MY 2 6 2005 ABSTRACT APPLICATIONS OF CYCLIC VOLTAMMETRY TO THE STUDY OF TRANSIENT INTERMEDIATES by Paul Joseph Kudirka Reduction of a group of sulfonephthalein acid-base indicators in aqueous solutions has been used to evaluate the theory of cyclic volt- ammetry for disproportionation reactions initiated electrolytically. Predictions of the theory are in excellent agreement with experimental results, and rate constants were measured by cyclic voltammetry for nine sulfonephthalein radicals. Moreover, the rate constant measured for disproportionation of phenol red radical [k = (3.4 i 0.3) x 102 pg? - sec- ] agrees exactly with conventional spectrophotometric meas- urements. The effect of maximum suppressors on measured rate constants by cyclic voltammetry was investigated using several sulfonephthalein compounds. Gelatin was found to have no effect on measured dispropor- tionation rate constants. Thus, at least for these systems, use of gelatin is an expedient and acceptable approach. Qualitatively, the effects of gelatin and Triton X-100 were found to be the same. How- ever, apparent rate constants measured with cyclic voltammetry in the presence of Triton X-100 depended upon the concentration of this sup- pressor as well as on scan rate and sulfonephthalein concentration. Thus, Triton X-100 is unacceptable for quantitative measurements. APPLICATIONS OF CYCLIC VOLTAMMETRY TO THE STUDY OF TRANSIENT INTERMEDIATES By Paul Joseph Kudirka A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1971 'c ACKNOWLEDGMENT The author wishes to express his appreciation to Professor Richard S. Nicholson for his guidance and encouragement throughout this study. Thanks are also given to Janet M. Kudirka, the author's wife, for her encouragement and understanding. ii n... oi. 0"- -‘I. “up; fi-au o a c ~~n~ u A” II .. a.“ :1- hi "I f'AI l (i L11 .7" I1! W n: II. III. IV. TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . A. Some Aspects of Voltammetry . . . . . . . . . . . B. Description of Cyclic Voltammetry . . . . . C. Application of Cyclic Voltammetry to Chemical Kinetics D. Applications of Cyclic Voltammetry to the Study of Species of Transient Existence . . . . . . . . EXPERIMENTAL. A. Electrochemical Apparatus B. Procedures for Aqueous Electrochemical Experiments. C. Vacuum Line and Associated Apparatus . . . . . . .. D. Procedures for Nonaqueous Electrochemical Experiments. E. Spectroscopy Apparatus and Procedures . . . . . . . .. F. Materials. . . . . . . . . . . . . . . . . . . ELECTROCHEMICAL REDUCTION OF CARBON TETRACHLORIDE IN NON- AQUEOUS SOLUTIONS. . . . . . . . . . . . . . . . . . . . . A. Some Aspects of Dihalocarbene Chemistry. . B. Previous Studies on Electrochemical Reduction of Carbon Tetrachloride in Nonaqueous Solutions C. Electrochemical Reduction of Halomethanes in Non- aqueous Solutions. . . . . . . . . . . . . . . . ELECTROCHEMICAL REDUCTION OF HALOBENZENE COMPOUNDS IN N,N-DIMETHYLFORMAMIDE. . . . . . . . . . . . . . . . . . . A. Results. . . . . . . . . . . . . . . . . . . . . . . . B. Discussion . . . . . . . . . . . . REDUCTION OF SULFONEPHTHALEIN ACID—BASE INDICATORS . A. Compounds Investigated . . . . . . . . . . . . . . B. Mechanism. . . . . . . . . . . . . . . . . . . . . . . C. Polarography . . . . . . . . . . . . . . . . . . . . . D. Controlled Potential Reduction . . . . . . . . . . . . E. Cyclic Voltammetry. . . . . . . . . . . . . . . . . .. F. Chemical Reduction . . . . . . . . . . . . . . . . . G. Chemical Generation with Electrochemical Detection . . H. ESR Spectroscopy . . . . . . . . . . . . . . . . . . . I. Spectrophotometric Measurements. . . . . . . . . . . J. Electrochemical Measurement of Rate Constants. . . . . K. Effects of Maximum Suppressors. . . . . . . . . . . .. L Considerations of Kinetic Effects. . . . . . . . . . . iii Page CIDWUO l9 29 29 38 39 1:6 1+8 50 53 5h 61 6h 72 7h 80 85 85 92 93 99 100 103 106 109 118 150 163 ’9‘- c..‘ '9‘, ‘Q Q.‘ V- 1 g‘- ‘R § ‘ \ J C V ‘D ‘K‘ N . I”; .fi. Table II. III. IV. VI. VII. VIII. IX. XI. XII. XIII. XIV. XVI. XVII. XVIII. XIX. LIST OF TABLES Polarographic Data of Wawzonek and Duty (56). . . . . . . Polarographic Data of Wawzonek and Wagenknecht (58) Half-Wave Potentials and Diffusion Current Constants for Halobenzene Compounds . . . . . . . . . . . . . . . . . 1/2 Potentlals at wh1ch L1m1t1ng Current XE: h were Conducted for Halobenzene Compounds. Studies Half-Wave Potentials and Disproportionation Rate Constants for Sulfonephthalein Indicator Radicals Disproportionation Rate Constants for Cresol Purple Radical by Polarographic Monitoring of Decay of Oxidation Wave of the Radical . . . . DisprOportionation Rate Constants for Phenol Red Radicals in Acetate Buffered Aqueous Solutions from Conventional Second Order Plots. . . . . . . . . . . . . Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Data on Cresol Purple . . Data Data Data Data Data Data Data Data Data Data Data Data on on on on on on on on on on on on iv Cresol Purple . Thymol Blue . Thymol Blue . . . Bromothymol Blue. Bromocresol Green . Phenol Red. . Phenol Red. . . Phenol Red. . . . Phenol Red. . . . Phenol Red. . . Phenol Red. . Phenol Red. . . . Page 22 25 73 75 86 105 117 120 121 122 123 12h 125 126 127 128 129 130 131 132 97v» ‘5.-. ~v—v- It..- ~~v~—. -_s- - . *Ifi.-‘ l n ‘h‘- ”-9.." v . '5 ‘~.I. .,.._~‘ I ‘u... -. I..- ..".Q. 9.. I 2"" “'-o. a ‘1‘... 1...... "Q‘ I; : ‘~~¢~ Table XXI. XXII. XXIII. XXIV. XXVI. XXVII. XXVIII. XXIV. XXXI. XXXII. XXXIII. XXXIV. XXXVI. XXXVII. XXXVIII. Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Cyclic Effect LIST OF TABLES (continued) Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric Voltammetric of Gelatin on Disproportionation Rate Constants by Cyclic Voltammetry for Radicals of Cresol Red Data Data Data Data Data Data Data Data Data Data Data Data Data Data Data Data Data on on on on on on on on on on on on on on on on Phenol Red. . Phenol Red. Phenol Red. Phenol Red. Phenol Red. Phenol Red. Phenol Red. Cresol Red. Cresol Red. Cresol Red. Cresol Red. . . Cresol Red. Bromophenol Red . Bromophenol Red . Bromophenol Red . Chlorophenol Red on Bromocresol Purple. and Phenol Red in Aqueous Solutions . . Page 133 13h 135 136 137 138 139 1ho 1A1 1h2 1A3 1AA 1h5 1M6 1h7 1u8 1h9 160 F». + a Y. in Q. 'I“. LIST OF FIGURES FIGURE Page la. Variation of electrode potential with time for cyclic voltammetry. . . . . . . . . . . . . . . . . . . . . . . 6 lb. Cyclic current function for reversible charge transfer. . 6 2. Current-potential curves for various values of RAE for first order chemical reaction following reversible charge transfer. . . . . . . . . . . . . . . . . . . . . lO 3. Ratio of the anodic to cathodic peak current as a function of 51 for first order chemical reaction following reversible charge transfer. . . . . . . . . .. 13 h. Cyclic current-potential curves showing effect of charge transfer parameter, W. . . . . . . . . . . . . .. l6 5. Block diagram of circuit configuration. . . . . . . . .. 3O 6. Block diagram of circuit used for coulometry. . . . . .. 3h 7. Cell arrangement. . . . . . . . . . . . . . . . . . . .. 36 8. Electrochemical vacuum line apparatus. . . . . . . . . . ho 9. Vacuum line electrochemical cell. . . . . . . . . . . .. A2 10. Polarogram of 1.5 EM CCl in acetonitrile with 0.10 M tetrabutylammonium perchIorate as supporting electrolyte. 66 11' E1/2 l2. Polarographic waves for the controlled potential elect- rolysis of a sulfonephthalein indicator that is reduced to a moderately stable radical which subsequently dis- proportionates. . . . . . . . . . . . . . . . . . . . . . 9h XE pH behavior of Cresol purple. . . . . . . . . .. 88 13. Limiting current y§_time behavior of the polarographic waves following controlled potential electrolysis of a sulfonephthalein indicator that is reduced to a moder- ately stable radical which subsequently disproportion- ates. . . . . . . . . . . . . . . . . . . . . . . . . .. 97 1h. Comparison of theory (points) and experiment (curve) for disprOportionation reaction mechanism for cyclic VOltmmetry O O O O O O O O O O O I O I O O O O O O O O O 101 IS. Esr spectrum of cresol purple following controlled potential electrolysis on first reduction wave. . . . .. 107 vi WON-”7‘ - Afiv-~- \‘) I\) a. FIGURE 16. 17. 18. 19. 20. 21. 22. LIST OF FIGURES (continued) A typical experimental curve for reduction of phenol red with V(II), and the following generation of phenol red by disproportionation of the radical. A typical second order plot of phenol red radical y§_time. . . . . . . . . . . . . Cyclic voltammogram on unpurified sample of phenol red in the absence of gelatin. Cyclic voltammogram on purified sample of phenol red in the absence of gelatin. Conventional polarogram on unpurified sample of phenol red in the absence of gelatin. Conventional polarogram on purified sample of phenol red in the absence of gelatin. Variation of apparent rate constant with scan rate and phenol red concentration in the presence of 0.01h z Triton x-1oo. vii Page 112 115 151 153 155 157 161 ‘ o r;fip ignb-O' .awpfl- 1" l as. .u i “35““; .ivv. 1 Q r?“ a“; .-\r‘ -- v ‘Q ~ w a...“ “ ‘fi ‘- la) I. INTRODUCTION Truly spectacular advances have been made in electrochemical theory and methodology since the discovery of polarography by Heyrovsky in 1922 ( 1,2 ), and the first report of an automatically recording polarograph by Heyrovsky and Shikata in 1925 ( 3 ). Progress has been especially rapid over the last decade ( h-l2 and references contained therein ) for several reasons. First, the availability of high speed computers has permitted successful math- ematical modeling of very complex electrode processes. Second, advances in electronics have permitted corresponding advances in electrochemical instrumentation. For example, commercial availabil- ity of operational amplifiers has made it possible to construct instrumentation embodying features such as positive feedback, thereby making precise electrochemical measurements practicable in low conductivity media, including even ethers. Moreover, with modern instrumentation, measurements can be made on a time scale not achievable even a decade ago. Thus, it is now possible to observe directly with electrochemical methods extremely reactive species whose lifetimes are of the order of milliseconds, or less. Of course, during this same period other extremely powerful techniques and tools have been developed for studying chemical species of very short lifetimes. Nevertheless, modern electro- chemical methods possess some unique features in this context, and therefore complement other more standard methods. For example, controlled potential electrolysis permits the generation of a far wider variety of intermediates than any other single technique. This feature, coupled with the extremely rapid response of recently 1 -1 (U '1 U) 3.1. '.C 3.18 P; (D .9 . v.“ .4 1a-.A“‘ ‘sl'. ‘~ ,,,. ‘5 o‘c “lb "322.1‘ bt~~.“ .G“y- 5“ '.V $}. 35,- ‘ " a. ‘~'~. 2 developed methods, gives modern electrochemistry its power and versatility. To date, actual applications of modern electrochemistry to the generation and observance of highly reactive intermediates are few, but it is readily apparent that such methods provide an exciting tool for studying reactive entities. Although a number of related electrochemical techniques have received extensive attention and development ( h-l2 and references cited therein ), cyclic voltammetry is the most versatile developed to date, and has reached a popularity unequaled by any of the other electrochemical techniques. This predominance is due to concomitant developments in instrumentation, making the method simple to use due to the ease of interpreting the experimental data, and to suc- cessful mathematical modeling of the method for many important electrolysis mechanisms ( 13—52 ). Because of the predominance of this method, and because this laboratory has been responsible for most of the theory that has popularized it ( 2h,25,27,h2—hh,h8—52 ), cyclic voltammetry was embraced as the primary tool for measurements described in this thesis. Cyclic voltammetric data will be presented ubiquitously through- out this thesis, and therefore it may be useful to provide some dis- cussion of the technique for the uninitiated reader. Hence, the re- Inainder of this introduction is divided into the following areas: (A) Some General Aspects of Voltammetry; (B) Description of Cyclic ‘VoltammetrY; (C) Applications of Cyclic Voltammetry to Chemical Kin- estics; and (D) Applications of Cyclic Voltammetry to the Investigation of Species of Transient Existence. 1r . . r . a . . . 4.. A. A I\ ‘4 u c M I g H» A.» a V a v t. «c .r. T . r . .. c s .. a» a . a. . .. a E n3 n1 is of. .M». New «ha 1 s A. . r: v" . a .. . Q . a: . , If“ 5. 2.. hr . pr \ ~s \‘ an» 4.. u b. i o . .| .I \ I. N I, \tll ||II « I .l. I. ah\ A~ . u “a .C - va 5‘ .,. .C .«v I-» :1 I». 71 c. .p\ «5 .~ a n .‘ a 3. f. I. . . I . I y r I 1 n. a \\s a. . n- v‘ :‘ . V“(\l 4 \b- 3 A Some Aspects 2;.Voltammetry The current in an electrochemical cell is a function of the volt- age applied to the cell. Voltammetry encompasses methods by which such currents are studied when the current-voltage characteristics depend on the rate of the overall electrolysis mechanism. In general, a variety of processes must be considered when volt- ammetric techniques are employed, since electrolysis mechanisms are intricate and comprise many steps. Electrode processes may be regard- ed as occurring formally in the following sequence (53): (1) transport of the depolarizer (reactant) to the vicinity of the electrode; (2) chemical changes in this depolarizer leading to an electroactive form; (3) influence of the electric field leading to further changes in the electroactive form; (A) actual electron transfer between the electrode and this electro- active form leading to an activated product; (5) influence of the electric field on this activated product leading to an intermediate product; (6) chemical changes in this intermediate product leading to a final product; and (7) transport of this final product into the bulk of the solution. Steps 1 and 7 involve mass transport, Steps 2 and 6 are homogenous chemical reactions, and Steps 3, h, and S formally constitute the electrode reaction. Three modes of mass transport must be considered, (Sh), namely, migration, convection, and diffusion. Migration is defined as the '1 av. VOVOO i: ovoai ‘3‘ T! V. E. e‘ \- C'WRFF‘ V" - Ahwv'- b. a.\ a . Ox —h~ J - =~ =~ v... i. .r5 9. C a. a; h. C. a. .a V. r . ”A 2. at s .. a . x .7» a . .a a A: a. a a a A: «C p. n.“ Av ..~ 5. . pg 3 .. a. r . a. u . u .. Cy ~. . :\ .Q 0 a. t . .. a a. 4 . a A a 5 .. A?!“ "‘v t s 3‘ in ‘b\ h motion of charged particles in an electric field. Since electric cur- rent in a solution is carried by all ions in the solution, migration can be minimized by employing an excess ( usually in a ratio of lOO/l, or greater ) of an electrochemically inert electrolyte, termed the supporting electrolyte. Tetraalkylammonium and alkali metal salts are commonly employed for this purpose. Convection arises from stirring by mechanical or ultrasonic means, or by density differences which in turn are primarily caused by con- centration and temperature gradients. Mass transport of reactants and products by convection is minimized by conducting measurements in quiescent solutions. Diffusion results from concentration gradients, or more rigorously, differences in chemical potential. Hence, if electrochemical measure- ments are conducted in quiescent solutions containing a large excess of supporting electrolyte, diffusion is essentially the sole means of mass transport of reactants and products. These latter conditions are usually selected, because of the three mass transport processes, dif- fusion is easiest to describe mathematically. Traditionally, electrochemical terminology is in disarray, and presently no single terminology is standard. Nevertheless, voltam- metry is generally regarded as the generic term for such techniques. Furthermore, either the voltage or current applied to the electro- chemical cell is controlled. Hence, a clear distinction exsists be- tween voltammetry at controlled potential and voltammetry at controlled current (5h). Both polarography and cyclic voltammetry are controlled potential techniques, but they differ in several major respects. First, cyclic -.-‘o ““5: 5-..“- ' 4.‘ -v .— rt: (0 l 1 9 l P» ."»-._ 9...! 3 ."-\ u. .--" I: H 47.5 ft! "v’ ‘1. “4 § ' L +- . ~ \I ‘4“ ‘ ‘1 ‘ "Q a ‘5 C ‘5 n U, ‘ .-l ' IA, “‘ l‘ :_ ‘ 1 ‘\ ‘3.“ d. . A... ~'.: $“ ‘& :1- 5. ‘ 'v F . "u “"i "d 5 voltammetry employs a stationary electrode, whereas polarography, of course, employs a dropping mercury electrode (DME). Second, cyclic voltammetric experiments are conducted with a linearly changing poten- tial, whereas polarography is a constant potential technique. Although cyclic voltammetry is similar to polarography, it is even more akin to another technique, commonly termed linear potential scan voltammetry ( LPSV ). The primary difference between cyclic voltammetry and LPSV is that with cyclic voltammetric experiments, a triangular-wave potential is used, whereas LPSV experiments are per- formed with a ramp potential function. Hence, cyclic voltammetry is a two-step technique, whereas LPSV is a one—step method. The greatly enhanced versatility and power of cyclic voltammetry over techniques such as LPSV is due primarily to this two-step feature (55). This important aspect of cyclic voltammetry will become apparent in the following discussion. B Description g§_Cyc1ic Voltammetry Figure la illustrates the variation of electrode potential, Eflp), with time for a cyclic voltammetric experiment. At the beginning of the experiment (t_= O) the electrode is at an initial potential, E1, which is anodic of the formal potential, Eé, for the given oxidation- reduction couple. During the time interval 0 < §_< A.the electrode potential is made to vary linearly with time in a negative direction so that it passes the formal potential. At time §_= A_the direction of the linear scan is reversed (the rate of change of potential,y, is held constant), and the potential is returned to the initial value. Figure la also includes mathematical relationships for the electrode potential in terms of the scan rate and the electrolysis time. Figure la. Figure 1b. Variation of electrode potential with time for cyclic voltammetry. Cyclic current function for reversible charge transfer. EU 1 l Current Function n, j o‘- i“h$‘;v‘.‘ v“. ups-J - o. s ‘Lq...‘.p syn-Avovc '\ . \rcyuc‘ «54‘. K-) . 1“»r;r‘ v‘n- 5...! 5"- vi- w o‘, - ‘ C. {:5 '“‘ v“~ ‘. ’ wan "ny .t'c. “r :. “P1.“ '..\, 9... ._ ‘ ¥ U I‘. ‘Q\“ ' \ a“ 'e a—I,‘ ‘Q A: P;_ . . u‘..‘4“.‘ - 5 V 1 C ‘ “‘ .~ ‘ - Q I'D A 2..» . . ‘1 . . ~,.. Awn- "~ ‘q'I-r ‘\ p, v ‘1"Q“ . 5“ re ~.. i'u - 'N: AL. ,. ‘Nz‘nl . .‘ O . h ‘ u.‘ " S m. ‘VC 'M ’1‘. '\( 'hQ ‘ u ”\s 'l ‘I h ”vs. . 'l »\l . ._‘ - ‘V ‘ \. N 8 Figure lb is a plot of the electrolysis current function (the current function is directly proportional to current density) for a reducible depolarizer (reactant), 93 which undergoes reversible electron (charge) transfer. During the initial linear scan the electrolysis current varies from zero to a maximum, and then decreases. This de- crease in current occurs because the amount of depolarizer available for electrolysis decreases, since as electrolysis proceeds depolarizer is progressively removed from solution in the immediate vicinity of the electrode. On the initial linear scan and the beginning part of the reverse scan, the current arises from the reduction process, Q_+ n§_+ R) and the reduced substance, R, is continuously generated at the electrode. At some potential sufficiently positive of E1 on the reverse scan some of the accumulated 3,15 oxidized back to Q, The current for the R_+ Q_+ n§_portion of the current-potential curve exhibits a minimum for reasons analogous to the origin of the cathodic maximum. C Applications gf_chlic Voltammetry tg_Chemical Kinetics In reality, electrolysis mechanisms are generally not as simple as the one discussed in connection with Figure 1. There the depolar- izer was assumed to be identical with species Q_in the bulk of the solution. Furthermore, the reduced species, 3, did not undergo any other reactions following charge transfer. In addition, the rate of the charge transfer step was limited only by the mass transfer rates for substances Q_and 5, Thus, the current-potential curve of Figure lb reflects only the diffusion processes for Q_and R, Complications of the electrolysis mechanism depicted in Figure 1b, and discussed in the above paragraph, can take many forms. For example, ‘Ur‘ ,_ .— V ‘aflov‘v '9‘. s v . . ‘ 1r:-- ‘9 y unavén-gv . . I nun, a N \ s ., "-.. a. - . _ A"‘¥r 5 l1” .“ a v‘ ‘- En , ‘7‘ '“VoAs. . ‘ a ‘f‘ .. ~ ‘7‘ C." " “~. V., ‘“~;_““ ( .'"-t-A., Inv'f“ 9 the depolarizer may be an intermediate resulting from a chemical reaction which precedes the charge transfer, and/or the product of the charge transfer step may be an intermediate which undergoes a chemical reaction in the bulk of solution. In addition, the charge transfer rate may vary from reversible to irreversible, and thus also may be a rate-limiting step in the electrolysis. Other complications such as adsorption on the electrode surface may participate in the overall electrolysis mechanism. Consider the following complication of a reversible electron transfer, where a chemical reaction of R_forms B, an electroinactive product ( 2h,hh ). P (II) Initially, if the scan duration ( time of the experiment ) is small with respect to the half life of Reaction II, an insignificant amount of §_will have time to react to form §_before R_is oxidized to Q, Quantitatively, this idea is equivalent to stating that the kinetic parameter §/§;( where k_is the rate constant in sec—l, 2.: EEK/El also in sec-l, and 1,18 scan rate ) is less than about 0.01 (2A). In this case no measurement of k.is possible and, in fact, the current-potential curve is essentially the one for a reversible charge transfer ( where El; is zero ) without any perturbation by Reaction II. A current- potential curve for this value Of.§(§ is shown in Figure 2. At the other extreme, current-potential curves can be obtained by employing such large scan durations with respect to the half life of Reaction II that all of B_has sufficient time to form P_before any 5. can be oxidized to g, Quantitatively, this case corresponds to values lO .smmmsmsp mmsmgo manwmsm>ms mcflzoaaom coflpowms HmoflSmno smcso pmsfim sow .m\m mo mosam> mSOHsm> sow mm>H50 HwflpcmpomlpcmsHSQ .m mssmflm 11 issue 0 om ON. Om. _ _ _ Nd: N. o ¢.O uoglound iuaung F. ‘:‘m A“ 71 £an a. ‘1 - ‘Vl-“F tfi A v.-U. V \l ‘. T‘C‘o: 2"nr nu..- V‘.‘.‘ a V - VA“ .'S on ‘V- n. .‘ 5‘ - - 1 V "N . . ,VJ“- . «__ “r; v.A\._ L.. t, " I -‘ l§“~ ‘ I~.n. _ ’\ .x-.,,_ v. \- 'N“ . ‘ '1”~~ A ~- --,~, g _ ¢ ‘ I ‘l‘ 1: .LD I ‘- -' is.» . . " ‘ v“ 3'55.“ ‘, -~ y». t.‘ § 315:” N.~‘_. 1': ' ' “"' if T . ' \ 1‘. " "‘1 if; 4 sh. ‘ - vx.‘ L" Q! 17 ‘ , “IA ‘1“; . H‘.‘ ‘ I‘Lw I‘v... 'n‘ (I ‘- u’, '- I‘.e "\-. h kgr. Q. §ne- . . 1' . .‘, a: ' « "P r. “‘\ A ~- n \“ ‘- d‘, 5‘!va .QQfi .V‘ . ‘ .v- ‘\ ‘~ "'Q N.‘ h: \ ‘. .IA‘ 'V ‘F ‘v~%“ \- \‘; ’\ 5‘. u; ~- 12 of k/a greater than about 10 (2A). A current—potential curve for this value of k/a_also is shown in Figure 2. To determine 5) scan rates must be selected so that only a frac- tion of.R ( ideally from about 10 to 50 % of R_) reacts to form P prior to oxidation. Quantitatively, this idea corresponds to scan rates such that k/§_is between 0.01 and l. A current-potential curve for k/a_equal to 0.10 also is included in Figure 2. Cyclic voltammetric theory (2h,hh) for determining rates of first- order irreversible succeeding chemical reactions is presented as a working curve, shown in Figure 3, which relates ratios of anodic to cathodic peak currents ( ia/ic ) to a parameter designated kl, Here k'is the rate constant in sec—l and 1 the elapsed time during the scan between potentials E_ and E? (see Figure l). A Experimentally, a rate constant is measured with cyclic voltam- metry in the following manner: (1) the ia/ic ratio is determined from the experimental current— potential curve ( see Figure 2 ); (2) the value of kl_corresponding to this value of ia/ic is read from the curve shown in Figure 3; (3) the value of 1_(in seconds) is determined from the experimental current-potential curve and the scan rate used for the particular experiment; and (A) finally, division of kl_by l_gives k, A simple calculation shows that by varying scan rate between 10 mV/sec and 10h V/sec, one can determine first—order rate constants in the range 10.2 to 106 sec-l. Similar calculations based on cyclic voltammetric theory for second—order chemical reactions, such as dimer— i2iation (51) and disproportionation (52), show that rate constants 13 .pmwmcwhp mwswno mapflmsm>mh mcflsoaaom soflpomms HmOflszo pmpho pmsfim sow MM.mo coflposdm n ma psmshdo anon oflwoapmo ow oflwocm mnp mo oflpnm .m mssmfla 1’4 I | I 0. 00. to. V. C\! O, — O O O O ' Q (onconivofl /(0IOONV)d! 0.0 -|.O Iog(kr) 0. ‘1‘ 1 SK .3 :u e u.“ r“ u! 94 wm w. p . l. . . .Q” A"! fidbo‘ qu v .5 \\ M rs fuck s v3- M." ‘ in» h§.~‘ - - A Int t“ F‘A ona... AL. '5‘ Q‘ . Va M: ..... 15 in the range 10 to 108 Mil-sec—l can be obtained with scan rates from 10 mV/sec to 10h V/sec. As implied by the above discussion, scan rate is the primary vari- able in cyclic voltammetry. Experimentally, scan rates generally can be varied from about 10 mV/sec to lOLt V/sec. This range is set by several experimental limitations, the lower is set by the fact that at long times convection becomes significant, making diffusion no longer the sole means of mass transport. The upper limit is set by the fact that as scan rate is progressively increased the fraction of current used to charge the electrical double-layer becomes dominant, because capacative current is directly proportional to scan rate whereas faradaic current is prOportional to the square root of scan rate (5h). Information on faradaic processes can still be obtained in such cases, but generally such data are only of qualitative value. Another limitation in using cyclic voltammetry to determine rates of coupled chemical reactions is that kinetics of charge transfer steps at the electrode alter the shape of current—potential curves as scan rate is varied. Obviously, information on rates of coupled chemical reactions can only be garnered when charge transfer kinetics are not completely rate—determining. A significant effect of charge transfer kinetics is that the anodic-to-cathodic peak separation ( about 60 mV for reversible charge transfer ) becomes progressively larger as scan rate is increased (27). This effect is shown in Figure A, and is frequently observed when scan rates larger than about 100 V/sec are employed. For example, if the charge transfer rate constant is 0.1 cm/sec (a typical value), then the morphology of the current-potential curve will essentially 16 Figure A. Cyclic current-potential curves showing effect of charge transfer parameter, W. 0.5 W = 7.0; a l I l I I *6 II 0.25; a 0.5 17 0.5 "‘ 0.4 *- O.3 ’ _ H . J 2 o o. a 1. across“. 29:50 ’60 ’IZO O IZO 60 (E- E'/2)n,mv . ~ v. .v.- n’r*c p . a Na. :«a i a 4‘ vl .2! n c , x . a . l .w .. ‘ . . . l s a: . . rs : . . . I. 1. 3: Wu . )2 a; : a a. .n. A c “1. up. a» Q. an. Mia . I . , r . .. \ . .rb F. D». yv “A. «My. .»b L; I: Q» a: a; G. .M“ Q» r» A,» hm a.» 1‘ ya H: V». ‘— 2. a: w . r . : .. 2 t. .7. .c . . v» : . a . r : a 2. h. .. .. .n. a; r. ‘ . a. 2 a.» a a...» N» F5. 0 v p.» I! h A u a k .a .a g .‘u an ‘1 fix FA 3. 3 \ : fl 5.. L. C. r: . 2 . h. 5 .ruas . a C» . . L. ‘ .n g 1.. u .4 t \ . .nu .3 q. u» .... .v . I a :1» a: v.. .1 a .. v N sx C. 2 . . s; ., . 1 y c v nix” .. Q h a . 6 fl . .. H .r a .. . .ua. \ w» .I\ «max ~u.\‘ MA» 18 correspond to reversible electron transfer at a scan rate of l V/sec. However, at a scan rate of 10 V/sec the peak potential separation will be about 80 mV, and at a scan rate of 100 V/sec the peak potential separation will be about 130 mV. At this point it may be useful to illustrate how cyclic voltam- metry can be used as a diagnostic tool for elucidation of electrolysis mechanisms. As shown in Figure 2, for small values of k/a_the morphol- ogy of the current-potential curve corresponds to that for a simple re— versible charge transfer. However, as the kinetic parameter k/a_in- creases the cathodic peak ( Q_+ n3.+ R.) shifts toward positive poten- tials, and the peak height increases by about 10 %. Moreover, when kfia is greater than about 1, an anodic peak ( R +~Q_+ g§_) no longer is observed. Hence, when k[a is greater than about 1 peak current ratios ( ia/ic ) cannot be used to measure rate constants since only one peak is present. Nevertheless, when k[a is greater than about 10, cathodic peak potentials shift 30/n mV anodically for a 10 fold in- crease in k/a, or, equivalently, shift 30/g_mV cathodically for a 10 fold increase in scan rate. This fact indicates the presence of a first—order irreversible succeeding chemical reaction. Anodic peaks also are absent for second-order reactions such as dimerization when k/agis greater than about 1. In this case, however, the cathodic peak potentials shift 20/n_mV for a 10 fold increase in scan rate when k/a is greater than about 10 (2h). Thus, even when anodic peaks are absent, cyclic voltammetry can be employed as a diag- nostic tool to distinguish between first and second—order chemical re— actions simply on the basis of cathodic peak potential X§_scan rate shifts (2h,51). n‘j ~.--p- r 1* m. L..v-' s «a- ‘ - C ’r. "“'\- 1.. u ”‘ ID» 'd-« . (n (I! t J: “~. -\ .._v v - x _" ;\-~ “. 19 For systems which display any or all of these complications, the morphology of the current-potential curves will be related to the elementary steps in the electrolysis mechanism. Hence, cyclic voltam— metry can be used as an extremely valuable diagnostic tool to eluci- date the nature of these steps and to measure their rates. Cyclic voltammetry has two outstanding features which make it especially suited for such mechanistic studies. First, a cyclic potential scan makes possible the investigation of both the reactants and products of the electron (charge) transfer in a single experiment. Second, the time of the experiment can be varied over a wide range ( about six orders of magnitude ). Thus, intermediates with a variety of half lives can be detected and their reaction rates measured. D Applications p£_Cyclic Voltammetry_pp_£h§_§£pgyngf Species pf_Transient Existence The original objectives of this research were to evaluate pos— sible applications of modern electrochemistry to the investigation of chemically interesting species of transient existence. Although many choices were possible, it seemed logical to consider only systems for which some prior electrochemical data and observations ( primarily polarography ) already existed. In this regard, three classes of com— pounds seemed most amenable to further study: (1) reduction of tetra— halomethanes in nonaqueous solvents, since there is some evidence for the intermediacy of dihalocarbenes (56,57); (2) reduction of pfdi- halobenzene compounds in nonaqueous solvents, since there is some evi- dence for the intermediacy of dehydrobenzene (58); and (3) reduction of certain common acid-base indicators in aqueous solutions, since there is reasonable evidence for the intermediacy of free radicals v¢£flk are .M‘ V.) O ‘avsa nap-p .v—a..-...&... . r.-. ‘ " "‘ Q‘ "»~ ‘.~ ~- ‘ ~ " “3‘. fl. .- u. ‘5“‘_ ‘q . n I- - a .y- a. r .~.~. ¥-‘ ‘ ‘ "qu q: '. .k ”'i.* ~ m. ..._ V ‘3‘”5‘. A?“ . 4» P . .~:~: .6 “‘~v, m, ‘ 2“ ‘r “ N—v‘ 4‘ i. n h gfi‘fia‘. ‘ ~s‘ v. ”I"; h r ‘ o -~§ H -\ ‘~“ ,‘ . .v‘ s. . "A :: uh: ‘ V a“ "s‘l \ ‘L-Q U‘J‘ ‘C1. n“~- 3': a I 20 which are moderately stable in aqueous media at room temperature (59). A brief survey of the prior electrochemical research in each of these three areas constitutes the remainder of this introduction, and the remainder of this thesis is devoted to results and discussion of re- search in these three areas. Two of the above areas for which the literature indicates pos- sible electrochemical generation of chemically interesting intermedi- ates (1:23, dihalocarbenes and dehydrobenzene) involve electrochemi— cal reductions of organic halogen compounds in nonaqueous solutions. Interestingly, both reports differ from the generally accepted electro- chemical behavior of organic halogen compounds ( 60—76 ). The electrochemical literature literally abounds with reports of reductions of both alkyl and aryl halogen compounds. In the majority of cases, unless some other group of the molecule is more readily re- duced or unless organo—metallic compounds are formed, the overall electrolysis results in proton substition for the halide. For ex— ample, for 3.: I» B3, C1) the overall electrochemical reduction is generally: + EE;_§» R _ H + x‘ R — X Thus, the electrochemical reduction of CClh in 75 % dioxane-water pro— ceeds as follows (60): + 0011; 9542—) 001311 + 01’ Wave I 4.. 2e, H - CC13H ——————9 CCl2H2 + C1 Wave II 2 H+ CC12H2 —e-’——> CC1H3 + 01‘ Wave III + 2e, H - CClH3 ——————9 CH1‘ + Cl Wave IV 3- 3 I33 :1 I) 1 4‘ 1“ ~ ".G V H 1 . . "‘ V90. ‘- -’ J. v... "L . n o. . ~“ ~P ~p'vc\ :1; .. an -~ ‘\ . .1 In." G “.“I‘yy .' . ‘ Liar} rflrgn .‘-... aqua. . ~‘ ‘ l? a I -" ‘abgi ‘ v‘.' .‘_ . P e ‘d L'ETQ n-a Q .‘.¥V» y“ ud‘. I u- 2 W:,._‘H a.“ u g .‘ n,- ‘- ~ in fig "c‘ Fr ~¥~.‘\ . q a~ .“ h. . ‘ «was N -..." ‘.‘ . “hi" .P‘ «a \ \4 'V x- "h ‘ s“ e QIN‘y‘A‘I - \‘,~.‘J ‘1 l“ ‘ ' Enter” A. “.‘ v‘ \_‘ .§r‘~ ‘4‘»: ( ‘r"~ I “5N1! 6‘ ,1"; :q,‘ ‘ Fwy. \ «‘V. r 21 Similarly, electrochemical studies of aryl halogen compounds such as meta-dibromobenzene have found the following overall reduction stoichiometry in both aqueous and nonaqueous media: Br Br Br 2e H+ _ \‘ ——l——> ./ + Br Wave I Br + ./ 28’ H . + Br- Wave II ——_’ Nevertheless, Wawzonek and Duty (56,57), who used polarography to investigate the electrochemical reduction of halomethanes in aceto- nitrile (AN) and N,N—dimethylforamide (DMF), reported entirely dif- ferent behavior. Wawzonek and Duty studied six halomethanes, and 'Table I contains the polarographic behavior reported by these authors lfor these compounds. These authors used a two—electrode polarograph Viith a mercury pool combination reference and counter electrode; l1ence the half-wave potentials in Table I are with respect to the Exatential of a mercury pool. Their polarographic solutions were re- INDrted to contain 0.175 M tetrabutylammonium bromide as the support— i ng electrolyte . Based on largely indirect evidence from polarographic experi- meEnts ( Table I in this introduction ), Wawzonek and Duty proposed t11€3 following mechanism for the electrochemical reduction of CClh irl .AN: r k K .Arxr..w >3.:- 3:: {323.3303 rNC QGSAN UW-.~:.~M&CL:.W.J.~ .N 333:.5 22 _ 3 a a u “ Am.aav ma.a " ass ” mamaoo " ©m>smmpo m>m> coapodmma o: “ mzm “ mmaomo Ame.av am.s m Aom.mv wm.a “ Ama.ev mo.o " man " same " Amm.mv am.a “ Ace.ev mm.o " tape: soanza _ save " Aoa.av am.o “ Aoa.av mm.o _ za “ aaoo “ Ama.mv wm.a “ u za _ maomo “ ©m>sompo m>m3 coflposmon oc " z¢ " Naommo Ame.av mm.m m Amm.av Hm.a “ Aom.av em.o “ tapes eoauazm “ ease Amm.av ea.m " Ama.av ma.a " Ame.mv mm.o m man " sage Aam.mv aH.m “ Ama.mv ma.a " " man " maomo Aem.ev aH.m " u 0 age _ Naommo " _ _ _ 6H A>Vm\amn “ 6H A>Vm\amn " eH A>Vm\fimu " pem>aom "eeaoaaoo » rs _ _ .Ammv hde was Mocoaawz mo spam ofinmwhwoadaom .H manna r ‘ hu- ...528 1.. -~ fir‘A’Aqr \ ‘, ’- vknvad‘,‘ 7 "I ‘oA n .‘ 11'” 'F" --~‘-‘ ‘4 9 p 1‘.» 23 CClh + 2e + 0013 + c1' (I) 001; : cc12 + c1" (II) cc12 + 2e + 001: (III) CCl: + 2 CH3CN + H2CC12 + 2 [CH2CN]- (1v) These authors also proposed that CClh reduction in DMF produces di— chlorocarbene, but to a lesser extent than in AN. In an attempt to provide direct evidence for the formation of dichlorocarbene suggested by this mechanism, Wawzonek and Duty also performed macroscale electrolytic reductions of CClh in the presence of reagents such as tetramethylethylene and cyclohexene which are ex— pected to trap any dichlorocarbene. Thus, Wawzonek and Duty found that gas chromatographic analysis of the final solution of the macroscale electrolysis of CClh in AN in the presence of tetramethylethylene gave a component ( no yields mentioned ) possessing the same retention time as a pure sample of l,l-dichloro—2,2,3,3-tetramethylcyclopropane, and regarded this gas chromatographic peak as proof of dichlorocarbene generation. How— ever, when these authors performed a similar macroscale electrolysis in the presence of cyclohexene, none of the expected product ( 7,7-dichlorobicyclo[h.l.0]heptane ) was observed, which makes the presence of dichlorocarbene less than certain. This ambiguity is heightened when one considers the way electrolyses were performed. The experimental procedures described by Wawzonek and Duty for their macroscale electrolyses are, to say the least, a bit unclear. For example, they stated that "Direct current was used for all electrolyses, and the voltage output was 120 volts." For the macroscale electrolysis a? (“’7 V‘ “U: in» n... . . b-¢».. .. ~‘.l“" FP- .—~§‘ ‘ 5.5... ‘ . ”nu-v- . ‘ a Lru .C ‘h‘ . - I1 .- I“ "An .' 'kn. ‘ g“ .Q a. .H 'v‘ ‘I fly». 1 r. b._"‘~:. 'W- . .‘v- ,. ‘hh-‘F' . . ‘ a . .‘AA ‘.~ 0‘ 7". "M,",'E‘ y. “ :PA" w.“ ., _. . “! \‘ ~-VV ‘_ F‘A (U ‘ ‘6" u. ‘c‘ a \““I I y. *.. ~N . - ‘s s \$\ ‘1. t. 2h of CClh in AN in the presence of cyclohexene, they reported that "Current was controlled at 0.3 amp. After 21 hOurs, the current reading was 0.16 amp., and the reduction was stOpped." And, for the macroscale electrolysis of CClb in AN in the presence of tetra- methylethylene, they reported that "Direct current was allowed to pass through the solution for 37 hours during which no visible color change was observed. The direct current was controlled at 0.30 amp and fell off to a final value of 0.15 amp." Based on these statements, it appears that the macroscale elect- rolyses were conducted by impressing 120 V across the electrolysis cell, and then halting the electrolyses after the current had de— creased to about one-half of its initial value. In other words, it appears that neither controlled current nor potential was employed. In summary, although the analytical data of Wawzonek and Duty are, at best, not very convincing, their results suggest that modern electrochemical techniques might be used to generate and then directly observe species such as dichlorocarbene. This possibility is quite interesting since carbenes are known to be highly reactive, and gen- erally are studied only by indirect methods. Therefore, in spite of doubts about the proposed mechanism of Wawzonek and Duty, reduction of tetrahalomethanes using cyclic voltammetry was investigated. In another paper from Wawzonek's laboratory, Wawzonek and Wagenknecht (58) described the polarographic reduction of several halobenzene compounds, among them pfdibromobenzene and pybromochloro- Imenzene, in nonaqueous media. These polarographic data are reproduced ill Table II. Again, half-wave potentials are is a mercury pool. ~‘-H~«u .w—yv u‘ o " ! 1"‘Il- ,- .1. a..- a, ‘II q. - "‘ “nu "‘ “Unh — - 25 Table II. Polarographic Data of Wawzonek and Wagenknecht (58). grdibromobenzene 82% Dioxane 1.ho (h.15) I I Compound E Solvent : ~El/2(V) Id : -El/2(V) Id : : : Bromobenzene I DMF I : 1.85 (3.28) pfdibromobenzene I DMF I 1.28 (6.53) : mfdibromobenzene I DMF I 1.ho (3.79) I 1.88 (2.68) | I I prdibromobenzene I DMF | 1.53 (3.02) 7 1.82 (3.h9) Chlorobenzene : DMF : I 2.00 (2.95) pybromochlorobenzene : DMF I 1.38 (b.50) I 2.02 (1.hh) mrbromochlorobenzene I DMF I l.h9 (3.32) I 2.0M (2.83) prbromochlorobenzene I DMF I 1.60 (3.09) I 2.02 (2.61) Chlorobenzene : AN I no reduction wate observed prbromochlorobenzene : AN I 1.58 (5-50) I gybromochlorobenzene I AN I 1.65 (h.85) I pfbromochlorobenzene I AN : 1.81 (h. 50) I Bromobenzene : AN : not investigated pfdibromobenzene : AN l 1.81 (9-77) pedibromobenzene : AN : 1.55 (5.00) 2.05 (3.90) pfdibromobenzene I AN : 1.70 (h.12) 2.02 (h.l2) I I ' I Erdibromobenzene : 82% Dioxane : 1.56 (2.66) Qrdibromobenzene I 75% Dioxane : l.h0 (3.8h) Erdibromobenzene I 75% Dioxane I 1.56 (2.57) .Q-dibromobenzene : 60% Dioxane I l.h0 (3.h2) Erwiibromobenzene : 60% Dioxane : 1.56 (2.28) 1 1 ”55‘ unv ll) C '7‘ (I) c..- ~ ‘-.\I I 4‘. q. “» A, h". v ‘1 b‘ 7 n. \ ~ \‘ ‘0. 1 . - '\ ‘ a ‘._~¥ ‘ 26 Although Wawzonek and Wagenknecht apparently performed no coulo— metric experiments, they nevertheless stated that the reduction of gfdibromobenzene in both AN and DMF results in a single h-electron reduction. 0n the basis of their polarographic data (Table II), these authors proposed the following mechanism, which involves the intermediacy of dehydrobenzene in DMF and AN. Br .. 2e ._ ——->- + B1" (I) Br Br k _. ___.~. . + Br (11) Br 2e soIvght (III) They also proposed that, under similar electrolysis conditions, re- duction of gfbromochlorobenzene results in dehydrobenzene intermediacy. In an attempt to provide direct evidence for the intermediacy of dehydrobenzene suggested by this mechanism, they also performed a macro- scale electrolytic reduction of gfdibromobenzene in the presence of furan, which is expected to trap dehydrobenzene. Specifically, they found that a macroscale electrolytic reduction of gfdibromobenzene (at -35° C, DMF as solvent, tetrabutylammonium iodide as supporting electrolyte) in the presence of furan yielded products which on acid hydrolysis gave a l % yield of a compound which possessed the same 513 retention time as a peak due to a pure sample of a—naptho1, and interpreted this as direct proof of dehydrobenzene generation. 27 Although the analytical data of Wawzonek and Wagenknecht are not conclusive, their results suggest that cyclic voltammetry could be used first to generate and then directly observe dehydrobenzene. Such a study would be highly informative since dehydrobenzene, like dichlorocarbene, is a highly reactive species which generally is studied only by indirect methods. Of all the possibly reactive species that can be studied electro- chemically, free radicals have received by far the greatest attention (8-12 and references contained therein). In the majority of studies nonaqueous solvents have been used because organic free radicals gen- erally are less reactive in aprotic, or nearly aprotic media, where protonation of electrogenerated radical ions is greatly diminished. Nevertheless, meaningful experiments in nonaqueous solutions are very difficult, as will be illustrated repeatedly throughout this thesis, and therefore much of the extant research on free radical electro— chemistry is not an unambigous evaluation of the potential of electro- chemical relaxation techniques for studying reactive intermediates. Hence, the recent experiments (primarily polarography) of Senne and Marple (59) on the electrochemical reduction of phenol red, the struc- turally simplest of the sulfonephthalein acid-base indicators, are particularly interesting because their data suggest that free radicals of moderate stability can be produced electrochemically in aqueous solution. These authors postulated a mechanism which rationalizes the results of their experiments. Essentially, this mechanism con- sists of a one-electron reduction yielding a product which dispropor- tionates to form starting material and the fully-reduced (two-electron) product, phenolsulfonephthalein. .¢' A '_. r n—QV «a-» cat‘s”- 2L--.— “£43,. iv'.. .Awa ~ ‘ V ‘M - Q n A ‘7: 7' “has ‘ 9 ‘>. If.“ a. ’ K. -- . .. g :- ~‘r‘ '-'- x- \nt :9- .53. . ~ - ' ‘FA ‘“ --‘: (D ‘ a. lo 28 Hence, it appeared that compounds such as phenol red could be em- ployed to evaluate the efficacy of modern electrochemical relaxation methods for studying free radical reactions, without the complications accompanying the use of nonaqueous solutions. Thus, sulfonephthalein compounds were selected as the third class for investigation in this thesis research. The research described in this thesis was initiated by applying recently-developed electrochemical techniques to all three of the sys- tems mentioned above, with the intention of devoting the major effort to the systems that appeared most promising in the terms of the major objectives stated above. As results in following chapters will show, the greatest promise seemed to lie in the study of the sulfonephthalein indicators, so that the most detailed investigations were made in this area. Nevertheless, very interesting preliminary results were obtained in the other two areas, and therefore the most salient features of ex— periments on halomethanes and.gfdihalobenzene compounds are also includ— ed in this thesis. OIII l l of nona‘ -1 a 4 .u. -_ .\ l .is “b -..L n A I ‘vq 5-4 u.' nw' 5-K LLS F; Fv II. EXPERIMENTAL A Electrochemical Apparatus Ohmic potential losses are a possible source of serious error in all of the electrochemical measurements made in this study. A conven— tional three electrode potentiostatic circuit does not compensate for resistance between the working electrode and reference electrode probe. This iR drop is especially significant when a conventional dropping mercury electrode (DME) is used. Also, the solution resistance is significant in low conductance solvents and even in high conductance solvents when fast scan rates (large currents) are employed. Fort- unately, it is possible to compensate electronically for ohmic potential losses by using a three electrode potentiostatic circuit with positive voltage feedback. Figure 5 is a block diagram of the circuit used here; this circuit is similar to one described by Smith and Deford (77). The control amplifier (C.A.) was a commercial instrument (Wenking potentiostat, Model 61RS, Brinkmann Instruments, Westbury, N.Y.). The two voltage followers (F1 and F2) and inverter (I) were solid state op- erational amplifiers (Philbrick Researches, Inc., Model P25AU, Dedham, Mass.). Cell current was determined from the potential drop across the load resistor, RL. To measure this potential drop two different recording devices were used depending on the time scale of the experiment. For experiments less than about one second duration a storage oscilloscope (Tektronix, Inc., Beaverton, 0re., Type 564 with 2A63 (vertical) and 2867 (horizontal) plug-in units) with Polaroid camera attachment (Tektronix Type C-12) was employed. For longer time experiments an 29 Figure 5. 30 Block diagram of circuit configuration Reference electrode Counter electrode : Working electrode Load resistor Control amplifier Voltage follower amplifier Voltage follower amplifier Inverter amplifier Function Generator Initial potential WWW L“ WWHWWOCUZOCU o MI—J o o C) H. 31 Einitial If _ 50 K9.” CE. )I’ @3RI. IF| :L—WW‘ c I we EG'N 4K I IKQ. RL HELIPOT r. nu. . s a. . . .7” v. at. .aa VJ V. a :2 .J. v...» V Q run Le n. - a: O . .1; .7“ a. . w . "I n E a . 2 8 . “I V r;v‘ b‘ .O‘ V“... A ‘p~. ~.:. V ‘6. - a. a. a a- _ n _ d 4‘. la n. .~‘ A . r1 ‘ v 2* a. h. 2. \ ~— w-_ ‘- L‘V‘“ .. v- ‘ ~ .- ‘ .- v s . w . \ ..\ r. a . T. a t .. . 5.» a . . . 2. 2. .5.» L t . . . .1 E .N.‘ h. ~N‘ 1h 32 x-y recorder (Honeywell, Inc., Model 520, San Diego, Calif.) was used. The various electrochemical techniques used in this study (polar- ography, cyclic voltammetry, constant potential electrolyses) require different potential-time functions. For some cyclic voltammetry ex- periments involving measurements of anodic currents it was necessary to have a base line constructed from the extension of the cathodic wave. This was obtained by scanning just past the cathodic peak, and then holding the potential constant to record the current—time curve. This required a voltage waveform consisting of the first leg of the triangular wave, followed by a constant potential. The function gen- erator for experiments of this nature has been previously described (78). In addition to this function generator, two commercial function gen- erators (Exact Electronics, Inc., Model 255, Hillsboro, 0re., and Interstate Electronics Corp., Model F52, Anaheim, Calif.) were used to generate triangular waves for cyclic voltammetry and polarography. No distinction could be made between experimental results obtained with any of these three function generators. For constant potential electrolyses the initial pcfiential source, a battery-operated low voltage power supply, of the Wenking potentiostat was used to set the potential of the working electrode. Coulometric measurements were made by means of a voltage-to— frequenoy conversion technique previously described by Bard and Solon (79). Instrumentation consisted of a commercial potentiostat (Wenking, describ— ed above) and a Heath Universal Digital Instrument (UDI, Model EU—805A) used simultaneously as a voltage-to-frequency converter and a frequency counter. The voltage measured by the UDI was the ohmic drop across a load resistor, 3L, placed in the current—carrying loop of the three A ‘1‘ “‘I boo v 3“" . .-. v a a Dr ,«r di~mt I“ v. M ‘ ~ 33 electrode potentiostatic circuit. Figure 6 is a block diagram of this circuit. The cell arrangement employed for most aqueous studies is shown in Figure 7. The cell consists of a glass weighing bottle with a g 60/20 ground glass joint on the top. The cell lid is made of Teflon and is machined to fit this joint. Holes were drilled in the lid to allow in- sertion of the various cell components (Figure 7). The counter electrode consists of a one foot length of 26 guage platinum wire wound around 6 mm soft glass tubing. One end of the platinum was sealed in the glass tubing, and electrical contact was made through mercury contained in the glass tubing. The reference electrode contained three separate sections separ— ated by 10 mm fine porosity glass discs. The left hand compartment (Figure 7) was a saturated calomel electrode, and the right hand com- partment contained the solution being investigated. Because, on occa- sion, this latter solution contained perchlorate, a l M_aqueous sodium nitrate solution was used in the middle compartment to prevent precip— itation of potassium perchlorate. Liquid junction potentials clearly arise when different solutions are contacted in this manner. However, in this work only relative potential measurements were critical. Pre— sumably the liquid junction potential for a given solution would remain constant during the course of an experiment. In fact, half—wave poten- tials for a given system never varied more than about 10 mV during the Course of an entire investigation. The right hand compartment of the reference electrode was a tube with a 10 mm fine porosity glass disc in Contact with the solution in the cell. For experiments requiring a stationary mercury electrode, a hanging 3h Figure 6. Block diagram of circuit used for coulometry. *IZI<.":UO£’;UHO H. r‘ Fifi-ltd r-zj O o ' 0 Control amplifier (Wenking potentiostat) ' Initial potential : Reference electrode : Working electrode : Counter electrode Load resistor : Voltage-to-frequency converter : Frequency counter 35 36 Figure 7. Cell arrangement. 'IJL'EIUOtfii-D ' Reference electrode Scoop Hanging mercury drop electrode : Dropping mercury electrode : Deaerator Counter electrode ”Fr ti. r4 me. - «we “H“ ; r{‘ Y‘ a. s5“‘ 7" 0 ~ — p. ~— ‘ OB mmdmm mwpwgomflw endow> oooa maze .o.>.o Hwom HmmeIOpImmmHo Haoo Hmoflamnoospomam m mmwnp z eflsaaq :oaovmz mdam m>mflm saddomaoz pmzawpwo Hm>oamm mo mam mm Hmowamnooppomam .m madman a: a: L) éi a: a: c: a: Al A2 Figure 9. Vacuum line electrochemical cell. Combination scoop and mercury pool contact Hanging mercury drop electrode (HMDE) : Reference electrode (sce) : Dropping mercury electrode (DME) Counter electrode (platinum wire) Cell bottom (male portion of a Q 60/50 joint sealed flat 5 cm from ground glass portion of joint) G: g 60/50 joint H: Adapter, threaded (Ace Glass Incorporated, part no. 5029-10) I: 3 1h/2O joint 'TJIFJUOUj3> J: h mm vacuum stopcock (ground glass) All ground glass joints, except the main g 60/50 joint, are Q 1h/2O joints. c’:*¢* ‘v‘y- u... r RFQV.” 3* .~~U hr - rye bA’ I...- W v- 3. Q .,. ‘0 . AU . “"a "f‘. “V ' -..E"" HPJ- ~. R 1“»... A‘v - . 4..." I... 7" e Hg? Q nu reference electrode had a three compartment configuration identical to the one described above for aqueous solution studies. The arm of this reference electrode configuration that fits into the cell is closed at the bottom with a fine porosity glass frit. The counter electrode chamber was also closed at the bottom with a fine porosity glass frit. A.mercury pool could be introduced at the bottom of the cell for coulo— metry and other macroscale electrolysis experiments. Electrical con- tact to this pool was made by a platinim wire that was an integral part of the scoop which was inserted through the top of the cell. The electrochemical cell is designed so that the bottom half of it can be immersed in another solution. For example, immersion of the cell into a dewar containing liquid nitrogen is necessary to transfer solvent from the vacuum line manifold to the cell, and to degas solvents. The vacuum stopcocks and joints used on this apparatus were of the conventional ground-glass type. Both silicone lubricant (Dow Corning "High Vacuum Grease") and Apiezon N grease were used to lubricate these ground-glass joints. The chemical inertness of these two lubricants appeared to be about equal towards solvents used in these studies. However, Apiezon N was found to be superior in that it appeared to make longer-lasting seals, which is important since a great deal of time and work is involved in dismantling, regreasing, and reassembling the vacuum apparatus. Furthermore, manipulations of stopcocks and other ground- glass joints appeared to require less effort when Apiezon N, instead of Dow Corning lubricant, was used. The use of ground-glass joints on the vacuum apparatus results in one unsatisfactory feature. Namely, solvents such as 1,2—dimethoxyethane (Glyme) and tetrahydrofuran (THF) rapidly dissolve both of the above .a C. .0 r“ ‘J flea. I < nnraw oned V A). ‘ I a. avhs w an‘% »~ 6. ‘Vb-.\---A- . .1 9%“ «NM .& 9M- '1' IL .3 .IL 5*,- \.¢ . I’" I APP av“. art Y‘" ‘ .Hv“ v. :- - =a (I ’r‘ ‘ ‘QP‘~ \v . n 1‘ s‘ - y‘a‘y. V. U\ a» flu he 'ZT‘M “I “I I45 mentioned vacuum lubricants. However, it was found that this vacuum apparatus was quite suitable for use with solvents such as N,N—dimethyl- foramide and acetonitrile. Thus, since the amides and nitriles are the most versatile and commonly-employed nonaqueous electrochemical solvents (because of their relatively high dielectric constants), the vacuum apparatus constructed for this research is suitable for most nonaqueous electrochemical studies. Commercial vacuum pumps were employed for the vacuum line. An air— cooled oil diffusion pump (Veeco Instruments Inc., Model EP 2A-l, Plainview, N.Y.) was operated in conjunction with a mechanical forepump (Central Scientific, Model 91138 Hyvac &, Chicago, 111.). The heater of the oil diffusion pump was rigged with an air flow switch (Rotron Mgf. Co., Model 2A, Woodstock, N.Y.) so that the heater shut off if the air supply stopped. The oil used in this diffusion pump was silicone fluid (Central Scientific, Type 93262-003, Chicago, 111.). Pressure in the vacuum line was measured by means of a cold cathode discharge guage (consolidated Vacuum Corp., Model GPH-lOOC, Rochester, N.Y.). This guage was capable of monitoring the pressure between 10-7 and 25 torr, and was isolated from the manifold of the vacuum line by a 2 mm vacuum stopcock. A g 2h/h0 ground glass joint was situated above this stopcock for ease of removal of the entire guage assembly for periodic cleaning. As depicted in Figure 8, a gas train is an integral part of the vacuum apparatus. This train directs inert gas through U-tubes contain- ing anhydrous magnesium perchlorate, molecular sieve (Matheson Coleman & Bell, Activated "Linde" Type 3A), and an oxygen removal catalyst (BTS catalyst, Badische Anilin und Soda Fabrik, A.G., Ludwigshafen am Rhein, W. Ger.) prior to entry into the electrochemical cell. The U-tube D6 containing the oxygen removal catalyst was wrapped with heating tape and the temperature kept at approximately 200°C. The reason for the inclusion of this gas train is that it is necessary to have a positive pressure of an inert gas (nitrogen or argon) blanketing the electrolysis cell when measurements are being made. This is done to prevent intro- duction of water and oxygen from the atmosphere into the cell during electrode manipulations that arise in the course of a series of experi- ments. D Procedures for Nonaqueous Electrochemical Experiments The primary reason for constructing a vacuum apparatus for electro- chemical experiments was to obtain a dry environment for conducting ex— periments using nonaqueous solutions. Solvents, supporting electrolytes, and other materials were purified (discussed below) prior to introduction into the vacuum line apparatus. The electrochemical cell (Figure 9) was designed so that measurements could be made with solution volumes from 30 to 100 m1. Taking acetonitrile as an example, the following statements list the sequence of operations carried out using this vacuum apparatus for a nonaqueous electrochemical experiment. First, a weighed amount of purified tetrabutylammonium perchlorate, the supporting electrolyte usually employed, was placed in the electrochemical cell. Then the cell was assembled with the counter electrode and compartment, the HMDE, the scoop, and two plugs placed in the tOp of the cell (the DME and refer- ence electrode were placed in the cell in a later Operation). The cell was then attached to the vacuum line, the proper stopcocks manipulated, and the contents of the cell exposed to the vacuum manifold. The cell was pumped until a pressure of about 10.6 torr was obtained for at least ‘r 7' .A‘ 9.. “a q ‘A ' abEV'vn. ~‘ \ ”‘43.. w.‘7 L4 . "~. 9" .K v‘ V. \ ~ ‘ § u s-‘n ' 5 .2: K 7“ tu‘ “ ‘ ‘0 § ,1 ‘ Ln one hour. Then, by further stopcock manipulation, the cell was isolated from the vacuum line manifold. Next, from 30 to 50 m1 of purified acetonitrile, contained in a round bottomed flask, was attached to the vacuum line manifold, the contents of this flask being separated from the manifold by a vacuum stopcock. The solvent was frozen with liquid nitrogen, and this stopcock turned and the solvent degassed at least four times. Then, by further stopcock manipulations, the electro- chemical cell was reconnected to the vacuum line manifold and the solvent transferred directly into the cell which was maintained at liquid nitro- gen temperature. The cell and vacuum manifold were then reisolated by stopcock manipulations, and the liquid nitrogen bath removed from around the electrochemical cell. Then the cell and contents were allow- ed to warm to room temperature (about two hours for acetonitrile solu— tions), and a positive pressure of argon introduced into the cell by turning the proper stopcocks. The reference electrode arm and DME were then placed into the cell, with the positive pressure of argon (hopefully) preventing contamination by oxygen and water from the atmosphere. Then electrochemical experiments were carried out on this blank solution. After these blanks were run, a measured amount of electroactive material was added, and the electrochemical experiments initiated. As indicated by the above statements, operations performed with this apparatus were quite time consuming. For example, it always re- quired at least 18 hours to run even preliminary experiments when aceto- nitrile was employed as the solvent. Furthermore, experiments conducted using N,N—dimethylforamide (DMF) as the solvent required far longer times because of the much lower vapor pressure of DMF. For example, transfer of 30 m1 of acetonitrile from the storage flask on the manifold A8 into the cell required about one hour whereas transfer of an equivalent volume of DMF required about six hours. E Spectroscopy Apparatus and Procedures Sulfonephthalein indicators absorb strongly in the visible region (Amax for phenol red is A30 nm) whereas the one and two electron reduc— tion products are colorless, absorbing in the uv region (Amax for phenol red is 265 nm). Thus, absorbance measured at A30 nm can be used to 5 determine the concentration of phenol red. For example, if a A x 10- 5 solution of phenol red is reduced chemically to 2 x 10- M5 then at 5 M, equilibrium the absorbance at A30 nm corresponds to 3 x 10- M, Of course, experiments of this type performed with rapid mixing provide a means of measuring the rate of disprOportionation. Experiments along these lines were performed using two procedures, one involving the use of a stop flow apparatus and the other a conventional spectrophotometric apparatus. A11 phenol red solutions in these studies were aqueous solu- tions containing 0.20 M acetic acid and 0.20 M sodium acetate. The preparation of solutions of chemical reductants is discussed below. Approximately half of the rate measurements were conducted with a stop flow apparatus which is described by Beckwith and Crouch (81). In a typical experiment a solution of phenol red was contained in one syringe and a solution containing approximately half an equivalent of V(II) was contained in the other syringe. Mixing into a 2.00 cm length cell was complete within a few milliseconds and transmittance through the cell was recorded as a function of time with a Heath (Model EU—201V) multi-speed strip chart recorder, Operated at a speed of one inch per minute. Zero % Transmittance was set with the shutter closed, and 100 % Transmittance with a blank (everything present in the solutions except A9 phenol red and V(II)) solution in the cell. Measurements were made at ambient temperatures of 23-2A°C. Other rate measurements were made with a Heath (Model EU—70l) single-beam.Spectrophotometer in conjunction with a Heath (Model EU- 703-31) photometric readout module. The current output of this photo- metric readout module was connected to a Heath (Model EU-20-28) log/linear current module which was incorporated in a Heath (Model EU- 201) multi-speed strip chart recorder, operated at a speed of one inch per minute. The reaction cell was a standard 1.00 cm spectrophotometric cell surrounded by a brass jacket through which water at 23 : 0.01°C was circulated. Zero % Transmittance was set with the Shutter closed and 100 % Transmittance with a blank solution in the cell. 2 m1 of phenol red solution was placed in the cell and the absorbance recorded. Then 0.50 ml of a V(II) solution, about half an equivalent of V(II), was injected into this cell using a 0.50 ml tuberculin syringe. Rapid mix- ing was obtained by this technique, as seen by the fact that these ex— perimental curves were identical in appearance to those obtained using the stop flow apparatus. This syringe injection technique was found to be necessary since the use of a magnetic stirring bar (Teflon coated) in the standard 1.00 cm spectrophotometric cell did not give sufficiently rapid mixing. No discernable differences were observed in measured rate constants obtained by this conventional procedure 13 the stop flow method. Other spectrophotometric measurements in the ultraviolet and visible regions were made with a Unicam SP 800 spectrophotometer. Studies were carried out in a conventional manner using standard 1.00 cm spectro— photometric cells, with a cell containing blank solution in the reference .. mayo" .3. ' ch opIfi‘A‘. v.66 " I" R J 1 c . . 1 i L ah. f.‘\ s.‘ ..‘ . * :a .a .1‘ 4‘ {x ‘ .~4. 1W4 1 ‘ AIH G a I .NK 1‘ \ ‘fi‘ a“. — +0 «I. :a :4 ~ .. s s >1. 3 p I. 3 3 3. IC AC .6 . a a . c. s \a .. s 3 s .w \a ... \a h. s .1; h... .«a .. I at. a la. x». Ia T. ‘« 50 beam. Both sample and reference cells were surrounded by a brass jacket through which water at 23 : 0.01°C was circulated. Spectra were re- corded from 200 to 850 nm. Infrared measurements were made with a Perkin-Elmer 237 B grating spectrophotometer. Spectra were recorded between 625 and A000 nm. Conventional techniques employing nujol mulls and KBr pellets were used. EPR experiments were run in a constant temperature room at an amr bient temperature of about 23°C using a Varian E-A EPR spectrometer system. An aqueous solution cell was used. The instrument was first calibrated with a strong pitch sample (Varian part no. 90AA50-01) and a weak pitch sample (Varian part no. 90AA50-02). F Materials Nonaqueous solvents were purified according to the following pro- cedures. Acetonitrile (Fisher, B.P. 8l.A - 81.7 0C) was purified by distillation under nitrogen according to the procedure of Mann (82). N,N—dimethylformamide (Fisher, B.P. 152.5 — 153.2 °C) was purified by distillation under reduced pressure according to the procedure of Visco (6). N,N-dimethylacetamide (Fisher, B.P. 165.9 - 166.6 00) was purified in the same manner as DMF. Tetrahydrofuran (Baker, B.P. 65.9 - 66.A 0C) was purified according to the procedure of O'Donnell (83). 1,2-dimethoxyethane (glyme) (Eastman red label, B.P. 83 — 8A °C) was purified according to the procedure of Mann(6). Tetrahydrothiophene- l,l-dioxide (Sulfolane) (Shell, Sulfolane-W) was purified by distilla- tion under reduced pressure from either calcium hydride or molecular sieve (Matheson Coleman & Bell, Activated "Linde" Type 5A). Nitro- methane (Matheson Coleman & Bell, B.P. 100 — 102 °C) was purified by storage over barium oxide for several days followed by a reduced pressure 51 distillation from barium oxide. Aqueous solutions were prepared using distilled water. Methanol (Matheson Coleman & Bell, B.P. 6A.A - 65.0 °C), acetone (Matheson Coleman & Bell, B.P. 55.9 — 56.0 °C), and methylene chloride (Matheson Coleman & Bell, B.P. 38.3 - A1.3 0C) were used without further purifica- tion. Tetraalkylammonium perchlorate salts were generally used for studies conducted in nonaqueous solvents. Both the tetraethyl and tetrabutyl salts were made by the metathesis of their respective chlorides (Eastman red label) with sodium perchlorate (G.F. Smith Chemical Co.). These salts were then purified by crystallization according to published pro- cedures (6). Furthermore, shortly before usage in electrochemical ex- periments, the salts were dried under vacuum at 80 0C for at least 2A hours. Other supporting electrolytes such as sodium perchlorate, lith— ium perchlorate, potassium chloride, and sodium sulfate were used with- out further purification. A11 halomethane and halobenzene compounds were obtained commercially (Eastman Organic Chemicals, Aldrich Chenical, J.T. Baker Chemical, K & K Laboratories, and Fisher Scientific) and were used without further puri- fication, except for carbon tetrachloride which was distilled under nitrogen at atmoshperic pressure. The purification procedure employed for phenol red and other sulfonephthalein indicators has been described previously (8A). Phenolsulfonephthalein (phenol red) was obtained from Aldrich Chemical Co., Inc. and from the stockroom (Allied Chemical). 3',3",5',5"-tetra- bromoemrcresolsulfonephthalein (bromocresol green), 5',5"—dibromos97 cresolsulfonephthalein (bromocresol purple), Qrcresolsulfonephthalein 52 (cresol red), and 3',3"-dibromo—5',5"-dichlorophenolsulfonephthalein (bromochlorOphenol blue) were obtained from Aldrich Chemical Co., Inc. Meta-cresolsulfonephthalein (cresol purple), thymolsulfonephthalein (thymol blue), 3',3"-dibromothymolsulfonephthalein (bromothymol blue), 3',3"-dibromophenolsu1fonephthalein (bromophenol red), 3',3"-dichloro- phenolsulfonephthalein (chlorophenol red), 3',3",5',5"-tetrabromophenol- sulfonephthalein (bromophenol blue), and 3',3",5',5"-tetrachlorophenol- sulfonephthalein (chlorophenol blue) were obtained from Matheson Coleman & Bell. Gelatin used in these studies was purified calfskin (Eastman red label). Triton X-100 was obtained from LaPine Scientific. Ti(III) solutions were made by dilution with 0.10 M HCl of a commercially-available (Matheson Coleman & Bell) 20 % titanium (III) chloride solution. V(II) solutions were prepared by dissolving vanadyl sulfate in 0.10 M HCl, reducing this solution with amalgamated zinc, and then diluting with 0.10 M_HC1 to the desired concentration. The amalgamated zinc (8-30 mesh) was obtained from Fisher Scientific Co. 7" rag" ‘p‘c-vl‘ . -‘ dv-‘V , 0" - d‘ “5‘ 4- -» 4.A-‘ I Q . . . ..I I . . u . q .0. . s . h v .- t» v u .3‘ .‘.$ “0‘ .~\ as fl» F.“ hflh .fi 1‘ F1 3c 2‘ a. p . .2 A» i. s x a: .. a 4‘ ix .5» .2 3 L.. at A a . a 2,. 4 .p . . .5. r1 .7 r a. a. u a: a: a. .. a. h. a. S «u r. as WK s . a; a. 3‘ 2.. a. “a a» c. 5 a . o r b. a u a v a: 2 . I“ u .. $3. C \ ‘.\s \ ‘ .s. n III ELECTROCHEMICAL REDUCTION OF CARBON TETRACHLORIDE IN NONAQUEOUS SOLUTIONS The work of Wawzonek and Duty on the reduction of several halo- methanes in acetonitrile (AN) and N,N-dimethylformamide (DMF), which was briefly summarized above, suggested the intermediacy of dichloro- carbene during electrolytic reduction of carbon tetrachloride. One potentially significant aspect of this report is that electrolytic carbene generation possesses several advantages over any other single synthetic procedure for producing carbenes. For example, as opposed to basic hydrolysis techniques, base sensitive substrates could be present in the generative solutions. Also, as opposed to thermolysis techniques, syntheses could be carried out at relatively low tempera- tures so that thermally sensitive or labile substrates could be toler- ated in the generative solutions. Moreover, as opposed to techniques employing more exotic compounds (diazo compounds, diazirines, and ketenes) as carbene precursors, the starting material could be a safely handled, readily available, and inexpensive compound such as carbon tetrachloride. Furthermore, syntheses could be conducted without inter- ferring side reactions from products of the carbene generative reaction, such as the alcohol produced in the basic haloform hydrolysis technique. In addition, the potential of the reducing agent (the electrode) can be very carefully controlled, thereby minimizing possible side reactions, such as further reductions of carbene products which may be reduced by alkali metals. Finally, the possibility of direct detection of carbenes with the aid of modern rapid-response electrochemical techniques is of obvious signifance. Hence, experiments involving reduction of carbon tetrachloride in 53 5h nonaqueous solvents have been performed by several workers in this laboratory over the past several years. For example, prior to initia- tion of this thesis research, Lundquist and Nicholson (85) studied the reduction of carbon tetrachloride in tetrahydrofuran (THF); the most salient features of their research are included here because they com— plement this thesis research, and because they have not been published elsewhere. In addition, a brief discussion of dihalocarbene chemistry also is incorporated in this chapter to provide a foundation for fol- lowing discussions of electrochemical studies on carbon tetrachloride reduction. Hence, the remainder of this chapter is divided into the following sections: (A) Some Aspects of Dihalocarbene Chemistry; (B) Previous Studies on Electrochemical Reduction of Carbon Tetra— chloride in Nonaqueous Solutions; and (C) Electrochemical Reduction of Halomethanes in Nonaqueous Solutions. A Some Aspects pf Dihalocarbene Chemistry A great deal is now known about the chemistry of carbenes; most of this knowledge has been gained over the last two decades, and has been summarized in a number of excellent reviews (86-91). Briefly, a car- bene is a neutral divalent carbon species in which the carbon atom has two of its four electrons in nonbonding orbitals with the remaining two electrons forming covalent bonds. Two possibilities exist for the electronic state of a carbene, depending on the relative configuration of the nonbonding electrons of the carbon atom. These electrons may occupy either a single orbital with paired spins (singlet quantum state) or they may have unpaired spins and occupy separate orbitals (triplet quantum state). A carbene in the singlet state exhibits the properties of an electron-deficient species since electrons occupy 55 only three of the four atomic orbitals of the carbon atom. Hence, the electrophilic character of carbenes is usually a dominant feature. However, the presence of a nonbonded pair of electrons on the carbon atom of singlet state carbenes leads to a different reactivity pattern than for other electrOphilic species such as carbonium ions. Triplet carbenes generally behave as diradicals by virture of the two unpaired electrons of the carbon atom. Dihalocarbenes are very reactive species, but are far less ener- getic than the simplest carbene, methylene (:CH ). This reduced react- 2 ivity of dihalocarbenes is usually explained on the basis of ground state stabilization of dihalocarbenes by lone pairs of electrons of the halogens interacting with the empty p orbital of the carbon atom. The importance of these interactions is of the order of F >> Cl> Br> I. Hence, chlorofluorocarbene (:CClF) is less energetic than dichloro- carbene (:CClg), and so on. Modern dihalocarbene chemistry began with the work of Hine and co- workers (92—97) when they elucidated the basic hydrolysis mechanism of chloroform, and demonstrated that trichloromethide ion and dichloro- carbene are both intermediates in the hydrolysis reaction. Dichloro- carbene was Shown to be formed from the decomposition of trichlorometh- ide ion in the rate-determining step of the hydrolysis; 0C1; + :CCl2 + 01-. In addition, these workers showed that the presence of perchlorate, nitrate, and fluoride ions does not affect the hydrolysis rate of chloroform. However, they demonstrated that the hydrolysis rate is decreased by addition of iodide, bromide, or chloride ion. Iodide ion decreases the hydrolysis rate to the greatest extent, chloride to the least extent, and bromide ion exhibits an intermediate A. I~ “ ‘V‘u q 0 .g'.“ a v- .v .-~.“' rAv-yn “all a a . k _ 2 A}; A: as» I. n a m 9. to V. .5 Q8 71 h~ «\w Q.» {Iv Pk .rr 56 effect. This decreased rate is attributed to an acceleration of the back reaction: x" + :CCl2 + CCl2X‘. Hine and coworkers also determined that the relative ease of loss of halide ion in the rate-determining step is in the order Br=rI> Cl>> F. Hence, basic hydrolysis of dibromofluoromethane (CHBrQF) yields bromo- fluorocarbene (:CBrF), and so on. In addition, these workers found that CHXF2 haloforms undergo a concerted HX elimination on basic hy- drolysis so that difluorocarbene (:CF2) is generated directly without trihalomethide ion intermediacy. Successful generation of dihalocarbenes 11a the basic hydrolysis of haloforms has been found generally to require the use of aprotic and nonnucleophilic solvents such as benzene, toluene, tetrahydrofuran, 1,2—dimethoxyethane, etc. In addition, the use of a strong base which is also a poor nucleophile is required, potassium pfbutoxide being the base of choice. This generation technique has been widely used in or— ganic synthesis in spite of possessing several important disadvantages. Namely, only moderate yields are generally obtained because of forma- tion of t-butyl alcohol from.pybutoxide (the alcohol then reacts with the dihalocarbene). Moreover, the use of alkoxides restricts this technique to substrates which are not sensitive to strong bases. Other sources of trihalomethide ions have been found which avoid one or both of the disadvantages of the basic haloform hydrolysis tech- nique. For example, trichloromethide ion is formed by reaction of alkoxides with ethyl trichloroacetate. Carbene adducts are sometimes obtained in better yields with this technique because the reaction does not result in alcohol production as a by-product. Another technique for dihalocarbene generation involves the S7 thermolysis of trihaloacetate salts. As with the above two tech- niques, dihalocarbenes are generated gig decomposition of trihalo— methide ions. However, as opposed to the above two techniques, this method can be used for synthesis in the presence of base sensitive substances. Still another dihalocarbene generative method proceeds;1;a therm- ally induced a-elimination reactions. For example, thermolysis of phenylbromodichloromethyl mercury (PhCBrClgHg) yields dichlorocarbene and phenylbromo mercury (PhHgBr). Carbene generation has also been carried out using a wide variety of other techniques. For example, the photolysis or thermolysis of diazo compounds results in the loss of a stable molecule, nitrogen, and provides a common route to carbenes. Formation of carbenes by this method is believed to be quite general when reactions are con- ducted in aprotic solutions. Photolysis of substituted diazirines, the cyclic isomers of diazo compounds, also results in loss of nitrogen and generates carbenes. This procedure is now a well established car- bene generation technique, and is somewhat better than the use of equi— valent diazo isomers because the diazirines tend to be somewhat less hazardous than the corresponding diazo compounds. Halodiazirines have been used to generate halocarbenes. Ketenes have also been used as carbene precursors. This reaction, as with the above two techniques, involves the loss of a stable molecule which in this case is carbon monoxide. The carbene generative procedures that have been mentioned here constitute most of the common generative methods. Other tech— niques, such as pyrolysis of chloroform or carbon tetrachloride at 1500°C, are generally not very amenable to widespread synthetic 58 applications. Carbene intermediacy is generally postulated on the basis of product analysis, and it is now generally thought that in some cases free carbenes are not involved in the reactions. Moreover, some generative procedures are more likely to result in free carbenes than others. Comparative precursor techniques are sometimes used to in- dicate whether or not a free carbene is involved in a given reaction, and identical product chemistry is taken as presumptive evidence for free carbene intermediacy. For example, conparative precursor experi- ments have shown that the intermediate generated by the thermolysis of phenylbromodichloromethyl mercury and the sodium salt of trichloro- acetic acid result in identical product chemistry, and both of these methods are thought to yield free dichlorocarbene. The reaction of alkyl lithium compounds with carbon tetrahalides in the presence of olefins results in dihalocyclopropanes. Alkyl lithium compounds are thought first to undergo lithium-halogen ex- change with carbon tetrahalides so that dehalolithium complexes are formed, which subsequently lose lithium halide in forming cyclopro- panes with olefins. A question remains as to whether aphalolithium complexes react directly with olefins, or first lose lithium halides so that free carbenes react with the olefins. However, it is now generally thought that aphalolithiums, and not free dihalocarbenes, are the actual intermediates. Exact structures of such carbeneoid complexes are unknown, and all that can be said is that lithium halides are apparently associated with the transition state entities during the formation of dihalocyclopropanes. For example, in a recent review article on carbeneoid complexes, K6brich (88) stated the following: 59 "The formation of dichlorocyclopropanes from trichloromethyl lithium and olefins is more complex than was originally thought. The tri— chloromethyl lithium produced from bromotrichloromethane in diethyl ether undergoes several secondary reactions even at temperatures below -lOO°C. It is therefore difficult to decide whether the formation of 7,7-dichloronorcarane observed under these conditions with cyclohexene is due to a reaction of trichloromethyl lithium itself or of an initi- ally formed cleavage product. Trichloromethyl lithium is stable in tetrahydrofuran at —100°C, but it is also inert towards cyclohexene under these conditions. However, its slow decomposition in this sol- vent at —72°C is accelerated by cyclohexene and other olefins. More— over, the difference between the amount of decomposition with and with- out cyclohexene roughly corresponds to the yield of 7,7-dichloronor— carane. In view of these results it is probable that trichloromethyl lithium can react with olefins without first decomposing to dichloro- carbene, and that this cyclopropane formation competes successfully with other reactions of the carbeneoid." The principal reaction of dihalocarbenes with olefins is addition to form dihalocyclopropanes. Moss (98) has stated the following about this reaction: "The relative rates of addition of dihalocarbenes to olefinic carbon bonds usually increases with increasing alkyl substitu— tion at the olefinic group since carbenes are predominantly electro— philic in nature. However, it has also been shown that steric hindrance is an important factor in controlling the rate of dihalocarbene addition to even the most simple olefins. Adverse steric factors presumably cut down the rate enhancements obtainable by alkylation of the olefinic carbon bonds." v nhnk‘ " ‘ b¢¢§ \J \v‘ , . \fl ¢V.L‘ 5“ « jfl" F. q.. gfi-.n V a aflfi‘h QCvVv' D It.-.“ ‘1 t \ 1th 'NES s ‘5 I 60 Dichlorocarbene is so reactive that it also undergoes a wide var- iety of addition and insertion reactions. For example, dichlorocarbene can insert into some carbon-hydrogen bonds, in the absence of strong base. Moreover, dichlorocarbene insertions into carbon-mercury, silicon-hydrogen, and Silicon-carbon bonds have also been reported. Furthermore, secondary and tertiary amines have been shown to react with dichlorocarbene, giving amides on acid hydrolysis of the inter- mediates. In addition, it has been reported that dichlorocarbene adds across the nitrile bond of trifluoromethyl cyanide, and adds across the carbonyl bond of hexafluorodimethyl ketone. In essence, then, the literature on dichlorocarbene indicates that this species is so highly reactive that successful generative procedures depend on a variety of factors. Thus, there appear to be several cru- cial factors for successful dichlorocarbene generation gig decomposi— tion of trichloromethide ions. First, the presence of fluoride, nitrate, or perchlorate ion would have little or no effect, but the presence of iodide, bromide, or chloride ion would retard the rate of dichlorocar- bene generation. Second, if the solvent were not sufficiently aprotic, the trichloromethide ions would react to form chloroform instead of dismutating to dichlorocarbene and chloride ion. Third, if the solvent were not sufficiently nonnucleophilic, dichlorocarbene might well react with the solvent. Hence, choices of supporting electrolytes and sol- vents should be crucial in any attempt to generate dichlorocarbene by electrolytic reduction of carbon tetrachloride. 61 B Previous Studies 9p_Electrochemical Reduction pf Carbon Tetrachloride ip Nonaqueous Solutions Wawzonek and Duty (56) may have proposed an essentially correct electrolysis mechanism for carbon tetrachloride reduction based on in- conclusive analytical data. Therefore, Lundquist and Nicholson (85) reduced carbon tetrachloride electrolytically in THF (a commonly used solvent for carbene generative procedures) using lithium perchlorate as the supporting electrolyte to determine whether dichlorocarbene could be produced under such conditions. The remainder of this sec- tion summarizes this work. All electrolyses were performed in a cell of conventional design (see Figure 7) in which the anode compartment was separated from the cathode compartment by a fine porosity glass frit. The cathode was a mercury pool for macroscale electrolyses, and, of course, a dropping mercury electrode (DME) for polarographic electrolyses. The anode was a spool of platinum wire. The reference electrode was a saturated calomel electrode (sce) with a flowing junction compartment between the sce and the electrolysis solution. All electrolyses were performed with the aid of an electronic three-electrode potentiostat, so that precise control of the potential of the working electrode could be maintained even in solutions of very high resistance such as THF. Macroscale electrolyses were performed in THF, however, the anode compartment always contained AN because oxidation of THF resulted in rapid formation of tar. The supporting electrolyte was 1.0 M_lithium perchlorate. Macroscale electrolyses of carbon tetrachloride were performed in the presence of a 5 fold molar excess of both cyclohexene and fiKb . A vV v.\ YA Fa. a ‘ . t a; . .m“ .1.‘ my ”.4 F Wu!» A: v: . . w.» L... .rt. raw 2. arm n. a La .1 . ~IJ a: a». ”we “vi-[q 5., E. 62 l-methylcyclohexene (neither of these olefins are reducible in THF). These electrolyses were conducted by setting the potential of the working electrode at -1.A0 V y§_sce, which corresponds to the polaro- graphic limiting current for reduction of carbon tetrachloride in THF. This potential is considerably anodic of the half-wave potentials for reduction of 7,7-dichlorobicyclo[A.l.0]heptane and l—methyl—7,7—dichloro- bicyclo[A.l.O]heptane which are reduced at about -2.7 V.X§ sce in THF. Thus, reduction of neither of these dichloronorcaranes could take place. Gas chromatographic analysis with two different columns operated at several different temperatures indicated two well resolved major com- ponents with retention times always identical with those of authentic samples of 7,7—dichlorobicyclo[A.l.0]heptane and l—methyl—7,7-dichloro- bicyclo[A.l.0]heptane. These two gas chromatographic fractions were collected, and their IR and NMR spectra were identical in every re— spect with those of authentic samples of these dichloronorcaranes. All macroscale electrolyses solutions were analyzed after first washing the contents of the cathode compartment with water to remove lithium perchlorate. After this step, solutions were directly analyzed by gas chromatography. By similar treatment of standard solutions, it was established that none of the dichloronorcaranes was lost in this water washing step. To obtain IR spectra, the dichloronorcaranes were first isolated by vacuum distillation, and then purified by gas chroma- tography. Pure samples of dichloronorcaranes were prepared according to published procedures, (99) wherein dichlorocarbene is generated by thermal decarboxylation of sodium trichloroacetate. The dichloronor- caranes were all characterized by boiling point, IR spectra, and NMR ‘1 "V‘ qr: V-‘ L&.- “ H M ta. - . r’31‘ ('5‘ . 'Q. V d S’g i, ‘0'- v": ~"\" A. b' -"‘ b‘*V.‘ +F‘.» ~ n—: 'v I'M» 2 *‘ '7"\_ _ '3 u -... 4‘ 4 x) I I I ‘1, [—1 F'- r I- \ ‘. I v. ‘a ‘44 "Du A \ I wj. a I t6 \ brhhh - s~.yn. a m p‘: “v 63 spectra. In addition, for comparative purposes, a sample of 7,7—di- chlorobicyclo[A.l.0]heptane (7,7-dichloronorcarane) was purchased from the Aldrich Chemical Company (Milwaukee, Wis.). This commercial come pound was identical in every respect with the compound prepared in this laboratory. These macroscale electrolysis experiments were also conducted with the intention of using the method of competing reactions to determine relative rate constants of addition of the electrochemically generated intermediates to cyclohexene and l-methylcyclohexene. It was found that the rate of addition of the electrochemically generated intermedi— ate to l—methylcyclohexene was about 10 times the rate of addition to cyclohexene. This ratio of relative rate constants (1:10) agrees with the value obtained for the reaction of chloroform and lithium metal in THF in the presence of a 5 fold molar excess of both cyclohexene and 1-methylcyclohexene. This ratio is also in fairly good agreement with published values (about 8:1) (99). Therefore, it was concluded that electrolysis of carbon tetrachloride in THF had resulted in formation of an intermediate which was either dichlorocarbene or a Species such as a lithium—carbeneoid complex that gave dichlorocarbene—like products. In summary, this work demonstrated that controlled potential re- duction of carbon tetrachloride in THF must lead to initial generation of trichloromethide ion, and, possibly even to free dichlorocarbene. Furthermore, this work indicates that controlled potential reduction of carbon tetrachloride in THF may be a very useful synthetic route to preparation of substances such as dichloronorcaranes. 6A C Electrochemical Reduction 9£_Halomethanes ip_Nonaqueous Solutions With the exception of macroscale electrolyses in AN, the work of Wawzonek and Duty (56) consisted of conventional polarography in AN and DMF. Hence, prior to applying more modern electrochemical techniques to the study of halomethane reductions, conventional polarographic ex— periments were performed in these two solvents in an attempt to dupli— cate the data reported by these authors. All electrochemical experiments performed in this thesis study on halomethane reductions in AN and DMF were conducted with the aid of a vacuum line apparatus, described above, to obtain the driest conditions possible. In addition, tetrabutylammonium perchlorate (TBAP) was em- ployed as the supporting electrolyte except, when experiments were con- ducted in an attempt to rigorously duplicate the solution conditions reported by Wawzonek and Duty, tetrabutylammonium bromide (TBABr) was used as the supporting electrolyte. However, no differences in the electrochemical behavior of any of the halomethanes were observed by changing the supporting electrolyte from TBAP to TBABr. Unfortunately, the polarographic experiments performed in this laboratory in DMF differed grossly from those reported by Wawzonek and Duty (see Table I for the data of these authors). Specifically, three polarographic waves were obtained for reduction of carbon tetrachloride in DMF, the same number of waves reported by Wawzonek and Duty, but as opposed to their results, relative limiting currents of all three re- duction waves were equal within 10 %. In fact, reduction of carbon tetrachloride in DMF appeared very similar to the polarographic be— havior of this compound in aqueous solutions (60). Moreover, addition . . .1 f.\ r h N V I . . ., . 5.. a. 2. . N 2.. a. a. . m n- A; .3 a... S r: .5 a o L . T w . Wu :C w. 4. ”A . A J. 3*» f. r: a. as a: f. a 1 a o no . .1 Z .r . .3 .. . o C. .. E and of. AV .. . a. . ar. C» . . A» a . S h» . .0 4. a o v w ‘b 0.5 {‘erc ." a 65 of from 1 % to 10 % water had no significant effect on either limit- ing currents or half-wave potentials. Furthermore, addition of chloro- form and methylene chloride demonstrated that the second and third reduction waves for carbon tetrachloride in DMF occurred at exactly equal potentials for chloroform and methylene chloride, respectively. In addition, coulometry on the first reduction wave for carbon tetra- chloride in DMF gave an n-value of 2.0, the same value reported for this compound in aqueous solutions (61). Hence, these data strongly suggest conventional stepwise reduction of carbon tetrachloride in DMF both in the presence and absence of added proton donors. Thus, as opposed to the report of Wawzonek and Duty, these data suggest that no dichlorocarbene is generated by the electrolytic reduction of carbon tetrachloride in DMF. Conventional polarographic experiments were also performed in AN where reduction of carbon tetrachloride resulted in three, or possibly four, waves. All these reduction waves except the last were drawn out and poorly defined so that it is difficult to determine whether there are three or four waves present. A polarogram showing this behavior is given in Figure 10. The morphology of these waves is dramatically different from that described by Wawzonek and Duty, who reported, as shown in Table 1, two polarographic waves of equal height for carbon tetrachloride in AN. The polarographic behavior of carbon tetrachloride in AN was dramatically affected by additions of small amounts (about 1 %) of water. This behavior is also depicted in Figure 10. Specifically, three well defined waves were obtained for carbon tetrachloride in AN when a small amount of water was added. Additions of chloroform and 66 . H E as a m\apm\m p .80 om u pzmfloz nadaoo hfidopmam wobpm sown: &H AIIIIV V .mp%HOMpomHm mchMOQQSw mm 0pmsoaaommm adflcoaamahpSthpop_m.oa.o zpfis Umbbm ampma o: A mausoasoooom ma saoop.dmm m.H no enamosmaom .oa madmam 67 ¢.O 68 methylene chloride to these "wet" AN solutions demonstrated that the second and third waves for carbon tetrachloride are due to reduction of chloroform and methylene chloride, respectively. Moreover, addition of other proton donors such as phenol and benzoic acid has the same effect as adding water. Hence, polarographic reduction in AN in the presence of small amounts of proton donors results in conventional stepwise reduction for carbon tetrachloride. In other words, reduction of carbon tetrachloride under these conditions results in behavior identical to that in DMF or aqueous solutions. Again, as seen in Figure 10, three or four rather broad and poorly defined waves are obtained for carbon tetrachloride reduction in AN in the absence of added proton donors. Hence, the polarographic behavior of carbon tetrachloride in AN both in the presence and absence of added proton donors differs greatly from the morphology reported by Wawzonek and Duty. The poorly defined polarographic waves obtained for carbon tetra— chloride in AN in the absence of added proton donors suggested that something other than conventional stepwise reduction might be occurring. Hence, cyclic voltammetric experiments were performed on carbon tetra- chloride under these conditions. Surprisingly, only three irreversible cathodic peaks were observed for carbon tetrachloride in AN under these conditions. Addition of chloroform and methylene chloride demonstrated that the second and third cathodic peaks for carbon tetrachloride reduc— tion were due to reduction of chloroform and methylene chloride, respec- tively. Hence, in contrast to the conventional polarographic experi- ments, cyclic voltammetry suggested that reduction of carbon tetra- chloride in AN in the absence of added proton donors also occurs in a 69 conventional stepwise manner. Sometimes a fourth irreversible cathodic peak was observed for carbon tetrachloride reduction in AN solutions. Specifically, on addition of very small amounts of water (about 0.1 %), a new cathodic peak appeared at a potential midway between the reduction peaks for carbon tetrachloride and chloroform. This new peak was extremely non- reproducible, and, in fact, always vanished when the amount of added water was increased to about 1 %. Furthermore, addition of other proton donors such as phenol and benzoic acid had the same effect as small incremental additions of water. This unusual and interesting behavior was not investigated in any further detail, only because it did not appear that such behavior could be related to dichlorocarbene generation. The polarographic and cyclic voltammetric experiments conducted in this laboratory indicated that dichlorocarbene is not generated by electrolytic reduction of carbon tetrachloride in either DMF or AN. This conclusion was substantiated by conducting a macroscale electrol- ysis of carbon tetrachloride in AN in the presence of cyclohexene (this compound is not reducible under these conditions) with tetrabutyl- ammonium perchlorate as the supporting electrolyte. The electrolysis was performed at a mercury pool and controlled potential of -l.0 V.X§ see, which corresponds to the limiting current of the first reduction wave for carbon tetrachloride in AN. Only two polarographic waves remained following exhaustive electrolysis of the first wave. The two remaining waves corresponded to chloroform reduction (E1/2 = -l.6 V y§_sce) and.methylene chloride reduction (El/2 = -2.5 Viyg sce). No wave was observed for reduction of 7,7—dichloronorcarane (E = —2.8 V -i/2 7O y§_sce) implying that dichlorocarbene was not generated. The remarkable differences between results obtained in this laboratory and the laboratory of Wawzonek and Duty probably can be attributed to some or all of the following factors. First, nonaqueous electrochemical experiments performed for this thesis research were conducted with the aid of a vacuum line apparatus which minimized water contamination. In contrast, Wawzonek and Duty used a polaro- graphic cell which essentially consisted of a rubber-stoppered beaker which, in turn, was partially immersed in a thermostated water bath. Hence, the data of Wawzonek and Duty correspond to a far higher de- gree of water contamination than data reported here. In addition, electrochemical measurements performed for this thesis research were generally made using modern three-electrode instrumentation so that i§_dr0p was compensated electronically. On the other hand, Wawzonek and Duty used two-electrode electrolysis cells so that their data must reflect effects of uncompensated resistances. In summary, the electrochemical behavior of carbon tetrachloride in both DMF and AN was found to be completely different than reported by Wawzonek and Duty. For example, these authors reported that methyl- ene chloride was not reducible in AN, whereas it was found in this thesis research that this compound always gives a well defined polaro- graphic wave in AN (El/2 = —2.5 V XE sce). However, more significantly, the experiments performed in this thesis research strongly suggest that no significant amount, if any, of dichlorocarbene is generated by elec— trolytic reduction of carbon tetrachloride in either DMF or AN. Norw- generation of dichlorocarbene in these two solvents is probably dUEE‘tO the fact that trichloromethide ion, the initial electrode reductior1 71 product of carbon tetrachloride, is protonated forming chloroform at a far greater rate than the rate at which this ion can decompose to dichlorocarbene and chloride ion. In fact, Hine and coworkers (97) have reported that the rate of protonation of trichloromethide ion is about five orders of magnitude larger than the rate of decomposi— tion into dichlorocarbene and chloride ion. Thus, even in such rela— tively aprotic solvents as DMF and AN, the rate of protonation of trichloromethide ion is probably still significantly larger than the rate of decomposition of this ion. IV. ELECTROCHEMICAL REDUCTION OF HALOBENZENE COMPOUNDS IN N,N-DIMETHYLFORMAMIDE Wawzonek and Wagenknecht (58) proposed that dehydrobenzene is formed during electrolytic reduction of gfdibromobenzene and gybromo- chlorobenzene in both DMF and AN in the presence of 0.20 M tetrabutyl- ammonium bromide as supporting electrolyte. Their work is summarized in the introduction to this thesis, and their data are reproduced in Table II. Briefly, they prOposed that pydibromobenzene undergoes an initial two-electron reduction to form the gfbromophenyl anion which, instead of being protonated to form bromobenzene, loses bromide to form dehydrobenzene. Furthermore, primarily from the fact that a single reduction wave is obtained for gfdibromobenzene, they proposed that the formal potential for reduction of dehydrobenzene is anodic of that for grdibromobenzene so that when dehydrobenzene is formed, it is immediately reduced to benzene. On the basis of the results of Wawzonek and Wagenknecht, several gydihalobenzene and monohalobenzene compounds were studied, first to verify the electrolysis behavior reported by Wawzonek and Wagenknecht, and then to attempt to detect directly dehydrobenzene with the aid of cyclic voltammetry. In general, both conventional polarography (but with a three- electrode potentiostat) and cyclic voltammetry were used, and experi- ments were performed with the vacuum line apparatus described earlier. The compounds studied are listed in Table III; all experiments were conducted in DMF with 0.1 M tetrabutylammonium perchlorate as sup- porting electrolyte. 72 73 Table III. Half-wave Potentials and Diffusion Current Constants for Halobenzene Compounds. I I I _ a b I _ a b | EILL/2 Ed I E-l/2 Ed | I I I Iodobenzene l I 1.87 (A.8)C I I gfdiiodobenzene l 1.73 (5.1)d | =l.9 (8.0)C I I Bromobenzene I I 2.57 (A.7)C . | I gfdlbromobenzene I 2.05 (8.7)C I Chlorobenzene I I 2.79 (A-SIC I gydichlorobenzene I 2.53 (5.8)e I 2.79 (9.1)C pybromochlorobenzene I 2.10 (5.7)e I 2.79 (8.7)C I I a. b _ . * 2/3 1/6 in V.X§ sce Id — id/Com t Cmeasured at -2.90 V pp sce dmeasured at —l.80 y§_sce emeasured at -2.70 V.y§ sce 7A A Results Before discussing the results of these experiments for the indi- vidual compounds, a few general remarks are appropriate. First, all of the compounds exhibited complex behavior, so that in many cases polarography served only to indicate the number of reduction steps involved. For example, polarograms for iodobenzene, pfdiiodobenzene, and pfdibromobenzene exhibited maxima or prewaves which are indicative of adsorption and/or stirring phenoma (100). Unfortunately, these effects preclude any quantitative treatment of the polarographic data. However, these effects were absent or minimal for the other four com— pounds in Table III. The dependence of limiting current on mercury column height can be used to determine whether the reduction is dif- fusion controlled (100), and studies of this nature were conducted for all of the compounds except pydiiodobenzene. In every case, limiting current y§_h}/2 plots were linear, indicating that the elec- ‘trode reaction is diffusion controlled. Table IV lists the potentials at which these limiting current studies were conducted. The cyclic voltammetric data also was very complex. Most of Iihese compounds exhibited either one or two broad cathodic waves. Ifiaese waves generally exhibited decreasing values of ip/vl/2 with in- c3I‘easing scan rate. Moreover, peak potentials for all waves shifted <3€Ithodically with increasing scan rate. Data of this nature suggests 131163 presence of a succeeding chemical reaction following charge trans— 1?€31?. The data, however, was not amenable to further analysis with £1113? of the available theories for cyclic voltammetry because of compli— C3‘3--‘l:.ions such as adsorption and/or stirring phenomena. In essence, IrICDESt of the compounds behaved similarly, however, the differences are 75 Table IV. Potentials at which Limiting Current y§_p}/2 studies were conducted for Halobenzene Compounds. Compound -Ea chlorobenzene 2.90 gfdichlorobenzene 2.70 and 2.90 gybromochlorobenzene 2.50 and 2.90 pfdibromobenzene 2.AO bromobenzene 2.90 iodobenzene 2.10 8'. in V vs sce 76 sufficient that results are presented below for each compound individ— ually. 1. Iodobenzene Polarograms were obtained for this compound at concentrations of O.A7, 1.13, and 1.85 EM: One reduction wave, with E1/2 = -1.87 V.XE sce, was found. Each polarogram exhibited a maximum centered at -l.95 V y§_sce. Cyclic voltammetry also was performed on this compound at concen- trations of O.A7, 1.13, and 1.85 EM: The scan rate was varied from A0 mV/sec to 100 V/sec. Under these conditions, a single broad cathodic wave was always found. The value of (Ep — Ep/2) increased with scan rate from 90 mV at the slowest scan rate to 350 mV at 100 V/sec. These peak potential shifts also depend on concentration. For example, the peak potential for the 0.1+7 _n_1M_ solution shifted 80 mV cathodically for a lO—fold increase in scan rate, whereas, for the 1.85 EM solution, it shifted 200 mV cathodically for the same increase in scan rate. {Phe value of __i_p/vl/2 decreased with increasing scan rate. 2. pydiiodobenzene Polarograms were obtained for this compound at concentrations of C)-25, 0.50, 1.00, 1.51, and 2.08 mM, All polarograms were very com- IPJJex. First, a small prewave started at about -l.A V y§_sce on every IRCXlarogram. This prewave was concentration independent, indicating £3131rong adsorption of the electrode reaction product (100). After t:115.s prewave, a second more conventional wave appeared with E1/2 — -l.73 \, \rs sce. Following this, a third wave appeared which exhibited a 'J‘Ellfge maximum on the first portion of the wave. This third wave and 7t;}1€3 maximum both Showed small cathodic shifts with decreasing 77 gydiiodobenzene concentration. Because of the maximum, E1/2 for this third wave could only be estimated as -l.9 Vuyg see. In addition, because of the prewave, E1/2's and IdIs always were measured at the highest prdiiodobenzene concentration where the effect of adsorption is minimized. A few cyclic voltammetric experiments were performed on this compound at a concentration of 0.56 EM: However, as was the case for the conventional polarograms, these current-potential curves were very complex, and were unsuitable for purposes of obtaining additional quantitative information. 3. Bromobenzene Polarograms were obtained for this compound at concentrations of 0.39, 1.01, and 1.97 mM. One reduction wave, with E1/2 = -2.57 V.X§ sce, was found. No prewaves or maxima were observed. Cyclic voltammetry also was performed on this compound at concen- ‘tretions of 0.39, 1.01, 1.97 EM: The scan rate was varied from 50 mV/sec ‘tc> AO V/sec. Under these conditions, a single broad cathodic wave was ELlways observed. The value of (E — E ) was independent of concen- -‘p -'p/2 thration, but increased with scan rate from 130 mV at the lowest scan JFTIte to 200 mV at A0 V/sec. The peak potential shifted 150 mV cathod- ‘i‘3ELlly for a lO—fold increase in scan rate, and was independent of C“311centration. The variation of‘ip/vl/2 with scan rate was very er- lrerttic and showed no clear trends for any of the concentrations studied. A. gydibromobenzene Polarograms were obtained for this compound at concentrations of C)"23(3, O.AA, 1.00, and 1.88 EM: One reduction wave, with E1/2 = -2.05 V V “4§i. SSce, was found. The 0.20 EM solution exhibited no maximum, but the 78 other three solutions all showed a maximum centered at -2.1 V y§_sce. Cyclic voltammetry also was performed on this compound at concen- trations of 0.20, 0.7A, and 1.92 mM, The scan rate was varied from 75 mV/sec to 500 V/sec. Under these conditions, a single broad cath- odic wave was always observed. The value of (E - E ) increased ‘13 -'p/2 with scan rate from 100 mV at the lowest scan rate to A00 mV at 500 V/sec. Peak potential shifts were also a function of concentration. For ex- ample, the peak potential for the 0.20 mM solution shifted 100 mV cath— odically for a lO—fold increase in scan rate, whereas, for the 1.92 EM solution, it shifted 330 mV cathodically for the same increase in scan .rate. The value of _ip/vl/2 also was concentration dependent, but de— <1reased with increasing scan rate. 5. Chlorobenzene Polarograms were obtained for this compound at concentrations of O. 31, 0.78, and 1.68 EM: One reduction wave, with E1/2 = -2.79 V‘yg Escez, was found. No prewaves or maxima were observed. Cyclic voltammetry also was performed on this compound at concen- tIYatiions of 0.31, 0.78, and 1.68 mM, with a scan rate range from 25 .nIV/sec to 1 V/sec. One broad cathodic wave, which began to merge with the background current at scan rates of greater than 1 V/sec, was observed. The value of (E — E ) ranged from 80 to 100 mV, and -'p -I>/2 inf31?€eased slightly with increasing concentration. However, this value wars inadependent of scan rate. The value of ip/vl/2 was independent 0 . . . . f. cl‘Orlcentratlon, and decreased wlth lncreaSlng scan rate. 6. gfdichlorobenzene 3Polarograms were obtained for this compound at concentrations of (3 'EBSD, 0.56, 0,83, 1.13, 1.AO, 1.66, and 1.96 EM: Two reduction waves, 79 wlth E1 prewaves or maxima were observed. Addition of chlorobenzene to these /2 of -2.53 and -2.79 V y§_sce respectively, were found. No solutions indicated that the second wave is due to chlorobenzene re- duction. Cyclic voltammetry also was performed on this compound at a con- centration of 0.29 EM: The scan rate was varied from A0 mV/sec to 20 V/sec. Two broad cathodic waves were observed for scan rates be- §low 1 V/sec, whereas only one peak was found for faster scan rates. kHaen two peaks were observed, they both shifted cathodically with in- crweasing scan rate. The behavior of the second wave was identical to tliat found for chlorobenzene itself (cf., above). The value of (E3 - Ep/g) for the first wave was 100 mV, and was independent of scan ‘19 rerte. The peak potential of the first wave shifted 70 mV cathodically fcxr a lO-fold increase in scan rate. Finally, _i_‘p/vl/2 decreased with iricreasing scan rate. 7. pybromochlorobenzene Polarograms were obtained for this compound at a concentration of 0.5%5 Egg Two reduction waves, with 21/2 of —2.10 and —2.79 V XE sce respbectively, were found. No prewaves or maxima were observed. Addi- tiOrl of chlorobenzene to this solution indicated that the second wave is determine n-values, the diffusion current constant, Id, for gfdi- Iarxamobenzene indicates that reduction of this compound in DMF takes Ipljace by an overall four-electron process. Nevertheless, this fact nerither proves nor disproves that dehydrobenzene is involved as an irrtermediate during the electrolysis of pfdibromobenzene. As previously discussed, Wawzonek and Wagenknecht proposed that dekurdrobenzene is generated by the decomposition of gybromophenyl anion, imIXLying that the rate of decomposition is much greater than the rate or Errotonation to form bromobenzene. k _ -—---3- +131” 01‘ AlSCD it has been shown that the decomposition is reversible (101), 3 an] aétpect that Wawzonek and Wagenknecht apparently did not consider. This; could be an important point, since these authors con- ductued their experiments in the presence of 0.20 M_bromide ion. In- teres'tingly, this factor apparently is not very significant, since 82 the use of a perchlorate salt in the work described here resulted in, for most of the compounds studied, essentially the same polarographic behavior as reported by Wawzonek and Wagenknecht. The electrolysis mechanism proposed by Wawzonek and Wagenknecht for the reduction of gfdibromobenzene seems implausible for several reasons. Specifically, even if the electrogenerated gfbromophenyl anion does undergo significant decomposition to dehydrobenzene in IMdF, there are significant reasons for doubting whether dehydroben- zenie is subsequently reduced at a formal potential anodic of or very clx)se to that for gfdibromobenzene. For example, it is known that thus formal potential for reduction of benzene is so cathodic that it limes beyond the decomposition potential of most supporting electrolytes arud/or solvents that have been used. In fact, approximate molecular oriaital calculations indicate that the formal potential for the reduc- tixan of benzene is about -3.5 V.X§ ace in solvents such as DMF (102). Furthermore, it is known that activated acetylenes (eg., diphenyl- acestylene) are reduced at slightly more negative potentials than their oleiiinic counterparts (eg., diphenylethylene). In view of these facts, the.:formal potential for the reduction of dehydrobenzene should be 91£N3€3 to that of benzene. Hence, if dehydrobenzene is generated during the Gilectrolysis of gfdibromobenzene, some other mechanism would seem to I>€h necessary to account for the enhanced diffusion current constant. VJhen dehydrobenzene is generated by the decomposition of gfbromo- phenYl anion in the absence of competing substrates, the following reaction is known to proceed (103)' 83 " Br Br + —-—— ,etC. Nevertheless, this reaction was not considered by Wawzonek and Wagenknecht. Clearly, generation and subsequent reduction of bromobi— phenyl and homologous compounds would constitute a more plausible ex- planation than dehydrobenzene reduction for the increased diffusion current constant observed for gfdibromobenzene. In addition, Wawzonek and Wagenknecht reported that in the absence of furan, reduction of pfdibromobenzene led to fluorescent products which they could neither identify nor explain. A mechanism involving the generation and sub— sequent reduction of bromobiphenyl and higher homologs does, however, account for the formation of fluorescent products. Wawzonek and Wagenknecht did not report any experiments on the reduction of either gfdiiodobenzene or pydichlorobenzene. The data in Table III Show that these two compounds, unlike pedibromobenzene, undergo electrolysis in DMF in a conventional stepwise manner. For example, during the first polarographic wave, gfdichlorobenzene is reduced to chlorobenzene which, in turn, is reduced to benzene during the second reduction wave. Hence, both the grchlorophenyl anion and the ggiodophenyl anion must be protonated to form the respective halo— benzenes at a much faster rate than the rate at which either of these grhalophenyl anions can decompose to dehydrobenzene in DMF. Further— more, the data show that gfbromochlorobenzene is reduced to the 97 Chlorophenyl anion which is protonated to form chlorobenzene, without 8A formation of dehydrobenzene. In summary, these experiments substantiated the polarographic data reported by Wawzonek and Wagenknecht for the electrolytic re— duction of pfdibromobenzene in DMF. However, there is a more plaus- ible explanation for the enhanced current for this compound than the mechanism proposed by these authors. In addition, as opposed to the report of Wawzonek and Wagenknecht, these experiments indicated that F the electrolytic reduction of gfbromochlorobenzene in DMF occurs in a conventional stepwise manner. In other words, the electrolytic re— duction of pybromochlorobenzene is analogous to that of gfdichloro— benzene since the initial electrode reduction product of both of these compounds is pfchlorophenyl anion, which is protonated forming chlorobenzene at a far greater rate than the rate at which this ion can decompose to dehydrobenzene and chloride ion. The fact that electrolytic reduction of pfbromochlorobenzene and pydichlorobenzene in DMF results in chlorobenzene formation during the first polaro- graphic wave is not very surprising, since, even in liquid ammonia (a more basic solvent than DMF), pfchlorophenyl anion is protonated 1.5 times faster than it looses chloride ion (10A). Moreover, the fact that these two compounds exhibit behavior different from that of gydibromobenzene is reasonable since it is known that the rates of loss of halides from.97halophenyl anions is in the order Br > 01 > F (10A). Thus, this study showed that, except for pydibromobenzene, the electrolysis of gfdihalobenzene compounds in DMF generally results in conventional stepwise reduction because the initial electrode re— duction products (i.e., the gfhalophenyl anions) are protonated at a faster rate than the rate at which these ions can decompose to de- hydrobenzene. V. REDUCTION OF SULFONEPHTHALEIN ACID—BASE INDICATORS The work of Senne and Marple (59) on the reduction of phenol red, which was briefly summarized above, suggested that free radicals of mod— erate stability are produced in aqueous solutions. Hence, compounds such as phenol red provide an attractive system for the evaluation of the usefulness of electrochemical relaxation techniques for studying reactive intermediates, since the experiments can be conducted with aqueous solutions. Thus, cyclic voltammetry can be subjected to ex- perimental evaluation without the complications that accompany non- aqueous electrochemical studies. A Compounds Investigated The compounds that are reduced according to the disproportionation mechanism are given in Table V. Although phenol red and cresol purple were studied most extensively, each of the nine compounds for which a rate constant is given was investigated in sufficient detail to con- firm the mechanism. Moreover, the same mechanism appears to apply to chlorophenol blue even though the rate constant for this compound was too large to measure with cyclic voltammetry. Thus, although most subsequent discussions will pertain to phenol red and cresol purple, except for quantitative differences in thermodynamic and kinetic data, the discussions are applicable to all of the compounds listed in Table V. In addition to these ten compounds, bromophenol blue and bromo- chlorophenol blue were also investigated. These studies showed that the reduction mechanism for these two compounds is more complicated than for the compounds listed in Table V. The behavior of these two Compounds was not, however, studied in sufficient detail to determine 85 86 Table V. Half—wave Potentials and Disproportionation Rate Constants for Sulfonephthalein Indicator Radicals?"b Compound pHC _'§1/2 Rape Constint? V‘yg sce M7 - sec— cresol purple A.8 0.56 1.87 x 101 thymol blue A.8 0.66 3.25 x 101 bromothymol blue A.8 0.61 6.67 x 101 bromocresol green 2.5 0.38 6.91 x 101 phenol red 2.5 O.A7 1.62 x 102 phenol red A.8 0.60 1.65 x 102 phenol red 6.8 0.72 1'6A x 102 cresol red A.8 0.65 6.62 x 102 bromophenol red 2.5 O.A6 1.01 x 103 chlorophenol red 2.5 O.AO 1.06 x 103 bromocresol purple A.8 0.6A 2.0 x 105 chlorophenol blue 2.5 (0.32)8 ----- a 25% (by weight) methanol-water. Cyclic voltammetric data for these compounds are given in Tables VIII-XXXVII. C 0.10 M_total citrate buffer. First eight rate constants are within i 10% of stated values, and k for bromocresol purple is within 1 20% of stated value. e Apparent E 1/2 since k_is too large to measure. 87 the actual mechanism of reduction. B Mechanism According to Bates (105), sulfonephthalein indicators are repre- sented structurally as hybrid ions containing a central carbonium ion and a negative charge on the sulfonate group with color changes re- sulting from successive dissociations of the phenol groups. Thus, probable structures of the three indicator forms of cresol purple are I“ .5 " '.' 0 OH OH + + — ° 3 “*3 1'. - "3 "3 .5. "3 (”I3 0&0» ‘— 0 c+ <— H 0+ 303 so; so; (Ib) (Ia) (I) base form acid form strong acid form The color transformation intervals of cresol purple occur in the pH ranges of 1.2 to 2.8 (red to yellow) and 7.A to 9.0 (yellow to pur- ple) (105). Hence, form I is the predominant species in solution be- low pH 1.2, and so on. Polarograms were obtained for cresol purple at ten pH's over a range from 0.3 to 7.A. Two reduction waves of equal height were ob- tained at each pH. The E1/2 3§_pH behavior of both waves is shown in Figure 11. Below pH 1.2, 31/2 of the first wave is pH—independent and equals -O.A0 V'yg sce. Form 1 is the predominant species in solution in this pH region, and the polarographic data indicate that this form Figure 11. 88 a,b,C,d E'/2 XE pH behavior of Cresol purple. 1 [:1 H01 solutions. (:) Citrate buffer solutions. [:5 Phosphate buffer solutions. a All solutions approximately 10% methanol-water. b * CO = 1.0 mM at each pH. C Wave II is masked by hydrogen discharge below pH = 2. d . In pH reglon 2 to 8, E1/2 cathodically with increase in unit pH. of Wave I shifts 59 mV 89 I23- mom—owe e— @29r--&~ 7A,— WAVE JI-/ 0080— / é) ._ / / 072— fig/ A __ / a) // 80.64; A a _ a“ > / > 0.56I— 934— WAVE I I / V __ Q / 0.48— @/ ' / S ._ // LLJ G/ o.4oI3{§L—/—— _ I 1 I I I I I I o l 2 3 4 5 e 7 a 9 90 is reduced directly under these conditions. OPI ()H H + + le_-:: HO co (1) However, in the pH region 2.8 to 7.A, E1/2 of the first wave shifts 59 mV cathodically per unit increase in pH. Form la is the predominant species in solution in this pH region, and the polarographic data in— dicate that one proton is consumed in reducing species Ia to 11. Hence, in this pH region, the stoichiometry for the first wave is (3H 4. "0 04- + [9 ¢——' H c o (2) o; °3 (Ia) (II) The fact that the limiting current of the first reduction wave is 91 pH-independent, and that E1/2 shifts in the manner described above, indicates that the proton-transfer chemical equilibrium between species I and Ia is rapidly established. Moreover, it is known that undisso- ciated forms of organic acids are generally reduced anodic of the dis- sociated forms (106). Hence, these data indicate, but do not prove, that, in the region where species Ia predominates in solution, species I is the form of the indicator undergoing reduction. Thus, the prob- able reduction mechanism, which is essentially the one cited by Senne and Marple, is H H 4. HO 0+ 2 HOGC‘? +I62Ho©——co (3a) so3 so3 so3 (Ia) (I) (II) OII H ”'3 H3 __ CH3 CH3 Ho©— o 'I' le zno-Q— - (3b) 0; °3 (II) (III) 92 OH OH H (II) (I) (111) OH O HO c- + H -—+ Hog-— -H_ (3d) so; °3 (III) (IV) C Polarogpaphy The polarographic behavior of the compounds listed in Table V is similar to that reported by Senne and Marple (59) for phenol red and thymol blue. Two reduction waves are generally observed; for the most stable radicals the first reduction wave is reversible, but becomes progressively less so as the rate of Reaction 3c increases. The sec- ond reduction wave is always irreversible. The height of the first wave varies from apparently one electron for cresol purple to two electrons for chlorophenol blue, with the other eight compounds in 't 93 Table V having intermediate apparent pfvalues. In each case the first wave corresponds to Reaction 3a with enhancement from Reaction 3c for those compounds where Reaction 3c is rapid on the polarographic time scale. The second wave corresponds to Reaction 3b, and is always ir- reversible because of reaction 3d. Because of these kinetic complica— tions, half—wave potentials for either wave are not directly of thermo- dynamic significance. The exception is those compounds, such as cresol purple, for which Reaction 3c is sufficiently slow on the polarographic time scale that the first wave is unperturbed. It might be argued that Reaction 30 should be reversible, and indeed lie to the left. Of course, a thermodynamic analysis using half-wave potentials for these two waves would lead to this conclusion. However, such an analysis is not valid because the second wave is ir- reversible. In fact, the polarographic results, as well as all the other data presented below, suggest the reaction sequence as written. Reaction 3c is the rate-determining step in the disproportionation reaction, and is essentially irreversible because Reaction 3d is very rapid. D Controlled Potential Reduction Controlled potential electrolysis of the compounds in Table V on either the first or second reduction wave leads to the sulfonephthalin, IV. For cresol purple, thymol blue, bromothymol blue, and bromocresol green the corresponding radical, II, is sufficiently stable that a con- ventional polarographic wave is observed for II following controlled potential reduction on the first wave. Figure 12 depicts the overall polarographic behavior at various Figure 12. 9A Polarographic waves for the controlled potential electrolysisa of a sulfonephthalein indicator that is reduced to a moder- ately stableb radical which subsequently disproportionates. Curve Curve Curve Curve Curve Time prior to controlled potential electrolysis Time immediately following exhaustive electrolysis on first reduction wave Time at which 1/2 of the electro-generated radical has disproportionated Time at which 3/A of the electro-generated radical has disproportionated Time at which all the radical has dispropor- tionated a At potential on limiting current of first reduction wave. b Stable on macroscale electrolysis time scale. 95 potential I6 8 420 29.30 9,228 96 times prior to and following controlled potential electrolysis on the first reduction wave for a sulfonephthalein indicator that is reduced to a moderately stable radical, one which subsequently disproportion- ates on a time scale that is longer than the time required for electrol- ysis. Figure 13 shows that the electrogenerated radical, II, decays in a second—order fashion while the carbonium ion, 1, and carbanion, III, appear in a second-order manner. These simultaneous events constitute Reaction 3c. The total current height of the second reduction wave is equal to the sum of the currents of the oxidation wave plus twice that of the first reduction wave. Controlled potential electrolysis experiments were performed on cresol purple, and the behavior agreed qualitatively with the morphol- ogy depicted in Figures 12 and 13. Thus, although this method was not used to determine rate constants, it provided further evidence of the disproportionation mechanism for cresol purple. Coulometry was performed on cresol purple, and pyvalues of 1.8 and 1.9 were obtained by controlling the potential on the limiting current of the first and second reduction waves, respectively. These data also confirm the disproportionation mechanism for cresol purple, since exhaustive electrolysis on either reduction wave should result in an prvalue of 2 for this mechanism. No conventional polarographic oxidation waves were observed fol- lowing controlled potential electrolysis on the first reduction wave for those compounds in Table V whose radicals disproportionate rapidly 3 Mfl-sec-l). on a polarographic time scale (k_greater than about 10 Hence, only the two reduction waves corresponding to Reactions 3a and 3b were observed for these compounds. For example, if a radical at an lfl'gr- '3 \_ 17—1 Figure 13. 97 Limiting current XE time behavior of the polarographic waves following controlled potentiala electrolysis of a sulfonephthalein indicator that is reduced to a moder- ately stableb radical which subsequently disproportionates. Curve X: Curve Y: Curve Z: Limiting current of oxidation wave (11 + 1) Limiting current of first reduction wave (1 —> 11) Total limiting current at potential on second reduction wave (1 + II and II + III) Time immediately following exhaustive elect- rolysis on first reduction wave Time at which 1/2 of the electro—generated radical has disproportionated Time at which 3/A of the electro-generated radical has disproportionated Time at which all the radical has dispropor- tionated a At potential on limiting current of first reduction wave. b Stable on macroscale electrolysis time scale. 98 _ Y 1 6 4 2 O 35:30 328: 03.6.3 Time 99 initial concentration of l mM_possesses a disproportionation rate con- stant of 103‘Mfl—sec-l, then half of the radical reacts in one second. Clearly in this case no conventional polarographic oxidation wave can be detected because polarographic eXperiments involve time scales of at least a few minutes duration. However, even though an oxidation wave is not observed for these rapidly disproportionating radicals, the enhancement effects of Reaction 3c are still manifested by the two reduction waves when incremental controlled potential reductions are conducted on the first wave. For example, as the controlled potential reduction on the first wave progresses, the relative limiting current ratio of the two reduction wave changes because the half-life of Re— action 3c is concentration-dependent, and increases as the concentra- tion of species I decreases. E Cyclic Voltammetry At sufficiently high scan rates (less than about 10 V/sec), the first reduction wave for the first eight compounds listed in Table V is unperturbed by the disproportionation (Reaction 3c) and corresponds to a reversible (ks greater than about 0.02 cm/sec) one—electron charge transfer. Experiments of this type are the source of the thermodynam- ically meaningful half-wave potentials in Table V. Bromocresol purple required scan rates of 30 V/sec to determine the rate of Reaction 3c. Nevertheless, in this case it was still possible to obtain a thermody- namically meaningful half-wave potential. For every compound listed in Table V, the second reduction wave (Reaction 3b) is irreversible even at the highest scan rates employed (eibout 100 V/sec). Thus, direct evidence for Reaction 3d could not be 100 obtained, nor could meaningful half-wave potentials be measured. For compounds where disproportionation is appreciable on the cyclic voltammetric time scale, the behavior of the first wave agrees quanti— tatively with theory for an irreversible disproportionation reaction following reversible electron transfer (52). The morphology of ip/yé/Q (i? is cathodic peak current and 1 is scan rate) with scan rate agrees with theory. Moreover, individual cyclic voltammograms agree exactly with theory. For example, the curve of Figure 1A is an experimental cyclic voltammogram for reduction of phenol red in an acetate-buffered aqueous solution at pH A.8. The points are theoretical for dispropor- tionation (52) and a rate constant of 3.A0 x lOQIMTl-sec—l. Nonelectro— chemical experiments carried out under similar solution conditions, de— scribed below, also yield the same rate constant. Thus, the excellent agreement between cyclic voltammetric theory and experiment for phenol red was further substantiated by a completely different generation as well as measurement technique. Furthermore, measurements of the rate constant of Reaction 3c for cresol purple (in 25 % by weight methanol-water with a citrate buffer at pH A.8) by cyclic voltammetry agree with the rate constant obtained by polarographic monitoring of the oxidation wave of the radical fol- lowing prior reduction with V(II), which is described below. F Chemical Reduction Reduction of all compounds listed in Table V with amalgamated zinc leads to the sulfonephthalin, IV. This is equivalent to controlled LPOtential electrolysis on the limiting current of the second reduction Imave (Reaction 3b). Chemical reduction can also be accomplished using either V(II) or Ti(III), both of which possess formal potentials Figure 1A. 101 Comparison of theory (points) and experiment (curve) for disproportionation reaction mechanism for cyclic voltam- metry. Aqueous solution * CO = 3.0A mM_phenol red (unpurified sample) 0.020 % gelatin 0.020 M acetic acid 0.020 M sodium acetate Scan rate = 67 mV/sec Theoretical points for W = 0.350 Rate constant = 3.A x 102 M-1 - sec-l 102 _ \s 0 _ 0 _ 6 0 u ’ l 0 l O 0 0 u 9 rd _ ‘x W ,2 ‘x m n O m N e IF 1“ _ _ l _ _ F l _ Q m 8 6 4 0 “1 ml. 0....30 103 corresponding to the limiting current region of the first reduction wave (Reaction 3a). Reduction with these two chemical reductants also ultimately leads to the sulfonephthalin because of the disproportiona- tion reaction (Reaction 3c). Preliminary experiments showed that the rate of reduction of phenol red with V(II) occurred at about twice the rate as when Ti(III) was used as the chemical reductant. Hence, most chemical reductions cor- responding to reductions on the limiting current region of the first reduction wave were carried out with V(II). Relative rate of reduc- tion by V(II) and Ti(III) of compounds other than phenol red were not studied because this aspect, although interesting, is outside the scope of this investigation. G Chemical Generation with Electrochemical Detection Reduction of most of the compounds listed in Table V leads to rad— icals which disprOportionate so rapidly that no oxidation wave is ob- served following controlled potential macroscale electrolysis on the first reduction wave. However, the first four compounds listed in Table V are reduced to radicals of sufficient stability (k_less than l-sec.l) that oxidation waves are observed following con- about 107M? trolled potential electrolysis on the first reduction wave. Hence, in these cases, radical concentration can be directly monitored with a DME at a potential corresponding to the limiting current region of the oxidation wave. Nevertheless, exhaustive electrolytic reduction Of millimolar solutions of these compounds requires hours, even when Irapidly stirred mercury pools are used, and therefore polarographic nlonitoring of the oxidation wave is quite difficult because a signifi- ‘3ant fraction of the radical reacts before polarographic monitoring 10h can be initiated. Thus, polarographic monitoring would lead to easier rate constant determination if the radical could be generated homogene- ously during a relatively small time interval. As indicated above, such homogeneous reduction can be accomplished with V(II). Polarographic monitoring experiments were performed with 2‘mM solutions (about 0.0A millimoles) of cresol purple into which 1.00 ml (about 0.003 millimoles) of a V(II) solution was injected. (Rapid mixing is achieved by injecting the chemical reductant into the cresol purple solution with a syringe.) The decay of the oxidation wave (see Figures 12 and 13) was monitored with a DME set at -O.20 V‘yg sce. After the completion of the disproportionation reaction, the initial amount of radical formed (or V(II) added) was determined by measuring the decrease in the limiting current (the difference between curves A and E in Figure 12) of the first reduction wave for the cresol purple (with a correction for the volume change caused by adding the V(II) solution). As previously discussed, the decrease in the limiting current of the first reduction wave is directly proportional to the amount of radical that was initially generated. The results of these experiments on cresol purple are given in Table VI. Rate constants for the disproportionation of cresol purple radical were determined by measuring the first two half-lives of the decay of the oxidation wave of the radical (times C and D on Curve X of Figure 13). As shown in Table VI, the rate constant for the dis— proportionation of cresol purple determined polarographically is 21 I 2 ‘Mfl-sec-l. This value agrees reasonably well with the rate determined by cyclic voltammetry, namely, 19 i Mil—sec 105 Table VI. DisprOportionation Rate constants for Cresol Purple Radicala by Polarographic Monitoring of Decay of Oxidation Wave of the Radical. Initial Radic First c Second Rape Conspant?’f Concentration, TM, half-life, sec half—life, sec M — sec 292 --— 20 i 2 0.17 --- 908 19 1‘ 2 AA2 -—— 21 i 2 0.11 -—— 1,360 20 + 2 2A1 ——— 22 i 2 O. 19 --- 706 23 i 2 a 25% (by weight) methanol—water, 0.10 M total citrate buffer, pH A.8. Determined by measuring decrease in limiting current of first reduc- tion wave. c Time at which one—half of radical has reacted. d Time at which three-quarters of radical has reacted. e 5-’ l(31/2 xB—initial Rate constants given to within about i 10% since initial radical con— centrations are within about i 10% of stated values. 106 H ESR Spectroscopy_ On the basis of the generally accepted structures of the sulfone- phthaleins and sulfonephthalins (105), the reactant of Reaction 3a (I) and the product of Reaction 3b (IV) should both be diamagnetic, whereas the one-electron reduction product of Reaction 3a (II) should be para— magnetic. Esr measurements were used to confirm these assumptions. Since facilities for rapid mixing and observation by esr were not avail- able, cresol purple was used exclusively for these esr measurements be- cause its rate of disproportionation is the slowest of all compounds studied. Solutions of cresol purple gave no detectable esr signal, nor did solutions of this compound which were reduced over a mercury pool at potentials on the limiting current region of the second reduction wave. Moreover, cresol purple solutions reduced with amalgamated zinc gave no detectable esr signal. 0n the other hand, solutions reduced electrolytically at potentials on the limiting current of the first re- duction wave, as well as solutions reduced with either V(II) or Ti(III), all gave identical esr spectra. A representative spectrum is shown in Figure 15. These results prove that both chemical and electrochemical reduction, at potentials on the limiting current region of the first reduction wave, lead to exactly the same paramagnetic species, presum- ably II, the product of Reaction 3a. As shown in Figure 15, the esr spectrum of cresol purple radical consists of a single resolvable line about 22 gauss wide, with a g value of 2.00, very close to the free electron value. Only partially resolved hyperfine splitting is evident, and no attempt was made to assign the spectrum quantitatively. Nevertheless, the results are consistent with structure II, with the free electron localized on the 107 .Hamo mflmhaonpooam 8090 :ofipSHom mo Hm>osmp ampgm .nn :\m mammeHxOpmmm Umcfimppo Ezspoomm p .oon NM > 8.0- 3. m .Nm0 m;.m "hosmdvmnm m>mBosoHE .0 0mm ”mhdpmsmafime .me 00H "honodvosm soflpmadwoz .sfla 0H "meme snow .0 m.apm.m “pom UHmfim .>E om\m0 0H "sm3om o>m30soflz .0 HIGH x 0.: ”mwdpwams< coflpmfldvoz .omm om.o "pampmcoo osfle .0 on umwcmm smom mmcoflpwdcoo mcflsoaaom Moons Empmmm mam :Im cmwam> so Uoanmsmm pumaflnmmxm p.mo>m3 nofiposwms pmmflm no mwmhaospomam Hmfipsopom pmaaospcoo wsflzoaaom mamhdm HOmmho mo adspommm Hmm .ma madman 108 ‘3‘ 109 central carbon atom and undergoing only long range interactions with ring and methyl protons. This picture is consistent with the fact that none of the compounds containing the added electron (II, III, and IV) absorbs in the visible region of the spectrum (Amax for mycresol- sulfonephthalin, IV, is 275 nm), suggesting that the added electron is not highly delocalized on the radical, II. In contrast, the free electron on triphenylmethyl radical is highly delocalized into the rings (as shown, for example, by its well known esr spectrum), and this species absorbs strongly in the visible region. However, the fourth electron of the central carbon atom in triphenylmethane is localized on the central carbon atom, in the carbon-hydrogen bond, and this species does not absorb in the visible region. For those compounds where disproportionation is slow, esr pro- vides a convenient means of monitoring the concentration of free radi— cal with time. For the radical generated by the reduction of cresol purple, the esr signal was observed to decay with time in a second order manner which is consistent with the disproportionation mechanism. I SpectrOphotometric Measurements Sulfonephthalein indicators absorb strongly in the visible region (Amax for the acid form of phenol red is A30 nm), whereas the one and two—electron reduction products of these compounds are colorless, ab— sorbing in the uv region (Amax for phenolsulfonephthalin is 265 nm). Thus, absorbance at A30 nm can be used to determine the concentration of phenol red. for example, if a A x 10"5 M solution of phenol red is reduced to 2 x 10-5 M5 then from the stoichiometry for complete dispro— portionation, the equilibrium absorbance at A30 nm will correspond to 5 3 x 10- .M phenol red. Hence, the disproportionation of the colorless 110 radical can be studied by monitoring the reappearance of the colored starting material. Of course, if experiments of this type are per- formed with rapid mixing, they provide a means of measuring the rate of disproportionation, and provide a completely independent confirma— tion of electrochemically measured rate constants. Under conditions of electrochemical measurements the half-life of the free radical of phenol red is less than a few seconds, and therefore spectrophotometric experiments were initially performed with a stop-flow apparatus (81). In a typical experiment, a solution of phenol red in the same electrolyte used for electrochemical ex- periments was contained in one syringe, and a solution containing approximately half an equivalent of V(II) was present in the other syringe. Mixing into a 2 cm cell was complete in about 1 msec, and absorbance through the cell was recorded as a function of time. To have an absorbance of less than about one with this system, however, required the use of solutions of phenol red that were relatively di- lute compared with those studied electrochemically. Phenol red con- centrations in electrochemical experiments ranged from about 5 x 10- M_to 5 x 10-3 M5 whereas in these spectrophotometric experiments the 5 concentration averaged about 5 x 10- M} Since the disproportionation is second order, this necessity of using lower concentrations resulted in a corresponding increase in the half-life of the chemical reaction. Thus, the use of a stop-flow apparatus was actually unnecessary since identical results could be obtained by simply mixing solutions in the cell compartment of a conventional spectrOphotometer. Results pre- sented below are a composite of both kinds of experiments. A typical % Transmittance y§_time curve for reduction of phenol 111 red with V(II) and subsequent disproportionation generating phenol red is shown in Figure 16. Clearly the reduction by V(II) is not instan- taneous, and indeed could be studied conveniently by stop-flow, al- though this possibility was not pursued. Percent Transmittance XE time data were converted to concentration y§_time data using the measured molar extinction coefficient for the acid form of phenol red )4 M—l—Cm-l) . (1.81 x 10 Concentrations of phenol red were then con— verted to concentrations of radical, II, by assuming the stoichiometry of a disproportionation reaction. Likewise, initial concentrations of the radical, Bin’ were calculated from twice the long—time limiting absorbance after subtracting that absorbance due to the amount of phenol red which was not initially reduced. The fact that these as— sumptions resulted in self—consistent rate constants under a variety of conditions is indirect proof of the postulated mechanism. A conventional method of determining second—order rate constants is to plot the reciprocal of concentration y§_time. If a plot of this nature results in a straight line, the reaction is shown to be second order (107). The disproportionation reaction (Reaction 3c) can be re- written as 2 R + O + P (h) where §_is the radical, Q_the sulfonephthalein, and P_the sulfone— phthalin. This reaction can be described by the equations -d[R]/dt = Idng (5) dIOI/dt = kIRJ2/2 (6) d[P]/dt = k[R]2/2 (7) and, from the stoichiometry for complete disproportionation [R] = IRIin — 2[0] (8) 112 Figure 16. A typical experimental curve for reduction of phenol red with V(II), and the following generation of phenol red by disproportionation of the radical. 113 mocoEEmcEH o\o O O O 3 4 5 fl — A 0 I20 60 seconds 11h Equation 6 can be rearranged as follows [0] t 2JF?£%%—» = k dt (9) o 0 However, [R] is not directly determinable in the case of phenol red, so that Equations 8 and 9 must be combined to give [0] d[o] 2 = kt (10) 2 Integration of Equation 10 gives kt directly in terms of [O] and [R]in’ the experimentally determinable quantities. 1 _ 1 = kt (ll) {[R]in - 2[o]} [R]in Hence, if the concentration vs time data are plotted according to Equa— tion ll, the slope of this plot gives directly the second order rate constant for disproportionation. Experiments were conducted with nine different solutions, with 5 initial radical concentrations ranging from 3.7 to 7.7 x 10. g, A representative experiment, with the data plotted in accordance with IEqustion ll, is shown in Figure 17. Results of these nine experiments are summarized in Table VII, which lists ths measured rate constants along with the initial radical concentrations for each experiment. During the course of this investigation rate constants for dis- Proportionation of phenol red in acetate buffered aqueous solutions were measured electrochemically by cyclic voltammetry with various amounts of gelatin present. The average value of all of these meas— urements is 3.h x 102 Mil-sec-l, which is (fortuitously) identical with the average value of the spectrophotometrically determined rate p- v Figure 17. 115 A typical second order plot of phenol red radical vs time. 116 20 30 40 IO seconds 117 Table VII. Disproportionation Rate Constants for Phenol Red Radicals in Acetate Buffered Aqueous Solutions from Conventional Second Order Plots. Initial Radical R. x 105, M -1n - Concentration, Rate Constant? §_x 10-2, Mil — sec—l 7.7 3.2 6.14 3.3 6.2 3.h 6.0 3.h 6.0 3.14 5.9 3.6 5.0 3.3 h.9 3.5 3.7 3.5 2 a Average value 3.6 %. is 3.h x 10 gfl - see—1 with a standard deviation of 118 constant in Table VII. This excellent agreement is taken as further evidence of the correctness of the disproportionation mechanism, as well as convincing evidence of the validity of measuring homogenous rate constants with modern electrochemical techniques. Based on the above results it is apparent that phenol red pro- vided a convenient system for comparing electrochemically measured rate constants with those obtained by more classical means. Surpris— ingly, the literature contains virtually no other direct comparisons of this type, apparently because the time scale for electrochemical methods complements rather than overlaps classical approaches. Thus, previous comparisons involved an extrapolation of first order rate constants as a function of pH, dielectric constant, etc (8-12, 108, and references contained therein). The comparisons reported here for phenol red and cresol purple avoid this problem since the dispro- portionation reaction is second order and the time scale can be alter- ed by simply working at different concentrations. In general, this approach was not possible until very recently, since rigorous electro- chemical theory for higher than first order processes (50-52) is of very recent origin. J Electrochemical Measurement gthate Constants Clearly several techniques could be used to measure rate constants of compounds considered in this investigation. Nevertheless, cyclic voltammetry was found to be by far the most convenient and versatile technique. For example, a rate constant was obtained for cresol purple by cyclic voltammetry in spite of the fact that this compound exhibits two conventional polarographic reduction waves of equal height so that 119 a rate constant could not be obtained by conventional polarography (109). Furthermore, conventional polarography requires an analysis of both re- duction waves to determine the rate of the disproportionation reaction (109), whereas cyclic voltammetry requires data from only the first wave (52). This aspect of cyclic voltammetry is highly advantageous in that it permits measuring rate constants at low pH where the second re- duction wave is masked by hydrogen discharge. Thus, in addition to be- ing inherently more versatile and sensitive than conventional polaro— graphy, cyclic voltammetry is also applicable to a wider variety of sol- ution conditions. Briefly, the experimental procedure employed for the cyclic volt— ammetry of these acid—base indicators was as follows. Cyclic voltam— mograms were recorded for a series of scan rates, together with appro- priate scan-and—hold experiments to generate the proper baseline for measurement of the anodic peak current (78). Experimental values of ia/ic were then converted to values of EQCST with the aid of a large scale plot of the theoretical data of Olmstead and Nicholson (52). In each case spherical parameters were determined and used to identify the proper theoretical working curve. Although generally such spherical corrections are relatively small, they are significant for most of the scan rates required in this investigation. Experiments were performed with (EA - E1/2)E.Of approximately four, subsequently corrected with the empirical equation of Olmstead and Nicholson (52). From values of k2C*T, k2 was calculated for each scan rate employed. Tables VIII through XXXVIIlist these cyclic voltammetric data for each of these acid-base indicators. 120 Table VIII. Cyclic Voltammetric Data on Cresol Purpleé’b’c’d Scan Rate, Cathodic Peak ig/X}/2 Peak Current Rate Constante v, mV/sec Current, iC, Ratio, i /i M"1 — sec-l -p -a -c - nA. 2h.3 8.0 1.62 0.9h3 18.0 26.3 8.2 1.61 0.9h5 18.8 28.7 8.6 1.60 0.9h7 19.6 31.6 9.0 1.60 0.955 18.2 35.1 9.u 1.59 0.958 18.7 25% (by weight) methanol—water, 0.10 M_total citrate buffer, pH n.8, * CO = 5.92 mg, no gelatin or Triton X~lOO present. Diffusion coefficient, 20 = 2.1 x 10'6 cme/sec. Radius of drop, £_= O.hh mm. h Spherical Correction Factor, 0 = 80 x 10— . 52 = 18.7 i 1.7; 1.7 is three times the standard deviation of these five experiments. 121 Table IX. Cyclic Voltammetric Data on Cresol Purple£.a"b’c’d Scan Rate, Cathodic Peak ig/XI/Q Peak Current Rate Constante z, mV/sec Current, 1;, Ratio, _i_a/_i_c Mfl — sec—l 0A. 27.5 9.2 1.75 0.935 25.6 33.0 9.9 1.72 0.9% 26.2 are 38.9 10.6 1.70 0.9h9 27.8 uh.0 11.2 1.69 0.959 25.2 _ 50.8 12.1 1.69 0.963 25.7 E; 58.0 12.8 1.68 0.966 26.8 66.1 13.6 1.67 0.972 25.8 75.1 1h.5 1.67 0.976 25.6 a Aqueous solution, 0.20 M_acetic acid & 0.20 M_sodium acetate, pH 2 h.8, * CO = 5.01 EM» no gelatin or Triton X-100 present. Diffusion coefficient = 2.7 x 10.6 cm2/sec. ’20 Radius of drop, £_= O.Ah mm. h Spherical Correction Factor, 0 = 70 x 10‘ . .52 = 26.1 i 2.5; 2.5 is three times the standard deviation of these eight experiments. 122 Table X. Cyclic Voltammetric Data on Thymol Blue?"b’c"d Scan Rate, Cathodic Peak iC/X}/2 Peak Current Rate Constante v, mV/sec Current, ic, Ratio, 1 /i M.1 — sec-l _ _.p .3 ._c __ 0A. 22.6 7.6 1.61 0.910 32.2 2h.h 7.9 1.60 0.918 31.0 26.h 8.2 1.60 0.921 32.2 28.8 8.5 1.58 0.929 31.0 35.2 9.3 1.57 0.936 33.9 39.6 9.9 1.57 0.9104 32.6 145.3 10.14 1.55 0.9h7 311.6 a 25% (by weight) methanol-water, 0.10 M total citrate buffer, pH h.8, * CO = h.78 EM: no gelatin or Triton X-100 present. b Diffusion coefficient, 20 = 2.5 x 10_6 cm2/sec. c Radius of drop, 3.: 0.hh mm. d Spherical Correction Factor, 0 = 80 x lO—h. e '32 = 32.5 i 3.7; 3.7 is three times the standard deviation of these seven experiments. 123 Table XI. Cyclic Voltammetric Data on Thymol Blueg’b’c’d Scan Rate, Cathodic Peak i;[v}/2 Peak Current Rate Constante v, mV/sec Current, ic, Ratio, i /i M”1 - sec-l uA. 27.5 8.8 1.67 0.931 31.9 33.0 9.5 1.614 0.9142 31.14 38.9 10.1 1.63 0.9h7 33.3 hh.0 10.7 1.62 0.953 33.2 50.8 11.5 1.61 0.959 33.h 58.0 12.2 1.61 0.963 33.9 66.1 13.1 1.61 0.967 39.5 75.1 13.9 1.60 0.971 36.2 25% (by weight) methanol-water, 0.20 M_acetic acid & 0.20 M_sodium * acetate, pH = h.8, CO = h.99 mg, no gelatin or Triton X-100 present. . . . . _ -6 2 IDiffu51on coeff1c1ent, 20 - 2.6 x 10 cm /sec. Radius of drop, _r_‘_ = 0.1114 mm. £3pherical Correction Factor, 0 2 80 x lO-h. e I: = 33.5 i h.5; 9.5 is three times the standard deviation of these -%2 eight experiments. 1214 Table XII. Cyclic Voltammetric Data on Bromothymol Blue‘il’b’c’d c 1/2 Scan Rate, Cathodic Peak ip/X- Peak Current Rate Constante v, mV/sec Current, ic, Ratio, i /i M.1 - sec—l _. _p _a._c ._ 11A. 2h.h 3.9 0.789 0.919 66.h 26.h h.0 0.779 0.92h 65.9 28.8 h.15 0.773 0.928 67.6 31.7 b.35 0.772 0.93M 67.1 ig'fi 35.2 n.55 0.767 0.9u0 67.2 5: 39.6 9.75 0.755 0.987 65-7 .25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH n.8, *- C = 2.36 mM_, no gelatin or Triton X-100 present. 0 . . . . _ -6 2 Diffu31on coeff1c1ent, DO - 2.5 x 10 cm /sec. Radius of drop, 3 = 0.10; mm. d . . -h Spherical Correction Factor, 0 '—" 60 x 10 . 352 = 66.7 i 2.3; 2.3 is three times the standard deviation of these 8 ix experiments . 125 Table XIII. Cyclic Voltammetric Data on Bromocresol Green8’b’c’d Scan Rate, Cathodic Peak ig/g}/2 Peak Current Rate Constante v, mV/sec Current, iC, Ratio, i /i M.1 - sec—l _. -p -a _c _. 0A. 22.6 3.h5 0.726 0.920 68.8 25.h 3.6 0.71h 0.930 65.8 28.8 3.8 0.708 0.93h 69.3 31.7 3.95 0.702 0.937 71.7 35.2 1.15 0.699 0.9h6 68.0 39.6 h.h 0.699 0.950 70.8 h5.3 h.7 0.697 0.957 69.h 25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH h.8, * CO = 2.08 EM, no gelatin or Triton X-100 present. IDiffusion coefficient, 20 = 2.7 x 10-6 cmz/sec. IRadius of drop, r_= O.hh mm. ESpherical Correction Factor, 0 = 60 x lO_h. 1&2 = 69.1 f 5.3; 5.3 is three times the standard deviation of these seven experiments. 126 Table XIV. Cyclic Voltammetric Data on Phenol Redz?"b’c’d Scan Rate, Cathodic Peak ig/31/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M-1 - sec-l - -p —a -c pA. 61.1 5.35 0.68h 0.907 1h9 71.3 5.75 0.681 0.917 151 85.6 6.3 0.681 0.921 169 107 6.95 0.672 0.935 171 122 7.h 0.669 0.9h6 162 EL 25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH 2.5, * CO = 2.26 2%: no gelatin or Triton X-100 present. IDiffusion coefficient, DO = 2.2 x 10_6 cm2/sec. IRadius of drop, £_= 0.hh mm. d. . . -h Spherical Correction Factor, 0 = 25 x 10 . e 332 = 160 i 30; 30 is three times the standard deviation of these five experiments. 127 Table XV. Cyclic Voltammetric Data on Phenol RedE.3"b’c’d Scan Rate, Cathodic Peak ig/3}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M.1 - sec"l 0A. 61.1 5.9 0.690 0.907 199 71.3 5.8 0.687 0.919 158 85.6 6.2 0.670 0 919 176 107 6.9 0.667 0.935 172 122 7.35 0.665 0.996 163 a’.25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH 2.5, * CO = 2.25 mg, 0.05% gelatin added. lDiffusion Coefficient, 20 = 2.2 x 10"6 cm2/sec. IRadius of drop, £_= 0.99 mm. ESpherical Correction Factor, 0 = 25 x lO-h. 232 = 169 i 29; 29 is three times the standard deviation of these five experiments . 128 Table XVI. Cyclic Voltammetric Data on Phenol Redé’b’c’d Scan Rate, Cathodic Peak ig/v}/2 Peak Current Rate Constant? v, mV/sec Current, iC, Ratio, 1 /i M_1 - sec-l ._ _p _fl'_c ._ uA. 33.3 9.8 0.832 0.856 168 39.3 5.2 0.829 0.878 159 98.0 5.7 0.823 0.890 169 57.6 6.2 0.817 0.901 180 72.1 6.7 0.789 0.925 163 86.5 7.3 0.785 0.931 176 108 8.0 0.769 0.9u3 180 127 8.6 0.763 0.950 183 25% (by weight) methanol-water, 0.10 fl_total citrate buffer, pH 9.8, * CO = 2.32 mg, no gelatin or Triton X-100 present. IDiffusion coeficient, 90 = 2.2 x 10’6. Radius of drop, 3 = 0.99 mm. Eipherical Correction Factor, 0 = 25 x lO-M. 2%? = 172 i 26; 26 is three times the standard deviation of these eight experiments. 129 Table XVII. Cyclic Voltammetric Data on Phenol Redef’b’c’d Scan Rate, Cathodic Peak ip/X}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, 1 /i M—1 - sec--l _. -p -a —c ._ 0A. 53.9 5.3 0.722 0.906 196 66.3 5.8 0.712 0.913 166 78.9 6.3 0.712 0.928 156 95.8 6.9 0.705 0.992 150 123 7.8 0.703 0.999 170 a 25% (by weight) methanol-water, 0.10 E total citrate buffer, pH 9.8, * CO = 2.31 mg, 0.05% gelatin added. Diffusion coefficient, 20 = 2.2 x 10-6 cm2/sec. C Radius of drop, §_= 0.99 mm. h d Spherical Correction Factor, 0 = 25 x 10- . k2 = 158 i 30; 30 is three times the standard deviation of these five experiments. 130 Table XVIII. Cyclic Voltammetric Data on Phenol Red8’b’c’d Scan Rate, Cathodic Peak lg/31/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i Mml - sec-l _. -p .3 -c _. DA. 39.0 2.15 0.369 0.930 160 90.1 2.3 0.363 0.939 162 h8.6 2.5 0.359 0.999 163 56.7 2.7 0.359 0.959 171 68.1 2.95 0.357 0.963 169 a 25% (by weight) methanol-water, 0.10 M total citrate buffer, pH 6.8, * CO = 1.10 EM: no gelatin or Triton X—100 present. Diffusion coefficient, 20 = 2.9 x 10”6 cm2/sec. C Radius of drop, £_= 0.99 mm. h d Spherical Correction Factor, 0 = 50 x 10_ . e :2 experiments. = 169 i l2; 12 is three times the standard deviation of these five 131 Table XIX. Cyclic Voltammetric Data on Phenol Redé’b’c’d Scan Rate, Cathodic Peak ig/v}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M.1 - sec-l _. _? -a -c ._ DA. 39.0 7.25 1.16 0.809 288 57.2 8.5 1.13 0.850 270 69.2 9.3 1.12 0.860 292 85.8 10.2 1.10 0.882 290 107.3 11.3 1.09 0.909 288 * a Aqueous solution, 0.10 M_total citrate buffer, pH 9.8, CO = 2.97 EM: no gelatin or Triton X-100 present. b Diffusion coefficient, 20 = 9.6 x 10-6 cm2/sec. C Radius of drop, r_= 0.99 mm. d . . -9 Spherical Correction Factor, 0 = 25 x 10 . e ‘52 = 286 i l6; 16 is three times the standard deviation of these five experiments. 132 Table xx. Cyclic Voltammetric Data on Phenol liedEEL’b’C’d Scan Rate, Cathodic Peak ig/X1/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, 1 /i M-1 - sec.-l _. -p .3 _c ._ DA. 39.0 7.1 1.19 0.819 257 57.2 8.95 1.11 0 856 255 if 69.2 9.0 1.08 0.878 259 85.8 9.95 1.07 0.889 289 E3 107.3 11.0 1.06 0.906 280 0 * a Aqueous solution, 0.10 E total citrate buffer, pH 9.8, C = 5.96 EM» 0 0.05% gelatin added. b Diffusion Coefficient, 20 = 9.0 x 10_6 cme/sec. C Radius of drop, £'= 0.99 mm. 9 d Spherical Correction Factor, 0 = 25 x 10- . e k2 = 266 i 99; 99 is three times the standard deviation of these five experiments. 133 Table XXI. Cyclic Voltammetric Data on Phenol Redé’b’c’d Scan Rate, Cathodic Peak lg/31/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M-1 — sec-l ._ ‘P -a -c .. DA. 27.5 5.7 1.09 0.772 329 33.0 6.2 1.08 0.795 323 60.0 7.5 0.97 0.853 330 73.5 8.3 0.96 0.880 323 99.9 9.05 0.93 0.895 393 118 10.1 0.93 0.916 326 199 11.1 0.92 0.923 353 ll Aqueous solution, 0.20 M_acetic acid & 0.20 M_sodium acetate, pH 9.8, * CO = 2.29 mg, no gelatin or Triton X-100 present. Diffusion coefficient, 20 = 5.0 x 10-6 cm2/sec. Radius of drop, r'= 0.99 mm. 9 Spherical Correction Factor, 0 = 30 x 10_ . k2 = 332 t 35; 35 is three times the standard deviation of these seven experiments. 139 Table XXII. Cyclic Voltammetric Data on Phenol Redg’b’c’d Scan Rate, Cathodic Peak ig/x}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, 1 /i Mml - sec-l _. _p .a _c _. pA. 37.5 5.2 0.850 0.813 395 96.9 5.7 0.837 0.839 392 57.7 6.25 0.822 0.856 355 75.0 7.0 0.807 0.879 365 93.8 7.65 0.790 0.899 369 125 8.6 0.767 0.920 366 Aqueous solution, 0.20 M_acetic acid & 0.20 M sodium acetate, pH = 9.8, * CO = 1.96 mg, no gelatin or Triton X-100 present. Diffusion coefficient, D = 5.0 x 10.6 cm2/sec. Radius of drop, £_= 0.99 mm. 5 Spherical Correction Factor, 0 = 30 x 10- . k2 = 356 i 32; 32 is three times the standard deviation of these six experiments. 135 Table XXIII. Cyclic Voltammetric Data on Phenol Red£.3"b’c’d Scan Rate, Cathodic Peak 16/21/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M.1 - sec-l _. -p -a .C _. DA. 27.5 5.8 1.11 0.768 393 33.0 6.35 1.10 0.795 323 38.9 6.6 1.06 0.810 336 97.2 7.1 1.03 0.829 359 60.0 7.75 1.00 0.858 328 73.9 8.6 1.00 0.876 335 a Aqueous solution, 0.20 M_acetic acid & 0.20 M_sodium acetate, pH = 9.8, {- CO = 2.22 mg, 0.017% gelatin added. Diffusion coefficient, D = 9.9 x 10.6 cm2/sec. C Radius of drop, £_= 0.99 mm. d Spherical Correction Factor, 0 = 30 x lO-h. e 32 = 337 i 33; 33 is three times the standard deviation of these six experiments. 136 Table XXIV. Cyclic Voltammetric Data on Phenol Red Scan Rate, Cathodic Peak ig/X}/2 Peak Current Rate Constant? v, mV/sec Current, 10, Ratio, i /i - sec-l - -p -a -c DA. 97.2 6.8 0.99 0.831 395 55.0 7.3 0.98 0.899 339 66.1 7.95 0.97 0.868 336 132 10.7 0.93 0.925 335 199 11.1 0.925 0.930 339 165 11.8 0.92 0.990 325 a Aqueous solution, 0.20 M acetic acid & 0.20 M_sodium acetate, pH 2 9.8, it CO = 2.20 2!: 0.099% gelatin added. Diffusion coefficient, D = 9.9 x 10-6 cm2/sec. C Radius of drop, 3 = 0.99 mm. 9 Spherical Correction Factor, 0 = 30 x 10- . e k2 = 336 i 20; 20 is three times the standard deviation of these six experiments. 137 Table XXV. Cyclic Voltammetric Data on Phenol Red£.a"b’c’d Scan Rate, Cathodic Peak ig/21/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M.1 — sec-l _. _? .3 .C ._ uA. 39.5 9.9 1.58 0.759 370 98.0 10.7 1.59 0.783 351 56.0 11.9 1.52 0.800 355 67.2 12.9 1.51 0.821 356 89.0 13.3 1.95 0.851 397 129 15.7 1.38 0.890 398 153 17.0 1.37 0.899 368 198 18.6 1.32 0.923 353 Aqueous Solution, 0.10 g acetic acid & 0.10 M_sodium acetate, 1.0 M KCl, * pH = 9.8, CO = 3.09 EH: 0.020% gelatin added. Diffusion coefficient, D = 9.5 x 10.6 cm2/sec. Radius of drop, §_= 0.99 mm. 9 Spherical Correction Factor, 0 = 25 x 10- . 52 = 356 i 26; 26 is three times the standard deviation of these eight experiments. 138 Table XXVI. Cyclic Voltammetric Data on Phenol Red£.9"b’c’d Scan Rate, Cathodic Peak 13/33/2 Peak Current Rate Constant? x, mV/sec Current, 1;, Ratio, ia/ic Mil - sec-l DA. 33.6 8.95 1.59 0.737 365 39.5 9.65 1.53 0.762 351 98.0 10.9 1.50 0.782 359 56.0 10.9 1.96 0.802 353 .F 67.2 11.7 1.93 0.829 338 g. 129 15.6 1.38 0.888 360 153 16.3 1.32 0.908 336 177 17.3 1.30 0.920 339 198 18.2 1.29 0 925 395 a Aqueous solution, 0.10 M acetic acid 8: 0.10 M sodium acetate, 1.0 M KCl, 9(- pH = 9.8, CO = 3.00 mg, 0.032% gelatin added. 6 b Diffusion coefficient D = 9.1 x 10- cm2/sec. ’ ‘0 c Radius of drop, 3.: 0.99 mm. 9 Spherical Correction Factor, 0 = 25 x 10_ . e 52 = 399 i 39; 39 is three times the standard deviation of these nine experiments. 139 Table XXVII. Cyclic Voltammetric Data on Phenol RedajL’b’c’C1 Scan Rate, Cathodic Peak ig/31/2 Peak Current Rate Constant? y, mV/sec Current, 1;, Ratio, ia/ic Mfl — sec'l uA. 33.6 8.5 1.97 0.797 392 98.0 9.7 1.90 0.799 327 56.0 10.5 1.90 0.819 325 89.0 12.3 1.39 0.896 367 112 19.0 1.32 0.875 367 129 19.7 1.30 0.898 329 153 16.0 1.29 0.906 350 177 17.0 1.28 0.912 371 198 17.9 1.27 0.923 362 a Aqueous solution, 0.10 M acetic acid & 0.10 fl_sodium acetate, 1.0 M KCl, * pH = 9.8, CO = 2.96 mg, 0.95% gelatin added. 6 Diffusion coefficient D = 3.9 x 10' cm2/sec. ’ ‘0 C Radius of drop, 3.: 0.99 mm. d'Spherical Correction Factor, 0 = 25 x lO-h. e 122 experiments. = 399 i 56; 56 is three times the standard deviation for these nine 190 Table XXVIII.CyClic Voltammetric Data on Cresol Redz.3"b’c’d f Scan Rate, Cathodic Peak ig/11/2 Peak Current Rate Constant, v, mV/sec Current, ic, Ratio, 1 /i M”1 — sec-l .. -p —a -c .. DA. 50.9 2.35 0.330 0.890 682 95 . t; 60. 0 2. 5 0. 322 O . 860 671 “.17" 72.1 2.7 0.318 0.881 652 ‘ 86.5 2.9 0.312 0.897 663 ? ,~- 103 3.1 0.306 0.912 659 Ej 123 3.3 0.299 0.925 653 50.9 2.25 0.316 0.899 673 61.8 2.95 0.312 0.867 655 78.6 2.75 0.310 0.890 658 £1 The first six experiments in this table were conducted in the presence of 0.020% gelatin, whereas the last three experiments were performed in the presence of 0.090% gelatin. 25% (by weight) methanol-water, 0.10 M total Citrate buffer, pH 9.8, a CO = 0.90 EM (0.020% gelatin solution) and 0.89 mfi_(0.090% gelatin solution). Diffusion coefficient, 20 = 2.7 x 10—6 cm2/sec in both solutions. Radius of drOp, £_= 0.99 mm. Spherical Correction Factor, 0 = 16 x lO-h. };2 = 662 i 33; 33 is three times the standard deviation of these nine experiments . Table XXIX. Cyclic Voltammetric Data on Cresol Red. 191 a,b,c,d Scan Rate, Cathodic Peak ig/X}/2 v, mV/sec Current, ic, - ‘P DA. 101 13.5 1.39 135 15.2 1.31 162 16.0 1.27 189 16.9 1.29 213 17.8 1.22 238 18.6 1.20 270 19.5 1.19 Peak Current Rate Constant? Ratio, ia/ic yfl - sec—1 0.657 2.11 x 103 E1 0.689 2.16 x 103 7a. 0 703 2.31 x 103 0.726 2.16 x 103 ,.. 0.732 2.37 x 103 :1 0.752 2.29 x 103 0.763 2.38 x 103 a Aqueous solution, 0 * 1.0 1!; K01, CO = 2.7 Diffusion coefficient, D C Radius of drop, 5 = dSpherical Correctio e = + ‘52 (2.23 - 0.33) viation of these se .10 M_acetic acid & 0.10 M_sodium acetate, pH 0 mg, 0.013% gelatin added. -0 = 9.5 x 10 0.99 mm. cm2/sec. n Factor, 0 = 10 x lO-h. x 103; 0.33 x 103 ven experiments. is three times the standard de- 192 Table XXX. Cyclic Voltammetric Data on Cresol Red£.1’b’c’d Scan Rate, Cathodic Peak ig/3}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M-1 - sec-l _. .p ._a _C _. 0A. 101 13.9 1.33 0.695 2.39 x 103 H1 135 19.6 1.26 0.689 2.28 x 103 '"sr 162 15.8 1.29 0.701 2.33 x 103 189 16.5 1.22 0.728 2.13 x 103 i a. 213 17.5 1.20 0.792 2.18 x 103 a} 238 18.6 1.20 0.752 2.29 x 103 9.8, a Aqueous solution, 0.10 M acetic acid & 0.10 M sodium acetate, * 1.0 M_KC1, CO = 2.67 g3, 0.26% gelatin added. *6 :1: I2 Diffusion coefficient, 20 = 9.0 x 10.6 cm2/sec. c Radius of drop, £_= 0.99 mm. d Spherical Correction Factor, 0 = 10 x lO—h. e k2 = (2.27 I 0.17) x 103; 0.17 x 103 is three times the standard de- viation of these six experiments. 193 Table XXXI. Cyclic Voltammetric Data on Cresol Red?"b’c’d Scan Rate, Cathodic Peak ig/g}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M"1 - sec“l _. _p _a _c _. HA. 101 13.0 1.29 0.659 2.29 x 103 135 19.2 1.22 0.690 2.29 x 103 162 15.5 1.22 0.710 2.26 x 103 189 16.2 1.19 0.730 2.19 x 103 it“ £15 213 17.0 1.17 0.750 2.13 x 103 - 238 17.7 1.15 0.760 2.19 x 103 270 18.5 1.12 0.776 2.20 x 103 a Aqueous solution, 0.10 M_acetic acid & 0.10 M sodium acetate, pH = 9.8, * 1.0 .124. KCl, CO = 2.60 E13, 0.051% gelatin added. Diffusion coefficient, D = 3.9 x 10_6 cm2/sec. C Radius of dr0p, g = 0.99 mm. 9 Spherical Correction Factor, 0 = 10 x 10- . 52 = (2.22 i 0.26) x 103; 0.26 x 103 is three times the standard de- viation of these seven experiments. e 199 Table XXXII. Cyclic Voltammetric Data on Cresol Red?"b’c’d Scan Rate, Cathodic Peak lg/38/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M“1 - sec-l _. _? .3 .1 _. DA. 101 5.05 0.50 0.767 2.23 x 103 135 5.65 0.995 0.805 2.17 x 103 162 6.2 0.99 0.823 2.27 x 103 189 6.6 0.985 0.839 2.39 x 103 213 7.0 0.98 0.857 2.18 x 103 238 7.9 0.98 0.865 2.27 x 103 270 7.95 0.98 0.880 2.20 x 103 a Aqueous solution, 0.10 M acetic acid & 0.10 M_sodium acetate, pH = 9.8, *- 1.0 g K01, CO = 1.09 EM: 0.051% gelatin added. b Diffusion coefficient, 20 = 3.9 x 10—6 cm2/sec. C Radius of drOp, r_= 0.99 mm. 9 Spherical Correction Factor, 0 = 10 x 10- . ‘52 = (2.29 i 0.18) x 103; 0.18 x 103 is three times the standard de— viation of these seven experiments. e 195 Table XXXIII. Cyclic Voltammetric Data on Bromophenol Redé’b’c’d Scan Rate, Cathodic Peak ié/y}/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, i /i M-1 - sec-l _. -p -a -c .. DA. 50.1 7.9 1.05 0.705 9.7 x 102 60.8 8.0 1.03 0.725 9.9 x 102 73.9 8.6 1.00 0.799 10.3 x 102 88.7 9.35 0.99 0.770 10.0 x 102 106 10.2 0.99 0.789 10.7 x 102 a 25% (by weight) methanol—water, 0.10 M_tota1 citrate buffer, pH 9.8, * CO = 1.98 EM: 0.010% gelatin added. b Diffusion coefficient, D = 9.7 x 10-6 cm2/sec. c Radius of drop, 3.: 0.99 mm. d Spherical Correction Factor, 0 = 8 x 10-9. ‘52 = (1.01 i 0.10) x 103; 0.10 x 103 is three times the standard de- viation of these five experiments. e 196 Table XXXIV. Cyclic Voltammetric Data on Bromophenol Red‘i"’13’(:’d Scan Rate, Cathodic Peak ig/11/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, 1 /i M-1 - sec-l _. ‘p -a.-c .. uA. 50.1 7.2 1.02 0.701 9.9 x 102 60.8 7.85 1.01 0.726 9.9 x 102 1.. 73.9 8.5 0.99 0.753 9.6 x 102 88.7 9.3 0.99 0.769 10.2 x 102 ~' 106 10.1 0.98 0.788 10.5 x 102 &! a 25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH 9.8, * CO = l.96 mg, 0.020% gelatin added. Diffusion coefficient, 20 = 9.9 x 10—6 cmZ/sec. c Radius of drop, £_= 0.99 mm. d Spherical Correction Factor, 0 = 8 x 10_h. 3 3 e 0.10 x 10 e + - kg (1.00 _ 0.10) x 10 , viation of these five experiments. is three times the standard de- 197 Table XXXV. Cyclic Voltammetric Data on Bromophenol Redém’c’d Scan Rate, Cathodic Peak lg/X1/2 Peak Current Rate Constant? v, mV/sec Current, ic, Ratio, 1 /i M.1 — sec—l __ _p ..a._c ._ DA. 50.1 7.3 1.03 0.698 10.3 x 102 60.8 8.0 1.03 0.725 10.1 x 102 73.9 8.65 1.01 0.751 9.9 x 102 2 88.7 9.3 0.99 0.771 10.2 x 10 106 10.0 0.97 0.790 10.6 x 102 a 25% (by weight) methanol-water, 0.10 M_total citrate buffer, pH 9.8, * CO = 1.92 mg, 0.090% gelatin added. Diffusion coefficient, 20 = 3.9 x 10.6 cm2/sec. c Radius of drop,_£ = 0.99 mm. d Spherical Correction Factor, 0 = 8 x lO-h. 'E2 = (1.02 I .07) x 103; 0.07 x 103 is three times the standard de- e viation of these five experiments. 198 d Table 'XXXVI. Cyclic Voltammetric Data on Chlorophenol Redé’b’c’ Scan Rate, Cathodic Peak ig/33/2 Peak Current Rate Constant? v, mV/sec Current, 1C, Ratio, 1 /i M'-1 - sec-l _. _p .3..C ._ DA. 35.9 3.9 0.57 0.697 10.0 x 102 93.1 3.65 0.56 0.671 9.7 x 102 50.7 3.9 0.55 0.686 10.2 x 102 59.9 9.2 0.59 0.708 10.1 x 102 71.9 9.55 0.59 0.720 11.1 x 102 86.2 9.85 0.52 0.792 11.3 x 102 103 5.25 0.52 0 767 11.1 x 102 123 5.65 0.51 0 788 11.3 x 102 W: a 25% (by weight) methanol—water, 0.10 M total citrate buffer, pH 9.8, * CO = 1.96, no gelatin or Triton X-100 present. Diffusion coefficient, D = 1.9 x 10"6 cm2/sec. c Radius of drop, £_= 0.99 mm. d Spherical Correction Factor, 0 = 6 x 10_h. k? = (1.06 i 0.20) x 103; 0.20 x 103 is three times the standard de- viation of these eight experiments. e Table XXXVII. Cyclic Voltammetric Data on Bromocresol Purple. 1 99 a,b,c,d Scan Rate, V/sec X; 0-53 0.70 1.05 2.11 3.01 10.5 17.1 19.5 23.3 27.3 31.1 Cathodic Peak Current, ic, -p DA. 13. 15. 18. 25. 30. 56. 71. 71. 75. 80. 85. .c 1 /v -p .. 18. 18. 18. 17. 17. 17. 17. 16. 15. 15. 15. 1/2 Peak Current Ratio, ia/lc 0.63 0.69 0.67 0.69 0.72 Rate Constant?’f Mil - sec-l 2.0 x 10 5 2.0 x 105 2.1 x 105 1.9 x 105 5 a 25% (by weight) methanol-water, 0.10 M_tota1 Citrate buffer, pH 9.8, * C 0 = 0°77.EM: no gelatin or Triton X-100 present. Diffusion coefficient , .120 = 2.9 x 10-6 cm2/sec. c Radius of drop, £_= 0.99 mm. e = + k2 (2.0 - 0.2) x 10 '5. , 0.2 x 10 tion of these calculated rate constants. 5 No spherical correction factor used. is three times the standard devia- No rate constants obtainable below a scan rate of about 15 V/sec. 150 K Effects 22 Maximum Suppressors Although Senne and Marple did not comment specifically on the fact, all their experiments were performed in the presence of Triton X—100. Some of the polarograms recorded for this thesis investigation exhibit- ed maxima and phenomena indicative of adsorption and/or stirring. To some extent such phenomena appeared to depend on the molecular struc— ture of the compound investigated, for example sometimes being greater for phenol red than cresol purple. More surprisingly, such phenomena also depended on the purity of the compound under investigation. For example, Figure 18 shows a cyclic voltammogram for a stockroom sample of phenol red whereas Figure 19 shows a cyclic voltammogram for a puri- fied sample. Moreover, conventional polarograms also exhibited these dramatic differences. For example, Figure 20 shows a conventional polarogram for the same stockroom sample of phenol red whereas Figure 21 shows a conventional polarogram for a purified sample. These in— teresting results were not investigated further since it was found that reliable data could be obtained by working with pure compounds. Never- theless, it was observed that Triton X—100 and gelatin both qualitative- ly suppressed this anomolous behavior, and therefore these systems pre— sented the possibility of evaluating the effect of maximum suppressors on kinetic measurements. In an earlier investigation (108) on azobenzene it was found that gelatin suppressed adsorption without interferring with kinetic measure- ments. Since from a pragmatic viewpoint the use of a "gelatin" electrode is much simpler than for example using rigorous theory 31§_a_xi§_Wopshall and Shain (39-36), it seemed useful to test the generality of previous COnclusions obtained from azobenzene studies. Phenol red is an ideal 151 Figure 18. Cyclic voltammogram on unpurified sample of phenol red in the absence of gelatin. Aqueous solution C = 3.10 mM phenol red 0.20 M acetic acid 0.20 M_sodium acetate Scan rate = 67 mV/sec Rate constant unobtainable 153 omm H|.m NOH x :.m n pampmGOU mpmm oom\>8 m» u mums cmom mpwpmom abwoom Z om.o owow ofipmom E om.o one Hososa_mm sm.m n mo soHpSHom m5005d< .sflpnaom mo monompm map ca 609 Homosa mo mamaom oofiMflhdm no swnmosawpao> owaoho .mH enamea 159 Tl>e oo_lv_ 155 aa.a u o\asm\aa .80 am n pzmflon QEnHoo mm oeoeoos aaaaon_m om.o anon oaeoos .m_om.o .mm 06.0 u 06 * coflpdaom mdoosw< .cwpmamm %o monomnw may CH ooh Hocmnm mo mamadm oofimfladmcd so emnmonwaom choflpso>coo .om enamem 156 mom 2. > Illl'l'll'|'l|'|ll|-l|'|l||"|"'J 157 NH.H H w\Hpm\N8 .so em u Dawson assHoo mm mpmpoom afifioOm.m.om.o anon canton.m_om.o mm sm.m 9 Mo soap5H0m m5oo5d< .cflpoamm mo mocmmpm map cw ooh Honosm mo mamswm wofiwwaso so swamOMmHom quowpqo>coo .Hm banana 158 159 system because of the behavior cited above wherein measurements could be made in the absence of gelatin as well as on samples which provided useful polarograms only in the presence of gelatin. Moreover, for this system the correct rate constant was presumably known. Measurements were performed on both purified and unpurified sam- ples of phenol red with various concentrations of gelatin. In every case the systems behaved as would be expected for the disprOportiona— 88, tion reaction, with no other complications (presumably the value of ks would be altered, but for scan rates employed in this study electron - .. ‘— 1'- IL. 1“. transfer for the first wave remained Nernstian). In addition, rate ‘— constants measured from cyclic voltammograms in the manner described above were within experimental error of the value that was obtained by independent spectrophotometric measurements (3.9 x 10 regardless of gelatin concentration. Typical results from such experi— ments on phenol red, along with results from similar experiments on cresol red, are summarized in Table XXXVIII. Thus, at least for these systems, use of gelatin is an expedient and acceptable approach, pro- -vided, of course, that the interest is not in the adsorption phenomenon pg; 52: or in details of the heterogenous electron transfer reaction. The effect of Triton X-100 as a maximum suppressor for phenol red was also investigated since Senne and Marple employed Triton X-100 rather than gelatin. Qualitatively, the effects of Triton X-100 and gelatin were found to be the same. However, Triton X—100 is unaccept~ able for quantitative measurements. For, example, apparent rate con— stants measured with cyclic voltammetry in the presence of Triton X—100 depended on the concentration of Triton X—100 as well as on scan rate and phenol red concentration. Typical results are shown in Figure 22 160 Table XXXVIII. Effect of Gelatin on Disproportionation Rate Constants by Cyclic Voltammetry for Radicals of Cresol Red and Phenol Red in Aqueous Solution? Compound 0;, EM. Gelatin, % 5:: M- - sec- Cresol redC 2.79 0 unobtainable 2.70 0.013 2.23 x 103 2.67 0.026 2.27 x 103 2.60 0.051 2.22 x 103 1.09 0.051 2'29 x 103 Phenol redd 3.10 0 unobtainable 3.09 0.020 3.56 x 102 3.00 0.032 3.99 x 102 2.96 0.095 3.99 x 102 Phenol rede 2.29 0 3.9h x 102 2.22 0.017 3.37 x 102 2.20 0.099 3.36 x 102 Phenol redf 2.97 0 2.86 x 102 2.96 0.050 2.65 x 102 a Cyclic voltammetric data for these compounds are given in Tables VIII— XXXVII. b Rate constants to within i 10% of stated values. c Cresol red, unpurified, 1.0 y;KCI, 0.10 M acetic acid, 0.10 M_sodium acetate, pH = 9.8. d Phenol red, unpurified, 1.0 M_KCl, 0.10 M_acetic acid, 0.10 M sodium acetate, pH = 9.8. e Phenol red, purified, 0.20 M_acetic acid, 0.20 M_sodium acetate, es— sentially same solution conditions as in spectrophotometric study, see Table VII. f Phenol red, purified, 0.10 M total citrate buffer, pH = 9.8. 161 .soflpsaom mzooSdm thOflnpm QH mumpoom as .68 m 36 ass .396 03009.2 3.0 .80. .4 o; a .zspoaanpao> oeHoho hp eoadmmoa mm Hmowomn ooh Hocozm mo sofipmsowpaomonmmflo pom mpcmpmcoo opmm w II. o ssmo.mu o .x. 222.9" I C 'O Emerso‘ .I o .dH . U .2 9.0 so. .ooalx covflhe m:ao.o mo monomonm one ca coflpmapcoozoo ooh Hosonm and owns coon ape: p.wpnmpmaoo owns peopmmmw mo GOflpprw> .mm oasmflm 162 CON 0.... 00_ .omm\>E. 33 soon ON. Om 0.? a 163 where apparent rate constants are plotted XE scan rate for several bulk concentrations of phenol red. These data are reminiscent of data reported by Wopshall and Shain (36) for azobenzene. In fact, these authors suggested using plots like Figure 22 to obtain homogenous rate constants for systems showing weak adsorption; the value of k extrapo— lated to zero scan rate is ostensibly the correct k, Interestingly, the data of Figure 22 all extrapolate to a common rate constant close to the correct value for phenol red. This fact suggests that Triton X—100 suppresses adsorption enough to eliminate obvious anomalies, but that some residual adsorption still remains. Unfortunately, this ex— planation is not correct, since it was found that apparent rate con— stants measured in the presence of both gelatin and Triton X-100 sim— ultaneously are still a function of Triton X-100 concentration. 0b— viously, one has to be careful in using empirical approaches, such as "gelatin" electrodes, in electrochemical kinetics. L Considerations gf Kinetic Effects Because of solubility limitations, experiments were generally con- ducted in 25% (by weight) methanol—water solutions containing a total 0.10 M citrate buffer at a measured pH. For example, these conditions apply to the data in Table V. However, experiments were also conducted on a few of these compounds in strictly aqueous solutions. These experi- ments show that the dielectric constant of the medium has a significant effect on the rate of disproportionation, as can be seen by comparing the rate constants in Tables V and.XXXVIII. For example, with both solu- tions containing a total 0.10 M;citrate buffer at pH 9.8, the rate of disproportionation of phenol red radical is about 70% faster in the . q v. 1. r; .‘4 .r a .71. \r/ C c 07.1 N5 QU v. : . H . . a ._ C :1 e .0 6 CS e a 0 at n» «G l D» e m a.“ 9‘ n14 . 0 #9 U a.» P... (r. v . . n.“ h 9.1 the t e 8 .hi. C . an if... h l e . a Q; . e 1?. Q» C +0 AU p n H .¢ v 1. V 0 03 ab n film a hua Ty +0 3 .x h a t m S r. mi 5 . 169 strictly aqueous solution than in the 25% methanol-water mixture. The sulfonephthalein acid-base indicators can be viewed as sub- stituted triphenylmethyl compounds, with one of the phenyl moieties containing a sulfonate group, and the other two phenyl moieties con- taining a wide variety of possible substituents. Thus, explanations of differences in the relative rates of disproportionation of sulfone- phthalein radicals require a consideration of many factors. The influence of the sulfonate group on the relative rates of dis- prOportionation is probably minimal since all of these radicals have this group located at an identical position in the molecule. However, the sulfonate group may play an important role in the kinetics of all of these radicals; for example, its proximity to the central carbon atom where the free electron is presumably localized may result in a steric effect. Furthermore, the sulfonate group obviously carries a negative charge, and therefore as increase in rate with an increase in the dielectric constant of the medium would be expected for these re- actions(107). This dielectric effect is observed for these sulfone- phthalein radicals, and may be due to this reason. 0n the other hand, the observed dielectric effect also can be explained on another basis that is consistent with the proposed meshanism. Thus, Kirkwood (110, 111) has shown that if the activated complex is more polar than the reactants (as presumably would be the case if the products are ions), 'the rate of the reaction increases with the dielectric constant of the Inedium. This theory may be applicable to the disproportionation of ssulfonephthalein radicals since, except for the sulfonate group, the ireactants are radicals whereas the products are ions (Reaction 3c). UDhus, there are at least two possible explanations for the increased 165 disproportionation rates of these radicals with increased dielectric constant. The first four compounds in Table V all exhibit relatively low rates of disproportionation, and have a methyl group ortho to the cen- tral carbon atom. The other six compounds in Table V have a proton instead of a methyl group in this position, and show significantly higher rates of disproportionation. These relative rates suggest that the methyl group causes steric hindrance at the central carbon atom. This behavior appears reasonable since, with the free electron local- ized on the central carbon atom, this position of the molecule would appear to be a reactive site for at least one of the partners in Reac- tion 3c. Another effect of structure on the rate of disproportionation is that radicals having either alkyl or halo groups EEEE.t° the carbon atom that directly Joins the central carbon atom exhibit significantly higher rates than analogous compounds with protons at these positions. A possible reason for this behavior is that alkyl groups are electron— releasing, and as such, may promote alternative (to the central carbon atom) reactive sites on the radicals. 0n the other hand, halo groups, being electrophilic, may attract a portion of the charge of the free electron that is presumably localized on the central carbon atom. Hence, halo groups may partially delocalize this free electron into the rings, and therefore also promote alternative reactive sites on the radical. In other words, both electron—releasing and electron-donating groups may promote alternative reactive sites on these radicals. This behavior even holds for the four radicals that possess methyl groups which sterically hinder the reaction. I'\) 10. 11. 12. 13. 19. 15. 16. 17. 18. :19. 220. 221. 232. 166 LITERATURE CITED Heyrovsky, J., Chem.Listy, 16, 256 (1922). Heyrovsky, J., Phil. Mag., 95, 303 (1923). Heyrovsky, J., and Shikata, M., Rec. Trav. Chim., 99) 996 (1925). Bard, A. J., Ed., "Electroanalytical Chemistry," Vol. 1., Marcel Dekker, Inc., New York, N.Y., 1966. Bard, A. J., Ed., "Electroanalytical Chemistry," Vol. 2., Marcel Dekker, Inc., New York, N.Y., 1967. Bard, A. J., Ed., "Electroanalytical Chemistry," Vol. 3., Marcel Dekker, Inc., New York, N.Y., 1967. Bard, A. J., Ed., "Electroanalytical Chemistry," Vol. 9., Marcel Dekker, Inc., New York, N.Y., 1969. Nicholson, R. s., Anal. Chem., 9g, 130R (1970). Hume, D. N., Anal. Chem., 90, 179R (1968). Reinmuth, w. H., Anal. Chem., 99, 185R (1968). Hume, D. N., Anal. Chem., 38, 261R (1966). Reinmuth, w. H., Anal. Chem., 38, 270R (1966). Matheson, L. A., and Nichols, N., Trans. Electrochem. Soc., 13, 193 (1938)- Randles, J. E. B., Trans. Faraday Soc., 99) 327 (1998). Sevcik, A., Collect. Czech. Chem. Commun., 13) 399 (1998). Delahay, P. J., J. Amer. Chem. Soc., 75, 1190 (1953). Nicholson, M. M., J. Amer. Chem. Soc., 6, 2539 (1959). Frankenthal, R. P., and Shain, I., J. Amer. Chem. Soc., 18, 2969 (1956). Reinmuth, w. H., J. Amer. Chem. Soc., 19, 6358 (1957). DeMars, R. D., and Shain, I., J. Amer. Chem. Soc., 81, 2659 (1959)- Gokhshtein, A. Y., and Gokhshtein, Y. P., Dokl. Akad. Nauk SSSR, 131, 601 (1960). Reinmuth, w. H., Anal. Chem., 33, 1793 (1961). ,1 1...‘ ‘m 9.]- .v: o 23. 25. 26. 27. 28. 29. 30. 31. 32. 33. 39. 35. 36. 37. 38. 39. 90. 91. 92. 93. 99. 95. 96. 167 Reinmuth, w. H., Anal. Chem., 33, 185 (1961). Nicholson, R. 8., and Shain, 1., Anal. Chem., 36, 706 (1969). Olmstead, M. L., and Nicholson, R. 3., Anal. Chem., 38, 150 (1966). Matsuda, H., and Ayabe, Y., Z. Electrochem., 52, 999 (1955). Nicholson, R. 3., Anal. Chem., 37, 1351 (1965). Gokhshtein, Y. P., and Gokhshtein, A. Y., Dokl. Akad. Nauk SSSR, 128. 985 (1959). Gokhshtein, Y. P., and Gokhshtein, A. Y., in "Advances in "“ Polarography," Longmuir, I. 8., ed., Vol. II., Pergamon Press, New York, N.Y., 1960, p. 965. Polcyn, D. s., and Shain, 1., Anal. Chem., 38, 370 (1966). Polcyn, D. s., and Shain, I., Anal. Chem., 38, 376 (1966). Berzins, T., and Delahay, P., J. Amer. Chem. Soc., 15, 555 (1953). Nicholson, M. M., J. Amer. Chem. Soc., _9, 7 (1957). Wopschall, R. H., and Shain, 1., Anal. Chem., 39, 1519 (1967). Wopschall, R. H., and Shain, 1., Anal. Chem., 39, 1527 (1967). Wopschall, R. H., and Shain, I., Anal. Chem., 32, 1535 (1967). Hulbert, M. H., and Shain, 1., Anal. Chem., 92, 162 (1970). Saveant, J. M., and Vianello, E., in "Advances in Polarography," Longmuir, I. 8., ed., Vol.I., Pergamon Press, New York, N.Y., 1960, p. 367. Saveant, J. M., and Vianello, E., Electrochim. Acta, 10, 905 (1965). Saveant, J. M., and Vianello, E., Electrochim. Acta, 8, 905 (1963). Saveant, J. M., and Vianello, E., Compt. Rend., 226, 2597 (1963). Nicholson, R. S., and Shain, I., Anal. Chem., 31, 178 (1965). Nicholson, R. s., and Shain, I., Anal. Chem., 31, 190 (1965). Olmstead, M. L., and Nicholson, R. 8., J. Electroanal. Chem., 19, 133 (1967). Schuman, M. 8., Anal. Chem., 91, 192 (1969). Schuman, M. 3., Anal. Chem., 92, 521 (1970). 97. 98. 99. 50. 51. 52. 53. 59. 55. 56. 57. 58. 59. 60. 61. 62. 63. 69. 65. 66. 67. 168 Saveant, J. M., and Vianello, E., Electrochim. Acta, lg, 1595 (1967). Nicholson, R. 3., Anal. Chem., 31, 667 (1965). Kudirka, P. J., M. S. Thesis, Michigan State University, East Lansing, Michigan, 1968. Olmstead, M. L., Ph. D. Thesis, Michigan State University, East Lansing, Michigan, 1968. Olmstead, M. L., Hamilton, R. G., and Nicholson, R. 8., Anal. Chem., .91, 260 (1969). Olmstead, M. L., and Nicholson, R. 8., Anal. Chem., 9;, 862 (1969). Vlcek, A., in "Progress in Inorganic Chemistry," Zuman, P., ed., Vol. 5., Interscience, New York, N.Y., 1963, p. 211. Delahay, P., "New Instrumental Methods in Electrochemistry," Interscience, New York, N.Y., 1959. Schwarz, W. M., and Shain, I., J. Phys. Chem., 62, 30 (1965). Wawzonek, S., and Duty, R. C., J. Electrochem. Soc., 108, 1135 (1961). Duty, R. C., Ph. D. Thesis, State University of Iowa, Ames, Iowa, 1961. Wawzonek, S., and Wagenknecht, J. H., J. Electrochem. Soc., 110, 920 (1963). Senne, J. K., and Marple, L. w., Anal. Chem., 92, 1197 (1970). Von Stackelberg, M., and Strache, W., Z. Electrochem., 63, 118 (1999)- Kolthoff, I. M., Lee, T. S., Stocesova, D., and Parry, E. P., Anal. Chem., 22, 521 (1950) Ehlers, V. B., and Sease, J. W., Anal. Chem., 31, 16 (1959). Hush, N. S., and Oldham, K. B., J. Electroanal. Chem., 6, 39 (1963). Fukui, K., Morokuma, K., Kato, H., and Yonezawa, T., Bull. Chem. Soc. Japan, 22, 217 (1963). Wawzonek, 8., Duty, R. C., and Wagenknecht, J. H., J. Electrochem. Soc., 111, 79 (1969). Parkanyi, C., and Zahradnik, R., Coll. Czech. Chem. Comm., 39, 353 (1965). Hussey, W. W., and Diefenderfer, A. J., J. Amer. Chem. Soc., 69, 5359 (1967). 169 68. Fry, A. J., and Moore, R. H., J. Org. Chem., 33, 1283 (1968). 69. Sease, J. W., Burton, F. G., and Nickol, S. L., J. Amer. Chem. Soc., 0, 2595 (1968)- 70. Caldwell, R. A., and Hacobian, 8., Aust. J. Chem., 21, 1903 (1968). 71. Ulery, H. E., J. Electrochem. Soc., 116, 1201 (1969). 72. Caldwell, R. A., and Hacobian, s., Aust. J. Chem., 21, 1 (1968). 73. Matsui, Y., Kawakodo, R., and Dato, Y., Bull. Chem. Soc. Japan, 9;, 2913 (1968). 79. Lawless, J., and Hawley, M. D., J. Electroanal. Chem., 29, 365 (1969). 75. Butin, K. P., Stanko, V. I., Belokoneva, N. A., Anorova, G. A., Klimova, T. V., and Beletskaya, I. P., Dokl. Akad. Nauk SSSR, 188, 819 (1969). 76. Mairanovskii, S. G., Barashkova, N. V., and Vol'kenshtein, Y. B., Elektrokhimiya, ;, 60 (1965). 77. Brown, E. R., McCord, T. G., Smith, D. E., and DeFord, D. D., Anal. Chem., 66, 1119 (1966). 78. Underkofler, W. L., and Shain, I., Anal. Chem., 36, 1778 (1963). 79. Bard, A. J., and Solon, E., Anal. Chem., 69, 1181 (1962). 80. Ross, J. W., DeMars, R. D., and Shain, 1., Anal. Chem., 26, 1768 (1956). 81. Beckwith, P., and Crouch, S. R. (to be published). 82. O'Donnell, J. F., Ayres, J. T., and Mann, C. K., Anal. Chem., 31, 1161 (1965). 83. O'Donnell, J. F., Ph. D. Thesis, Florida State University, Tallahassee, Florida, 1966. 89. Orndorff, W. R., and Sherwood, F. W., J. Amer. Chem. Soc., 92, 986 (1923). 85. Lundquist, J. T., and Nicholson, R. S. (unpublished results). 86. Kirmse, W., Angew. Chem. Internat. Edit., 9, 1 (1965). 87. Nefedov, 0. M., and Manakov, M. N., Angew. Chem. Internat. Edit., _2, 1021 (1966). 88. Kbbrich, G., Angew. Chem. Internat. Edit., 6, 91 (1967). 89. 90. 91. 92. 93. 99. 95. 96. 97. 98. 100. 101. 102. 103. 109. 105. 106. 107. 108. 170 Bethell, D., in "Advances in Physical Organic Chemistry," Gold, V., ed., Vol. VII., Academic Press, New York, N.Y., 1969, PPo 153-209- Moss, R. A., Chem. Eng. News, June 16, 1969, pp. 60—68. Moss, R. A., Chem. Eng. News, June 30, 1969, pp. 50-58. Hine, J., and Dowell, A. M., J. Amer. Chem. Soc., 6, 2688 (1959). Hine, J., and Burske, N. W., J. Amer. Chem. Soc., 16, 3337 (1956). Hine, J., Burske, N. W., Hine, M., and Langford, P. B., J. Amer. Chem. Soc., 19, 1906 (1957). Hine, J., and Porter, J. J., J. Amer. Chem. Soc., 19, 5993 (1957). Hine, J., and Langford, P. B., J. Amer. Chem. Soc., 7 , 5997 Hine, J., and Ehrenson, S. J., J. Amer. Chem. Soc., 66, 829 (1958). Moss, R. A., and Mamantov, A., Tetrahedron Lett., 36, 3925 (1968). Nefedov, 0. M., and Shafran, R. N., Izv. AN SSSR, Ser. Khim., 538 (1965). Meites, L., "Polarographic Techniques," New York, N.Y., 1965. 2nd ed., Interscience, Gilchrist, T. L., and Rees, C. W., "Carbenes, Nitrenes, and Ar- ynes, " Appleton-Century-Crofts, New York, N.Y., 1969, p 98. Sternberg, H. W., Markby, R. E., Wender, I., and Mohilner, D. M., J. Electrochem. Soc., 113, 1060 (1966). Gilchrist, T. L., and Rees, C. W., "Carbenes, Nitrenes, and Ar- ynes," Appleton-Century-Crofts, New York, N.Y., 1969, p 109. Hoffmann, R. W., "Dehydrobenzene and Cycloalkynes," Academic Press, New York, N.Y., 1967, pp 99—95. Bates, R. G., "Determination of pH," J. Wiley & Sons, New York, N.Yo , 196h, pp 1.31-1th Kolthoff, I. M., and Lingane, J. J., "Polarography," Vol. 1., Interscience, New York, N.Y., 1952, pp 268—299. Benson, S. W., "The Foundations of Chemical Kinetics," McGraw- Hill, New York, N.Y., 1960. Lundquist, J. T., and Nicholson, R. S., J. Electroanal. Chem., 96, 995 (1968). 171 109. Orleman, E. F., and Kern, D. M. H., J. Amer. Chem. Soc., 12, 3058 (1953). 110. Kirkwood, J. G., J. Chem. Phys., 9, 351 (1939). 111. Frost, A. A., and Pearson, R. G., "Kinetics and Mechanism," 222. ed., J. Wiley & Sons, New York, N.Y., 1961, Chapter 7. "11191)919L911 If 31 0