreds of millivolts per second. The voltage increases and de- creases at the same rate. This makes possible the generation of a complete voltammogram with cathodic and anodic waves, which are re- corded or displayed on the same set of current-voltage axes. The signal may be composed of two parts: (1) the background current which is due to the charging of the electrical double layer, and (2) the Faradaic current which flows when the desired electrochemical process occurs. Figure 4 displays the perturbation applied (A) and the typical current-voltage response cycle (B). The differential pulse technique has been developed and used (1, 40, 43) to measure electroactive species because it is more sensi- tive than cyclic voltammetry. In differential pulse voltammetry (DPV), a linear voltage ramp is applied to an electrode. Superimposed on this ramp are uniform square voltage pulses which are synchronized so they will occur at definite time intervals (Figure 5). The magnitude and the period of the pulse are fixed during the run. Application of the pulse in the absence of a Faradaic reaction results in a current spike due to the charging of the double layer. This current decays almost to zero by the end of the pulse. When a pulse is applied in a po- tential region where a Faradaic reaction occurs, the current-time l6 Cathodic Anodic / APP TIME . f T ' Cathodic Anodlc fl Figure 4. Cyclic voltammetry; A. perturbation (voltage scan) applied, B. current-voltage response 17 TIME => Figure 5. Potential time sequence for differential pulse voltammograms; A. potential-time voltage pulse, B. current-time reSponse 18 behavior due to the Faradaic current occurs also. It decays according to the Cottrell equation, as 643. A sampling of the current is taken at t1, prior to the pulse application, and again at t3, just prior to the end of the pulse. Thus, the difference in currents,Ai, is dependent primarily on Faradaic current. By eliminating the unwanted contribu- tion of the charging current and the use of a high frequency pulse, the sensitivity is enhanced. Alpha-tocopherol quinone adsorbed on graphite exhibits readily detectable cyclic voltammograms because of its good surface coverage. For example, Figure 6A shows the cyclic voltammogram obtained when alpha-tocopherol quinone is adsorbed on the electrode. Figure 6B shows the DPV obtained with the same electrode. The relative improve- ment in the ratio of peak current to background current is apparent. Brown, Koval, and Anson (43) observed the electrochemistry of reactants which are irreversibly adsorbed on the surface of the elec- trodes by employing DPV. This allowed low surface coverages of reactants to be detected quantitatively and qualitatively. In later work by An- son and Brown (1), cyclic and differential pulse voltammetrics were employed to detect reactants which were chemically attached to elec- trodes. These two techniques were used to determine surface concentra— tions. Significant increases in sensitivity were attained when the magnitude of the DPV peak currents (ca. 0.05 uA) were correlated with the quantity of adsorbed reactant (ca. lO‘llmoles/cmz). D. Electrode Preparation and Characterization In general, solid electrodes are of particular utility at po- tentials beyond which mercury is oxidized (ca. 0.4V vs SCE, dependent upon the supporting electrolyte). Electroanalytical measurements using 19 l l [— l l r -0.1 -0.2 -0.1 -0.2 0 0 Figure 6. Sensitivity comparison of cyclic and differential pulse voltammetries, a 5 x 10’6M alpha-tocopherol quinone solution after two hours immersion time of edge pyrolytic graphite; A. cyclic voltammogram, scan rate 75 mV/sec, B. differential pulse voltammogram, scan rate 5 mV/sec. 20 carbon electrodes have been used and developed over the past two decades (36-39, 44, 45). The carbon electrode has been found to be a good electrode in studying the anodic processes in cases where the use of platinum or gold is complicated by surface oxide redox couples (45). Pure graphite has limited use as an electrode material because of its high porosity, which leads to high background currents. Wax im- pregnated and carbon paste electrode minimized this problem, as Adams (16) review states. Several different forms of carbon suitable for electrochemical analysis are readily available. Glass-like or vitreous carbon has been popular because of its low porosity and relatively reproducable per- formance. Its structure probably consists of layers of thin tangled ribbons of graphite—like sheets. These may be crosslinked to some ex- tent but without the extensive regular orientation of that found in graphite (46). Carbon film electrodes have recently appeared. Beilby and co- workers (47) compared the performance of wax impregnated graphite elec- trodes with the pyrolytic carbon film electrode. Mattson and co-workers (48, 49) and Blaedel (50) have done spectoelectrochemical studies using optically transparent carbon films. Pyrolytic graphite is formed when a methane-nitrogen mixture is pyrolyzed at low pressures (25—150 mm Hg) in a temperature range of lOOO—ZSOOOC (51). The basal planes of the graphite crystallites are deposited parallel to the deposition surface. This resulting structure has different chemical properties than conventional graphites. Pyrolytic graphite is anisotropic with the planes of the hexagonal graphite rings parallel to the surface of the substrate. It exhibits high strength properties parallel to the basal planes (52). It is quite 21 dense, has high purity, and exhibits high thermal and electrical con- ductivity along the plane direction. Perpendicular to the plane, the material exhibits insulator characteristics in regard to electrical conductivity. Pyrolytic graphite is highly resistant to oxidation and very impermeable to gases and liquids. Yeager and Randin have studied the differential capacitance of stress-annealed (53) and edge orienta- tion (54) pyrolytic graphite in aqueous solutions and contend that the differential capacitance of the edge pyrolytic graphite is abnormally high, 3 uF/cm2 vs. 50 uF/cm2, respectively. This was likely due to the surface roughness and surface group states, they concluded. The extensive body of literature, which describes the development of carbon electrodes during the past twenty years, was selectively re— viewed a few years ago (55), and dealt with the topic of surface oxides and other functional groupings of carbon. The previous history (thermal and chemical) and the sources both are factors which control the surface population and functional groupings. Graphite surface functions have been determined. Three widely pro- posed and supposed carbon surface functions of interest here are pheno- lic (—OH), carboxylic acid (-COOH), and quinone (=0) (56). These groups are present on carbon surfaces which have been exposed to chemical oxi- dents or to air during heating between ZOO-400°C. Additional reports of the quinone groupings are based on infrared spectra (57-59), charac- teristic quinone reactions (60), and electrochemical measurements (54, 61-63). In an electrochemical study by Miller and co—workers (64), cyclic voltammograms showed that when pyrolytic graphite was subjected to oxi- dizing conditions, a film was formed which changed the voltammetric 22 characteristics of the electrode. Blurton's study (63) on pressed mi- crocrystalline graphite concluded that the cathodic peak at ca. 0.2V (vs. the standard hydrogen electrode, SHE) was due to the reduction of quinone-like groups, as the phenol groups could not be titrated and the carboxylic acid groups were not electrochemically active in the region studied (-0.6 to +1.19V). Similar studies by Yeager and Randin (54, 65) on the edges of pyrolytic graphite concluded that the potential region at which the Faradaic surface wave occurs is similar to that observed by Blurton. Kuwana and Evans (66) studied the nature of pyrolytic graphite surfaces with surface and electroanalytical techniques. The electrodes were characterized prior to and following high frequency plasma treat- ment. A pH dependent semi-reversible surface couple was found on all electrodes. The pH dependence is consistent with a 2e-/2H+ process such as that expected for a quinone/hydroquinone type species. Plasma etch- ing increased the number of oxygen containing functional groups as seen by an increase in current in the potential region of the surface redox processes. E. Chemically Modified Electrodes by Covalent Attachment There has been considerable interest in the covalent attachment of molecules to the graphite electrode surface for the fabrication of so-called "chemically modified electrodes", (CME). Miller and co-workers (67) utilized the carboxylic acid function on oxidized graphite to im- mobilize an optically active amino acid ester through formation of an amide surface bond. After synthesis of this chiral electrode surface, it was used to perform an asymmetric reduction, where a new asymmetric center was created from the ester. Soon after this, Murray et a1. (68) demonstrated similar results with edge pyrolytic graphite which was ’7 ‘hl‘ ‘X‘ .J... _._ .__ 23 modified using the edge surface (because of the surface oxides present). Miller's amidization procedure was also used by Lennox and Murray (69) for binding tetraporphyrin on glassy carbon. Mazur and co-workers (70) described an alternative method of modi- fication for edge pyrolytic graphite electrodes. Very reactive sur- faces are produced by heating the carbon to a high temperature in a vacuum to have reactions occur at an oxide-free surface. Anson and co- workers (71) utilized both Mazur's idea and Evans' and Kuwana's (66) radiofrequency plasma etching. They prepared edge pyrolytic graphite for amide attachment to oxide free graphite by simply exposing the clean electrode to the vapor of the desired reagent and achieved promising results. In an extension of this work, Oyama and Anson (72) have covalently attached to edge pyrolytic graphite a ruthenium-EDTA complex. It readily underwent ligand substitution reactions and electron transfer was facile. Elliott and Murray (73) have examined the reactions of organo- silane with the surfaces of glassy carbon and spectroscopic graphite. The main thrust of their work to date has been directed toward the verification of the modification by surface analysis (eg. ESCA) and by electrochemical characterization of the redox behavior of the attached reactant. In a continuation of this work on graphite, Murray et a1. (74) examined the question of organosilane reagent monolayer coverage by the same methods as above. In addition to the amidization and silanization types of reactions schemes, Kuwana and co-workers (75) demonstrated the use of cyuranic chloride as.a linking agent for the attachment of redox groups to pyro- lytic graphite and metal oxide electrodes. The linkage has been shown to be quite stable. 24 The major aim of electrode modification is to fabricate a CME which can be coupled catalytically to the Faradaic turnover of solution spe- cies. Research has been directed toward the electrocatalytic reactions, specifically in the area of EC catalytic reactions in solution. Here the oxidized species is first reduced by a fast electron transfer step (reaction 9). The newly created reduced species then reacts with the electroactive species X in a chemical step to form the products Y and regeneration of the initial oxidized species Ox (reaction 10). There- fore, the name catalysis is appropriate. k S(l) (9) Ox + ne- \___——J Red at El T km 1 Red + x ————> Y + Ox (10) v— This class of reactions is based upon the principle that the electrode reaction k s 2 X + ne-—"(—")'\ Y atE v-—-—- 2 (11) is thermodynamically but not kinetically favorable at the electrode surface at potential E1. The minimum potential at which catalysis of the solution species will occur is governed by the potential of reaction (9). The potential of this reaction must be negative of the reversible redox potential for reaction (11); thus E2:> E1. However, the reduced species acts as a facile reducing agent, and reaction (10) is favored thermodynamically and kinetically. Thus AG° = -nF (E2 - E1)<0. (12) Electrocatalysts are currently being developed as a mediator species which are attached to the electrode surface for use in EC catalytic reactions. These electrodes will be of technological im- portance because of the possibility of their reversibility, reusability 25 and design to perform a specific task. Progress toward this goal was reported by Kuwana and co-workers (76). They observed an enhancement of the rate in the electrochemi- cal oxidation of a solution species. The EC catalysis of ascorbic acid was observed through the attachment of benzidine to radiofre- quency plasma treated pyrolytic graphite. Kuwana and Tse (77) have recently found that some o-quinoidal structures appear to be rather specific for the fast homogenious oxidation of NADH (dihydronicoti— namide adenosine diphosphate) to NAD+ (nicotinamide adenosine di- phosphate). NADH oxidation catalysis was through an EC catalytic mechanism where catecholamines covalently bound to graphite electrodes were used. F. Chemically_Modified Electrodes by Adsorption A number of schemes for the covalent attachment of a variety of molecules to graphite through amidization, silanization, etc. have been proposed and tested recently. In most cases, the molecules ex- hibited irreversible, shortlived electrochemical behavior. This is the Opposite of what is desired for effective catalysis of charge transfer reactions involving reactive substrates. A molecule which exhibits electrochemical reversibility is desirable because it must behave in the manner reactions (9) and (10) require for EC catalysis. The attach- ment to the electrode must be long lived for any kind of practical use. The most common approach taken in order to determine the electro- chemical properties of a molecule both easily and rapidly, and to see if these properties will solve the catalysis objective, has been by studying the electrochemical behavior of spontaneous adsorbed mole- cules in the attached state onto pyrolytic graphite. ‘w .. 26 Yeager et al (78) showed evidence for the irreversible adsorp— tion of water soluble cobalt (II) tetrasulfonated phthalocyanine in acidic and basic solutions on a stress-annealed pyrolytic graphite electrode. The kinetics of oxygen reduction were then examined, using the adsorbed electrode in the rotating disc method. Anson (43) has directed studies which relied upon the tendency of reactants, which contain multiple aromatic rings (e.g. 9,10- phenanthrene quinone, iron protoporphyrin IX, and iron tetraphenyl— porphyrin) to adsorb very strongly. These could then produce electro- active surface species which could be examined in the absence of any diffusing reactant. These species could be repeatedly cycled between oxidation states electrochemically for hours in media free of attached reactant, with little or no loss of reactant. It also proved possible to influence the electrochemistry of the adsorbed reactants by altering the composition (pH, ligand composition) of the reactant-free solutions in which their electrochemistry was inspected. The more sensitive differential pulse method proved to yield qualitative and quantitative electrochemical information concerning the adsorbed reactants. The apparent requirements for the strong adsorption to occur was the presence of a sufficiently large aromatic center of an extensive delocalized unsaturation as in the porphorin complexes. Anson and Brown (79) tested the assumption for an aromatic center by attaching a simple transitional metal complex, Ru(NH3)5L+, where L is a large aromatic ligand. The molecules electrochemical behavior was examined on basal plane pyrolytic graphite. Cyclic voltammograms showed an increase in adsorption of the complex with time exposure, until an adsorption of the complex reached a maximum value (maximum 27 surface concentration coverage). In the supporting electrolyte only, the adsorbed reactant electrode exhibited characteristic properties of the attached reactants, which underwent rapid electron transfer, but had a half-life of only one hour. Koval and Anson (80) devised experimentation to compare the electrochemical behavior and measure the surface concentrations of covalently bonded and irreversibly adsorbed attached reactants to pyrolytic graphite electrodes. The electrochemical behavior of the attached Ru(III)/Ru(II) couples for both types of attachment was virtually identical. But the difference was dramatic in terms of longevity of the attached reactants and surface concentrations. The half-lives for the spontaneously adsorbed and covalently linked species were ca. 5 and 18 hours, respectively. The adsorptive attachment was approximately an order of magnitude larger than the covalent attachment. In general, attachment by irreversible adsorption yields larger quantities of reactant on the electrode surface, but the covalently attached reactant is longer lived on the surface. Attachment by ir— reversible adsorption does have the advantage of being faster and simpler than other techniques, but is less specific in terms of attach— ment at the electrode surface. G. Conclusion Adsorbed species in charge transfer reactions have been examined from several new viewpoints in terms of attachment and electrochemical properties of immobilized redox reagents. Studies of the effects of adsorption have been important because the adsorption can inhibit or catalyze the electrode reactions. The .5» 28 development of diagnostic criteria has been important in characterizing new systems. Several methods have been recently developed which describe how reactants can be chemically attached to the graphite electrode surface. These include covalently bonded silyl, amide, and cyuranic chloride groups, adsorbed olefinic chelates, and strongly bonded aromatic com- pounds. Cyclic and differential pulse voltammetries are the most common electrochemical methods used for characterizing the attachment in terms of electrochemical behavior. This includes surface concentrations, attachment longevity, etc.. Attachment by irreversible adsorption has been found to be a simple and rapid procedure. One can characterize systems in a re- latively short period using this approach. Thus this method of attach- ment will permit rapid electrochemical screening of a molecule's po- tential catalytic usefulness. CHAPTER II EXPERIMENTAL EXPERIMENTAL A. Electrochemistry All cyclic voltammetric and chronoamperometric measurements were performed on a five amplifier potentiostat of conventional design. The circuit diagram is shown in Figure 7. There are external con- nections for the reference, auxiliary, and working electrodes with the amplifier A4 acting as a voltage follower. An external bias can be applied via amplifier Al as regulated by variable resistor R . External input, W l 1 various external potential wave-forms. A compensating circuit is , could be used to apply directed through control C2 compensating for iR drop between the working and auxiliary electrodes. Differential pulse voltammograms were obtained with a standard Princeton Applied Research Model 174A instrument. The following set- tings were used: Modulation amplitude (the magnitude of the pulse), 5 mV; d.c. potential scan rate, 5mV per sec.; clock, 0.5 sec., (i.e. pulse repetition rate 2 per sec.). In both voltammetric and chronoamperometric studies, current- voltage output was recorded with a Houston Series 2000 X-Y Recorder. For all measurements, potentials were measured and reported with re- spect to the Ag/AgCl, KCl (sat.) reference electrode and platinum auxiliary electrode. B. Electrochemical Cell The type of cell used is shown in Figure 8. The electrolysis cell was a handblown 50 ml. glass flat bottom flask. It consisted of five ground glass joint openings, three for the electrodes, and two for the gas entrance bubbler frit and gas vent. The latter two 30 C] ’\N‘ I L‘J‘w2 J T 9 b —-o ‘——’\/\/‘— . A1 W Aux Ref HI 9Worklng c2/ ‘4 Figure 7. Diagram of a five-operational amplifier potentiostat. 32 I 1“/40 Figure 8. The electrolysis cell, A. side view, B. top view. 33 openings could be sealed off by means of a stOpcock. The cell would then be airtight. The reference electrode consisted of a glass tube with a four millimeter thirsty quartz frit (Corning Glass Works, Corning, New York) sealed at one end with heat-shrinkable tubing. A silver wire with deposited silver chloride, dipped into a saturated potassium chloride solution acted as a Ag/AgCl reference. A short length of platinum wire, coiled and sealed in the cell, served as an auxiliary electrode. Male ground glass joints were on the end of both the reference and auxiliary electrodes. The working electrode was mounted in a teflon holder, similar to that described by Anson and Koval (80). A diagram of the teflon electrode holder is shown in Figure 9. The outside of the electrode holder was completely teflon. The holder consisted of two pieces, one of which is the bottom cap which can be screwed on and off for quick electrode disc mounting. Electrical contact was made via a brass rod, part of which protrudes through the top of the holder. This was pushed against the graphite disc by employing a spring. The graphite disc was seated in a teflon circular disc, which was open at the tOp and bottom. This made a tight fit and assured constant alignment of the electrode. The O-ring seal defined a reproducible area of ca. 0.54 C1112 , which was determined by measurement with calipers. There was no evidence of the solution leaking under the O-ring even after three hours of use. However, the sanding treatment given before each use of the electrode would cause the area to be larger than its geometric area. An average surface area of ca. 0.74 cm2 was determined by chronoamperometry. The method used was suggested by Adams (16) in which 34 _.._...--— Electrical Contact 75 £'— Spring —— Iran Rod «:13 cm —— Detachable lot tom / /— / / I 24/40 Taper / / / / / Graphlto Electrode Vlton O-Rlng l _ _ _ Figure 9. The teflon electrode holder design. 35 potassium ferrocyanide in 2.0 M KNO3 was employed. The electrode sur- face was recessed ca. one millimeter while stirring preceded each measurement. This insured that the solution at the electrode interface was representative of the bulk solution. C. Materials and Reagents The carbon electrodes employed were pyrolytic graphite (Union Carbide Corporation, Chicago, Illinois) obtained in the form of a cylindrical rod in which the layered (basal) planes of graphite were parallel to the axis of the rod. Surfaces cut perpendicular to the rod axis presumably contain a large number of "edge" carbon atoms and these discs will be called edge pyrolytic graphite (EPG) electrodes. Quinones used were p-benzoquinone (Matheson, Coleman, and Bell), vitamin K1 and alpha—tocopherol quinone (ICN Pharmaceuticals), and vitamin K3 (Nutritional Biochemicals Corporation). These were used without further purification. All were stored in a freezer when not in use. Sodium ferrocyanide (Coleman and Bell), ethanol (200 proof) and L(+) ascorbic acid, ACS grade (Matheson, Coleman, and Bell) were used without further purification. D. Aqueous Solutions All solutions were prepared using potassium permanganate doubly distilled deionized water and deoxygenated by bubbling the solution thoroughly with purified nitrogen gas. The pH 5.0 buffer was prepared using potassium hydrogen phthalate (KHP) and sodium hydroxide. It was ca. 0.4M in ionic strength. The pH 3.0 buffer was prepared using KHP and nitric acid, and was ca. 0.1 M ionic strength. Initial experiments were conducted in solutions containing 10’6M, 36 5 x 10‘6M, 1075M, and 5 x 10"5 M concentrations of the quinone. Be- cause of the various solubilities of the selected quinones in water, a 2 x 10-4M stock solution was prepared in pH 5.0 buffer: 200 proof ethanol mixture by volume with the following percentages for each quinone: p—benzoquinone, 100% pH 5.0; vitamin K 20:80; vitamin K 1’ 3’ 85:15; alpha-tOCOpherol quinone, 30:70. These ratios are the minimum solubilities of the quinones in aqueous solution. Subsequent dilutions to volume were with pH 5.0 buffer. E. Adsorption Method of Quinones onto EPG Graphite discs were cut from a carbon rod with a coping saw. The EPG electrode was then heated in an oven to 400°C for four hours (probably surface oxides were formed during this heat treatment), cooled, treated in a Soxhlet extractor with 200 proof ethanol for 24 hours, and dried under vacuum. It was then abraided with No. 400 silicon carbide paper to expose fresh graphite. Loose graphite particles were removed by vigorous washing with water and dried under vacuum. Used electrodes were regenerated by the above procedure, except that the heating step was eliminated. Quinones were adsorbed onto the electrode in the following manner. A 25.00 ml volume at a Specific concentration of the quinone was pipetted into the cell and deaerated with purified nitrogen for ca. 15 minutes. The electrode was then placed into the cell and the stopcocks were closed, to prevent air from seeping into the system. The electro- chemical behavior was then recorded at one minute, five minutes, ten minutes, and at ten minute intervals for the first hour, and twenty minute intervals the second hour. The solution was stirred very slowly between recordings. 37 F. Reactions at the Modified Electrode l. Electrochemistry in an oxygen saturated solution To observe if the adsorbed quinone pyrolytic graphite electrode would affect oxygen reduction, the following procedure was performed. A cyclic voltammogram of a clean EPG electrode was recorded every ten minutes for one hour in 25.00 ml of deaerated pH 5.0 buffer, to characterize the surface. At the end of the hour, the same elec- I trode was rinsed with doubly distilled water and dried by vacuum, but not sanded. It was then placed in an oxygen saturated pH 5.0 buffer for one hour and cyclic voltammograms recorded every ten minutes. The 5?. same electrode was again rinsed and dried. The quinone was then ad- w sorbed onto the EPG electrode using the procedure previously described in section E, using a 5 x 10’5M quinone solution. The adsorbed electrode was rinsed with doubly distilled water and dried in vacuum. The electrode was then placed in an oxygen saturated pH 5.0 buffer and a cyclic voltammogram was recorded every ten minutes for one hour. 2. Electrochemistry in an ascorbic acid solution To determine if an adsorbed quinone EPG electrode would affect ascorbic acid oxidation, the following procedure was used. A clean EPG electrode was characterized by cyclic voltammetry in 24.75 ml deoxygenated pH 3.0 buffer. Deoxygenated 10-4M and 10'2M ascorbic acid stock solutions were prepared in pH 3.0 buffer for later dilutions in the electrochemical cell, using a calibrated syringe. Cyclic voltam- mograms were recorded after each dilution to the following concentra- tions: 10-614, 5 x 10‘6M, 10‘5M, 5 x IO‘SM, 10'4M, 5 x 10‘4M, and 10-3M ascorbic acid. The same electrode was rinsed with doubly di- stilled water and vacuum dried. The appropriate quinone was then adsorbed 38 onto the electrode using a 5 x 10-5M solution, using the procedure described in section E. The adsorbed quinone EPG electrode was then placed into 24.75 ml of deoxygenated pH 3.0 buffer and characterized by cyclic voltammetry. The buffer is then diluted with the stock ascorbic acid solutions to the concentrations previously mentioned. After equilibrium has taken place at each concentration, cyclic voltammograms were recorded. CHAPTER III RESULTS AND DISCUSSION .a RESULTS AND DISCUSSION A. Attachment of Opinones to the Graphite Electrode Surface By Means of Adsorption The adsorption of aromatic compounds onto pyrolytic graphite has been of recent interest (43, 79, 80). Four quinones were chosen in this study to characterize the adsorption of the quinone family, in general, with respect to their aromaticity and/or their phytyl side chain. The structures of the quinones chosen are shown in Fi- gure 2. p—Benzoquinone, the basic structural species, is considered the standard quinone. Alpha-tocopherol quinone adds the aromatic dimension to the species, while vitamin K3 incorporates a phytyl side chain and aromatic character. The adsorption dependency of both these characteristics were studied. The edge pyrolytic graphite electrode was electrochemically characterized in pH 5.0 deoxygenated buffer with measurements made over a two hour period. Cyclic voltammetry indicated a semi-reversible surface couple present on this electrode and all others used in this study (Figure 10). The redox couple remains surface bound and electro- chemically active, irrespective of repeated cycling. The cathodic and anodic peak potentials were +0.13V and +0.21V, respectively. Thus the reaction was considered semi-reversible. The potential did not shift and the charging current did not measureably increase over the two hour period. Other reports (54, 56, 63, 65, 66) have shown that this Faradaic activity is in the region where the quinone/hydroquinone type species redox process occurs. Evans and Kuwana (66) reported a semi-reversible surface couple present on the EPG electrodes used in their study. Cyclic voltammetry showed an oxidation peak at +0.24V 40 41 209a 1 1 1 T l l V I l +400 +200 0 —200 -400 E (mV) Figure 10. Cyclic voltammogram of the edge pyrolytic graphite electrode in pH 5.0 deoxygenated buffer, scan rate 75 mV/sec. m: "94.4 42 (vs. Ag/AgCl (1.00M KCl) and 50 mV/sec scan rate) in pH 5.10 citrate buffer. The authors concur with Randin and Yeager (54) that the electro- active surface species was a 1,2 naphthoquinone—like structure, al- though there is no direct proof for this assignment at present. To determine if an adsorption of the selected quinones occurs onto the EPG electrode, cyclic and differential pulse voltammograms were taken at set times, following the immersion of the electrode into the solution. Peak area and shape, and characterization of the modi- fied quinone electrode in reactant-free buffer were factors in deter- mining if adsorption occurred. The p—benzoquinane/hydroquinone couple (Figure 11) gives rise to a cathodic wave centered at +180 mV in a 5 x 10'5M pH 5.0 buffered solution. The cathodic peak is very broad because of the overlapping peak from the quinone-like surface species which is electroactive at +130 mV. The peak shape has a sharp rise appearance, but diffusion control is impossible to determine because of the peak interference by the quinone-like surface species. Thus, adsorption by wave shape criteria can not be determined. Both differential pulse and cyclic voltammograms indicated no increase in peak area with time. Also, when the electrode was taken out of the solution after two hours, and placed in reactant-free pH 5.0 buffer after being rinsed with distilled water, voltammetry indicated only the quinone-like surface species present. On this basis, the conclusion that p-benzoquinone is a non-adsorbing quinone, is reached. This non-adsorbing tendency was expected because of its lack of aroma- ticity and phytyl character. 43 l T I l 1 +400 +200 0 E (mV) Figure 11. Cyclic voltammetry of p-benzoquinone in pH 5.0 buffer at 5 x 10’5M, scan rate 75 mV/sec. 44 The alpha-tocopherol quinone, vitamin K , and vitamin K3 couples 1 all give rise to a voltammetric peak in the negative potential region, as the electrode potential is scanned in a negative direction. The peak current of the cathodic process increases in magnitude as more and more of the quinone diffuses to the electrode and adsorbs on its surface. This is true for all four solution concentrations used. Linear sweep and differential pulse voltammograms show this for alpha- tocopherol quinone (Figure 12 and 13), vitamin K1 (Figure 15 and 16), and vitamin K3 (Figures 18 and 19), respectively. Adsorption is quantitatively indicated in cyclic and differential pulse voltam- metries by the observance of increasing Faradaic peak current with time, which corresponds to an increase in peak area. These data sug- gest that these three quinones readily adsorb to the EPG electrode surface. Linear sweep voltammetry shows a cathodic peak centered at -120 mV (Figure 12) for alpha-tocopherol quinone. The area under the catho- dic peak increased with time. Differential pulse voltammetry (Figure 13) shows the cathodic peak at -120mV, and also suggests that more and more of the quinone diffuses to the electrode and adsorbs. Cyclic voltammetry indicates a species (Figure 14) which gives rise to an anodic peak centered at +70mV, AEp for the redox waves is ca. 200 mV, which indicates an electrochemically irreversible reaction. Linear sweep voltammetry shows a cathodic peak centered at -375 mV for vitamin K1 (Figure 15). The area under the peak increases with each recording. Differential pulse voltammetry (Figure 16) shows the cathodic peak at ~360mV, and it too increases in peak area. Cyclic voltammetry of the couple (Figure 17) shows an anodic peak at +90 mV. 45 5\c 20m \‘~ MINUTES F .L l I 0 -200 -400 E (mV) Figure 12. Lingar sweep voltammogram of alpha-tocopherol quinone 10- M solution in pH 5.0, scan rate 75 mV/sec. 46 MINUTES Figure 13. I ‘I T -200 -400 E (mV) Differential pu se voltammogram of alpha-tocopherol quinone 5 x 10- solution in pH 5.0, scan rate 5 mV/sec. 47 40w T T T T 1 I I T T +400 +200 0 -200 -400 E (mV) Figure 14. Cyclic voltammogram of alpha-tocopherol quinone 5 x 10-5M solution, pH 5.0, scan rate 75 mV/sec. 48 I20 100 50 I . // I 5x ,/ 40 pa MINUTES J" 4% T 47m 1 -6OIO E (mV) Figure 15. Linear sweep voltammogram of vitamin K1 10’5M solution pH 5.0, scan rate 75 mV/sec. 49 § IO pa l0 MINUTES —r 1 1 l -200 -400 E (mV) Figure 16. Differential pulse voltammogram of vitamin K1 5 x 10-6M solution in pH 5.0, scan rate 5 mV/sec. 50 50 P0 +250 0 -200 -400 -600 E (mV) Figure 17. Cyclic voltammogram of vitamin K1 5 x lO'SM solution in pH 5.0, scan rate 75 mV/sec. 51 1P / 20 no MINUTES «l- \ I l l I l 0 -200 E (mV) Figure 18. Linear sweep voltammogram of vitamin K 10-5M solution in pH 5.0, scan rate 75 mV/sec. 52 1009a L__ ___ l 1 _' T I ] I TI +200 0 -200 -400 E(mV) Figure 19. Cyclic voltammogram of vitamin K 5 x 10‘5M solution in pH 5.0, scan rate 75 mV/sec. 53 Thus the AEp is 530 mV, which indicates a very electrochemically irreversible reaction. Linear sweep voltammetry shows a cathodic peak centered at —100 mV for vitamin K3 (Figure 18). The peak area increases with each recording. Cyclic voltammetry of the couple (Figure 19) shows an anodic peak at -65 mV. AEp is 35mV, which indicates an electro- chemically reversible reaction. The symmetric waves exhibit the characteristic properties of attached reactants which undergo rapid electron transfer. Adsorption is suggested in linear sweep (Figure 18) and cyclic (Figure 19) voltammetries by the symmetric shape of the cathodic peaks. This can be a qualitative indication for adsorp- tion (6, 81, 82). The differential pulse method's resolution was not employed here because the cyclic voltammograms were uncomplicated, and the sensitivity good. B. Rate of Loss of Attached Reactants From the Electrode's Surface If the adsorbed electrode (an EPG electrode which was placed in a specific quinone solution and adsorbed some of the quinone) is removed from the solution, washed with doubly-distilled water, vacuum dried and then placed in reactant-free electrolyte, cyclic voltam- metry indicates identical voltammograms to those taken at two hours in the quinone solution. These electrodes were then repeatedly cycled over a two hour period. The electrochemical response showed that the adsorbed alpha-tocopherol quinone and vitamin K1 molecules both left the electrode surface so slowly that no half-life could be calculated in the two hour interval scanned. The adsorbed vitamin K3 couple's peaks gradually diminished with time while the potential was repeatedly scanned. Approximately 80% of the adsorbed species 54 was desorbed in two hours. Two factors could account for this observation. Vitamin K3 was the most water soluble of the quinones which adsorbed. Its desorb- tion rate is probably quite large. Structurally, vitamin K3 does not have a long alkane side chain, unlike the irreversibly adsorbed quinones, alpha-tocopherol quinone and vitamin K1. It has been frequently found that larger molecules are more ad- sorbable than smaller molecules of similar chemical nature (83). Although various factors such as solubility often play a part, the behavior is in line with the Langmuir concept that adsorption is a function of the time lag between initial attachment and subsequent desorption of a molecule. Small molecules which may be attached at a single point, can be desorbed as soon as this bond is broken. Larger molecules can be adsorbed initially by becoming attached through a single atom, and the binding force is strengthened when other points of contact are made. In subsequent desorption, however, the larger molecule will not be released until all points of attachment are broken simultaneously, and this will happen less often than the break- ing of a single bond holding a smaller molecule. As observed, the effect of a substituent group is associated with the change in adsorption prOperties. A polar group such as OH will diminish the adsorption of a solute from an aqueous solution (83). This polar group makes a solute more soluble in water and cause the solute to be pulled away from the surface of the carbon. In corre- lating absorbability with orientation, data suggest that the hydro- carbon part of the molecule in vitamin K and alpha-tocopherol is at- l tached to the surface of the carbon, probably at more than one point 55 of contact because of their irreversible adsorption. Whereas, vitamin K3, the smallest of the molecules characterized here, is most likely adsorbed as a result of weak dispersion forces. Thus when the mole- cule is reduced, it becomes more soluble because of the OH polar groups, and desorption occurs. C. Evaluation of Surface Concentration The peak areas from the cyclic voltammograms were determined, when possible, and related to concentrations of adsorbed reactants. Cyclic voltammograms were recorded every ten minutes the first hour and every twenty minutes the second hour. Three concentrations of vitamin K3 could be determined. Figure 20 shows the concentration-time curve for the adsorption of the quinone from the solution to the electrode. (10-6M solution was not detect- able using cyclic voltammetry, and differential pulse voltammetry was not employed here). Each of the three concentrations eventually reach its unique sur- face saturation value. This value is 1.1 x 10-9 moles/cm2 (5 x 10-6M), 9 moles/cm2(1O-SM), and 2.3 x 10-9 moles/cm2(5 x 10-5M) for 1.9 x 10' the given solution concentrations in parenthesis. The 5 x 10-6M con- centration-time curve approaches the maximum surface coverage after one hour. The two more concentrated solutions reach this saturation value in approximately twenty minutes of electrode immersion. The rate of surface saturation is directly related to increasing solution concentration of quinOne. An adsorption isotherm for vitamin K3 is shown in Figure 21. The slope of the isotherm is concave to the concentration axis, indicat- ing that it is of the Langmuirian type. Unfortunately, other data 56 2.4 7 - . . I; I, O 2 2.4 0 o 2.0.q ‘ 1.8.... a I 1.6.4 o a l 1.4 _l w A P C‘ 1.2 E 1I? U \ M II . . 3 ‘0 E v 0 I O F X 0 h) T T l I I —I 80 100 120 TIME (minutes) Figure 20. Concentr tion-time plot 0 vitamin K , AstO‘ M, u1.Ox10' M, 05x 0'5M. CONC X 10-9 (molas/cmz) 57 1.0 I M F 10'6 10‘5 SOLN CONC (M) Figure 21. Adsorption isotherm of vitamin K3. 58 points are not available. But it is believed that the isotherm would plateau very soon after the last data point. This reasoning is be— cause a monolayer of planar vitamin K molecules with the estimated 3 ' ’ -10 dimensions of the molecule (6.2A x 5.4A) corresponds to 5 x 10 moles/cm2 so the maximum surface concentration measured at 5 x 10-5M probably approximates a fully covered surface. Adsorption onto the non-polar solid, graphite, is facilitated by the hydrophobic benzene 1 ring. The Langmuir isotherm corresponds to a flat orientation of the adsorbed quinone molecule on the graphite surface, thus suggesting that the benzene ring is parallel to the surface (11, 12, 84). .‘ The alpha-tocopherol quinone concentration-time plots (Figure 22) for the two lower concentrations attain maximum surface coverage. (Again, the 10-6M solution was not detectable by cyclic voltammetry). Saturation values were attained in ca. 80 and 50 minutes for the 6M and 10'5M solutions, respectively. These values are 4.9 x 5 x 10- 10"11 moles/cm2 for the 5 x 10-6 M solution and 1.6 x 10-10 moles/cm2 for the 10-5M solution. Since differential pulse peak area can not be directly related to concentration, just peak areas are reported here. Areas from the 10-6M and 5 x 10-6M solutions are reported in Table I. In the 10-6M solution, it is observed that the alpha- tocopherol quinone is barely detectable until after one hour of immer- sion time. Then the area of the peak increases slowly and plateaus at 80 minutes immersion time. The area increases a little faster for the 5 x 10-6M solution but finally reaches a steady value at 50 minutes, with a larger area than that for the 10-6M solution. This follows the trend noticed in the vitamin K3's electrode coverage that the surface coverage rate is related to the initial solution concentration. 59 12.. 11.- 10.. CONC X lO"°(molos / cmz) T 0 40 TIME (minutes) Figure 22. Concentration-time plot 0 alpha-tocopherol quinone, A5 x 10’6H, a 1.0 x 10‘ , o 5 x 10-514. 60 TABLE I DPV Areas from alpha-Tocopherol Quinone Solutions TIME CONCENTRATION MINUTES 10'6M 5 x 10‘6M 1 x 0.56 10 0.01 0.79 20 0.01 0.87 30 0.01 1.10 40 0.01 1.00 50 0.06 1.00 60 0.29 1.05 80 0.42 1.25 100 0.46 1.29 120 0.40 1.30 61 The highest concentration measured, 5 x 10-5M, does not reach a plateau during the two hours but has a constant increasing concentra- tion-time slope. Thus an adsorption isotherm could not be plotted. The following is an explanation for these observations concerning the concentration-time curves. The adsorbed quinone molecule is most likely in a flat orientation on the graphite surface at the lower con- centrations where equilibrium is attained. However, at the higher concentration these molecules could possibly reorient themselves and thus more new adsorption sites to be made available which corresponds to more adsorption occuring during this reorientation. The phenol molecule has been reported to act in a similar manner (11). Its attraction for adsorption arises from its hydroxy group, but is initially not monofunctional toward graphite. (Monofunctional refers to a solvent molecules large hydrophobic residue (:>CS) and is characterized by localization of forces of attraction for an ad- sorbent over a short section of its periphery). Phenol's attraction for the adsorbent lies probably in non-polar forces operating over the whole phenol nucleus. At low concentrations, the phenol molecule in water on graphite surfaces is oriented flat to the surface. At high concentrations, it is oriented normal to the surface. Also, because the 5 x 10-5M concentration-time curve increases and does not attain a plateau during the two hour time interval, the flat orientation of the molecule must be shifting to a perpendicular one and thus permits more adsorption sites to be available. With more sites open, the surface will attract more molecules. This assumption becomes more reasonable when the role of wett- ability is considered in the orientation of the adsorbed layer. Carbon 62 is hydrophobic (adverse to water). From this it has been deduced (11) that when organic substances are adsorbed from water the hydrocarbon part of the molecule is attached to the surface of the carbon and any polar groups extend into the liquid phase. This deduction can be modified because all carbons are not completely hydrophobic. The nature of the carbon surface suggests that many varied orienta- tions can occur with polar groups being held at hydrophillic centers, and hydrocarbons at other centers. The vitamin K1 concentration-time curves (Figure 23) are similar to those of alpha-tocopherol quinone. The lower concentration (5 x 10-6M) curve slowly approaches its maximum surface coverage (2.4 x 10-11 moles/cmz) which occurs at 80 minutes. Differential pulse voltammograms were recorded for 10'6M and 5 x 10-6M vitamin K1 solutions. Peak areas are reported in Table II. It is observed that in both solutions the quinone is barely detectable until after one hour immersion time. Then the areas increase and plateau for the 10'6M solution but keep increasing for the 5 x 10'6M solution. At the upper three concentrations, the curves indicate all three have increasing surface concentrations on the electrode, with a steeper slope corresponding to the increasing concentration of the solutions. The rate of adsorption corresponds to the initial concentration of the quinone solution. None attain a plateau region during the two hour interval, thus an adsorption isotherm could not be plotted. Maximum surface coverage is not attained with vitamin K1, similar to that of the alpha-tocopherol quinone curves, at the higher con- centrations. Because these two quinones share in common the long alkane side chain, it is believed that the same type of adsorption process 63 12- 11“ 1.. / 9- 3- a X 1040 (moles/c012) LI" CONC b o I 20 4O 60 80 100 120 TIME (mlnotos) 6 Figure 23. Concentrati n-time plot_gf vitamin K1, A 5 x 10- M, 01.0x10- ,05x10 M. 64 TABLE II DPV Areas from Vitamin K Solutions 1 TIME CONCENTRATION MINUTES 10'6M 5 x 10‘6 1 x x 10 0.16 0.17 20 0.16 0.13 30 0.20 0.14 40 x 0.13 50 x 0.12 60 0.44 0.26 80 0.50 0.32 100 0.50 0.52 120 0.64 0.50 65 occurs here as with the alpha-tocopherol quinone. At higher con- centrations of the quinone, the molecules shift position on the elec— trode from a flat to a normal position and thus allow more quinone to adsorb onto the electrode. D. Electrochemistry in an oxygen saturated solution Oxygen reduction was studied using the "adsorbed" EPG electrode. The investigation was done by recording linear sweep voltammograms in a cathodic direction with the same electrode in an oxygenless pH 5.0 buffer, and then an oxygen saturated pH 5.0 buffer. The adsorbed electrode was then prepared, as described earlier by using a 5 x 10’5M quinone solution, and placed in an oxygen saturated pH 5.0 buffer. Alpha-tocopherol quinone and vitamins K1 and K3 adsorbed electrodes were tested to see if any had an effect on oxygen reduction. This was a quick and easy method to observe if the attached quinone could change the energy requirements of a reaction. The adsorbed quinone electrode would be expected to catalyze oxygen reduction through a surface EC catalytic mechanism. The thermodynamic requirement for the EC catalytic sequence is that the redox potential of the quinone/hydroquinone must be negative in value to that of the reversible redox potential for oxygen. The thermo- dynamic equilibrium value for oxygen at pH 5.0 is: Ee,02/OH'= +0.73V (vs Ag/AgCl (Sat.)). Oxygen reduction is thermodynamically favored. However, the kinetics are not favorable, as the reduction potential is ca -0.6V on the EPG electrode (Figure 24). Therefore, conditions are met for EC catalysis of oxygen reduction. Only the alpha-tocopherol quinone adsorbed electrode voltammo— gram indicated a change; an inhibition in the rate of electrochemical 66 ouo 9000 Im a m moomw>s m wows “own a OCHHI cccHI .u0wwan amumuaumm awwmxo :H muouuuofio ococfiav onuomcm .m .u0uusn mOumusumm cowxxo a“ ovouuuofio 0mm :moHu .< nowwan o.n ma CH sewuuavou aowmxo mo mEmeOEEmuHo> aoo3m umoafia .eN shaman 95: seal cocI ooqI comI o ..I _ IF. — w I» .< 67 reduction of the dissolved oxygen (Figure 24). For the clean EPG electrode, the i-E response showed no Faradaic activity in the region scanned (0.0 to ca. -l.15 V. The clean electrode in oxygen saturated buffer voltammogram (Curve A) indicated oxygen peaks at ca. -0.58 and —0.90V. When the adsorbed alpha-tocopherol quinone electrode was placed in the oxygen saturated buffer, the resulting voltammogram is shown in Curve B. The cathodic peak is present at —0.225V for alpha- tocopherol quinone. The oxygen peaks, originally present at —0.58 and -0.90V are now shifted to -0.62 and -l.l7V respectively. The former peak is now very broad, while the latter one almost approaches the cathodic limit. Thus the alpha-tocopherol quinone adsorbed electrode moderately inhibits oxygen reduction. A very large number of reaction intermediates, hence possible mechanisms and rate-determining steps, may occur in the electro- chemical reduction of molecular oxygen in aqueous solutions. Two mechanistic schemes (86) are proposed for the reduction process. One is based on an initial one-electron reduction of oxygen to super— oxide ion, followed by chemical reactions which involve superoxide ion and are governed by solution conditions. The other involves a preceding chemical reaction with the electrode instead of direct electron transfer to molecular oxygen. Scheme one supports the initial one-electron reduction of oxy- gen to superoxide ion 02 + e-fi 0; (13) where the standard potential is -0.6V (vs. Ag/AgCl (sat.)) in aqueous media. This reaction is then followed by chemical reactions which involve the superoxide ion and are governed by solution conditions. 68 A combination of these one-electron processes give an apparent irreversible two-electron process + - __A 0_ + 02 + 2H + 2e ‘—- H202 E - 0.45V (14) which accounts for the positive shift in reduction potential. Scheme 2 involves reactions which are dependent upon the electrode material and results in two possible net reactions: 02 + 2H20 + 4e" {—223 4OH' E°= +0.17.v (15) or 02 + 4H+ + 4e- ._——_:__—3 2H20 E°= +1.0V (16) and does not involve the formation of peroxide. If oxygen reduction follows the scheme 1 mechanism, the for- mation of superoxide occurs at -0.6V. Hence the search for catalysts in the EC catalytic mechanism should be for those which have equili- brium potentials negative of -0.6V. This is a possible reason for the non-catalytic behavior of the adsorbed quinone electrodes. E. Electrochemistry in an Ascorbic Acid Solution Ascorbic acid oxidation was studied using an adsorbed EPG elec- trode. Experimental data were taken by recording cyclic voltammograms initially scanned in the anodic direction. They were then characterized in deoxygenated pH 3.0 buffer and voltammograms were recorded succes- sively on the same electrode after introduction of increasing quanti- ties of ascorbic acid. The same electrode was then modified by the adsorption process previously described. After placement of the adsorbed electrode into a dilute solution of ascorbic acid, voltammo- grams were again taken successively after the introduction of increas- ing amounts of ascorbic acid. Voltammograms of ascorbic acid in the presence of the clean and adsorbed electrode were then compared. 69 The choice of ascorbic acid was dictated by previous work in which it could be oxidized in an EC sequence (76, 85). Thus, it is candidate for demonstration of EC surface catalysis. Results indicate that the adsorbed vitamin K1 and K3 electrodes both changed the oxidation wave potential of ascorbic acid (Figures 25 and 26, respectively). No change was observed for the alpha- tocopherol quinone adsorbed electrode. The wave area in the ascorbic 1 acid peak increased as the ascorbic acid concentration was increased L from 0.1 to 1.0mM, while both the vitamin K1 and K3 anodic waves re- mained unaffected. In comparing the ascorbic acid voltammograms with the ones in which the adsorbed vitamin K1 electrode was used, the ascorbic acid oxidation took place at a potential about 50 mV more positive at the adsorbed quinone electrode; with the vitamin K electrode, the ascor- 3 bic acid peak occurs at potentials about 30mV more positive. Thus the adsorbed quinone electrodes appear to provide a modest inhibi- tion of the ascorbic acid reaction. The unreactivity of the vitamin K and alpha-tocopherol quinone l adsorbed electrodes was expected because of their electrochemical irreversibility on EPG electrodes. For the EC catalytic mechanism to occur, the catalyst (the bound quinone) must have fast electron transfer, i.e. be reversible. The total unreactivity of the reversible vitamin K3 adsorbed electrode is possibly due to the structural orientation of the quinone on the electrode surface. The vitamin K3 molecule is believed to be adsorbed on the graphite surface with the plane of the ring lying parallel to the surface. The vitamin K3 electrode could possibly be less accessible for the oxidation of the solution species. 70 .u« cu wonuomvo HM cqewu«> cow: ovouuoo~o mean .m .Aocga oozmmvv opouuuofio cmoao .< "oom\>E as much caum .ummwan o.m ma CH meow ownuoomo tea «0 msmuwossmuao> ufiaozu .mN muawfim 5:: m oomI o oo~+ ooq+ ooc+ oow+ » h I p L F r r P T . P H.— .uw cu conuomco mx Gasmufi> Luw3 occuuumao 06mm .m .Aoawa conmch oncpuoofio cacao .< "oom\>E mm mum» zoom .uoumss o.m ma cw vuum cannoomm 25 ~ mo mEmeOEEmuHo> owauzu .oN ouswwm 9.5 w CON+ ooq+ ooc+ _ .? 72 This conclusion is reached after comparing reports of Brown and Anson (79) and Tse and Kuwana (77) concerning quinone CMEs. Brown et al. reported graphite electrode modification by strong surface adsorption of an o-quinone. This quinone was supposedly adsorbed on the surface with the plane of the ring lying parallel to the surface. This o-quinone/hydroquinone did not catalytically oxidize NADH, even though conditions for EC catalysis were favorable. Whereas, Tse and Kuwana reported graphite electrode modification with a covalently bound o-quinone. The fast oxidation of NADH with this electrode was reported. It was believed that oxidation occurred here, instead of at the CME described by Brown et. a1. because the adsorbed quinone was very accessible to the NADH, i.e. better struc— turally oriented. Orientation of a molecule on a surface of an electrode can be important in its use as a catalyst. 73 F. Conclusion Adsorption onto EPG electrodes was a fast and rapid method com— pared to covalent attachment. The surface coverage rate of all quinones increased with increasing solution concentration. The vitamin K3 adsorbed electrode most closely met the criteria for use as an adsorbed modified electrode. It exhibited fast electrochemical reversible electron transfer. However, the quinone electrode was very short-lived in reactant-free solution. This could be due to the large effect of this molecule's polar groups and weaker attractive forces between the aromatic portion of the molecule and the graphite sur— face, i.e. fewer places for adsorption contact. Vitamin K1 and alpha-tocopherol quinone both exhibited electrochemically irreversi- ble reactions when adsorbed, therefore making them poor choices for use as modified electrode catalysts. However, both exhibited a very desirable characteristic, i.e. strong adsorption to the electrode, as indicated by very little loss of adsorbed reactant in reactant-free solution. The aliphatic portion of the molecule most likely contri- butes the major attractive forces. The ideal molecule is one that would strongly adsorb to the gra- phite and still have electrochemically reversible behavior. 0f the molecules that were investigated, vitamin K3 was the most promising. A molecule similar to vitamin K3 but slightly less water soluble and had more adsorption sites would appear to be more suitable. The fact that vitamin K1 (aliphatic substitution) does not provide reversible behavior would suggest that a molecule with aromatic substitution would be more appropriate. The purpose of electrode preparation was to study their effect 74 on catalyzing reactions. The vitamin K3, as well as the vitamin K1, electrode was shown to inhibit rather than catalyze ascorbic acid oxidation. Alpha-tocopherol quinone was shown to inhibit oxygen reduction. The electrochemical irreversibility of the alpha-toco- pherol quinone and vitamin K1 could account for their non-catalytic behavior. The vitamin K3 non-catalytic behavior may be due to the inaccessability of the quinone for easy electron transfer due to its molecular orientation on the electrode. Also, mechanistic step potentials more negative than that of the quinone (catalyst) could cause non-catalysis of the redox couple. Even though vitamin K3 showed the most promise for this electrode modification, it failed to show any catalysis behavior, and therefore makes it unsuitable for this application. Perhaps if the ideal molecule described above, was identified, the catalytic behavior would also be achieved. IV LIST OF REFERENCES l. 2. 3. 10. ll. 12. 13. 14. 15. l6. l7. l8. 19. 20. 21. LIST OF REFERENCES A.P. Brown and F.C. Anson, Anal. Chem., 32, 1514 (1977). R. Brdicka, Z. Elektrochem., 48, 278 (1942). I.M. Kolthoff and Y.O. Okinaka, J. Am. Chem. Soc., 82, 3528 (1960). A.M. Bond and G. Hefter, J. Electroanal. Chem., 35, 343 (1972). A.M. Bond and G. Hefter, J. Electroanal. Chem., 18, 209 (1968). R.H. Wopshall and I. Shain, Anal. Chem., 32, 1514 (1967). K. Holub and L. Nemec, J. Electroanal. Chem., 11, (1966). K. Holub and L. Nemec, J. Electroanal. Chem., 18, 209 (1968). M. Sluyters-Rehbach and J.H. Sluyters, J. Electroanal. Chem., _65, 831 (1975). G. Wedler, Chemisorption: An Experimental Approach, Buttersworth, London (1976). C.H. Giles and S.N. Nakjwa, and D. Smith, J. Chem. Soc., 3973 (1960). C.H. Giles and S.N. Nakjwa, J. Appl. Chem., 12, 266 (1962). N.K. Adam, The Physics and Chemistry of Surfaces, Oxford University Press, London (1941). A.W. Adamson, Physical Chemistry of Surfaces, Interscience Publishers Inc., New York (1960). J.M. Hale and R. Parsons, Trans. Faraday Soc., 59, 1429 (1963). R.N. Adams, Electrochemistry at Solid Electrodes, Marcel Dekker, New York (1969). K.J. Vetter, Electrochemical Kinetics, p. 483, Academic Press, New York (1967). V.D. Parker, Chem. Comm., 716 (1969). Y.A. Gawargious and S.W. Beshara, 2. Anal. Chem., 245, 366 (1969). L.F. Fieser and M. Fieser, J. Am. Chem. Soc., 51, 44 (1935). G.J. Patriarche and J.J. Lingane, Anal. Chim. Acta, 42, 241 (1970). 76 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 77 J.C. Vire, G.J. Patriarche, and G.D. Christian, Mikrochemica Acta, 1, 17 (1977). J. Lindquist and S.M. Farroha, Analyst, 100, 377 (1975). Princeton Applied Research, Application Brief Vl (Vitamins). J. Bessard, G. Cauquis, and D. Serve, Tetrahedron Lett., 33, 3103 (1970). B.R. Eggins and J.Q. Chambers, J. Electrochem. Soc., 311, 186 (1970). L. Jeftic and G. Manning, J. Electroanal. Chem., 36, 195 (1970). J.L. Brisset, J. Electroanal. Chem.,iq, 217 (1975). P.H. Given and M.E. Peover, J. Chem. Soc., 385 (1960). S. Wawzonek, R. Berkey, E.W. Blaha, and M.E. Runner, J. Electrochem. Soc., 103, 456 (1956). W.R. Turner and P.J. Elving, Anal. Chem., 31, 467 (1965). W.R. Turner and P.J. Elving, J. Electrochem. SOC".llZ. 1215 (1965). I.M. Kolthoff and T.B. Reddy, J. Electrochem. Soc., 393, 217 (1975). K.J. Vetter, Z. Elektrochem., 36, 797 (1952). J.B. Morris and J.M. Schempf, Anal. Chem., 33, 257 (1959). P.J. Elving and A.F. Krivis, Anal. Chem., 39, 1645 (1958). P.J. Elving and D.L. Smith, Anal. Chem. 33, 1849 (1960). R.D. DeMars and I. Shain, Anal. Chem., 32, 1825 (1957). J. Lindquist, J. Electroanal. Chem., 33, 37 (1974) E.P. Parry and R.A. Osteryoung, Anal. Chem., 36, 1366 (1964). E.P. Parry and R.A. Osteryoung, Anal. Chem., 31, 1634 (1964) F.C. Anson, J.B. Flanagan, K. Tukahashi, and A. Yamada, J. Electroanal. Chem., 31, 255 (1976). A.P. Brown, C. Koval, and F. Anson, J. Electroanal. Chem., 13, 379 (1976). H.E. Zittel and F.J. Miller, Anal. Chem. 31, 200 (1965). R.E. Panzer and P.J. Elving, J. Electrochem. Soc., 119, 864 (1972). G.M. Jenkins and K. Kawamura, Nature, London, 231, 175 (1971). 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 78 A.L. Beilby, W. Brooks, Jr., and G.L. Lawrence, Anal. Chem., 32, 22 (1964). J.S. Mattson and C.A. Smith, Anal. Chem., 22, 1395 (1977). T.P. DeAngelis, R.W. Hurst, A.M. Yacynych, H.B. Mark, W.R. Heineman, and J.S. Matteson, Anal. Chan., £2, 1395 (1977). W.J. Blaedel and C.A. Mabbott, Anal. Chem., 32, 933 (1978). R.E. Panzer and P.J. Elving, Electrochim. Acta, 22, 635 (1975). T. Noda, M. Inagaki, and S. Yamada, J. Non-Crystalline Solids, 1, 285 (1966). J.P. Randin and E. Yeager, J. Electrochem. Soc., 118, 711 (1971). J.P. Randin and E. Yeager, J. Electroanal. Chem., 33, 313 (1975). B.R. Puri, Surface Complexes on Carbon, from Chemistry and Physics of Carbon, vol. 6, P.L. Walker, Jr., Ed., Marcel Dekker, New York (1970). H.P. Boehm, Advances in Catalysis, D.D. Eley, H. Pines, and P.B. Weisz, Editors, vol. 16, p. 179, Academic Press, New York (1966). J.S. Mattson, H.B. Mark, Jr., and W.J. Weber, Jr., Anal. Chem., 31, 355 (1969). J.S. Mattson, and H.B. Mark, Jr., J. Colloid Interface Sci., 31, 131 (1970). J.S. Mattson, L. Lee, H.B. Mark, Jr., and W.J. Weber, Jr., J. Colloid Sci., 33, 284 (1970). M.L. Studebaker, E.W.D. Huffman, A.C. Walfe, and L.G. Nabors, Ind. Eng. Chem., 23, 162 (1956). I.F. Jones and R.C. Kaye, J. Electroanal. Chem., 22, 213 (1969). J.V. Hallum and H.V. Drushel, J. Phys. Chem., 22, 1502 (1968). K.F. Blurton, Electrochim. Acta, 13, 869 (1973). G. Mamantov, D.B. Freeman, F.J. Miller, and H.B. Zittel, J. Electroanal. Chem., 2, 305 (1965). J.P. Randin and E. Yeager, J. Electroanal. Chem., 32, 257 (1972). J.F. Evans and T. Kuwana, Anal. Chem., 32, 1632 (1977). B.F. Watkins, J.R. Behling, E. Kariv, and L.L. Miller, J. Am. Chem. Soc., 21, 3549 (1975). 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 79 B.E. Firth, L.L. Miller, M. Mitani, T. Rogers, J.C. Lennox and R.W. Murray, J. Am. Chem. Soc., 23, 8271 (1976). J.C. Lennox and R.W. Murray, J. Electroanal. Chem., 13, 395 (1977). S. Mazur, T. Matusinovic, and K. Cammann, J. Am. Chem. Soc., 22, 3888 (1977). N. Oyama, A.P. Brown, and F.C. Anson, J. Electroanal. Chem., 31, 835 (1978). N. Oyama and F.C. Anson, J. Electroanal. Chem., 33, 289 (1978). C.M. Elliott and R.W. Murray, Anal. Chem., 33, 1247 (1976). E D.F. Untereker, J.C. Lennox, L.M. Wier, P.R. Moses, and R.W. 2 Murray, J. Electroanal. Chem., 31, 309 (1977). A.W.C. Lin, P. Yeh, A.M. Yacynych, and T. Kuwana, J. Electroanal. . Chem., §4, 411 (1977). ’ J.F. Evans, T. Kuwana, M.J. Henne, and G.P. Roger, J. Electroanal. Chem., 33, 409 (1977). D.C. Tse and T. Kuwana, Anal. Chem., 33, 1315 (1978). J. Zagal, R.K. Sen, and E. Yeager, J. Electroanal. Chem., 33, 207 (1977). A.P. Brown and F.C. Anson, J. Electroanal. Chem., 33, 203 (1977). C.A. Koval and F.C. Anson, Anal. Chem., 32, 223 (1978). E. Laviron, Bull. Soc. Chem. Fr., 3717 (1967). H.A. Laitenen, C.A. Vincent, and T.J. Bednaraski, J. Electrochem. Soc., 115, 1024 (1968). J.W. Hassler, Activated Carbon, p. 229, Chemical Publishing Co., Inc., New York (1963). J.S. Kipling, Adsorption From Solutions of Non-Electro1ytes, Academic Press, New York (1965). N. Winograd, H.N. Blout, and T. Kuwana, J. Phys. Chem., 13, 3456 (1969). M.A. Andrews, Inorg. Chem., 13, 499 (1977). "I19.111111171711010s