A STOPPED-FLOW STU-DY OF THE PROTONATION 0F AROMATl-C RADICAL ANIONS Thesis for the Degree of Ph. 'D. MICHKEAN STATE UNIVERSITY E. RANDALL MINNICH 1970 l. 3 7232.3 R Y Michigan Sufi University ‘. THEng This is to certify that the thesis entitled A STOPPED-FLOW STUDY OF THE PROTONATION OF AROMATIC RADICAL ANIONS presented by E. Randall Minnich has been accepted towards fulfillment of the requirements for Ph .D . degree inflhfimim Date August 5, 1970 0-169 v ‘ BINDING BY HMS 8: SUN! ' 9’ WDK‘BIIIDERY 1 It. .3. L nnARY BIND! ‘5 mil-Fungi.“ III 'M" T 1 s I I .I i ABSTRACT A STOPPED-FLOW STUDY OF THE PROTONATION OF AROMATIC RADICAL ANIONS BY E. Randall Minnich A study of the reactions of the radical anions of anthracene and terphenyl with a series of proton donors is reported. Previous investigations (1-5) employed polar solventsanuiconcentrations of the proton donors which were high enough to affect the solvent composition significantly. In the present investigation, lower concentrations of proton donors and less polar solvents were employed. The reactions were followed spectrophotometrically by using the stopped— flow method. In tetrahydrofuran (THF), the reactions of potassium anthracenide with a series of alcohols and with water were second-order in the anion and about half-order in ROH. .The reaction rates were independent of the concentration of RO-. Similar results. were observed for the reactions in di- methoxyethane (DME), but the pseudo-second-order rate constants were lower (kTHF 3:_1 x 104M'1sec71; kDME321 1 x 103M‘1sec21). E. Randall Minnich The reactions of sodium anthracenide in THF appear to be similar to those of the potassium salt in the same sol- vent. However, in DME they are a factor of ten slower than those of potassium anthracenide. Moreover, the decay of sodium anthracenide is first—order in DME. The reaction of potassium terphenylide with ethanol in THF appears to be first-order in each reactant and exhibits a rate constant of about 1 x lOsM'isectl, These results are consistent with a mechanism in which both first- and second-order paths are available for the reaction between a radical anion and a proton donor: k M+Ar'°' + M+Ar7 «7%: (M+Ar—°-) 2 (:L) (M+Ar7)2 + ROH —kJ-> M+ArH' + Mac” + Ar (2) M+Ar7 + .ROH —k2-> ArHo + M+RO— (5) MfAr7 + ArH° -£3>- MfArH_ + Ar (4) M+ArH- + ROH —k—‘*> 1:.ng + M+RO-. (5) In the case of potassium anthracenide, the second-order process (Equations 1, 2, and 5) is thought to predominate in situations which favor the formation of contact ion-pairs. When such ion-pairing is not favored, the first-order process (Equations 5-5) is observed. E. Randall Minnich REFERENCES 1. S. Arai and L..M. Dorfman, J. Chem. Phys,, $1, 2190 (1964). 2. A. P. Krapcho and A. A. Bothner-By, J. Am. Chem. Soc., §§, 751 (1960). 5. K. Umemoto, Bull. Chem. Soc. Japan, $9, 1058 (1967). A STOPPED-FLOW STUDY OF THE PROTONATION OF AROMATIC RADICAL ANIONS BY E. Randall Minnich A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1970 G - (£95126: 1 [”536-7/ To Claudia ii ACKNOWLEDGEMENTS The author wishes to express his appreciation for the assistance he has received during the course of this work. Professor James L. Dye suggested the problem and provided valuable guidance during the study of it. Dr. Vincent A. Nicely contributed creative ideas, constructive criticism, and invaluable assistance in the data analysis. Dr. Earl M. Hansen and Marc G. DeBacker helped with both the acqui— sition and the analysis of the data. Financial assistance from the United States Atomic Energy Commission is gratefully acknowledged. iii TABLE OF CONTENTS I. INTRODUCTION . . . . . . . . . . . . II. III. HI STORI CAL C O O O O O O O O O O O O A. Introduction. . . . . . . . . . . B. General Properties of Aromatic Anions 1. Early Studies. . . . . . . . . 2. Electrochemical Studies. ... . 5. Studies of the Optical Spectra 4. E.S.R. Studies . . . . . . . . 5. Studies of Ion-pairing . . . . a. Evidence for a dissociative b. Evidence for a non-dissociative equi- librium . . . . . . . . . . c. Conclusions regarding ion-pairing Kinetic Studies . . . . . . . . . 1. An Early Mechanism . . . . . . 2. Kinetic Studies in Ammonia . . O O O O O O 0 equilibrium 5. Kinetic Studies by Pulse Radiolysis. 4. A Kinetic Study Using ESR and Polarography 5. Summary. . . . . . . . . . . . EXPERI MNTAL 0 O O O O O O O O O O O A. B. .1. Alkali Metals. . . . . . : (The Stopped-Flow-Experiment . . General Laboratory Techniques Purification Techniques . . 2. Anthracene and Terphenyl 5. Alcohols and.Water . . . . 4. Tetrahydrofuran and Dimethoxy 5. Helium . . . ... . . . . . . Preparations. .'. . . . . . 1. Potassium Hydroxide and Potas 2. ROH Solutions. . . . . . . . 5. Anion Solutions. . . . . . . 1. Introduction . . . . . . . . 2. The Mixing System. . . . . . 3. Data Acquisition . . . . . . 00.00030000000- iv oooogoos‘ooooo n e mEt O O O O O O O I hoxide Page comments“:- P I—‘ TABLE OF CONTENTS—-continued Page 4. Data Analysis. . . . . . . . . . . . . . . 57 a. Program PUNDAT. . . . . . . . . . . . . 57 b. Program KINET . . . . . . . . . . . . . 58 IV. RESULTS AND CONCLUSIONS. . . . . . . . . . . . . 62 A. A Survey of the Data. . . . . . . . . . . . . 62 1. Preliminary Experiments. . . . . . . . . 62 2. Principal Experiments. . . . . . . . . . . 65 5. Discussion of the Data . . . . . . . . . . 66 4. Summary. . . . . . . . . . . . . . . . . . 69 B. Results . . . . . . . . . . . . . . . . . . . 70 1. Dependence on the Anion Concentration. . . 70 2. Dependence on ROH. . . . . . . . . . . . . 74 5. Dependence on R0”. . . . . . . . . . . . . 80 4. Search for Intermediates . . . . . . . . . 84 5. Dependence on Solvent. . . . . . . . . . . 84 6. Dependence on the Cation ; . . . . . . . . 88 .a. The reactions of Nafln‘ in THF. . . . . 89 b. The reactions of Na+AnT in DME. . . . . 95 c. Conclusions . . . . . . . . . . . . . 98 7. Dependence on the Anion. . . . . . . . . . 98 C. A Discussion of Mechanisms. . . . . . . . . . 101 1. Inapplicable Mechanisms. . . . . . . . . . 101 2. A Suggested Mechanism. . . . . . . . 106 a. Formulation . . . . . . . . . . . . . . 106 b. Experimental basis. . . . . . . 108 c. Implications. . . . . . . . . . . . 112 D O ConCIUSions O O O O O O O O C O O O O O O O O '114 “FERENCES O . O O O O O 0 O O 0 O O O O O O O O O C O O 116 APPENDIX--Representative Sets of Kinetic Data . . . . 120 LIST OF TABLES TABLE 1. 2. 5. 10. 11A. 118. List of Experiments. . . . . . . . . . . . . . . Pseudo-second-order Rate Constants for the Reac- tion 2KfAfi7 + 2ROH._2§E,.AnH2 + An + ZKtOR'. . . Values of the Order, n, of ROH in the Reactions of KIAnT with ROH in THF . . . . . . . . . . . . Rate Constants Calculated from Equation 46 for the Reactions of Potassium Anthracenide with Various Proton Donors in THF . . . . . . . . . . Pseudo-second-order Rate Constants for the Reac- tions of Potassium Anthracenide with (a) Ethanol and (b) Water in DME . . . . . . . . . . . . . . ,Pseudo-second-order Rate Constants for the Reac- tion Na+AnT'+ EtOH in THF. . . . . . . . . . . . Rate Constants Which Result From Application of a Parallel First- and Second-order Rate Law (Equation 58) to the Reaction of Na+AnT with Water in THF . . . . . . . . . . . . . . . . . . Bimolecular Second-order Rate Constants for the Reaction of Sodium Anthracenide with Ethanol in DEC 0 O O O O O O O O O O O O O C O O O O O O O Pseudo-first—order Rate Constants for the Reac- tion of Sodium Anthracenide with Water in DME. . Bimolecular Second-order Rate Constants for the Reaction of Potassium Terphenylide with Ethanol in THF 0 O O O C O O O O O O O O O C O O O O C 0 .Rate Constants Which Result from Application of Equation 71 to the Protonations of KIAnT and Na'l'An'T in THF and in DME . . . . . . . . . . . . Estimates of Uncertainty for the Rate Constants in Table 11A . . . . . . . . . . . . . . . . . . vi Page 64 75 76 79 87 89 92 96 96 101 110 110a LIST OF FIGURES FIGURE 1. 2. 10. A syringe used in the stopped-flow system. . . Two vessels used in the preparation of RO—. Weighed samples of potassium were prepared in vessel (a). The bulblets of potassium were broken in the presence of ROH in vessel (b). . Vessels used for the preparation and storage of (a) ROH and (b) anion solutions . . . . . . Optical spectra of (a) a yellow-green solution of potassium terphenylide in THF at 25°C and (b and c) similar solutions after partial de- composition to a red color. “SpeCtra b and c were obtained by using RCA 6199 and 7102 multiplier phototubes, respectively, with the scanning monochromator . . . . . . . . . . . . E.S.R. spectrum of a red (partially decomposed) soéution of potassium p-terphenylide in THF at 25 C O O O C o C . ° 0 o O 0 . O O . . O C O . Stopped-flow system. . . . . . . . . . . . . . Block diagram of the data-acquisition system . Decay of the 714 nm. peak of potassium anthra— cenide after mixing with 0.00175 M ethanol in THF 0 O O O O O O O O O O C C O C 9 O O O O O 0 Block diagram of the apparatus used to trans- form analog data to digital form . . . . . . . The decay of sodium anthracenide upon mixing with 0.000498M ethanol in THF. .The problems of mechanical vibration and poor stopping are illustrated. The spectrum was monitored at 714 nm . . . . . . . . . . . . . . . . . . . . vii Page 51 57 59 42 44 46 49 55 54 67 LIST OF FIGURES--continued FIGURE 11. 12. 15. 14. 15. 16. 17. 18a. 18b. 19. The decay of sodium anthracenide upon mixing with 0.00282M ethanol in THF. This push is representative of the majority of the data. A segment of the 714 nm. peak was scanned. . . . Pseudo-first-order (a) and pseudo-second-order (b) kinetics applied to the reaction of potas- sium anthracenide with 0.225 M.water in THF. . . Pseudo-second-order plots of the reactions of potassium anthracenide with (a) 5.08 x 10'3 M ethanol and with (b) 5.40 x 10‘3 M t—butanol in THF . . . . . . . . . . . . . . . . . . . . . Determination of the [ROH] 100 ’24. Simultaneous fitting of Equation 71 to three data sets. ,The reactions of KIAnT'with (a) 0.008 M, (b) 0.02 M, and (c) 0.0054 M t-BuOH in THF are treated . . . . . . . . . . . 109 ix I. INTRODUCTION The study to be presented is an examination of certain reactions of ionic species known generically as the aromatic radical anions. The properties of these anions and the con- tributions of several investigators toward the understanding of these properties are discussed in some detail in the Historical section of this thesis. For purposes of orienta— tion, however, a very general discussion of the aromatic anions and our study of them is presented at this point. The reader is referred to the Historical section for more details and for references. Aromatic radical anions are highly conjugated organic molecules to which one or more extra electrons have been added. Three very common methods for forming these anions are: reduction yi§_reaction with alkali metals, whereby the outer "s" electron of the metal is transferred to the organic molecule, electrolytic reduction, and reduction by solvated electrons produced by ionizing radiation. Once formed, they are, themselves, strong reducing Species which react with air and various proton-donating compounds. If kept in aprotic solvents under either vacuum or an inert atmosphere, their strongly colored solutions can be kept for months. In addition to the property of deep coloration, solutions of the mononegative anions, with which we are chiefly concerned, are paramagnetic and conducting. In recent years, pulse radiolysis studies of solutions of various aromatic compounds in several different alcohols have shown that the anions thus formed react rapidly (pseudo- first-order rate constants of the order of 103 sec’l) with the alcoholic medium. The disappearance of the anionic species was shown to be first-order in the concentration of anions, but, of course, the order with respect to the alcohol could not be determined in the pure alcohol. The present study was initiated on the basis of these pulse radiolysis results. It was reasoned that, by using low alcohol concentrations (approximately 10"2 M) in an inert solvent, the reactions could be slowed sufficiently to be studied by the stopped-flow method. The anion was to be formed by reaction with an alkali metal; its reaction with alcohol was then to be studied spectrophotometrically. There were two main objectives. The first was to obtain a comparison between the results of two different methods applied to the same reaction. The second was to extend the kinetic data to lower and varied alcohol concentrations in order to aid in the establishment of a mechanism for the reaction. Initially, our studies were limited to the reaction of the mononegative anion of anthracene (formed by reaction with potassium) with ethanol or water in tetrahydrofuran (THF). This system was chosen because the physical proper— ties of potassium anthracenide in THF had been studied rather extensively by other workers previous to our study. It very quickly became apparent that the results could not be compared directly with those from pulse radiolysis experiments. The anion decay was not first order. ,The main objective then became the discovery of a mechanism which would explain this difference. The scope of the research was expanded to include the effects of changing the cation, anion, and solvent, as well as the alcohol. An internally consistent mechanism for the protonation reaction of potassium anthracenide has resulted from these studies. In the following pages, we will present the back- ground of ideas and the experimental results which led us to propose the mechanism. Finally, the implications of the mechanism will be discussed, and experiments which might be done to test its validity will be suggested. II. HISTORICAL A. Introduction The background information which forms the basis for an understanding of our experiments is divided into two sections. The first part concerns the production and properties of stable solutions of aromatic radical anions. Among these properties, of course, are the types of reactions which they undergo. These are discussed briefly, but the emphasis is placed on the basic nature of the anions under non-reactive conditions. In the second section, direct kinetic studies of the protonation reactions of the Species are examined. B. General Properties of Aromatic Anions 1. Early Studies In 1914, Schlenk and co-workers (1) first produced aromatic radical anions when they mixed sodium and anthracene in diethyl ether. Later, Scott §£_§l, (2) were able to extend the reaction to other metals and aromatic molecules by using other ethereal solvents. The latter investigators studied the reaction: Na + Naphthalene _dimethy1 ether:_ Na-Naphthalene. (1) ‘diethyl ether Dark green conducting solutions were produced in dimethyl ether. The reversibility and the importance of solvation effects were shown by the observation that the addition of diethyl ether restored the starting materials. .The stoichiometry of the reaction is given by Equation 1. Two classes of reactions were reported (2). A reversible addition was observed when mercury, oxygen, benzyl chloride, or sodium was added to a solution of naphthalene. The irre— versible reaction of naphthalene in the presence of an alkali metal with water, alcohols, and a range of other pro— ton donors was described by C10H8°Na2°C10H3 + ZRH —+C;QH10 + CgoHe + 2NaR, (2) where R is the conjugate base. The formula for the naphthalene—sodium complex given by Equation 2 was chosen to explain the stoichiometry of the overall reaction 2Na + choaa + 2RH ———> 2NaR + CloHlo + €3,ng (3) even though the solution exhibited high conductivity, which would indicate the presence of ionic Species. 2. Electrochemical Studies Another method of production of the aromatic anions is via electrochemical reduction of the parent molecules. Polarographic reductions indicated that two distinct one- election additions occurred (5,4,5), i.e., Ar + e' —->'Ar-°_, followed by Ar? + e_ ——> Ar=, (4) where Ar indicates an aromatic molecule. Later, potentio- metric titrations of various aromatic compounds were carried out by using the anion produced from biphenyl and sodium (Biph7) as a reducing agent (6). The results of these titra- tions were interpreted to show that both the mono- and the di-negative anions could be produced, but that the dianion was formed in two one—electron additions, Blph + Ar 77+~ Biph + Ar’ (5a) and —>- <+—- Biph" + Ar? Biph + Ar: . (5b) The difference in the reduction potentials of the hydroe carbons and the univalent anions agreed well with the polaro- graphic data. Since the polarographic studies were done in dioxane- water mixtures, the final products were the dihydro adducts of the parent molecules. In order to test the proposed mechanism of the reduction (4) Ar + e- -—-9" Ar? (6a) Ar7 + e_ --+- Ar= (6b) Ar: + 2H20 -—)-Ang + 20H" , (6c) Maccoll (5) plotted the observed reduction potentials of some aromatic hydrocarbons against the energies of the low— est unoccupied w-electron levels which had been obtained from zero-order molecular orbital calculations. A linear relationship was both predicted and observed. His results were verified by Hoijtink and Van Schooten (5) , who noted, however, that the data were also consistent with the mechanism Ar + e- 2 Ar? (7a) Ar? + H20 ——> ArH- + OH- (7b) ArH- + e- --+- ArH- (7c) ArH— + H20 —-—+ ArH2 + OH- . (7d) Polarographic experiments with varying dioxane-water mixtures and with varying concentrations of HI Showed that the applicability of Equations 6a-c or 7a-d was dependent upon a competition between steps 6b and 7b (7). High concen- trations of water or the addition of HI resulted in a fast protonation step, while for solutions in 96% dioxane-4%*water mixtures, the addition of a second electron was faster than the protonation. In the former case, one two-electron wave resulted; in the latter, two one-electron waves were seen. These data also indicate that the protonated radical (ArH-) has a higher electron affinity than does the parent molecule (Ar). Regardless of the reduction mechanism, however, the success of the molecular orbital calculations provided a good model for the electronic structure of the anions. 5. Studies of the Optical Spectra In a comprehensive study of the properties of the anions, Paul, Lipkin and Weissman (8) verified the previously published (2) stoichiometry of formation of the anions with metals and measured the conductivity of the solutions thus formed. They studied the optical spectra of a few compounds which had been reduced with sodium in 1,2-dimethoxyethane (DME) and in tetrahydrofuran (THF), noting strong absorptions in the ultraviolet, as well as in the visible region. The Beer-Lambert law was Shown to apply to sodium anthracenide (4.6 x 10"4 to 5.4 x 10’3 M) and sodium naphthalenide (1.3 x 10“4 to 8.67 x 10-3 m) in THF. Later studies involved both the mono- and di-anions of a large number of compounds which were formed with Li, Na, and K in both DME and THE (9,10). Extinction coefficients of the order of 104 M‘1 cm'."1 were generally found. The polarizationscmfthe electronic transitions were studied by Hoijtink gg_g;, (11,12). 4. EJS.R. Studies Since the aromatic mono-anions are odd-electron systems, they are, of course, paramagnetic. The dianions, on the other hand, have been found to be diamagnetic (15). Early electron Spin resonance studies of the mono-anions of naph- thalene (14), anthracene, and biphenyl (15) showed both paramagnetism and hyperfine splitting of the electron signal due to magnetic interactions between the electron and the protons on the aromatic molecule. The spectra correlate surprisingly well with the predictions of simple Hfickel molecular orbital calculations. Comparisons between the calculated and the observed Spectra, as well as the calcu- lated spin densities, have been reported by DeBoer and Weissman (16). Calculated Spin densities at the various carbon atoms of the anthracene and terphenyl anions are reproduced below: .096 .195 .008 .049 .062 l J L I I \ / \\> / \ / \ (8), .048—9 \ f \ 89 . / Z/’ \ / ——— .——- .0 .004 ._ .125 .055 .— Anthracene' Terphenyl’ AS would be expected, the electron density is distributed over the whole aromatic system. The species thus formed can be extremely stable and non-reactive towards the solvent. Some radical anion solutions last for months. Still, the electron is readily transferred to other acceptors, as was shown by Ward and Weissman (17). They added excess naphtha~ lene (Naph) to solutions containing naphthalenide anion (NaphT). A broadening of the e.s.r. spectrum was found to accompany addition of naphthalene. This behaviour was attrib— uted to the reaction Naph + NaphT' -r—+- NaphT. + Naph . (9) Depending on the cation and solvent used, second-order rate 10 constants ranging from 1 x 107 to 1 x 109 M“:L sect:L were measured. The cation dependence indicated some sort of cation participation in the structure of the free radical Species or in the electron-transfer process. Although Li, K, Rb, and CS plus naphthalene in either DME or THF gave the same spectra, the sodium salt in DME produced much wider lines (17). In THF, it displayed a completely different spectrum. Addition of KI to sodium naphthalenide in DME caused a narrowing of the Spectrum to that typical of the potassium salt. This suggested an equilibrium condition Ke _. q _. ' Na+Naph’ + KI 'T-**- KfNaph' + NaI, Keq 27 104 (10) in which the species Na+NaphT and K+NaphT were considered to be different enough to be spectroscopically distinguish- able. Interest in such Species became intense, as witnessed by the large number of investigators who are still trying to determine the structures of the ion-pairs (18). 5. Studies of Ion-pairing Early data indicated that, even in solutions containing only one cationic and one anionic species, at least two spectroscopically distinguishable species might exist. Atherton and Weissman (19) found that, in THF, sodium naph- thalenide (Na+Naph7) exhibited the same 25 line e.s.r. spectrum as was seen in DME, but with each of the lines further split into a quadruplet. This suggested the 11 influence of the 5/2 spin of the sodium nucleus. Cooling of the solution from 25°C to -70°C caused a diminishing of the sodium splitting and a simultaneous growth of the unsplit 25 line Spectrum which was superimposed on the split spectrum. An ion-pairing equilibrium was postulated Na+Napth --+- Na+ + NaphT. (11) At low concentrations (less than 10-3 M), the superposition of the two distinct spectra could be observed even at room temperature. This further supported the idea of a dissoci- ative equilibrium. .Atherton and Weissman (19) also considered the Structure of the ion-paired complex. The observation of splitting due to sodium implied a certain amount of spin density at the sodium nucleus. This requirement could be met if the sodium were directly above the center of either of the benzene rings at a distance of 2.5 R. Sufficient overlap of the sodium 58 orbital and the singly occupied v orbital of naph- thalene would then result. If, however, the equilibrium position of the sodium ion were to Shift with temperature toward a spot above the intersection of the naphthalene rings, the splitting would decrease. This initial description of ion-pairing was shown to be inadequate by later experiments. These studies also indi- cated that aromatic anions could exist in two or more forms in solution and that these forms were in equilibrium (24~28). 12 However, the nature of these forms was in Question. The results of the various studies generally fell into one of two distinct categories. One class of results could be explained by a dissociative equilibrium M+Ar7 —*= M+ + Ar? . (12) The second class was characterized by its explanation via a non-dissociative equilibrium between two kinds of ion pairs, (I‘ll-*-}\12‘—;)J~ -—"+ (M+AI‘T)2 . (15) Further data which indicate the presence of a dissociative equilibrium will be discussed before consideration of the nature of the ion pairs. a. Evidence for a dissociative equilibrium: A super- position of two distinct e.s.r. spectra was observed in solutions of sodium anthracenide in methyl-tetrahydrofuran (MTHF) at -950C (20). A change in the relative intensities of the two spectra with concentration of the anion was in- terpreted to indicate the occurrence of a dissociative process. Dodson and Reddoch (21) extended the study of the e.s.r. spectra of the naphthalene radical anions to include those formed with all of the alkali metals in both THF and DME. Sodium and rubidium naphthalenide in DME and both the lithium and the sodium salts in THF exhibited concentration dependent 15 superpositions of spectra at room temperature. Again, a dissociative equilibrium was suggested by the concentration dependence. The authors noted that, in order to see super- imposed spectra such as these, the Species must have life- times of at least 10’8 seconds. The e.s.r. spectrum of lithium naphthalenide in DME exhibited neither a metal splitting nor a concentration dependent proton shift. This was interpreted as evidence for completely dissociated lithium naphthalenide. Some conductance data also indicate dissociative equi— libria. Buschow, Dieleman, and Hoijtink (22) measured the temperature dependence of the conductance for solutions of several aromatic anions which had been reduced with a variety of alkali metals in THF. For potassium anthracenide (KfAnT), for example, the conductance was measured from -750C to +200C. An initial increase from about 10 mho. cm? mole-l at -750C to a maximum of approximately 25 mho. cm? mole-l at OOC was observed. Increasing the temperature from 00C to 250C resulted in a slight decrease in conductivity. .The behavior from -750C to 00C was considered to be caused by the normal increase in conductivity of the ionic species with a decrease in the viscosity of the solvent. The de- crease of the conductance above 00C was postulated to be due to a shift to the left of the equilibrium K+An7 --—>i K+ + AnT (14) 14 which was pronounced enough to overcome the effect of vis— cosity on conductance. The conductance of sodium and lithium anthracenide solutions increased continuously from -750C to +250C, indicating that these solutions have appreciable concentrations of free ions, even at room temperature. Optical spectra of these same solutions (22) Showed that, for the smaller cations, the absorption peaks occurred at higher energies. The Spectral shifts were usually of the order of 200 to 500 cm‘1 for peaks in the ultraviolet or visible region. The conductance data indicated that the free ions were responsible for the absorptions at higher energies. .Examination of the results for a series of compounds led to the following suggestion (22): dissociation of ion— pairs into free ions is enhanced by small cations, large anions, and low temperatures. By using a combination of potentiometric, polarographic, conductometric, and spectroscopic techniques, Slates and Szwarc (25) were able to measure the dissociation constants for a number of anions. These constants increased with the size of the anion. The mobilities of the ions were also measured; radical anions were shown to have much larger mobilities than the cations. Thus, coordination of the nega— tive ions by solvent (THF) does not appear to be favored. b. Evidence for a non-dissociative equilibrium: While some anion solutions probably do contain free ions, as shown 15 in the previous section, other data suggest that Equation 15 may also apply in many instances. For example, the optical Spectrum of sodium fluorenyl (Na+FlT) in THF was shown to contain a number of peaks whose relative intensities could be varied, reversibly, by changing the temperature (24). For instance, a peak at 556 nm. gradually diminished in size as the temperature was changed from +250C to -500C. At the same time, a peak at 575 nm. grew in size. This interconver- sion was not concentration dependent and was not affected by the addition of a common ion. Therefore, the authors elimi- nated the dissociative equilibria + 7 + 7 (Na Fl )2 jf—+’ 2 (Na Fl ) (15) and Na+F17I -——*' Na+ + F17. (16) «s——- as possible explanations of the temperature dependence. A study of the effects of various solvents on the spectrum of lithium fluorenyl (24) suggested an explanation for the temperature dependence. In dioxane at 250C, only a single peak at 546 nm. was observed. In THF, of course, two peaks were seen, whereas in DME, only the absorption at 575 nm was observed. Lower temperatures and more strongly solvating solvents seemed to favor the Species which was responsible for the lower energy peak. These observations indicated an equilibrium between two differently solvated ion-pairs: 16 1.1+, Fl' ——->E Li+|| F17 , (17) where (Li+ H F17) represents a more strongly solvated, or "solvent-separated“, ion-pair. The "contact ion-pair" (Li+, F1?) was considered to be a pair of ions held together by coulombic attraction. Insertion of one or more solvent molecules between the ions would produce a solvent-separated ion-pair. The driving force for such an insertion might be a lowering of the energy through solvation of the cation. In order to test for both the existence and the structure of such species, Hogen-Esch and Smid (24) added small (2 x 10’3 M) quantities of dimethyl sulfoxide (DMSO) to a solution of lithium fluorenyl in dioxane. The addition of such a strong solvating agent would be expected to shift the equilibrium .+ . .+ 7 .L1 , Fl + n DMSO -——9- L1 H Fl (18) to the right. The 546 nm.band was converted to the peak at 575 nm.by this addition and a plot of log (546 nm.band/575 nm. band) vs. log [DMSO] produced a straight line with a slope of 1.15. Therefore, the authors suggested that the insertion of one DMSO molecule would form a solvent—separated ion—pair. Other evidence for the existence of two types of ion— pairs included the studies of Biloen, Fransen, Tulp, and Hoijtink (25). These authors reported a lack of concentra- tion dependence in the temperature dependent conversion of ultraviolet absorption bands of sodium terphenylide in THF. 17 Also, Hirota and co-workers (26) were able to explain the temperature dependence of the sodium splitting of the e.s.r. spectrum of sodium naphthalenide in various solvents by using the concept of solvent-separated ion-pairs. c. Conclusions regarding ion:pairing: Neither a model based entirely on a dissociative nor on a non-dissociative equilibrium could fit all of the data. Hogen-Esch and Smid (27) found that both their spectroscopic and conductance data could be described by the set of equilibria 14+, Ar? ————>: M+ ll Ar": § fl (19) M+ + Ar? in which two kinds of ion-pairs were in equilibrium with the dissociated ions. The Species M+ is an alkali metal cation; M+, ArT is a contact ion-pair. .Another model was presented by Hirota (28). On the basis of measurements of the temperature dependence of the alkali metal splitting and the line width dependence on the magnetic quantum numbers of the alkali metals, the following set of equilibria was proposed: + + 7' 7' 'I' 7 (M,Ar)1——*’=(M.Ar)22M HAr \ +H _/ (20) M + Ar“ . The species (M+, Ar7)1 and (M+, Ar7)2 were both designated "contact ion-pairs", but were considered to be solvated differently. 18 Chang, Slates, and Szwarc (29) suggested a model which did not require two or more distinct kinds of ion-pairs in equilibrium with each other. .Rather, the ion-pairing was considered to be described by a potential well whose Shape was temperature and solvent dependent. However, none of the above models is totally satisfac— tory. At this time, it is probably safe to acknowledge the existence of both contact and solvent-separated.ion—pairs in ethereal solutions of aromatic anions. Their detailed struc- ture is not yet known. Of course, in this study, the detailed structure of ion pairs and, indeed, even the existence of ion-pairs is of major concern only insofar as it affects kinetics. Certain studies do indicate such an effect. Hirota §£__l, (26) and Ward and Weissman (17) have shown that, in the electron exchange reaction between sodium naphthalenide and naphtha— lene (Equation 9), the rate of exchange is strongly dependent upothhe solvent used. The latter authors report rate con- stants of 1 x 107 M“1 sectl in THF and 1 x109 M'i sec?1 in ‘DME. (Since there appears to be a large difference in the ion-pairing in these solvents, these results provide evidence of a kinetic effect of ion-pairing. .Hogen-Esch and Smid (24) presented evidence-that the anionic homopolymerization reac- tion of styrene in THF proceedle3 times faster through the free ion form than through the ion-paired form. Thus, ion- pairing can be kinetically significant. 19 C. Kinetic Studies 1. An Early Mechanism The reactions of interest are the protonations of aromatic radical anions in ethereal solvents. The stoichi— ometry and products of such reactions have been shown to be (2,50,51) 210.167 + 2ROH ——>— Ar +.ArH2 + 2RO- . (21) Paul, Lipkin, and Weissman (8), basing their suggestions on the stoichiometry only, came forward with a mechanism for the reaction of either C02 or various proton donors with the naphthalene radical anion. Written for the reaction with an alcohol, the mechanism was: (22a) . (22c) 20 The electron transfer (22b) was assumed to be very rapid compared with the proton transfers. .The first proton trans- fer (22a) was assumed to be the rate—determining step. On the basis of this suggestion, one would expect the decay of the anion to proceed according to a first—order rate law. Of course, confirmation of this hypothesis awaited actual kinetic studies. 2. Kinetic Studies in Ammonia The early studies of aromatic anion protonation were carried out under conditions which were very different from ours. .In the studies to be discussed (52,55) benzene and its alkyl derivatives were reduced with an alkali metal in liquid ammonia. The differences between the anions, the temperature, and especially the solvent properties tend to make direct comparisons of mechanisms and rates in this sys— tem with those in ethereal solutions hazardous. .Nevertheless, there are enough Similarities to make the comparisons inter— esting. In these early studies, a hydrocarbon, then an alkali metal, and finally an alcohol were added to liquid ammonia. Aliquots were periodically extracted, quenched, and analyzed. Such a method was possible because half-times were of the order of hundreds of seconds. The observed stoichiometry followed the equation: H H 0 + 2M + 2ROH —->- Q + 2MOR. (25) H H 21 Analysis showed that unreacted benzene was also present at the end of the reaction, but this was considered to be un— reacted starting material. Therefore, it is possible that the stoichiometries of the reactions in ammonia and in ethers are the same. The authors found that their results were consistent with a third-order rate law, 1E5]- = kIAr] [M] [ROH] at . (24) This led them to propose the following mechanism: \ M + Solvent 7——*' M+(solv.)°°° e-(Solv.) (25a) M+(Solv.)--¢e-(solv.)+Ar 1—r'Mf(solv.)~°oArT(solv.) (25b) M+(Solv.)'°°Ar7(solv.)+ROH -9-ArH°(solv.)+ ROM(Solv.) (25c) M+(Solv.)°°°e-(solv.)+ArH°(solv.)-f>-M+(solv.)°°°ArH-(solv.) (25d) M+(solv.)°'°ArH_(Solv.)+ROH -—)-ArH2 + ROM(Solv.) . (25e) Except for the implied participation of the cation and the use of solvated electrons as an electron source, the mechan- ism is very Similar to that given by Equation 22. -Equation 25c was considered to be the rate—determining step. .The inclusion of the cation in the mechanism was based on the following observations. The sodium adduct reacted more Slowly than did the hydrocarbon which had been reduced with lithium. Addition of sodium bromide to the solution of +7. ' Na Ar dld not affect the rate of the reaction, whereas the 22 addition of lithium bromide to such a solution caused the reaction rate to increase. Equilibrium 26 was proposed to account for the observation: Li+(solv.)°°°Br-(solv.) + Na+(solv.)oo-Ar7(solv.) -—:;: Na+(solv.)°°-Br-(solv.) + Li+(solv.)°°° Ar7(solv.). (26) Another interesting observation was that bulky substit- uents slowed the reaction significantly. Hindrance of both the solvation of the anion and the accessibility of the anion to ROH were suggested causes of this effect. Other investigators (54), however, found different kinetics for the same system. They used the expression flaif—l = k[Ar][M]2[ROH] (27) to describe their results. .This fourth-order rate law led them to mechanism (28): C6H6 + e" Z °C5H5- (28a) °C5H5 + ROH :3 HoceHe + R0- (28b) HoCeHe + e" ZH:C6H5- (28c) H:C5H5-‘ F1335 C6H7- (28d) c6117" + ROH a}? C6H8 + 110' . (28e) Addition of ethoxide ion inhibited the reaction. Consequently, a reversible proton addition (28b) was postulated. Step 28d was considered to be rate determining. The authors suggested 25 that the intermediate (HzceH3_) could be a w-complex which would slowly rearrange to form another, more readily pro- tonated intermediate (C5H7-). 5. Kinetic Studies by Pulse Radiolysis Later investigations of these protonation reactions were accomplished by utilizing the method of pulse radiolysis. In this method, electrons in solution are produced by high energy electrons from a linear accelerator. In the most rele- vant experiments, aromatic hydrocarbons were dissolved in cyclohexane (55) and in ethanol (56), bombarded with electrons, and observed spectroscopically. ~Absorbing species with life- times in the microsecond time range formed and then decayed. Mapping of the spectral region (400-8001nmJ gave Spectra which were very similar to those obtained when the same hydrocarbons are reduced with alkali metals. Therefore, the transient species were identified as the familiar radical anions, formed this time by the reaction e-(eolv.) + Ar -——+' Ar° . (29) Arai and Dorfman (56) found the rate constant of this reac- tion to be approximately 109 M‘1 sectl, depending upon the particular hydrocarbon used. The formation of the radical anions was accompanied by, but was not itself, the formation of a species in the triplet state (56,57). Anthracene, biphenyl, and terphenyl radical anions were produced in a series of four alcohols (56,58). In each case, 24 the decay of the anion fOIlowed the rate law, — o :_d_.[A_r__L_= kn [Ar7] at , (so) with k' designating a pseudo-first-order rate constant. For a given anion, the rates of reaction with the various alcohols increased in the same order as did the acidity of the alcohols. This was strong evidence that the reaction being investigated was a protonation of the anion. A likely mechanism was pro- posed to be: Ar? + ROH -——>- ArH° + RO- . (51) Second-order rate constants ranged from 2 x 102 M"1 sec::- to 8 x104 M'l sectl. Activation energies were calculated to vary from.2 to 7 kcal. mole'fi depending on the alcohol and anion studied (59). No complete mechanism for the reactions in these systems has been published. However, an interesting possibility has been suggested (40). .When a biphenyl solution was pulsed, a relatively long-lived tranSient species with a peak at 560 nm. was observed (56). The decay of this species was a second-order process. It was tentatively identified as 012H11°, the intermediate product of the protonation. Hoijtink and Velthorst (41,42) have shown that the protonated anion (Aer) is often very stable. .Therefore, the inter- mediate ArH° might be stable enough to be observed. Thus, a possibility for the mechanism is the sequence: 25 Ar' + ROH ——'-ArH~ .+ RO (52a) 2ARH° —-*-Ar + ArH2 . (52b) 4. A Kinetic Study Using ESR and Polarography Umemoto (45) studied the protonation reactions by yet another means in still another solvent. The anthracene radical anion (AnT), among others, was produced electro- chemically in N,N'-dimethylformamide (DMF). AS in the studies mentioned earlier (5-5), d.c. polarograms indicated that two reversible one—electron additions to the aromatic hydrocarbon took place in the dry solvent. Addition of water resulted in an increase of the height of the first wave at the expense of the second. .Alternating current polarography also indicated that two reversible reductions occurred. In dry DMF, the peak currents were 0.55 and 0.04 mix for the first and second waves, respectively. AS the concentration of water in DMF was increased from 0 to 40%, the height of both of these peaks decreased. The results of both the a.c. and the d.c. polarography experiments were interpreted as evidence for a protonation reaction between water and anthracenide. The mechanism which was proposed to explain these results was similar to equation 7: _. (I .— An° + H2O ?£-*- AnH° + OH (55a) AnH° + Art-'- -——>-E AnH’ + An (55b) AnH- + H20 ——->‘ AnH2 + OH_ . (55c) 26 The steady-state assumption was made for the concentration of AnH° and the proton addition step (55a) was considered to be irreversible. The rate law which was derived on the basis of these assumptions predicted a decay which would be first-order in anthracenide: 393%]- = 2 k' [AnT] [H2O] . (34) The experimental arrangement in this study allowed the author to pass electrochemically produced anions into an e.s.r. cavity. Since the protonation reaction was relatively slow (of the order of minutes), it was possible to observe the decay of the e.s.r. signal of anthracenide in this way. .A first-order curve was observed in H2O-DMF mixtures which contained from 1 to 6% water. However, the pseudo-first- order rate constant did not vary linearly with the water concentration as equation 54 would predict. This was attrib- uted tx> changes in the solvation because of the changing solvent composition. An alternative explanation might be that the role of water in the proposed mechanism is incorrect. The data given for the anthracenide-water reaction indicate that, between 1 and 5% of water, a rate law which is second- order in water would be more nearly correct. 5. Summary The kinetic experiments to date have been conducted under very different conditions. Certainly, many of the differences 27 in the experimental results can be attributed to this fact. Nevertheless, there are many points of agreement in spite of the differences in the experimental conditions. In each case, the anion was observed to decay according to a first- order rate law. There seems to be general agreement that the first step is the addition of a proton to the radical anion, although the reversibility of this step is questionable. In both the studies with liquid ammonia and the pulse radi- olysis studies, the rate of reaction with a series of alcohols increasesas the acidic character of the alcoholsin— creases. The ion-pairing data indicate that different cations and solvents should affect the kinetics. The kinetic data do Show such effects. Beyond this point, physical and chemical effects such as differences in temperature, solvent polarity, cationic species, hydrogen bonding, etc. probably mask other chemical similarities which may exist. I I I . EXPERI MENTAL A. General Laboratory Techniques Because of the extreme reactivity of the aromatic radical anions, special care must be taken if stable solu- tions are to be prepared. Since the solutions decompose when exposed to air, high vacuum techniques are necessary. Rigorous cleaning of all glassware is also important. .The following cleaning procedure was used for all stor- age vessels and related glassware. ,The item to be washed was soaked in a cleaning solution which consisted of the following reagents, measured by volume: 60% distilled water 55% Fisher reagent grade HNOs 5% Baker reagent grade 48% HF 2% acid-soluble detergent (Tide) After a few minutes of soaking in the HF cleaner, the vessel was rinsed with distilled water and soaked in aqua regia for several hours. .Finally, it was rinsed ten times with dis- tilled water, followed by five rinses with conductance water. (All vessels were pumped to pressures of the order of 1 x 10‘5 torr for several days before they were used. The cleaning procedure for the stopped-flow system was less rigorous because of the difficulty involved in 28 29 dismantling and reassembling it. .Aqua regia was poured into the system and left for half a day. The system was then rinsed out with distilled water, followed by conductance water as before. The entire stopped-flow system was pumped to at least 4 x 10'5 torr for several days before a run. The anions' sensitivity to air requires that all liquid transfers be done in a closed system. Vacuum distillation was the most common method of transferring pure liquids from one place to another, while solutions were transferred under a helium atmosphere. (In either case, a vacuum System was necessary. Pumping was done with a Cenco Hyvac 7 mechanical pump and a Veeco air-cooled diffusion pump. Dow Corning 704 diffusion pump oil was used in the latter. The pumps were separated from the vacuum system by a liquid nitrogen trap. Pressures of 7 x 10-6 torr could be regularly attained in the vacuum manifold. Tetrahydrofuran, the main solvent used, dissolves most stopcock greases and attacks Viton-A ”o-rings" which are used in many greaseless stopcocks. Therefore, a system in which liquids and vapors could contact only Teflon and glass was desifable.. Two commercially available needle-valve type stopcocks were found to be acceptable after some modification. The Fischer-Porter Lab-Crest 4 mm. Quick Opening Valve uses a glass-to-Teflon seat and isolates the "o-ring" from the vacuum system with a Teflon ridge. It can be made leak~tight by wrapping the threaded part of the Teflon insert with 5O Teflon tape. The Delmar-Urry valve employs four "o-rings", two of which are susceptible to attack by the solvent. .New Teflon inserts, in which these two "o-rings" were re— placed by Teflon ridges, were made and used very successfully. Design of a syringe for the stopped-flow system which would be both greaseless and air-tight posed a special prob- lem which was solved by using a design which was first suggested by Dr..V. A. Nicely. Hamilton No. 1010 Gastight syringe barrels were modified by attaching a 5 mm. Fischer- Porter Solv-Seal joint and a sidearm. Teflon plungers having two sets of double Teflon-ridges were made (see Figure 1). Teflon ridges form an adequate seal for liquids, but not always for gases. By pumping between the pairs of ridges, the gas leakage problem was eliminated. B. Purification Techniques 1. Alkali Metals Sodium and potassium were both handled in the same way. In a helium-filled dryfbag, lumps of metal were cut into small cubes and put into a tall vessel along with several two-foot-long tubes that had been sealed at the top end. The vessel was then connected to a vacuum line and evacuated. The vessel was heated until the metal melted and formed a pool which covered the open ends of the tubes. At this time, helium was admitted, which forced the molten metal up into the tubes, where it was allowed to cool and solidify. 51 Figure 1. A syringe used in the stopped-flow system. 52 These tubes then served as convenient containers which could be cut to the desired length when a small amount of metal was needed. rMetal was introduced into a vessel by inserting the appropriate length of metal-filled tube into a sidearm, pumping the vessel to < 5 x 10‘5 torr, and distilling the metal into the vessel. The sidearms were constricted in four or five places so that, after the metal had been distilled through a particular constriction, that constriction could be sealed off. In this way, solid residue was eliminated and several independent distillations were effected. 2. Anthracene and Terphenyl Zone-refined anthracene and p-terphenyl were obtained from James Hinton Co., Valparaiso, Florida. No further puri- fication was attempted. Samples were weighed and transferred to tubes that were constricted at one end and had a break- seal at the other. The air was then pumped out, the constric- tion sealed off, and the sample stored in the dark. (The sample was put into a vessel for use by sealing the tube to a sidearm on the vessel, pumping out the vessel, and breaking the break—seal with a glass- or Teflon-enclosed magnet (Figure 5). 5. Alcohols and Water Distilled water was passed through a Deminizer ion exchange column made by Crystolab, then redistilled through 55 a three-foot column filled with glass tubing. Degassing was accomplished by freezing with liquid nitrogen, pumping to approximately 1 x 10"5 torr, thawing, and repeating this cycle until the pressure did not jump to more than 5 x 10"5 torr when the frozen sample was opened to the pump. About 200 ml. were vacuum distilled into a storage vessel. Whenever new samples were to be prepared, milliliter amounts were vacuum distilled into very small, weighed vessels in which the degassing procedure was repeated. These small vessels were equipped with a constriction and a break-seal so that the water sample could be admitted into a vessel for mixing with solvent. .All of the alcohols were purified by distillation on a Nester-Faust annular Teflon spinning band distillation column. The alcohols so treated and the fractions retained were: anhydrous, A.C.S. analyzed reagent grade methanol from Matheson, Coleman, and Bell (b.p. = 64.2-64.5OC), spectral grade absolute ethanol (purified by E.M. Hansen), Fisher certified reagent grade 2-propanol (b.p. = 82.00C), and Fisher certified reagent grade t—butanol (b.p. = 82.00C). After distillation, the same degassing and handling proce- dures were used for the alcohols as described previously for water. 4. Tetrahydrofuran and Dimethoxyethane Burdick and Jackson "Distilled in Glass" tetrahydrofuran (THF) was siphoned from the bottle into an evacuated vessel 54 which contained CaH2 (“purified" grade from Fisher Scientific Company). Since the THF was bottled under nitrogen, it had only a very brief contact with air. After several days of refluxing over CaH2, the THF was forced through evacuated tubes into a one-liter distillation pot. From there, it was distilled through a four foot Vigreux column under about 0.75 atmosphere of helium into a vessel which contained NaK (a 1:5 sodium-potassium alloy) and benzophenone from Eastmanf Kodak Company. The purple color of the benzophenone ketyl formed immediately when the THF contacted the NaK and benzo- phenone. The benzophenone ketyl is not stable in the presence of water. It served as both a drying agent and an indicator of dryness. This purple solution could be stored for months at room temperature without decomposing. The THF used for runs KR1 through KR5 was vacuum distilled from the purple benzophenone ketyl solution onto a potassium mirror in another vessel. .After about two days, the clear liquid would turn pale blue. This iS thought to be due to the solution of the metal in THF and was the ultimate cri— terion for dryness. This blue color was more stable at low temperatures than at room temperature. The THF—potassium solutions stayed blue for two days, at most. In the runs noted, the THF was distilled from clear solutions which had been blue with potassium. For all runs after KR5, solutions were prepared from THF which had been distilled directly from the purple solution 55 of the benzophenone ketyl. (The treatment with potassium was found to be time-consuming and was not required for the preparation of stable solutions. When a thermal conductivity detector was used, gas chromatography of THF both before and after distillation on the Vigreux column indicated the presence of only one com- ponent. A more sensitive flame-ionization detector Showed one other component in the undistilled THF. An electron- capture detector indicated that this minor component had a large electron-capture cross-section. We therefore~suspected it to be the stabilizer, butylated hydroxytoluene (BHT), which had been put into the solvent by Burdick and Jackson. New methods of purification were sought in order to make larger scale purifications possible. A zone purification method was tried, unsuccessfully. -Although passing THF directly from the CaH2 vessel into a vessel containing NaK- benzophenone did not produce the stable ketyl, two or three successive vacuum distillations onto NaK-benzophenone did eventually lead to the stable, purple solution. This method of purification was used for runs KR9 through KR14. It eliminated the distillation through the Vigreux column. 5. Helium Helium used for any purpose was first passed over hot (550°C) copper shavings and then over hot cpper oxide. It was then.passed through a tube containing Ascarite and finally through a liquid nitrogen trap. All helium.was grade A helium from Liquid Carbonics Company. 56 C. Preparations 1. Potassium Hydroxide and Potassium Ethoxide These compounds were prepared by vacuum distilling water (or ethanol) onto metallic potassium in order to assure the absence of contaminants. All starting materials were purified in the manner described previously. In order to produce known amounts of product, weighed amounts of potassium were required. The method of Watt and Sowards (44) was used for this purpose. First, small bulblets of approximately 1.5 cm. diameter were made from 5 mm. Pyrex tubing. The tubing was cleaned by the method described earlier, drawn into thin capillaries, and blown into fragile bulblets. All blowing was done through a plastic diaphragm to avoid contamination by the breath. The bulblets were weighed and then placed in the vessel Shown in Figure 2. Potassium was distilled into the vessel through the sidearm until a pool of metal formed around the open ends of the bulblets. When a small pressure of helium was admitted, the potassium was forced up into the bulblets. The vessel was opened in a helium-filled dry-bag where the bulblets were removed. The bulblets,'now filled with helium, were quickly sealed off in the open room. This last step afforded a chance for contamination by air. However, only three of the six bulblets showed any sign of contamination, that indica- tion being a very slight bluish tinge on the metallic surface. Finally, the filled bulblets were reweighed. 57 /} l j I . tgflfl!k:i*’ _, ,1: 4 \_—— I (I I a Figure 2. Two vessels used in the preparation of RO-. Weighed samples of potassium were prepared in vessel (a). The bulblets of potassium were broken in the presence of ROH in vessel (b). 58 The actual reaction took place in the second vessel shown in Figure 2. The bulblet was cleaned, rinsed, and placed in the tube along with a Teflon—coated magnet. When the system had been pumped to,1 x 10-5 torr, the bulblet was broken and water (or ethanol) was distilled onto the exposed potassium. On completion of the reaction, the excess water (or ethanol) was pumped out and the vessel was sealed off. The break~seal vessel could then be attached to a solvent vessel as shown in Figure 5. 2. ROH Solutions After an ROH (or K+RO-) sample had been sealed to the main vessel, as shown in Figure 5, the vessel was pumped to 1 x 10'5 torr or less for several days. Solvent (THF or DME) was vacuum distilled, usually from the purple solution of benzophenone ketyl, into the vessel. .Degassing was accomplished by the freeze-pump-thaw‘method described earlier. After the vessel had been weighed, helium was added and the solvent was frozen and stored under liquid nitrogen and in the dark until shortly before a run. Storage for one or two weeks was not uncommon. Storage at liquid nitrogen or dry ice temperatures was initiated when the addition of pure THF, which had been stored for two days in the dark at'room temperature, caused potassium anthracenide solutions to decompose. When a solution was to be used, it was thawed, the break- seal was broken, and the solvent was mixed with the ROH. 59 .mcoHusaom sodas AQV ppm mom Amy mo mmmnoum paw coflumummmum msu mom poms mammm0> .m musmam Q m _ 1.3,. . .‘y ltl } ...... . 1, d e. \ 4 . \ .., km: W... 1 \\ (2 Ln T w. _ 4..“ 4O 5. Anion Solutions Solutions of the aromatic radical anions were prepared in vessels similar to that shown in Figure 5. The procedure was much the same as that described for ROH solutions. After metal had been distilled into one Side of the vessel, solvent was distilled into the other side. .The break-seal was broken, the aromatic compound dissolved (slowly), and the resulting solution poured through the coarse frit onto the metal. A colored solution formed immediately and Slowly increased in depth of color. .When the color indicated that the absorbance would be approximately unity in the flow system, the solution was poured from the metal surface, helium was admitted into the vessel, and kinetic experiments were begun. Although the solutions were usually used immediately after their formation, instant use was not necessary. Anthracenide solutions were stable for months, even at room temperature. -Anthracene is only slowly soluble in THF in the quanti- ties we used (approximately 4 x 10'3 M). The concentration of anions was usually around 5 x 10"4 M, so that up to a ten-fold excess of unreacted anthracene was present. This excess prevented the formation of appreciable concentrations of dianions. .The spectra of the blue anthracenide solutions were identical to the published spectra of the mono-anions (9). The anion concentration for each run was calculated from the absorbance by using the Beer-Lambert law and extinction coefficients given in the previous reference. 41 Solutions of potasSium p-terphenylide anions were pro- duced in the same way and at about the same concentrations. The yellow-green terphenylide anion seemed to form more rapidly than did the anthracenide anion, but this probably only resulted from the more rapid rate of dissolution of p-terphenyl in THF. This anion was also characterized by its optical Spectrum (9) and solution concentrations were de- termined from the absorbance. The p-terphenylide anion was found to be a rather un- stable species in our system. When solutions of potassium p-terphenylide were stored at room temperature for a few days, red crystals were found growing on the vessel walls. However, the color of the solution was still yellow-green. In the flow system, contamination by the red compound was even more obvious on most occasions. Immediately upon entering the system, the solution turned red. Although rinsing the flow system several times with anion solution finally gave green solution in the burettes, the green color turned to a rusty color and, finally, red, after only minutes of stability. The change in color seemed to occur uniformly throughout the solution. .A spectrum of such an orange-red solution, taken on the Cary model 14 spectrophotometer from540 to 1140 nm., showed the expected spectrum of the terphenylide anion. The only indication of the existence of an impurity was an extraneous peak at 550 nm. .This appeared as a small shoulder on the 480 nm. peak of the p-terphenylide anion (see Figure 4). 42 .uoumEonno locoE msflccmom mnu Sufi? .maw>fluommmmu .mmnsuouonm umflamwuase Noah was mmam 40m mchS an pmcamuno muoB 0 ppm Q muuowmm .uoaoo own m Op GOA» (HmomEoomp Hmfluumm umumm msowusaow HMHHEHm A0 cam QV pom 00mm um mma SH unflamcwnmumu Esammmuom mo coflusaom cmwuml3oaam> m Amv mo muuowmm HHUHDQO .d musmflm flan: once on w moaflmamaflmaoasa mamaom ammmmmamm om a —:—.dfl_a_ j I a _ . 4 a I _ aiO N a O D O O 1- N D O b 6 €_or x saunoo 45 Such red-orange solutions were paramagnetic, as would be expected. An e.s.r. spectrum of such a solution is given in Figure 5. The symmetry of the Spectrum is striking and probably indicates the existence of only one paramagnetic species. The published splitting values for Na+Ter7 in DME (45) correspond well with splittings between major peaks in the spectrum of red-orange sqution. There is a good, but not exact, correspondence between the spectrum of Na+Ter7 and that shown in Figure 5. When a non-correspondence occurs, it can be related to the observation that the published spectrum contains more lines. The Spectrum of the unknown solution was measured at room temperature in THF with a Varian E4 EPR spectrometer, while the known Spectrum was measured at -400C in DME. The differences in temperature, cation, and solvent would all lead to lower resolution in our Spectrum. Thus, we conclude that e.s.r. Spectrum of the orange-red solution is that of potassium p-terphenylide. »D. The Stopped-Flow Experiment '1. Introduction Although the stopped-flow apparatus shown in Figures 6 and 7 may be complex in appearance, its purpose is to solve a very simple problem. For a reaction which occurs in the time range from a few seconds down to a few milliseconds, the reaction is over before more conventional methods of mixing and analysis can be applied. A stopped-flow apparatus 44 .00mm um mma SH mUHH>Smnmumutm Esflmmmuom mo coHDSHOm Aommomfioump madeunmmv 00H S no Esuuummm .m.m.m .m musmflm mlllv. Tldm 00d+ mmsmw Dawn p . w m P p L _ P L _ _ _ C _ L . 00H! L— )- p. 45 is designed to both mix and permit analysis of chemicals in only a few milliseconds and, therefore, allows one to study very fast reactions. Since this apparatus is designed to solve the twofold problem of fast mixing and analysis, a discussion of it is conveniently divided into two parts deal- ing respectively with the mixing and the analysis. 2. The Mixing System The mixing system, Shown in Figure 6, consisted of four parts: burettes and mixing vessels used to dilute solutions of the reactants, a pair of syringes and a pushing block for initiating the flow, a cell for mixing the two reactant solu- tions, and a third syringe designed to abruptly stop the flow of liquid. The glassware was held rigidly in place by a set of aluminum plates which were connected by threaded 5/8 inch brass rods. This whole system was bolted to a table made of 1/4 inch angle iron. Solutions were introduced into the system yia_ports above the burettes. The burettes were standard 50 ml. Pyrex burettes calibrated to 0.1 ml. We estimated the readings to 0.01 ml. The burettes and mixing vessels were always thermo- statted at room temperature since the lack of thermostatting elsewhere in the system required that all reactions be studied near room temperature. Vessels containing solutions of radical anion and of ROH were connected to the system yia 5 mm. Fischer-Porter Solv-Seal joints above the left and right burettes, respectively. A vessel of pure solvent was also attached to each side. 46 Fischer—Porter 5 mm Teflon joints Stopping block Teflon//27 valves 2) II Mixing chamber and optical cell /1(5) .W_Burette ( A ll7lI/I/IIIII/Il [- llllll/IIf/lN/lI/l-IIII/llllll/ll-lllll Mixing L— +—— Z///// l_ Waste vessel 3 \ i : \ \ \ N : 4 t i : N __ : C D E __+ Waste . E A B E . :‘jgg: Aluminum : l . plate X) Threaded E : brass rod “—‘_*: .12: . : =~\‘\_Syr1nges Vacuum «fi——— : (see Figure 1) N N : 3 ‘ Pushing block [J_LI - a——- Figure 6. Stopped-flow system. \ 47 5 torr The entire flow system was pumped to 4 x 10- before solutions were admitted. Dilutions were made by admitting a measured volume of a solution into the mixing vessel, rinsing the burette with pure solvent, and finally admitting a measured amount of pure solvent into the mixing vessel. .Mixing was accomplished by boiling the solution. At these reduced pressures, the heat from one's hand sufficed to cause boiling. The plungers for the bottom syringes were set into an aluminum pushing block which was attached to a lever. (By opening stopcocks A and B and pulling the plungers down, the solutions could be drawn into the syringes. Closing A and B, opening C and-D, and pushing the syringes up caused the solutions to flow through the mixing cell. Mixing times for similar cells had been previously estimated (46) to be less than 2 msecs. ,A description of the construction and design of the mix- ing cells is given by Hansen (47). The particular cell used in this study was made of 1.0 mm. quartz capillary tubing with two optically flat faces. The four tangential jets in its base were drilled by Dr. E. M. Hansen as described by him in the previous reference. When the top syringe filled to a preset volume, which was usually 5 to 4 ml., the plunger struck the upper plate, stopping the flow abruptly. Actually, the design of the system as shown in Figure 6 proved to be somewhat faulty because of the violence of this collision between the plunger 48 and the stopping plate. In some of the early kinetic studies, especially for fast decays, an oscillation of the signal amounting to about 0.05 absorbance units was noticed. The oscillation was traced to vibrations set up in the System by the stopping mechanism. Changing the physical arrangement of the top three plates so that they were no longer attached to the rest of the flow system yia the brass rods greatly diminished this oscillation. However, the older version is shown in Figure 6 because of its simplicity. 5. Data Acquisition The reactions between aromatic radical anions and various proton donors were studied by monitoring the disappear— ance of the anions spectrophotometrically. The analysis system consisted of a light source, monochromator, beam- splitter, sample cell (which in this case was the mixing cell), reference cell, multiplier phototubes, and the data storage system. A block diagram of this network is given in Figure 7. Two different light sources were used, depending on the wavelength range to be studied. Over the range from 250 to 400 nm., a Bausch and-Lomb xenon lamp was employed. A Bausch and Lomb tungsten-iodine lamp was used for work between 400 and 850 nm. The monochromator, a Perkin-Elmer Model 108 Rapid Scan Monochromator, has been described in detail by Feldman (46). This monochromator is capable of scanning over any desired spectral width compatible with the source and detector at .Empmwm :oHuHmHDqUMImumo map mo Emnmmflp onHm .> musmwm 49 Hoxmmmm ocm muonmouoflz monsuouonm .1 cmmouuflz l umflfimsmuoum oe Boo mnsu Hams r .l .2. ---/ _ )lL - wllluasuuao kuuflamm EmmmmVN+.:.Il.. Houmfionnoosoz Hmpuoomm . man _ cmomtowmmm ll meme Hams h 2m easy It). moss t...ul.u\\\ touonm (momma. 9 _ Homoanu " Hmsuflm . . _ umSOHHom mousom semen smoomoaaflumo moonumo msmflm ummmflua 50 rates up to 150 scans per second. We usually observed a complete absorption peak of an anion (c’150 nm.) at a scan-' ning rate of 60 scans per second (16.7 msec. per spectrum). By observing successive Spectra, the decay of an entire peak could be followed. Alternatively, a single wavelength could be continuously monitored simply by turning off the motor and choosing the desired wavelength. This method was used for very fast reactions. When the scanning capability was used, a sharp Spike, used for triggering an oscilloscope or other observational device‘, was; produced at the beginning of each spectrum. The method used to produce this signal was changed from that described earlier (46). A slotted wheel was attached to the drive mechanism such that it turned once per revolution of the mirror. -A small neon bulb, the slot, and a CdS photo- cell were positioned so that a Spike which rose to approxi- mately 1 volt in.1 millisecond was produced each time a new scan began. ,This signal was passed through a cathode follower in order to match the impedance of the trigger system to that of the tape recorder. (A stacked-mirror beam- splitter of the type used in the Bausch and Lomb Spectronic 505 Spectrophotometer split the single beam of light coming from the monochromator into two beams. These were focused on the mixing cell and on a solvent—filled reference cell, respectively, by spherical focusing mirrors of 98 mm. focal length. These mirrors were obtained from the Karl Lambrecht 51 Company. Finally, the light fell onto a pair of multiplier phototubes. The type of'phototube used depended on the wavelength region which was being studied at the time. The RCA multiplier phototubes 6905, 6199, and 7102 were employed in the wavelength regions between 250 to 400, 400 to 600, and 600-900 nm., respectively. ’In order to minimize the effect of thermal noise, the RCA 7102 phototubes were cooled with a stream of nitrogen which had been passed through copper tub- ing immersed in liquid nitrogen. The anode currents from the phototubes were matched by controlling the amount of light which fell onto each phototube when both reference and mixing cell were filled with solvent. The multiplier output currents were then passed into a log circuit which converted the intensity measurements to an absorbance reading. .The components of the circuit include a high impedance input amplifier (P25A), a dual logarithmic transconductor (Pl1—P), an operational output amplifier (P65AU), and a power supply (PR-50C), all available from Philbrick Researches, Inc. The output voltage could be ad- justed to 1, 2, 5 or 10 volts per absorbance unit. Prior to each-run, the linearity of the absorbance measurement was checked with neutral density filters purchased from the Oriel Corp. Data storage during a run was accomplished with an Ampex SP-500 FM.Direct Recorder/Reproducer which had four channels available. .The absorbance data, trigger signal, and a verbal 52 explanation were recorded on three channels. The fourth channel was left blank for later common—mode noise rejection. As the absorbance data were recorded, they were simultaneously read from the tape and displayed on a Tektronix 564 Storage Oscilloscope which employed a Type 2A65 Differential Ampli- fier Unit and a 584 Time Base Unit. This display was strictly for monitoring purposes; no attempt was made to analyze the data from the oscilloscope. After a run had been completed, the kinetic data were read from the tape recorder onto computer cards in digital form and onto graph paper in analog form. A sample graph is given in Figure 8. Such pictures were used for purposes of visualization only; the data which had been punched on cards were analyzed. The apparatus used to effect this data con- version is shown in Figure 9. Basically, the arrangement shown allowed us to solve the following problem. The kinetic data were in the form of discrete spectra stored on magnetic tape as recorded voltage fluctuations. After the time of flow stop, each spectrum was a slightly decayed version of the spectrum preceding it. In order to study the kinetics of the decay, identical spectral regions had to be abstracted from successive spectra and stored in digital form. This desired result was obtained by using a Varian C-1024 Computer of Average Transients (CAT). The CAT sampled volt- ages from the tape at prescribed time intervals, transforming 55 .mmB CH Hocmnuw z_m>fioo.o :33 97.038 Houwm mpwcoumnnucm Esammmuom mo xmmm .E: «VS. 93 mo ~3qu .m wusmflm AEsuuowmm use .Umm mmo.ov wEHB H.O N.o m.o «.0 m.o soueqxosqv 54 FM Tape Recorder Speaker J Audio ' Blank Channel External Absorbance Data .Trigger - Oscilloscope Input + Input Pulse Gate Out Delay Circuit Differential Amplifier A, Trigger Input “a“ Wavetek V ‘ I Address “I1 Advance C°A°T° J l Bin ry Analog Output Ou put To Keypunch To X-Y Recorder Figure 9. Block diagram of the apparatus used to transform analog data to digital form. 55 each voltage measurement into a number of counts and storing .the result in a channel. .The 1024 channels were triggered successively, but independently. By triggering the channels in the way to be discussed next, the spectra could be sampled as desired. Suppose, for example, that we wish to study eight points near the middle of each of 128 successive peaks. As noted earlier, a trigger signal is produced at the beginning of each spectrum. If this signal is fed into a pulse delay circuit, that circuit can be made to produce yet another trigger signal after a desired time interval has elapsed. In this case, the interval chosen will be the time necessary for the recorder to have played back half of the spectrum. The second trigger commands a waveform generator to produce eight Square-wave pulses which cause eight channels to be opened successively for analysis. The time between successive channel-advance pulses relates directly to the Spacing, in wavelength between the measured points on the peak and is determined by the frequency of the square-wave generator. The whole system then waits for the next trigger signal from the recorder. .Thus, eight measurements of the absorbance at eight different wavelengths in the same Spectrum have been made and stored. Of course, the number and separation of the points may be changed by changing the number and fre- quency of the square waves produced by the waveform generator. When the main trigger again signals the beginning of another 56 spectrum, the sampling process is repeated. In this example, the sampling procedure will be repeated until eight points from each of 128 successive spectra have been measured. At this point, all 1024 channels in CAT will have been filled. With this background information, it is possible to describe the system as a whole. The output from two of the recorder's tracks, one containing the absorbance data and the other containing only tape recorder noise, were played into a calibrated differential amplifier. Common mode recorder noise was diminished in this way. The resultant signal was sent into the CAT for conversion and into the Tektronix Storage Oscilloscope for visual observation. Simultaneously, the trigger signal entered the oscilloscope. A gated pulse could be sent out from the oscilloscope either after each spectrum or after two or more spectra had passed. Thus, not every spectrum had to be studied. This gated pulse was fed into the time base of a Tektronix Type 545 A Oscil- loscope, which was used as a pulse-delay circuit. The square- waves were produced by a Model 116 Wavetek Signal Generator and were fed into both the CAT and the storage oscilloscope to be superimposed on the spectrum displayed there. In this way, it was possible to see which portion of the spectrum was being analyzed. Verbal identification of the displayed data was played through a speaker through the fourth channel. The data in the CAT were automatically punched on cards by an IBM type 526 Keypunch equipped with a Varian C-1001 coupler. 57 The data on cards were then analyzed with a CDC-6500 computer by using programs which will now be described. 4. Data Analysis The basic goal in the analysis was, of course, to analyze the ability of a given equation (in this case, a rate law) to describe the data. -A computer program which can decide which of two or more rate laws most closely applies to a set of data has not been written. However, Dr. V. A. Nicely has written a program (KINET) which provides criteria upon which the analyst can make such a judgment. A secondary problem was to transform the data punched by the CAT into a form acceptable to KINET. A program (PUNDAT) which was originally written by Dr. Nicely and modified by the author was used for this purpose. a. Program PUNDAT: The input to this program consisted of a set of organizational constants, the set of spectra of the reacting aromatic anion, and a set of calibrated neutral density filter data. This latter set was included to allow corrections to be made in case the voltage measurements were not linearly dependent upon the actual absorbances over the wavelength range studied. Actually, in the present study, linearity was observed for absorbances less than 1.5 absorb— ance units. Therefore, since the measured absorbances were never greater than 0.7 absorbance units, such corrections were not necessary. 58 After this first (optional) correction for non-linearity, the program analyzed each individual spectrum for spurious electronic noise which was sometimes generated by the CAT. A polynomial was fit to each spectrum and points which deviated from this curve by more than four standard deviations were discarded. From each calculated spectrum, the absorbance at one or more wavelengths and the time from flow-stop to the time of measurement of that spectrum were calculated. Thus, absorb- ance versus time curves were obtained for one or more wave- lengths from the set of Spectra. In order to speed the convergence process in KINET, each absorbance-time curve was smoothed with two passes of a five- point smoothing formula. The absorbances were converted to concentrations. Finally, the program punched the resulting concentration-time curves, along with a ratio of the esti- mated variances of the concentration and time measurements, in a format acceptable to KINET. The variances were estimated to be 1 x 10'8 sec.2 and 25 x 10’5/E moles2 liter-2, where E is the product of the extinction coefficient times the path length. b. Program KINET: .As stated earlier, this program cannot make decisions regarding the applicability of a particular rate law. Rather, it estimates the set of parameters (rate constants, initial concentrations, etc.) which will give the "best" correlation between a given rate law and a set of data. 59 The significance of the results must be decided by the user. Since the author of the program discusses both the philosophy and the methods applied in KINET (48), the emphasis in this discussion will be placed on the type of results which are obtained from.KINET and the criteria that we used to interpret them. In this program, the "best“ estimates of the parameters are defined to be those obtained from a least-squares treat- ment of the data, i.e., S = Z W. F.2 = minimum points . (55) The necessity for including the weights, Wi' and the method for obtaining them are discussed by Wentworth (49). In our case, the weights are given by w. = {L (56) 1. OF 2 i where or 2 or 2 2 ._ 2 2 615.1 (g—t)i citi + (“6:91 6C1 . (57) The quantities ti and c1 are the values of the time and con- centration for the ith point. The quantity Fi is defined by x ) (58) where (x‘.) = f (Y'-) . (59) 60 Generally, in our study, the experimental values for the th point were used for (Xu) concentration and time at the i 1 exp ) and (Y-) 1 exp' respectively. The dependent variable, (X i calc' then became a calculated value of the concentration and was obtained from a rate law in which the independent variable was the experimental value of the time. The result of the minimization procedure was a set of values for the parameters within the rate law which allowed the best fit of the data to a given rate law as defined by Equation 55. The results which are obtained from these calculations are values for S, Fi (the residual at point i) for all values of i, aj (the jth parameter). and oj (the estimated standard deviation for that parameter) for all j, and an estimate of the correlation between parameters. In addition, the com- puter prints comparisons of the calculated and experimental concentrations as functions of time. Several criteria can be used to evaluate such data (50). As a first indication, the magnitude of the sum of the squares of the residuals (S) could be used to compare the fgoodness of fit" of two or more rate laws applied to the same data. ,Since the Size of S is strongly dependent upon the weighting, however, such a comparison can be misleading. A better indication is the magnitude of the individual residuals, Fi.and the randomness of the signs of the residuals. -Another potentially useful criterion is the size of the values of the parameters (aj) relative to their estimated 61 standard deviations (dj). However, for all non-linear equations, these values of the standard deviations are only estimates and should be treated as such. The minimization method is based on a Taylor's Series expansion of the equa- tion with truncation after the linear term. The values of 6j are therefore dependent upon the accuracy of this lineari— zation. .Nevertheless, in general, as the value of Oj approaches that of aj, the applicability of the equation under investigation becomes more doubtful. The multiple correlation coefficients give an indication of the linear dependence among the adjusted parameters. As the values of these coefficients approach unity, the ability of the program to obtain unique solutions diminishes. Often, this problem can be solved by a change in the form of the equation which is being used. Finally, the values of the parameters must make sense in terms of the chemistry which is involved. This final criterion is simply an affirmation of the fact that a back- ground of chemical knowledge must be applied in the interpre- tation of any calculation. While the application of these criteria has not been given explicitly for each of the mechanisms studied, they have been considered in the treatment of the data. The re- sults of the analysis are discussed in the following section. IV. RESULTS AND CONCLUSIONS A. A Survey of the Data 1. Preliminary Experiments As the historical section indicates, the protonation of aromatic anions by alcohols and water has been studied by several investigators. However, the experiments were carried out in media which varied in composition from one to one hundred per cent of the proton source. Even for the case in which the concentration of ROH was lowest (45), the author sug— gested that its concentration was large enough to alter the properties of the medium. One of our objectives was to study the reactions at very low ROH concentrations and, consequently, in an "inert" medium. In this respect, we were entering an unexplored area. Our first problem, then, was to decide upon a system for study. For reasons previously cited (p. 3.), THF was selected as the primary solvent. The reaction between potassium biphenylide (K+Biph7) and methanol (Run KR1) proved to be too fast for us to study quantitatively. The pulse radiolysis studies mentioned previously (p. 25) sug- gested that the terphenylide anion might react more slowly, especially if less acidic proton sources were used. 62 65 Our studies of the reactions of potassium terphenylide (K+Ter?) with ethanol and with water (Runs KR2, 5, and 6) showed that this was indeed a slower reaction. However, solutions of the terphenylide anion were rather unstable and reacted, in the absence of ROH, to form a product with an absorption peak at 550 nm. Since the effect of such a species on the protonation of terphenylide was unknown, the results of our studies of that reaction were of question- able reliability. We were unable, at that time, to prevent the appearance of this unknown species, so we sought a more stable system for study. In experiment KR7, the reactions of potassium anthracenide (KfAn7) with water and with ethanol were studied. Solutions of this anion in THF were stable and the reaction rates were well within our range of study. Therefore, this system (K+An7 in THF) was chosen for further investigation. 2. Principal Experiments The course of this investigation may be seen by inspec- tion of Table 1, which gives a list of all the experiments we carried out. Our choice of experiments was governed by the early observation that the reactions of anthracenide did not follow the simple rate law which was expected. In run KR8, we studied two spectral regions in an attempt to find possible intermediates and to test the reproducibility of the results of run KR7. Experiments 9 through 11 were 64 oomuomm m omoo.onomo.o use roam)» lunarm muo.oamx oomtomm m aao.onaaa.o was moumuoma .kearx mua.oamx oomuomm a ouoatho.ouo.mv are moum brass oauaa.mmx oomnomm m «woo.ou>¢o.o mma Omm prerx oalm.mmx oomnomm m onoaxfia.aum.mv are our heaa+mz mue.mmx oomnomm a onoaxAsm.ouo.mv mma scum he<.+mz oun.mmx oomuomm m oaoo.onsao.o use omm .he¢.+mz mua.mms oomuomm m mo.oumm.o Ems omm brass mua.mmx ooouooo m mo.onm.o are omm hears mna.mmx oomuomm a muoaxom.m has moum th¢.+mz s.>mx oomuomm m ouoaxxom.wnmm.ov are moss. brass mum.smx oomnomm a mmo.oumm.o are omm brass ana.emx UOCHMuhO MUMU 0C “OHQMDmCD OHO3 MGOHHDHOm COHGM PC4+M mlvaH UOCflMUQO MUMQ 0C “OHQMHmCD THO»? mCOHHDHOm COHCM IPHOB+M filmmvm ooouoos a o.a are owm .rume+x mauoa.mmx ooouooa a onoaxm.m are roam .rnma+x mum.mmx oomuomm m ouoaxxmm.ouo.av are roam .hume+x eua.mmx ooonomm m ammo.o-a.a was mom: hemeoam+x ass AECV mcoflumuucmu AZV mCOHumuucmocoo ucw>aom mom coflc4. cum soflmwu Icou mom mo mom m0 mmcmm numcmam>m3 HmQEDZ mucmfiaummxm mo uqu d mflmdB =3.“ 65 03 a mnoax as. 0:00. 3 ems. mops humane mum . 3mm oomuomm w mmoo.o:>mmo.o mza.. owm usdsmz, matmd.mdmx oomnomm m ouoaxfiam.0|mm.mv MEG moum (cramz «Ham.mdm& oomlomm m muoaxfimm.oumw.mv man. moan ht4¥M muw.mdmx oomuomm m mmmoo. on ammo. 0 man omm tears mum . ma UmsHmqu sump on .manmumcs mum32m:0HuDHOm sodas hMma¥x Ndmx oomlomm s mmmo.onmmao.o mme 0mm tseamz ma|0d.adm& oomuomm m mloaerm.oumm.mv mma moum (seamz m|>.admx mtoaxAmm.aums.mc -mo+ oomuomm a 8.0 are Our -, leafs mufiéfimx muoaxxmo.o-so.mv -oom+ oomlomm m mtoaxfimm.finmw.mv mus m0um. .Pscsx. mfilfifi.0HmM oomuomm m ouoax B. 0:53 are moo: . humps mus 6me C O «a l 39 .14 av . --- .12 -... - I - I) I 66 examinations of the effects of the proton donor, the counter- ion, and RO-. In KR15, the effect of changing to a different solvent (dimethoxyethane) was studied. Finally, in KR14, we were able to obtain a stable solution of potassium terphenylide. Thus, the effect of the proton acceptor was also studied. 5.-Discussion of the Data Table 1 gives a comprehensive list of the experiments which were performed. With the following exceptions, the results of all of these experiments were used in the data analysis. Two sets of experiments contained flaws which precluded a meaningful interpretation. Consequently, they were not considered in the analysis. The first set consisted of the early attempt to study potassium terphenylide (KR2). As stated earlier, these solu- tions contained observable (approximately 10'4 M) quantities of an unidentified impurity. The presence of this impurity caused us to reject this run. The second set comprised much of the first part of run KR9. At least two‘mechanical difficulties combined to cause the decay curves to be unusable.. The stopping mechanism was not functioning properly, which caused the initial part of the curve to be poorly defined. At the same time, mechanical vibration in the flow system produced a sinusoidal variation in the observed signal. Figure 10 shows one of these decay curves. It should be contrasted with the kinetic decay shown in Figure 11, which is representative of the majority 67 .8c «do um pououflcos mm3 Esupommm one .pmumuumsaaw mum mcwmm0um uoom cam coflumunw> Hmowsmzoms mo meHQoum one .mmB CH Hocm£um Smmoooo.o Suez mcHxHE com: spasmomueucm ESHUOm mo mouse was A.omev mafia oom 0mm oom oma ooa om _ .OH musmflm t «.0 . N. m o I q. a 1 m.o .o e l $00 1 m.o .omccmom mmB xmom .5: «as wnu mo ucmEmmm 4 .mumo mzu mo muauohme may mo m>aumusomwummn ma swam mane .mma SH Hosmnum 2Nmm00.0 SDHB maHxHE poms spasmomnnucm ESHUOm mo amuse one .afi musmflm A.ommv mEHB o.¢ m.m o.m m.N O.N m.d O.H m.o 68 l _ . A a _ _ ,a.o -m.ow S O I d. D. -. w no. so 69 of our data. These-mechanical difficulties were noticed during the analysis of the data after the run and their causes were eliminated as far as possible. While the mechanical vibration was virtually eliminated, the flow-stop continued to be a problem in the study of reactions with half-times of 20 msecs. or less. It is possible that an abrupt flow-stop was not attained because of relaxation of the Teflon under pressure. With the elimination of these data sets, about half of the data which were to be used to examine the effect of changing the cation and the anion were lost. The data which remained will be discussed later. We emphasize this problem, not because of any deficiency in the remaining data, but rather because the checks for reproducibility between runs are absent. 4. Summary The bulk of our data deals with the reactions of potas- sium anthracenide with various proton donors. The dependence of this reaction on the particular proton source, the concen- tration of RO-, and the solvent were studied. Sufficient data are available to assure us of the internal consistency of the experiments. Several conclusions will be drawn from these results. However, the studies of the dependence on the anion and the counter-ion were not as complete. Their results Show some interesting trends, but confirmation of these trends must await further experimentation. 70 B. Results 1. Dependence on the Anion Concentration A plausible mechanism for the reaction 2K+An7 + 2ROH ———-> 2KOR + An + AnH2 (40) might have been derived from Weissman's suggested mechanism simply by assuming all steps to be irreversible, as shown below. AnT + ROH l3» AnH° + RO- (41a) Art7 + AnH° —£2#- AnH- + An (41b) slow AnH +‘ROH 753! Itan + R0 (41c) Polarographic studies (7) indicate that the second protonation is faster than the first. Assumption of a steady state con- centration for [AnH°] leads to a first-order decay in [AnTj. If Equation 41b is considered to be reversible and the steady state is assumed for both [AnHo] and [AnH-], a rate law for the disappearance of An7 may be derived. It is complex, but first-order in [An7]. Of course, replacement of Equation 41b by a fast equilibrium step would cause 41a to be the slow step. A simple rate law which is first-order in [An?] results (Equation 42): ~d|An71 =1< dt 1 [An-7] [ROH] . (42) Certainly, the results of earlier work in other Systems make such a mechanism plausible for those systems. 71 We expected to observe similar behavior. -Accordingly, our first efforts were directed toward the confirmation or nega- tion of this supposition. The first graph in Figure 12 con- tains a plot of the logarithm of the absorbance of [AnT] versus time for the reaction with water. The straight line was drawn by using the pseudo-first-order rate constant which was calculated from a linear least-squares fit of the data. A pseudo-order treatment was certainly justified, since the initial concentration of water was at least one hundred times that of the anion. Such non-adherence to first-order dependence was observed in the reactions of potassium anthracenide with each of the alcohols, as well as with water, in THF. Therefore, the first-order rate law was considered to be invalid for this system. The deviation from first-order shown in Figure 12 seemed to indicate a decay which was closer to second-order in {An7]. When the rate law -d|An7) = 7'2 dt kPS[An ] (45) was used to fit the same data set, the results shown in the second graph in Figure 12 were obtained. This ability of a pseudo-second-order rate law to describe individual decay curves over at least three to four half-lives was observed for every reaction of potassium anthracenide, providing, of course, that the ROH concentrations were high enough to per- mit a pseudo-order treatment. Figure 15 presents two more examples of this behavior. 72 OD Umflammm mufluwcflx AQV HoUHOIUCOUmmlopsmmm was Amv HmUHOIumHHMIopsmmm .mme CH umum3_z mmm.o zuHB spasmomunucm Esflmmmuom mo coHuommH opp A.mev mEHB w.o m.o $.o Q m.o N.o fi.o Al q 4 1 J uv] A50? [. m A.ommv weds w.o m.o $.O m.o «.0 H.o (a 1 musflom Hmuamfiwummxm .u mafia Umumasoamu.lll J .Na musmflm No.0 moo sod mod 86 55 oo.o moo uv] SOTXI 75 .mma a. Hosanna-» a once x os.m Ans e223 ocm Hocmrum z ouoa x mo.m Ame aufl3 wpficmumuspcm EDHmmMDom mo mcofluummn may no muoam “musclpcoommlopsmmm .ma musmwm Q m A.Ummv mafia H.0omv mafia om?“ omJu OTIHH Om. Oh. cm. on. 0.“. Com O.N OJ” O.—H . )4. A _ _ a d A) .- _ a q I ... 0 \e.&.N «ofm 0\4 (Rm 0\ \¢\ \0\ 4m d \o\0 .0.“ .4. \. .. -... \. .3 \h -nfi \o 1w.“ «\o \6 I (am Ty (\0 .NN Or 4 AM \u x, . -mm x, .mm W. \ m G \. a m. . .8 t._ ion . W . .mm . fin a . 15m awn W a lad .ma \\ q .me we \\ )mw 74 2. Dependence on ROH The applicability of pseudo-order kinetics allowed us to use the following method to determine the dependence of the reaction on the concentration of the proton donor. While equation 45 described individual decay curves for the reaction with a given proton donor, the pseudo-second-order rate constants, kps, varied with the initial concentration of the donor. Since the order in ROH may be expressed by _ n Eps — k[ROH] , (44) or log kps = log k + n log {ROH} . (45) the variation of log kp8 with the logarithm of [ROH]initial was treated by the method of least-squares. The values of the order, n, and plots of log kps vs. log [ROH]initial which resulted from this analysis are given in Tables 2 and 5 and in Figures 14 and 15. .No value for the reaction with methanol is reported, since pseudo-second-order kinetics does not apply to this case because of the low initial con- centrations of methanol. The notation i i.a represents the averaged values of the rate constant and standard deviation obtained from two or more individual experiments. Two important conclusiOns can be made on the basis of the results given in Table 5. First, the order with respect to [ROH] is certainly less than unity in all cases and prob- ably is less than 0.5 in most cases. Second, the order is not necessarily the same for all of the alcohols. Both of these observations imply a complex mechanism for the reaction. 75 TABLE 2 Pseudo-second-order Rate Constants for the Reaction 2K+An'°' + 2ROH lag-fAan +.An + 2K+0R’ Run [MeOH]ox104M. [An'10x104Mav kx10‘3 14-1 sec'1 KR10,7 4.48 5.4 --- * KR10,8 12.0 5.7 --- * KR10,9 20.1 5.2 2.01 i..02 [EtOHlxlOsM kx10’4 M‘l sec’l KR7,6 2.20 5.4 2.20.: .06 KR9,11 0.485 5.2 --- * KR9,12 1.00 6.0 --- * KR9,15 2.42 6.5 1.14 i °01 KR9,14 5.00 6.1 1.60 i .02 KR10,11 1.75 4.6 1.23 i .01 KR10,12 1.55 5.8 0.95.: .02 * KR10,15 3.08 5.8 1.82.1 .01 KR10,14 4.78 4.0 2.54.: .01 KR10,15 6.45 4.2 3.02.: .02 liso-PrOH1x10+lM KR10,1 1.14 3.7 5.94 i..1 KR10,2 0.55 5.2 4220.1 .05 KR10,5 0.12 5.2 2.86.: .05 [t-BuOnglOaM KR10,4 0.80 4.5 2.20.: .02 KR10,5 0.54 4.2 1.75.: .02 KR10,6 2.05 5.5 5.05.: .02 [H20]§101M KR7,1 3.60 1.5 8.91 i .44 KR7,2 1.85 1.9 6.05.: .13 KR7,5 1.11 2.2 4.40.1 .19 KR7,4 0.578 2.4 5.22.: .06 KR8,4 0.860 6.0 3.69.: .11 KR8,5 2.23 5.5 5.12.: .10 KR9,9 0.0414 5.6 1.07.: .01 KR9,10 0.471 5.8 3.62.: .07 * Pseudo-second-order treatment probably not valid because of low initial donor concentration. 76 TABLE 5 Values of the Order, n, of ROH in the Reactions of KfAHT with ROH in THF EtOH iso-PrOH t-BuOH H20 n 0.55 i .15 0.55 i .05 0.41 i_.04 0.44 1,.04 If a value for the order in [ROH] could be established such that one value of n applied to all of the alcohols, it would be possible to determine the dependence of the rate constant (k) on the acidity of the alcohol. Tabulation and ordering of the values of k would then indicate the importance of the protonation step (or steps) in determining the rate of reaction. This was done for the rates of decay of the anions produced by pulse radiolysis of solutions of biphenyl and of anthracene in various alcohols (56). The calculated bimole- cular rate constants were shown to increase with the acidity of the alcohol. Such results were interpreted to mean that the reaction being studied was, indeed, a proton transfer from the alcohol. Since this is such an important point, let us assume for the moment that the reaction follows the rate law 1%: k[An7]2[ROH]'l* . (46) This is probably incorrect, but it will enable us to calculate 'E‘ x 10"4 IE' x 10-4 k' x 10‘4 77 _ A 8.0 — /////// 600'” .. A 4.0 — 5.0” 2.0 _ 1.0 1 119111111 1 11111111 1 1111111 10.0 100.0 [H2O] x 103 b 5.0 - A 2.0 - ° /A 100 1 1 1 111111 1 1 1111 1.0 10.0 [t-BuOH] x 103 c 7.0 — 6.0 ~ a 5.0 _ 4.0L A 5.0 ’A 2'0 1 1 1111111 1 1 11111 1-0 10.0 [iso-PrOH] x 102 Figure 14. Determination of the [ROH] dependence. Log k. is plotted vs. 109 [ROH] for the reactions of KAnT with T5) water, (b) t-butanol, (c) iso-propanol in THF. and 78 -—Calculated line V'Experimental, KR9 AExperimental, KR10 ()Experimental, KR? .0 10.0 [EtOH] x 103 b 8.0: ? 6.0- 2 b 0 o X 4.0- o 54 5.0.— 2°0 1 1 1 1 1 1Ln_1 1 1 1 11114 1.0 10. [H2O] x 103 Figure 15. Determination of the [ROH] dependence. Log R“ is plotted vs. log [ROH] for the reactions of KIAnT'with TE) ethanol in THF and (b) water in DME. 79 and compare estimates of the rate constants for the various alcohols used. The results are shown in Table 4. TABLE 4 Rate Constants Calculated from Equation 46 for the Reactions of Potassium Anthracenide with Various Proton Donors in THF A ' MeOH EtOH iso-PrOH t-BuOHA H2O k(M'3 2sec-1) 3.28x105 3.76x1os 1.7911105 2.14x1o5 1.48x105 .Realizing that these results are likely to be only quali- tativelylmeaningful, we can only say that the smaller, more acidic alcohols do react more rapidly than do the larger alcohols. Thus, it becomes necessary to retain a protonation step in any mechanism we might propose. Furthermore, protona- tion must either compete with, or come before, the major rate- determining step. The protonation reaction cannot be the only rate-limiting step, since this would give first-order behavior. Later, a mechanism which accounts for the variation in dependence on alcohol concentration will be proposed. At that time a more quantitative treatment of the data will be given. 8O 5. Dependence on RO- Since acid-base reactions are often very rapid, we might expect the protonation step 14117 + ROH ——+= AnHo + R0. (47) to be fast and reversible. This might be followed by consecu- tive electron and proton transfers as shown in Equations 48 and 49: 7 kw '- An +AnH W AnH +An (48) AnH + ROH 5;? AnH2 + R0 . (49) Regardless of the nature of the steps which follow Equation 47, however, the first reversible protonation of An7 requires that the overall rate be suppressed by the addition of RO-. Indeed, such an effect could have caused the observed devi- ation from first—order dependence on the alcohol. For instance, if the concentration of AnHo is assumed to be in a steady state, Equations 47 through 49 predict the following rate law: 14422.1 = leIAnT]2-I[%%{:}- . (50) In order to test for such a reversible protonation, the initial concentration of alkoxide was varied while that of the proton source was kept constant. This experiment was carried out for the reactions of potassium anthracenide with both water and ethanol. A pseudo-second-order equation (51) 81 was used to fit the data -d |An-°-| ___ T 2 at kps [An 1 . (51) Figure 16 gives a typical example of the calculated and experimental decay curves. The existence of a reversible proton transfer (Equation 47) requires that kps decrease with added RO-. Equations 47 through 49, for example, predict that this decrease should be described by Equations 52 and 55: g [R031 kps klx [R0‘] (52) log kpS = log(k1K[ROH]) - log [RO-J , (55) Figure 17 shows that changing the alkoxide concentra- tion does not affect the reaction rate. Therefore, the reac- tion mechanism may not include, before the rate-determining step, an equilibrium of the type given by Equation 47. As a result, it is necessary to discard the mechanism described by Equations 47 through 49. One possibility which cannot be tested by these experiments is the reversible attachment of ROH to An? without formation of R0- (Equation 54): T —'+ 000 00° - a An + ROH : An H OR (54) This possibility will be commented upon later. 82 4.0' [An° ]x 104 M ,.———v—(>’-"""> 2.0 ‘———Calculated curve 0 Experimental points 1.0- g ‘o. A \o..\ A \A\°\o \EX‘ 0 A A O + 1 1 4L 1 1 1 1 1 1 1 1 1 1 1 O 0.5 1.0 1.5 Time (sec.) Figure 16. A.pseudo-secogd-order curve and the observed decay of K An' after mixing with 0.00605M ethanol and 0.00214M potassium ethoxide in THF. 85 5.0 a r 4.0»— <3 a o 0 x o 0 (05.0 t m Ix 2.0 l l L J 1.0 2.0 5.0 4.0 [EtO—J x 103 b 7.0 m o E) 6.0+ e1 x 0 3.5 o~ Ix . o 400 1 l l 1 1 1.0 2.0 5.0 4.0 5.0 [on‘] x 103 Figure 17. Determination of the effect of RO-. The pseudo- second-order rate cogstants are plotted yg, the_ concentrations of R0 for the reactions of KIAno with (a).6.4 x-10‘3M ethanol and (b) 0.07 M water in THF. 84 4. Search for Intermediates It would be an important aid to the development of a mechanism if intermediate species could be observed during the reaction. Since our only means of observation is spectroscopic, direct evidence for intermediate species can be obtained in only one of two ways. If the intermediate has an absorption peak in a region which is free of absorp- tions due to other species and if this intermediate exists in quantities great enough to be observed, then we should be able to observe the appearance and decay of that peak. No new peaks were observed in the range from 500 to 750 nm. The second possibility is that the absorption peak of an intermediate might occur beneath one of the peaks of either anthracene or anthracenide. In this case, the rate of decay of the major peak should be wavelength dependent. No wave- length dependence was observed for the potassium anthracenide. bands at 569 and 714 nm. While some changes were seen in the series of anthracene peaks which occur from 292 to 574 nm., analysis indicated that these variations were due to residual absorbances of anthracenide bands. Of course, this does not disprove the existence of inter- mediates at concentrations which are too low to permit direct observation. .This suggests that the steadyestate treatment may be applied to postulated intermediates. 5. Dependence on Solvent Hirota e; pl, (26) obtained evidence that, in the reaction 85 “7+ ‘I' '7 Naph M + Naph ———a- Naph + M Naph , (55) where "Naph" represents naphthalene, the electron transfer proceeded more slowly for contact ion-pairs than for the solvent-separated form. Both cation and solvent effects were noted. Since ion-pairing has been shown to occur in the system under investigation and to be very dependent upon the solvent, it seemed important to observe the effect of a more polar and/or more strongly solvating medium on the reaction in order to determine whether or not ion-pairing affects proton-transfer rates. Dimethoxyethane (DME) was chosen to be the solvent for this experiment. .Although the dielectric constants of DME and THF are approximately the same (7.15 and 7.59 at 250C for DME and THF, resPectively), the ability of DME to solvate cations is greater (24). Consequently, the formation of solvent-separated ion-pairs is favored in DME. The expecta- tion was that the protonation reactions would proceed more rapidly in DME. As test cases, we mixed potassium anthracenide with water and with ethanol in DME. (As usual, the resulting reac- tions were second-order in [An7] (Figure 18). The rate, however, was much slower in DME. While in THF the pseudo- second-order rate constants were of the order of 5 to 6 x104 M‘l secfl, in DME they were around 1 x 103 M’1 sec:l The reaction appeared to be zero-order in the concentration of water, as shown in Table 5 and Figure 15. Because of the 86 A A A 15 '- / A A r-: A A/ I' / g 10~ a °/A I—a A O/ G)\ 0/ A/’ . S ,3/ —' Calculated line 5 h /A/ . . ,0 6 Experimental p01nts ara’o’o’ 1 J L l l l l l 1 l 1 l l l l P 0 1.0 2.0 5.0 4.0 Time (sec.) Figure 18a. Pseudo-sec nd:order kinetics applied to the reaction K.An‘ + .0118 M H2O in DME. A A 25» /;V///// 8 /° A 20~ a 13; . ,4: 1 5 r A/ .\ / $3 10.. /A{/K// A/"-" ‘*‘Calculated line 5__ '/A/’ A Experimental points A /’ IA’A A’A 4 1 2 5 4 5 10 15 Time (sec.) Figure 18b. Pseudo-second-order plot of the reaction of potassium anthracenide with .00265 M ethanol in‘DME. 87 TABLE 5 Pseudo—second-order Rate-Constants for the Reactions of Potassium Anthracenide with (a) Ethanol and (b) water in DME 5a ,Run [EtOHLgiOaM [AnTLglO4M ‘E(M*lsec-1)x10‘3 KR15,6* 0.68 5.5 0.96 1 .02 KR15,7* 1.55 6.2 0.79 1 .02 KR15,6 2.65 7.2 1.46 1 .02 5b Run |H20|§102M lAnTlgloéM 'ELM’lsec‘l)x10'3 KR15,2 0.56 1.4 4.65 1 .09 KR15,4 5.67 4.4 4.64 1 .09 KR15,5 1.18 5.9 5.51 1 .07 *- This concentration is so low that pseudo-second-order . kinetics certainly are not applicable. low concentrations of ethanol which were used, the pseudo- order method could not be used to determine the order with respect to [ROH]. The results shown in Table 5 indicate that it is greater than zero, however. Thus, it appears that stronger solvation does not enhance the rate of the reaction. .Rather, the results in DME sug— gest that the mechanism which applies in THF also applies in DME and that this mechanism utilizes an intermediate step, prior to the protonation step, which is hindered by stronger solvation. The zero-order dependence on water indicates that, in this case, the intermediate step is slower than the pro- tonation process. 88 This inhibition of the reaction by a change in the solvent was even more striking when potassium anthracenide was mixed with water in ethylenediamine. In this instance, no reaction was observed, even over a period of 5 to 10 minutes. ,Such a result might mean that the reaction with water requires an intermediate ion-paired form of the anion which was unavailable in the solvent of higher solvating abil- ity° An alternative explanation might be that the water was tied up with the solvent through hydrogen bonding and therefore did not react. 6. Dependence on the Cation In a further attempt to elucidate the importance of ion—pairing, the reactions of water and of ethanol with the sodium salts of the anthracene radical anion were studied. This cation was chosen on the basis of conductance data which indicated the sodium salts to be less strongly ion—paired than the potassium_salts (221. Although the lithium salt would have been a better choice from the standpoint of mini- mization of ion-pairing, metal purification would present greater experimental difficulties for lithium than for sodium. One would expect, however, that either cation would reduce the rate of reaction to something slower than that observed for potassium anthracenide if the reaction were to proceed preferentially gig a contact ion-pair. The general reaction which was studied was, of course, + '7 + - 2Na An + 2R0H ——-+ 2Na R0 + Aan + An . (56) 89 Two proton sources, ethanol and water, and two solvents, THF and DME, were employed. Unfortunately, mechanical dif- ficulties caused about half of the work in THF to be discarded. The reader is referred to section IV—A for a discussion of these difficulties. In the following sections, the data which remained are discussed. a. The reactions of Na+AnT in THF: The reaction of sodium anthracenide with ethanol in THF could be described by the rate law -ddlin o] = kps [An-5'] 2 . (57) A representative fit of this rate law to the data is shown in Figure 19. Rate constants calculated from Equation 57 are given in Table 6. TABLE 6 Pseudo-second-order Rate Constants for the Reaction NaJ'An'T + EtOH in THF Run [EtOHLxlOam [AnTlx103M 'Eésx10'4M-lsec'l KR11,7 0.845 0.40 0.86 * KR11,8 1.69 0.56 1.71 KR11,9 2.82 0.26 1.87 *Pseudo-second-order treatment not valid. 90 Calculated curve 1 i ! 5.0.. \ 0 Experimental points 1x104M 0 (An? 0/ \O\?~o~—o ~—O\__CL_O_ 5 15 25 55 45 55 65 Time (msec.) Figure 20. Parallel first- and ecgnd-order equation applied to the reaction: Na An° + 0.00256M H20 in THF. o 40 - O o 55_ O o 50» /9/o// o H 25 " /0/ lg 0/0 h\-'. 20 "' /O o /o O ,0 . .4 15_ /0 '—-Calculated line 0 / /0 c>Experimental points 10 r- 0/0 o/ / 50:0’0 9 1 1 1 A 1 1 1 1 1 l 1 I 0.2 0.4 0.6 0.8 1.0 2.0 Time (sec.) Figure 19. Pseudo-second-order kinetics applied to the reaction: NafAn7 + 0.00282M.Et0H in THF. 91 Surprisingly, a pseudo-order treatment gave a good fit to the data, even for the lowest concentration studied. This prob— ably indicates a low order for the dependence on the ethanol concentrations. Thus, the results seem to indicate no strik— ing differences between the reactions of K+An7 and of NafAn7 with ethanol in THF. This statement can not be made with regard to the reaction between NafAn7 and water in THF, however. In this case, the decay of the anion could not be described by either a first- or a second-order dependence on the anion. Rather, a fractional order between one and two seemed necessary. Moreover, the two sets of experiments with this system produced different results. While some of the decay curves were obviously unus- able because of the mechanical problems discussed earlier, many decay curves did not appear to be seriously affected, yet did not give consistent results. We were unable to disCOVer a reason for the inconsisten- cies. .Nevertheless, as we will show later, there was reason to suspect that a combination of first- and second-order decays, -d An? _ '7 7 2 —L—]-dt - kfIAn ] + ks[An ] . (58) might fit the data. A representative decay curve and the tabulated results of the application of Equation 58 are given .in Figure 20 and Table 7, respectively. The rate constants for experiments KR9 indicate a reaction in which both the 92 TABLE 7 Rate Constants Which Result From Application of a Parallel First— and Second-order Rate Law (Equation 58) to the Reaction of Na+An° with Water in THF % First Run [HZOLfiM) k3(M‘1sec?1)x10‘5 kf(sec31) Order KR11, 10.4 0.0165 1.0.8210.04 -1.2910.19 7 KR11, 10.8 0.0165 1.5910.05 -6.7110.45 17 KR11, 11.5 0.0525 2.5410.06 -4.9510.52 KR11, 11.4 0.0525 2.4510.06 -5.1110.55 KR11, 12.4 0.0599 2.8910.04 -9.0510.41 10 KR9, 8.5.' 0.00114 0.7710.04 59.2 10.7 57 KR9, 8.5 0.00114 0.9510.07 40.0 11.5 52 KR9, 7.4. 0.00256 0.6910.05 52.7 11.0 60 KR9, 2.5 0.00467 1.1210.16 15.5 11.9 55 KR9, 2.6 0.00467 1.1210.14 18.7 12.5 55 * The "per cent first-order" is the per cent of the initial rate which is due to the first-order term. 95 first- and_second-order components contribute equally to the initial rate of decay. Experiments KR11, however, resulted in data which can best be described by a nearly second-order decay with a systematic negative contribution from a first- order term. This term contributes little to the initial rate of decay, which makes its sign less alarming than if it were a major contributor. Two conclusions might be drawn from the data given in Table 7. First, changing the concentration of water did not cause large variations in the rate of the reaction. This is in agreement with our observations for the reactions of KfAnT; Second, however, the rate constants for the second- order decay are consistently three to four times those calcu- lated for the reactions of KfAnT with comparable concentra- tions of water. Since the order of the decay was neither reproducible nor consistent with the results of other reac- tions of Na+An75 we attributed these results to some unknown and therefore uncontrolled chemical cause and did not include them in our considerations of a general mechanism. b. The reactions of NafAn7 in DME: Two major points can be made with regard to the reaction of NafAn7 with ethanol in DME. First, the reaction is very slow when compared with the same reaction in THF. Since solvation of the cation is greater in DME, that solvent should also diminish the degree of contact-ion-pair formation. Thus, one would expect the rate of reaction to be lower. Second, the data could be fit 94 by the bimolecular second-order mechanism "LEI-$11 = 1%[An7] [EtOH] . (59) Figure 21 contains a representative data set and the compar- ison with the calculated curve. The rate constants obtained from the least-squares analyses of the data are given in Table 8. The one value of kb which seems out of line comes from the one experiment of the three which was most likely to be in error, since it was done after a series of reactions of potassium anthracenide. Contamination by potassium ions would be expected to increase the rate and cause a deviation from Equation 59. .It.is possible, therefore, that Equation 59 does, in fact, describe this experiment. However, the small number of data points involved does not permit evalua- tion of a valid rate constant. -The first-order nature of the decay with respect to {AnTJ is, however, evident in all cases. For the reaction of Na+AnT with water in DME, a pseudo- first-order rate law was able to describe the decay of anthracenide. Figure 22 shows an example of this behavior. Once again, the reaction in DME was found to be slow compared to its counterpart in THF. The rate constants are listed in Table 9. Surprisingly, the dependence on the water concentra- tion is low. A possible explanation for this phenomenon will be given later. 95 .mmomo may no uflm HmUHOIUGOme amasomaoeflm .ZmIOd x 6m.a n Hmoumw .mza a“ moan + hbmymz mo sofluommm A.ommv mafia mm ON ma oa m _ 4 _ _ s q _ Ikrl o -IIIOIIII 0/ 1 Oil. /0 Jj/ 1 o IAY/ o/, Q/d// J 9/0 muswom Hmucmfifiummxm o // o m>nsu Umumasoamo III // n .dm musmflm d.o N.o m.o «.0 m.o .o: x [Luv] 96 TABLE 8 Bimolecular Second-order‘Rate-Constants for the Reaction of Sodium Anthracenide with Ethanol in DME Run [EtOH]3<103M' [An7]ox104M k(M‘1sec-l)x10'2 KR15,9 2.65 6.7 72.14 i..07 KR15.10 1.54 4.4 0.94.: .004 KR15,11 0.91 4.8 0.95.i .01 TABLE 9 Pseudo-first-order Rate Constants for the Reaction of Sodium Anthracenide with Water in DME Run , [H20]g<102M [An7]0x104M k(sec"1) KR15,12 0.85 5.0 0.418.i .004 KR15,15 .1.24 ' 6.5 0.251.i .005 KR15,14 1.80 5.5 0.225.: .004 KR15,15 5.57 6.5 0.550.: .006 97 ii 0.51% b ——-Calculated curve 2 0-4‘ Rb 0 Experimental points m . 9 \o >< 0.5 - \ I: O s. 0 .2 — \O \ O\O 0 .1 4 \O\o \O \O ‘*\~0 1 1 1 1 1 T—\0 1 0\10‘ 1.0 2.0 5.0 4.0 5.0 6.0 7.0 8.0 Time (sec.) Figure 22. Reaction of NafAnT + 0.0085M H20 in DME. A pseudo-first-order fit of the decay. '98 c. Conclusions: While the results of the experiments with sodium anthracenide contain more uncertainty then we would like, some conclusions seem valid. The solvent di- methoxyethane certainly inhibits the rate of the reactions. Furthermore, in DME the reactions proceed more slowly when sodium, rather than potassium, is the cation. In order to.. show this, let us apply the pseudo-second-order rate law (Equation 12) to the data for the reactions of both‘NafAn'7 and K+AnT with water in DME. Of course, the fit is not good for the sodium salt, but it should be good enough to permit an estimate of relative rates.> The calculated rates-are of the order of 4 x102 and 4 x103 M‘l sec‘1 for the sodium and potassium salts, respectively. -Lastly, there does appear to be a change in the rate law which describes the disappear- ance of the anthracenide anion. There is a progression from a strictly second-order decay with~K+AnT in THF to one which is first order for NafAnT in DME. These results are con- sistent with a mechanism in which ion-pairing leads to a fast, second-order reaction. Under some circumstances, this mode of reaction predominates over a slower, first-order reaction which does not require such pairing. Under circumstances unfavorable to ion-pairing, the-first—order reaction becomes apparent. 7. Dependence on the Anion Another factor which can affect the extent of ion-pairing .is the nature of the anion. .In order to study this effect 99 and, at the same time, to test the generality of the results obtained for the anthracenide system, the reactions of potassium terphenylide with water and with ethanol in THF were studied. The reaction of'KfTerT with ethanol in THF appears to be another example of a reaction which proceeds yig_the first-order mechanism. Figure 25 is an example of the results obtained by fitting Equation 60 1% = k[TerT] [EtOH] (60) to the data. Table 10 contains a list of the rate constants which were obtained at various concentrations of ethanol. The concentration independence of the rate constant is further evidence that Equation 60 does apply. When a solution of potassium terphenylide was mixed with water in THF, no decay of the anion was observed. .While this K+TerT solution was one in which decomposition to the red solution had occurred, e.s.r. and optical evidence indicate that significant quantities of terphenylide were still present in such solutions. .Thus, the terphenylide anion does not appear to react with water in THF. .The results obtained with ethanol indicate that p-terphenylide is an example Of an anion which has access to a fast first-order reaction with ROH. Conductance and optical studies of solutions of Na+TerT in THF indicate that approxi- mately equimolar concentrations of free ions, solvent—separated .mme a“ Hocmnum z_¢IOd x m>.m + .HmB M "cofluommn may 0» Umfiammm moflpmcwx HmUHOIUcoomm HMHDQMHOE*m .mm musmflm A.ummEv mafia 100 5N mm mm Hm ON ma ma Na ma ma 9d ma Nd Ha 0d m m fl N r r _ r _ _ 1 q r _ d _ A A A _ A O / 1 Foo o 1 w.o 9/ 1 m.o / 1 o.H mucflom Hmucmemexm o / ..a.d m>usu Umumasoamo 1|II /y 1 N.H w .0? X [_. 18.1.] 101 TABLE 10 Bimolecular Second-order Rate Constants for the Reaction of Potassium Terphenylide with Ethanol in THF Run [EtOHinOSM [TerTLx104M 'k(M-1sec-l)x105 KR14,2 2.57 0.50 0.57.: .01 KR14,5 0.975 0.70 1.58 1,.05 KR14,4 0.575 1.40 1.22,: .01 KR14,5 0.418 . .1.70 1.24.1 .01 ion-pairs, and contact ion-pairs are present in solutions of low (approximately 1 x 10-4 M) total anion concentration at room temperature (50). The e.s.r. spectra of both the lithium and the sodium salts in THF show no structure due to the cation (45). These results suggest that potassium terphenylide solutions in THF contain appreciable quantities of both contact and solvent-separated ion-pairs. If this is true, then the first-order process in the reaction with ethanol must be fast enough to compete favorably with the second-order process, if, indeed, such a process exists for terphenylide. C. A Discussion of Mechanisms 1. Inapplicable Mechanisms Before considering a final mechanism, let us digress for a moment to review some preliminary ideas which proved not to be in accord with the results. This evolutionary 102 approach will outline the basic problems which arose and the steps which led to development of a consistent mechanism. We have seen that no previously postulated mechanism predicts a second-order disappearance of the anion. The existence of an equilibrium such as An° + ROH ——> AnHo + R0 (61) could explain the low order in ROH. However, the absence of an alkoxide effect ruled out such an equilibrium. Thus, the previously postulated mechanisms were inconsistent with' two of our most basic observations. Using the stoichiometry of the reaction as a basis, Weissman had suggested a mechanism which contained consecu— tive proton and electron transfers (p. .19). Another possible sequence might be one in which an electron transfer precedes the proton attachment, thereby creating a species with an enhanced proton affinity. This species might then react with a proton much faster than could the anthracenide anion itself. Several studies have indicated the existence of an equilibrium between mono- and di-negative anions (51-55). Other studies suggest that, in the case of anthracene anions, the di-anion has a higher proton affinity than does the mono- anion. This conclusion is based on the observations that the reaction 2An? + 2ROH —->-2R0" + An + AnH2 (62) 105 appears to be irreversible, yet a proton can be abstracted from the dihydroproduct in the reaction An- + AnH2 ——+-E 2AnH' . (65) These ideas suggested the following mechanism for the reac- tion under investigation: An° + An? 43"- An= + An (64a) El = - k _ _ An + ROH Tic??? AnH + R0 (64b) AnH' +y'ROH 73%;?- AnH2 + Ro' . (64c) Since no observable Concentrations of the dianion were present in our experiments, application of the steady-state assump- tion for [An:] seemed justified. This led to the rate law: "9113:? = 2k1[An-:]2 12:. [An]' . (65) 1 k2 [ROH] Although such a form seemed capable of explaining both the second-order decay of An7 and the low dependence on the concentration of ROH, it was unable to fit the data. The values of both {An} and [ROH] were known and had been varied from run to run, so that an adequate test of this mechanism was available. Therefore, it was necessary to discard the mechanism. There is evidence for the effect of clusters in ROH in the kinetics of proton exchange reactions in aqueous solutions 104 of imidazolium.ions 154]. In this instance, hydrogen bonding within the clusters appears to compete with the protonation process.' It seemed reasonable to expect that such clusters could also exist in media of low dielectric constant. If hydrogen bonding were competitive with proton transfer, the effective concentration of available protons and, consequently, the order of the ROH dependence would be lowered. A mechanism.which describes this situation is given below: 2ROH 35» (Roma (66a) An7+ ROH —-1->k [An°°°H°‘°OR]- (66b) k—i [An°°°H°°°OR]- + An7 :55? AnH- + R0_ + An (66c) Ann“ + ROH %? AnH2 + no” . (66d) The intermediate in Equation 66b was suggested by the absence of an alkoxide effect. If we make the steady-state assumption for the concentration of this intermediate, the rate law given by Equation 67 can be derived: -d An? _ k k _ [1111712 1 _ __a%_l - .5122 R2 [An7]+k_1 {(8K[ROH]t + 1) 1 }. (67) The term, [ROH]t, refers to the total concentration of alcohol, [ROH]t = [ROH] + [(ROH)2] . (68) 105 This mechanism contained two very attractive aspects. First of all, depending on the relative magnitudes of k—l and k2, the order of the dependence on [An7j could be first, second, or something in between these. One would expect this competition between k_1 and kg to be dependent on the particular anion used. An effect due to the cation would be allowed, but its cause-would not be obvious. The order with respect to the anion would also be expected to be dependent upon the acidity of the alcohol, however. That particular dependence was not observed. Since the magnitude of that effect is not known, its apparent absence should not destroy the credibility of the mechanism. The second facet of the mechanism which interested us was its prediction that the order with respect to the alcohol concentration should depend upon the particular alcohol. 7 However, since the hydrogen bonding should be greatest for the smaller alcohols, these are the ones which should exhibit the lowest order. The reverse of this is observed. Another failure of this mechanism resided in our in- ability to fit it to the data. .At least three explanations are possible. We have found that, as the number of para- meters and the non-linear dependence on them increase, the curve-fitting program demands better initial estimates of the parameters. We may have failed to produce sufficient estimates. Secondly, a correct mechanism might require formation of higher aggregates of alcohol molecules. 106 HOwever, this would lead to more parameters with no obvious increase in the correctness. In fact, the reactions in DME exhibited the lowest ROH dependence of all. Thus, this explanation would require that larger clusters exist in DME than in THF, which is the reverse of what would be expected on the basis of solvating ability. Lastly, the mechanism might simply be incorrect. In view of a later mechanism which seemed both quantitatively and qualitatively superior to this one, we favor the third explanation. 2. A Suggested Mechanism a. Formulation: While the two previously discussed mechanisms were deemed to be incorrect, they contained ideas which led to yet another scheme for the reaction process. It seemed that, if neutral molecules could form clusters, so also might neutral ionhpairs. As far as this author knows, no direct evidence for ion-quadruples exists for aromatic radical anion systems, but evidence for such Species is available for other systems (55,56). Current theories sug- gest the existence of several kinds of ion-pairs in solutions of anions. Consequently, the presence of paired ion—pairs in solvents of low dielectric constant seems possible. The reaction with a proton donor might proceed preferentially through such a species since the necessary electron transfer would already be at least partially completed. Our proposed mechanism is: 107 k. + '7 + ,7 ,_+ + T,- M ,An + M ,An ' *1: (M An )2 (69a) + 7. k ’ + - + — (M An )2 + ROH 13—15? [M AnH ] + M ,R0 + An (69b) [MfA H'] + ROH -53—+- AnH + M+ RO- (69c) 8 ’ fast 2 ’ ’ Since there.seemed to be no g_priori reason to assume that the contact ion-pair, M7,An7; could not also react with the alcohol, a first-order reaction was included in the general mechanism. we expected this to contribute very little to the overall rate of reaction and added it only for completeness. Specifically, this part of the mechanism con- sisted of the sequence: M+,An7v+. ROH J‘s—>- M+R0" + AnH° (69d) .slow + ? k + - M ,An + mm 754—» M AnH + An (69e) ast + - k + — M AnH + ROH E-Eg?’ M R0 + AnH2 . (69c) If a steady state concentration is assumed for the species, (MfAn?)2, the general rate law for steps 69a—e becomes k -d1M+18n1 = _ 4,- k_+k1[ROH] dt 2 k+[MTAn7]2 1 + ( 7 0) k 3:. [M+,An7.] [ROH] . The form which was actually used by the curve-fitting program was derived from this one by minor algebraic manipulations: 108 + 7 1 + k1 [ROH] (71) k3 [M+,An"'] [ROH] . Three parameters were defined for the computer. They were u(1) = 2 k+, u(2) = k_/k1. and u(5) = k3, where u(I) designates the Ith parameter. We could find no way to determine more than the ratio of k1 and k_. b. Experimental basis: In order to obtain the most statistically rigorous treatment of the data, three or more data sets were fit simultaneously by the program. .Each data set consisted of a concentration versus time curve for the reaction of anthracenide with one concentration of a particu- lar alcohol. .If, for example, three data sets were used, each set represented a different initial concentration of ROH. Therefore, when the program minimized the sum of the squares of the residuals of all three data sets, it was required to handle both the anthracenide and the alcohol dependence simultaneOusly. Some representative results of this procedure are shoWn in Figure 24. Table 11 contains the resultant rate constants. .Although the godd agreement between the shapes of the predicted and experimental decay curves speaks strongly for the mechanism, other criteria must also be met. The rate constants k+ and k_ should be constant for the reactions of potassium anthracenide in THF. The value of k_ is hidden in a ratio, but the value of k+ is, within one standard ]x103 [A117 109 0.4Q ——— Calculated curves E) 0 ° Experimental points 0.5 K) 3 0.2.. Q Q O\ O 0.1 " \O\O “006 A O\..o‘o ~—o a ‘7‘h‘o‘_“‘°‘-——O o o— 0.0 “0.5 “004 A \ \> -0.2 \. 004 q)- I \A‘o \ \.\ . 0.1 0.5 .0 A\°\o \O \A\0_§A__ A A A 0* \> -1O.O 0.2 _ b. Q 0\0 \ 0 1 _ 0‘0 o O \O C \O\O~Q-O o 0 0.0 l l l l l l l l l l L l 1 1 L 0.5 100 105 Time (sec.) Figure 24. Simultaneous fitting of Equation 71 to three data sets. (a) 0.008 M, t—BuOH in THF are treated. The reactions of KfAnT'with (b) 0.02 M, and (c) 0.0054 M 110 To H mé 6105.880 H «min; monAmoo H 3.3 83 CH 0mm + PEWM >.m H n.08 v10023.30 H ~m.0v MOHxAmN.o H mm.¢v mzn CH mon + FCCFM 0.: H 0337 m10023.0 H 3.3 «0280.0 H 3.3 BE. CH COHC + Ic¢+mz m.H H N.HH muonAmH.o H mm.Hv «oaxAmo.o H >>.HV mma CH owm + FCCHM m.N H m.mm mIOHxAmN.o H dd.mv 40HxA¢o.o H #w.dv mma CH mosmlu + PCCHx m.H H o.mH m1OHxAmw.o H mm.mv ¢oax1mo.o H mo.mv mma CH mOHCIOmH + htCHfi «0.0 H mm.m mIOHxAmH.o H mo.¢v vOHxA>0.0 H om.mv mme CH movm + hCCHM «.OH H Ho.m1 m1oaxflmm.o H ww.mv ¢OHxAOH.O.H mm.mv mus CH mom: + hrmrx Aauomma1szx sz Hx\1x HanommH12v+x CoHuommm «Ha mqméa mzn CH 0C0 mma CH.PC¢+MZ UCm .C¢¥& Ho MCoHHmCOHOHm map 0H an CoHumsqm Ho CoHumuHHmmfi EOHM.HH5mmm CUHCZ mquumCoo mumm {‘11.}. It‘ll 110a mum mmsoum HCmemmHv EOHH UmCHmuno muHCmmH 0:» Ho mCoHHMH>mU UHMUCmHm .muHsmmu 02H mo moCmHmHmCOU wCu mmumoHUCH mHH anme mHHC3 .=HHH mo mmmeoom: mnu mwumoHUCH Hm .mdd OHQMB CH .mHmsomCmuHCEHm HHH mumm mumv m>HH 0H 03u Ho wumHMCoo msoum m 0H0£3 .msoum m CHCHHB COHHMH>00 wumvaHm 0:» mo mmumEHHmm Ummmnm>m mum «Ha anma CH muHEHH quHmunmoCs 0C9 m EACH «.3 x 26H 60H x 86... BE CH Omm + 1..C<+x m séHH .10H x meH N0H x HH.0H mzCCH CBC + bC<+C H .. . . .. Cue. CH moum + 1.51.62 m m¢.NH ~10H x mm.0H 40H x >H.OH mma CH 0mm +.PC¢%M m omHH 010H x mH.mH «OH x 3.3 ..Ee CH mosmuu +PC<+C m OMAOH. 010H x 0m.mH vOH x 00.0 mma CH mOHQIOmH +.PC¢H& N $0.0H 010H x >m.0H 90H x 0N.0H mus CH m0um +.PC4+M N .... 010H x m0.HH «OH x 00.0 hue CH mom: +.PC¢+M mmsouw Amxv o Aax\1xv o A+xv o CoHuommm > k_/(kl[ROH]), the reaction should be virtually independent of the concentration of ROH. Such a situation would be most likely to occur at high concentra- tions of the alcohol. “Conversely, low concentrations of alcohol would be expected to cause the order with respect to [ROH] to increase until a limiting order of unity were reached. A value of k_/k1 of 10'2 ought to allow one to observe a change in order from one to zero by changing the alcohol concentration OVer a range of 10-3 to 1 M. Although, in our experiments, the alcohol concentration was not varied over such a wide range, we have indications that the phenomenon may apply. In our studies of the reaction KfAn? + ROH in THF, the initial concentrations of ethanol 115 were of the order of 10—3 M, whereas those of isopropanol were‘10-1 to 10‘2 M. The order in ethanol was 0.5 or possibly higher, whereas for isopropanol the order was only 0.5. The second major implication is that the order of the dependence on the concentration of the anion should decrease as the extent of ion—pairing diminishes. This effect should be dependent on the solvatiOn properties of the medium, the nature of either the cation or the anion, and the tempera- ture. The kinetic studies which were discussed in the histori— cal section are probably instances in which both the cation and solvent effeCts reduced ion-pairing. Another such in- stance would be our study of the reactions of sodium anthra- cenide in DME, in which the decay was first-order in [NafAn?]. The first-order dependence on terphenylide in its reaction with ethanol in THF represents an example in which the nature of the anion is apparently the deciding factor. Several other combinations of variables should be good tests of the effect of ion-pairing. The extension of studies of potassium anthracenide to more polar solvents, the use of lithium, rather than potassium, anthracenide in THF, and a study of the reactions ongfAn'7 in THF at lower tempera- tures are instances which should result in a first-order dependence on anthracenide. On the other hand, a study of the reactions of K+TerT in methyltetrahydrofuran or some 114 other weakly-solvating medium might result in a second- order dependence on [TerTj. This would be a good test of the generality of the anthracenide results. D. Conclusions The reactions of potassium anthracenide with various proton donors in THF proceed yia_a mechanism which is second- order in anthracenide.’ The importance of this second-order process appears to be diminished by conditions which decrease ion-pairing. While the reaction of NafAnT with ethanol in THF is probably similar to that of KfAnT, the rate of the reactions of KfAn? with ethanol and water in a more strongly solvating medium.(DME) is greatly reduced. When the same reactions of Na+AnT are studied in.DME, the rates are even slower and the dependence on [AnTj is first order. This observation indicates that a first-order path for the protonation of anthracenide does exist, but that it is much slower than the second-order process and only predominates when contact ion-pairs are absent. The rate of this first-order process may well be dependent upon the nature of the anion. A fast, first-order decay of the anion was observed in the reaction of potassium terphenylide with ethanol in THF. In this case, the first- order process may be fast enough to compete favorably with the second-order path, even in the presence of an appreciable quantity of contact ion-pairs. 115 While the rates of the reactions of K+AnT with various alcohols in THF appear to increase with the acidity of the alcohol, the effect of acidity is a small one. This is thought to be caused by the dependence of the total rate on both the rate of formation of a reactive, ion-paired species and the rate of its subsequent protonation. The reactions with water are somewhat anomalous. While the kinetics observed for the reactions with water are the same as those for the alcohols, the rates with water are slower than might be expected on the basis of relative acidities. Indeed, in ethylenediamine, water does not react with potassium anthracenide. .This may indicate that water interacts with the solvent more strongly than do the alco- hols. An interesting test of this would be a study of the reaction (or lack of a reaction) of ethanol with KfAnT in ethylenediamine. Several tests of these ideas have been suggested in a previous section. .An extension of our studies to a wider range of alcohol concentrations would examine the validity of Equation 71. A more basic study would be one which fur- ther examines the effects of ion+pairing upon the reactions. .Also, the studies should be extended to other aromatic anions to determine the generality of our observations. REFERENCES ll 1 [.11 Ill 1 5. 4. 10. 11. 12. 15. 14. REFERENCES W. Schlenk, J. Appenrodt, A. Michael, and A. Thal, Ber., 41, 479 (1914). N. D. Scott, J. F. Walker, and V. L. Hansley, J. Am. Chem. Soc., §§, 2442 (1956). A. Maccoll, Nature, 165, 178 (1949). H. A. Laitinen and S. Wawzonek, J. Am. Chem. Soc., 62, 1765 (1942). G. J. Hoijtink and J. van Schooten, Rec. Trav. Chim., .11, 1089 (1952). G. J. Hoijtink, E. DeBoer, P. H. Van Der Meij, and W. P. Weijland, Rec. Trav. Chim., 1§, 487 (1956). G. J. Hoijtink, J. van Schooten, E. DeBoer, and W. Ij. Aalbersberg, Rec. Trav. Chim., 1Q, 555 (1954). D. E. Paul, D. Lipkin, and s. I. Weissman, J. Am. Chem. Soc., 18, 116 (1956). P. Balk, G. J..Hoijtink, and J. W. H. Schreurs, Rec..Tr§v. Chim., 16, 815 (1957). E. DeBoer and S. I. Weissman, Rec. Trav. Chim., 16, 824 (1957). G. J. Hoijtink and P. J. Zandstra, Mol. Phys., Q, 571 (1960). K. H. J. Buschow, J. Dieleman, and G. J. Hoijtink, Mol. Phys., 1, 1 (1965-64). T. L. Chu and S. C. Yu, J. Am. Chem. Soc., 16, 5567 (1954). T. R. Tuttle, Jr., R. L. Ward, and s. I. Weissman, J. Chem. Phys.,.§§, 189 (1956). 116 15. 16. 17. 18. 19. 20. 21. 22. 25. 24. 25. 26. 27. 28. 29. 50. 51. 52.' 55. 117 E..DeBoer, J. Chem. Phys., 88, 190 (1956). E. DeBoer and S. I. Weissman, J. Am. Chem. Soc., 88, 4549 (1958). R. L. Ward and S. I. Weissman, J. Am. Chem. Soc., 18, 2086 (1957). M. Szwarc, Accounts Chem. Res., 8, 87 (1969). .N. M. Atherton and S. I. Weissman, J. Am. Chem. Soc., 8;, 1550 (1961). A. Crowley, N. Hirota, and R..Kreilick, J. Chem. Phys., .49. 4815 (1967). C. L. Dodson and A. H. Reddoch, J. Chem. Phys., 88, 5226 (1968). K..H. J. Buschow, J. Dieleman, and G. J. Hoijtink, J. Chem. Phys., 18, 1995 (1965). R. V. Slates and M. Szwarc, J. Phys. Chem., _8, 4124 (1965). T. E. Hogen—Esch and J. Smid, J. Am. Chem. Soc., 88, 507 (1966). P. Biloen, T. Fransen, A. Tulp, and G. J. Hoijtink, J. Phys. Chem., 18, 1581 (1969). N. Hirota, R. Carraway, and W. Schook, J. Am. Chem. Soc., .29, 5611 (1968). T. E..Hogen—Esch and J. Smid, J. Am. Chem. Soc., 88, 518 (1966). .N. Hirota, J. Am. Chem. Soc., 88, 5605 (1968). P. Chang, R. V. Slates and M. Szwarc, J. Phys. Chem., 19, 5180 (1966). W. E. Bachmann, J. Org, Chem., 1, 547 (1956). J. F. Walker and N. D. Scott, J. Am. Chem. Soc., 88, 951 (1958). A. P..Krapcho and A. A. Bothner-By, J. Am. Chem. Soc., '81, 5658 (1959). A. P. Krapcho and A. A. Bothner-By, J. Am. Chem. Soc., .§§. 751 (1960). 54. 55. 56. 57. 58. 59. 40.- 41. 42. 45. 44. 45. 46. 47. 48. 4 9 O 50 O 51 O 118 O. J. Jacobus and J. F. Eastham, J. Am. Chem. Soc., 81, 5799 (1965). J. P. Keene, E. J. Land, and A. J. Swallow, J. Am. Chem. Soc., 81, 5284 (1965). S. Arai and L. M. Dorfman, J. Chem. Phys., 11, 2190 (1964). S. Arai and L. M. Dorfman, J. Phys. Chem., 88, 2259 (1965). L. M. Dorfman,.N. E. Shank, and S. Arai, Advances in Chemistry Series, 88, "Radiation Chemistry-2", R. F. Gould, Ed., American Chemical Society, Washington, D. C., 1968, p. 58. S. Arai, E. L. Tremba, J. R. Brandon, and L. M. Dorfman, Can. J. Chem., 18, 1119 (1967). L. M. Dorfman, personal communication. N. H. Velthorst and G. J. Hoijtink, J. Am. Chem. Soc., .Ql: 4529 (1965). .N. H. Velthorst and G. J. Hoijtink, J. Am. Chem. Soc., 88, 209 (1967). K. Umemoto, Bull. Chem. Soc. Japan, 18, 1058 (1967). G. W. Watt and D. M. Sowards, J. Am. Chem. Soc., 18, 4742 (1954). K. H. Hausser, L. Mongini, and R. van Steenwinkel, Z. Naturforsch., 19a, 777 (1964). L. H. Feldman, Ph. D. Thesis, Michigan State University, 1966. E. M. Hansen, Ph. D. Thesis, Michigan State University, >1970. V. A. Nicely, Ph. D. Thesis, Michigan State University, 1969. W. E. Wentworth, J. Chem. Educ., $8, 96 (1965). J. D. Olson, personal communication. .K. H. J. Buschow and G. J. Hoijtink, J. Chem. Phys., 28, 2501 (1964). 119 52. G. Henrici-Olivé’and S. Olive: Z. physik.Chem. Neue Folge, $8, 527 (1964). 55. K. K. Brandes and R. J. Gerdes, J. Phys. Chem., 11, 508 (1967) . 54. E. K. Ralph, III and E. Grunwald, J. Am. Chem. Soc., 88, 517 (1968). 55. S. Bruckenstein and Saito, J. Am. Chem. Soc., 81, 698 (1965). 56. S. Bruckenstein and-D..F. Untereker, J. Am. Chem. Soc., 81, 5741 (1969) . APPENDIX REPRESENTATIVE SETS OF KINETIC DATA REACTION OF K+Am- MF0H=.000448M 120 + MEOH IN THF MFOH=.00120M MEOH=.00376M AR- CONC. TIMF(SEC.) AP- CONC. TIMF(SEC.) AR- CONC. TIMF(9FC.) .585F-03 O. .QQOF-01 0. .3845-03 0. .44?F-01 .667F-0) .374F-03 .667E-0) .229F-01 .667F-fll .367F-0? .133F+00 .3025-“3 .133F+00 .1666-03 .133F+00 OBIRF"03 CPODF+00 o?67E"03 0200E’On 01325-03 o?00E*00 .28OF-03 .267F+00 .235F-03 .267E+00 .lllE-O3 .267F+nn .2545—01 .333F+00 .210F-03 .333F+00 .964E-04 .133F+00 .PWWF-03 .400E+00 .IQRE-03 .400E¢00 .8285-04 .AOOF+00 .21?E~03 .4675900 .1785-03 .467E+00 .7512-04 .467E+OO .194F—03 .533F+00 .1625-08 .533E+oo .6705-04 .533F+nn .180F~03 .600F+00 .ISPE-U? .600F000 .6045-04 .600E+00 .170E003 .667F+00 .IWRE-U3 .700E+00 .549E-04 .667E+OO .1595093 .731F+00 .1236-01 .833E+00 .505E-04 .733E+00 .1495-03 .800F+00 .1176-03 .967E000 .4715-04 .800F+00 .lth-01 .867F+00 .lOIEvO? .IIOE+01 .471E-04 .867F+On .13?F-01 .933F+00 .QPSE-nq .1?3E+01 .430E-04 .933E+00 o126F-03 .100F+Ol .RRRF-Oh .137E+01 .3725-04 .100F+Ol .IPOF-O3 .107F+01 .RIUE-04 .150E+01 .369E-04 .107F+nl .114F-03 .11?F+01 .7655-04 .1635+01 .349E-04 .113E+Ol .110F-03 .1?OF+0) .7PPF-04 .177E+01 .1435-04 .120F+0] .IHAF-O3 .177E+01 .681E-04 .IQOE+01 .3016-04 .127F+01 CIGOF-03 .133F+0] .6135-04 .210500] .295F-04 .133F+nl .9555~Oh .140F+01 .539F~04 .237F+01 .287E-04 .140F+0) .975F-04 .147E*()l .517E-04 0263E*01 o?66E"04 0147F4’0] .HRIF-04 .153F+O) .QRIE-OQ .290E+01 .260E-04 .153F+01 .fi7hF-04 .160F+0) .4415-04 .317E+01 .2396-04 .160F+01 .826F~04 .167F+0) .415E-04 .343E+01 .2425-04 .167F+Ol .8175—04 .171F+0) .383E-04 .370E¢01 .226E-04 .171F+0) .765F-04 .180F+01 .3576-04 .397E+01 .199E-04 .180F+01 074:;E'04 olg7F+01 0. 00 0220E‘04 01R7F+01 .7llE-O4 .193E+01 0. 0. .190E-04 .193F+0) 0702F’04 0200F*01 0. 0“ .18§E’04 CPOOF§01 REACTION OF K+AN- + ETOH 1N THF FTOH=.00175~ FTOH=.00308M ETOH=.OOHARM .SlflF-oa o, .RDWE-n? 0. .APHF-n? n. .SAIF-Oi .h67F-nl .2665—03 .667E-0) .7QSF—01 .313F—n) .261F-09 .1316+nn .1955-n1 .133E+nn .224E-03 .6676-01 .217F-oa .2onE+oo .1605-01 .2006onn .1836—08 .100F+00 .186F-08 .267F+on .l32E-03 .267E+oo .153E-03 .133F+00 .166E-01 .1336+on .1142—03 .313£+nn .1295-03 .167E+nn .1495-93 .4onr+on .1025-03 .gogg.on .1206-08 .7nor+no 121 AP- CONC. TIME(§FC.) AQ- CONF. TIMF(5FC.) AQ— CONC. TIMF(SFC.) .1575-03 .567F+00 .916F-04 .467E+00 .IOSE-O? .753F+00 .122F-01 .533F+00 .805F-04 .533E*00 .9765-04 .267F+00 .115F-03 .600F+00 .7PHE-04 .600E+00 .8665-04 .300E+00 .1055'03 .667H+00 .673E-04 .700E*00 .RIZF-04 .5505*00 .9735-04 .732F+on .SASE-04 .833E+00 .7096-04 .417F+00 .930F-04 .800F*00 .snnE-oa .967E+00 .6025-04 .ARWanfl .874F-04 .867F+00 .4585-04 allOE+01 .5395-04 .55OE+00 .829F-04 .933F+00 .hnfiF-Dh .123E+01 .5165-04 .617F+00 .776F-04 .100F+01 .198E—n4 .137E+01 .3805-04 .683F+00 .738F-04 .107F+01 .3415-“4 .150E*0) .389E-04 .750F+00 .694F-04 .31?F+01 .?lOE-04 .163E+Ol .171E-04 .817E+00 i656E’04 .l?fiF+01 .2655-04 .177F*0) .3?2E-04 .885F+00 0637F-04 o‘27F*01 .PEQE-04 O‘QOE*0] .356E'04 .QSOF+00 oleE-Oh .175F+01 .245E-04 .2IOE+01 .3076-04 .105F+01 .567F-04 .140F+Ol .2I7E-04 .237500) .2545-04 .118F+01 .SRRE-Oh .147F+0) .IQRE-na .263E*01 .226E-O4 .13?F+01 .527F-nh .l53F+01 .169E-04 .290E*01 .PPIE-04 .145F+01 65145-04 olF‘OF’nl 0161E"()4 .317F“()] .201E-04 o]SHF+0] .SDOF-Oh .157F+U) .145F-04 .343E*01 .174E-04 .172F+0) .48?F-04 .17‘F+01 .1295-04 .370F'0] .197E-04 .185E+0) 0471F'04 0180F+01 .114E‘04 .397E*0) 0. 0. OQSEF-04 .1R7F+01 no 00 0. 00 .454E-04 .1938+Ol 0. 0. 0. 0. RFACTION 0F K+AN- + ISO-PROH IN THF PDOH=.1)4M PROH=.0555M PR0H=.01165M 05641:-03 no 33355-07 00 oQOOE-03 00 .520F-03 .115F-01 .721E-0? .3336-01 .258E-03 .667F—01 .172F-0q .6h7F-01 .735E-01 .667E-01 .178E-03 .133F+nn .132F-03 .1006+UO .173E-03 .100E+0n .lqu-OW .?00F+00 .106Fc01 .131¢+On .IAOE-n3 .131F+00 .lnfiF-03 .767F+nn .9715-04 .167F+00 .llqE-OW .167F+“n .BRSE-04 .331F+00 .771F-na .?nnF+00 .QQVE-04 .200F+00 .7456-04 .annF+nn .qqqpsna .731F+00 .RARE-Oa .233Eonfl .654E-04 .467F+00 .finRF-Oa .?67E+00 .7R7F-04 .267F+f)n .661F-04 .613F+00 .903E-04 .300F+00 .6QHE-Oa .300E+00 .604E-04 .600F+00 0400F¢04 .753F*00 .656F‘04 0333F*00 oleE’OA .667F+00 .AGOF-Oh .167F+00 .618F-04 .367E+00 .4006-04 .733F+00 .3ROF-Oh .400F+00 .524F-04 .4OOE+OO .1816-04 .ROOF+00 .376E-04 .460F*00 .aaqE-na .433F+00 .1176-04 .867F+00 .240F-04 .617F400 .aanF-na .467E400 .319F-04 .913F+00 .ZQOF-OA .RRRF+OH .A‘fiF-04 .SHOE+OU .107F—04 .100F+0] .203F-04 .650F+00 .1946—04 .531E+un .PQQF-Oh .107E+01 .207E-Oa .717E+00 .3?OE-04 .567F+00 .PSQF-na .111E+01 .1725-04 .7H?E+On .337F-04 .600E+00 .P64F-04 .120E+01 .186E304 .RSOE+OO .127E-n4 .633E+nn .P36F-04 .127F+nl .182F-OA .917F+flfi .PQaF-Oa .667E*nn .196E-04 .133F+n) .568E106 .983F+00 .871F—04 .7OOE+00 .2226-04 .140F+01 0. ‘ 0. .?40E-04 .733E+no .2146-04 .147F+nl AQ- CONF. TIMF(SEC.) AD— CONC. TIMF(SFC.) AR— CONC. TIME(SFC.) “. 0. .351F-04 .767E+00 .203F-o4 .IS3F+01 0. 0. .24SE—04 .ROOE+00 .1876-04 .160E+01 0. 0. .PShF—Oa .8335+00 .160E-04 .167E+nl O. 0. .EOGF—OA .867E+00 .17SF-04 .173E+01 0. 0. .IOPF-Oa .900E000 .l67E-04 .IROE+01 0. 0. .Pfiqfi-Oa .933E+00 .128E-04 .187E+01 0. 0. 0. 0. .1576-04 .193E+01 O. O. O. 0. .1375-04 .?00E+n) no 0. 0. no 01146-04 QPOTE+01 122 QFACTION 0F K+AN- + HHOH IN THF RHOPC,HOQCM HHOH=.OOHM RUOH=.0204M .PARF-O? .6875-01 .973E-01 .667E-01 .268E-03 .667F-01 .20?F-01 .137F+00 .IQRE-OW .133E+00 .169Enn3 .133E+00 .156E-01 .200F+nn .ISAF-03 .200E+00 .lROF-03 .200F+n0 .131F-01 .?67F+00 .1979-03 .267E+nn .104E-03 .267F+nn .112F-01 .333E+00 .IHRF—n1 .333F+On .RREF-oa .333E+00 .970F-06 .400F+00 .940F-04 .400F+nn .711E-na .400E+00 .QO4F—Oa .467F+00 .611F~na .467E+00 .676F-04 .467F+On .716F-oa .600F+00 .711F-04 .533E+00 .SO7F-04 .613F+nn .671F-04 .800E+nn .650F-HA .6HOE+00 .444F-04 .600E+nn HASRFona .100F+01 .6025-04 .667E+00 .413E-04 .667F+00 .376F~Oh .12OE+0] .537E-“4 .733E+00 .171E-04 .7336+nn .3ORF-na .140F+0] .4115-04 .933E+nn .325E-04 .800F+nn 027QF’04 .160“+0] 95:0E-()4 .113E2'0) 0‘?4F-04 0867E*On .260E—04 .180F+0] .286E-na .133E+01 .ZfiaE-na .100E+01 .PO7E-Oa .POOF+HI .PilF-na .163E+0) .246E-04 .1?nF+n) 0. 0. .23lF-04 .173F+01 .l7SE-04 .I4nF+01 0. no oquE"()/+ .19?F+()1 01506‘04 .160F+0) REACTION OF K+AN- + H20 IN THF H?0:.]RSM H?0:.00414M H20=,047]M o170F-03 0. .5485-03 00 og87E-03 00 .1305-03 .333E-01 .4?OE-01 .333F-01 .154E-03 .133F-01 .921E-04 .667F-01 .fifilE-n? .667E-01 .?57F-03 .667F-nl .897F-04 .100E*00 o3nSE-nq 01006.00 0187E-03 .100F000 .764FHOA .IRRFonn .262F-nw .133E+On .lSlE-03 .133E+On .6SSF-Ok .167E+00 .24SE—n3 .167E+00 .1295-03 .167E+On .6026énk .?OOF+0” .221F-03 .200E+00 .lnlE-03 .?OOE+00 .464F«Da .231F+00 .198E-03 .233E+On .QIUF-04 .233F+nn .4065404 .267E+00 .182F—03 .267E+00 .796E-04 .267E+00 .435E.na .300E+00 .175E-03 .3005+on .797E-04 .3005+no .3695Hoa .333F+00 .164E—U3 .333E+nn .627F-04 .813F+nn .319F204 .367E+00 .144E-03 .367E+nn .SPSE-Oa .367F+nn .3136w04 .4OOF+00 .1476-0? .400F1HH .6246-04 .400F+00 125 AP- CONC. TIMF(SFC.) AR- CONC. T1MF(SEC.) AP- CONC. TIMF1SFC.) .249F-04 .433F+00 .127E-03 .433F+nn .189E-04 .500F+00 .306Fv04 .467F+HO .125F-03 .467F+00 .4325-04 .600F+00 .260fivoh .SOOF+OO .llBE—n? .533F+00 .364F-04 .700F+nn .EIQF-OA .533F+00 .lORF-n3 .633F+nn .236E-04 .ROOF+00 .236F-04 .S67E+00 .797F-04 .731E+00 .233E-04 .QOOF+00 .244F-04 .600E+0“ .760F-04 .811F+00 .213E-04 .100F+01 .136E-04 .63?F+00 .696E-04 .9336000 .185E-04 .IInF+nI .2116~04 .667F+00 .694E-04 .103Et01 .158E-04 .131F+01 .196F-Oa .700F+00 .4066-04 .113E+01 .949E-OS .157F+n) .152F-04 .733E+00 .5905-04 .123E901 .1436-04 .IROF+O] .IRIF-Oa .767E+OO .SIQE-na .13?E+01 0. 0. .IBBF-Oa .ROHF+00 .470E-04 .143E+01 0. 0. .129F-04 .933F+00 .APBF-Oa .1636+01 O. 0. .147F~OA .967F+00 .3RHE-04 .163E001 0. 0. .164F-04 .900F+00 .1?9E-04 .173E001 O. 0. 0. O. .246E—04 .193E+OI O. 0. REACIIQM OF K+AN- + ETOH IM DME ETOH=.00068% ETOH=.00133M ET0H=.00263M ogan-Oq ()0 OSQOE-01 0. 06907;-01 0. .502F-O3 .767F+00 .507E-03 .267E+00 .5285-03 .P67F+fln .443F-03 .513F+00 .443E-03 .5335’00 .434E-03 .533E*00 .396F-01 .800F+00 .hOIE-n3 .800E§00 .376E-03 .ROOF+00 .36?F-03 .107F+01 .369E-0? .107E*Ol .3305-03 .107F+0] .3295-05 5131F+01 .3425-0? .133F+0) .2955-03 .133E+0] .306F-0? .160F+01 .3185-03 .160F+01 .266E-O? .I73F+fl] OZRRF-O1 .187E*0) 03006‘03 0187F*nl o?29F'03 0227F+07 .277F-0? .2]1F+01 .285E-01 .213E+0) .192E-0? .PROF+01 .2445-03 .267E+01 .2575-03 .267F+01 .151E-03 .187F+01 .731F-fl3 .293F+01 .2445-03 .2936‘01 .135E-03 .440F+01 .219F-03 .320F+01 .234E-03 .3?0F*01 .llqE-O? .495E401 .195F-01 .400F+Ol .2235-03 .347E+O) .llBF-O? .547E+OI .ITSF-03 .480E+Ol .211E-03 .3715001 .lOlF-03 .600F+01 .160F-n? .560F+01 .IQQF-01 .427E#0) .954F-04 .653E+01 .147F-05 .640F+01 .l76E-03 .507E+01 .762E-04 .773F+01 .l‘fiF-O1 .7POF+0) .162E-03 .587E001 .5655-04 .960F+n) .1745-01 .HOOF+01 .1485-0? .667F+Ol .4305-04 .1)5F+0? .IIRF-O? .890F+01 .139F-03 .747F+Ol .360E-04 .131F+O? .llnF-07 .9606+01 .1?7E-01 .827E*Ol .255E-04 .152E+0? .104F-01 .104F+O? ollfiE-03 .907500] 0. 0. .971F-04 .l]?F+0? .102E-03 .104E§02 0. 0. .9075-04 .120F+02 .909E-04 .121F‘02 0. 0. .8712-04 .128F+02 .7766-04 .141E+02 o. n. .849F-04 .136F+O? .6965-04 .160F+02 0. 0. 68065-04 .144F+02 O. O.‘ 0. 0. 6773F-04 .15?E+O? 0. O. 0. 0. 07055-0“ ol60F’0? O. 0. 0. 0. REACTION OF K+AN~ 124 + H20 IN DME H?0=.00360M H20=.0357M H?O=.0118M AP- CONC} TIME(SFC.) AP- CONC. TIME(SEC.) AR- CONC. TIME(SEC.) .IBOE?01 0. OSISE'03 no 0630F-03 0. .IIQE-03 .700E+00 .401E-03 .1335+00 .S4OE-03 .667F-nl .IIOF-O? .400E+00 .3325-03 .267E+00 .484E-03 .133E+00 .9935604 .600F+00 .2845-03 .400E’00 .4295-03 .POOF+00 .QOQF-O4 .800F+00 .243E-03 .533E*00 .389E-03 .767E*nn .825F-04 .IIOE+0] .PIRE-03 .667E*00 .3525-03 .333F+00 .724F-04 .150E*01 .IQSE-03 .BOOE+OO .3175-03 .400F+00 .676F-04 .190E+Ol .171E-03 .933E+00 02955-03 .467F+00 .559E~da .230E+01 .147E-03 .1205+01 .2765-03 .533E+on .SllE-flh .270E+01 .125E-03 .147E+01 .258E-03 .6nnE+nn .457F-04 .310E+01 .lllE-OW .173E+Ol .229E-03 .73?F+00 .4216-04 .350F+01 .963F-04 .200E0Ol .193E-03 .933F+00 .380F-04 .390E*01 .877E-04 .2?7E+Ol .170E-03 .113F+01 .356F-04 .430E*01 Q7RBE'04 OZSJE’OI .1496’03 0133F+01 .33lF-04 .47OE+01 .703E-04 .280E+01 .133E-03 .151F+fl) .305E-04 .510F+01 .655F-04 .307E+Ol .122E-03 .I73F+0] .2835-04 .550E+01 .569E-04 .347F601 .IIOE’OB .I91F+nl .266F-04 .590F+01 .486E-O4 .400E+01 .990E-04 .213F+OI .252F-04 .630E+01 .416E-04 .453E*01 .949E-04 .273F+01 .241F-04 .6708+01 .187E-04 .SO7E+0] .893E-04 .253E+01 .227F-04 .7]OE+01 .364E-04 .5605+01 .8085-04 .283F+01 .223F-04 .7SOE+01 .3215-04 .6135+01 .7185-04 .3PBE+OI .202E-04 .790E*01 0294E-04 o667E’01 oq83E’04 0363F*nl .187E-04 .830F*01 02696-04 0720£*nl 0. no 0174E-04 .870E+01 00 00 00 no .ITOF-Oh .910E+0] 0. O. 0. 0. .167F-04 .960F+01 0. n. 0. 0. oquF—‘0‘Q .990E*01 0. ()0 O. ”o olalF-Oq .10?E+02 00 0. 0. no .138F-06 .107F+O? 0. O. 0. 0. .133F-04 .111F*0? 0. 00 O. 00 .IISE+02 0. O. 0. 0. 0132F'04 REACTION OF NA+AN~ + ETnH IN THF ETOH=.000845M FTOH=.00169M ETnH=.00282M ,3Qns-03 n, .1inF-01 0. .2976-03 0. .ROIF-Ofi .BCIV-H) .PQQF-n3 .333E-01 .216E-n3 .313F-nl .292F~n? .133C+Ofl .267F-“3 .667E-O) .186F-03 .667F-01 .211F-01 .200F+Ou .PPSF-ni .IOOE+OO .167E-03 .100F+nn .IQWF-OI .?A7F+nn .198E—03 .133E+00 .lSSF-OB .133F+nn .17SF-01 .331F+00 .17RF-03 .167E+00 .144E-03 .167F+00 .157F~03 .400E+00 .167F-03 .200E+On .1326-03 .200F+On .147F-01 .467E+00 .147E-03 .2336400 .1266-01 .P33E+00 .114F901 .631E+OH .1165—03 .267E+00 .119E-03 .767F+nn .123F—03 .600E+00 .lPfiF-OB .3OOEi00 .1126-03 .300F100 125 AP- F0MC. TTV519EC.) AP- COMC. TIMF(9FC.) AR- CONC. TIMF(SFP.) .IIGF-Oi .667F+00 .I?”E-U? .333t+00 .1065-0? .333F+00 .ll’F-03 .711F+Ufl .lllF-03 .367E+00 .IOOE-HT .367F+00 .IHWF—03 .800F+00 .107F-01 .QOOE+00 .9525-04 .400F+00 .9636-04 .933F+00 .86lF-UQ .500E+00 .93BE-04 .433E*00 .871F-04 .107F+01 .80?F-04 .600F+00 .859E-04 .467F+00 .802F—04 .120E+01 .730E-04 .700E+0” .818E-04 .500E+00 .739F-04 .133E+01 .6116-04 .800E+00 .772E-04 .533F+00 C701F-04 0‘47F*Ol QQQHE’OQ .900E+00 o77lE—04 .567E+nn ofiSIF-Oh .160F+01 .QQHE-OQ .IUOE+01 .7795-04 .600F+00 .601F-04 .171F+Ol .4347-04 .110E*0) .7115-04 .633F+00 .5456-04 .187E+01 .AAOE—Oh .IZOE+0] .647F-04 .700E+00 .505E-04 .200F+nl .QQSF-Oa .1105001 .5945-04 .800F+00 .4975-04 .211F+HI .170F-04 .140E*01 .542E-04 .900E+00 .459F-04 .??7F+”1 .3555-04 .150E+01 .487E-04 .IOOF+01 04?2F“a4 024(1E"01 01)!)F’0’4 0160E*nl 04‘37E*04 ollnE*n] .387F-04 .267F+01 .707E-04 .170F+01 .419E-04 .1?flF+01 .368F-Oh .PQ7E+U] .PRPE-04 oiKOE+01 .401E-04 .130E*01 “34017—04 .110F+(H 0. 11. .373E-04 .140F+OI .306F-04 .357F+01 0. 0. .34lE-04 .150F+01 .P76F-04 .107F+01 n. 0. .332E-04 .160F+OI O. 0. 0. 0. .307E-04 .I77F+0) 0. 0. no 00 o?§9F’04 .200F+01 ”o 00 “a 00 .2295-04 0773F*01 O. 0. 0. 00 02495’04 o?47F‘0‘ 0. 0. 0. 0. .173E-04 .?70F+0] HFACTI0N OF NA+AN- + H20 TN THF H?n=.0165M H?0=.0325M H?0=.0399M .220F-01 n. ,weqF-nl 0. .5145'03 0. .IREF-01 .3111-01 .PISE-ni .QHQF-O? .742E-03 .391F-n? .Ian—01 .flk/F-“I .PnnE-n1 .QVRE-OP .193E-03 .7RIF-n? .HQQF-Oa .1001+nn .lauF-u) .147E-01 .164E-03 .IlTF-nl .7SRF-Oh .1??F+OU .lfihE-O3 .196F-01 .140E-03 .156F-OI .YHIF-Oa .167F+OU .IlRF-03 .244F-01 .1765-03 .19SF-OI .h??F-Oh .PO0F+O“ .107E-01 .293E-OI .114E-03 .2346-01 .SGIF-na .267F+00 .QRPF-na .342E-01 .InSE-n3 .773F-nl .460F-0a .367F+nn .73IF-na .538F-01 .0936-04 .315E-nl .379F-04 .467E+00 .6I7F-na .734E-OI .RISE-04 .430F-nl .RAWF-Oa .867F+00 .SQPE-na .929E-01 .6956—04 .6?SF-OI .316F-04 .667F+OU .4786-04 .IIPEtnn .GSQE-O4 .RPOE-Ol .?RhF-ba .767F+0n .471F-04 .1325000 .806E-04 .102F*on .Ph7F-na .R67F+OH .4436-04 .152E+OO .4875-04 .121F+00 .ZPTF-Oa .967F+00 .3745—04 .I71E+00 .401E-04 .141F+nn 126 AP- CONC. IIHF(SEC.) AU- CONC. TIME(SEC.) AR- CONC. IIMF(SFC.) .194F-04 .107F+01 .179F-04 .191E*nO .AIQE-04 .160F+nn .2]?F-04 .II7F+nl .3669-04 .210F+00 .3656-04 .IROF+00 ,IQBE-Oa .12/F+01 .2995-04 .259E+00 .1IOE-04 .IQQE+OD .155E-04 .137F+01 .270E-04 .3OBE+00 .332E-04 .219F+00 .IARFy04 .147F+01 .243E-04 .357E+00 O. 0. olg6E-04 o157F+01 o?14E’04 o406E*00 Go Go .132F—04 .167E+Ol .194E-04 .455E000 0. 0. .1236-04 .177E+OI .IQOE-fla .503F+On 0. n. .lllF-Oa .187F+Ol .172E-04 .SS3E+00 0. 0. .131F-nh o197E201 o146E’04 o601E*00 0. 0o REACTION.0F NA+AN- + ETOH IN DME FTOH=.OO?63M ETOH=.00091M ETOH=.00134M .670F-01 0. .6706-03 U. .400E-03 .333F+00 .5865-03 .200F+UO .5266-03 .120E+Ol .353E-03 .133F+Ol .SOQF-03 .400F+UU .ARiF-Oi .2406+01 .3115-03 .P33E+01 .440E-01 .600F+00 .447E-03 .360E+01 .276E-03 .333E+01 .3R3E~03 .800F+00 .392E-01 .599E+OI .24SE-03 .433F+OI .334E—03 .IonF+HI .346F-0? .839E+0) .2185-03 .533E+01 .PQAF-OB .120E+01 .106F-01 .IORF+OP .197E-03 .633F+01 .264F-01 .140F+O] .PYPF-03 .13?E+0? .1776-03 .713E+OI .2386-03 .160F+01 .PhlE-OW .156E+0? .158E-03 .R33F+01 .216F-01 .190F+01 .215E-03 .lHOE+OP .144E-03 .933F+OI .IQQF—03 .2001+01 .IRSE-OR .2?OF+OP .131E-03 .103F+0? .183F-03 .??0F+n) .ISSE-03 .260E+02 .1172-03 .1135+n2 .ITIF-O3 .P4OF+OI .131F-01 .300E¢O? .108F-03 .171F+0? .ISQF-03 .P60F+01 .IlOE-O? .34OF+O? .9656-04 .I33E+0? .153F-03 .?80F+UI .QOOE-OA .BROE+0? .869F-04 .141F+n? .IhaF-n3 .300€+OI .790E-04 .AIBE+0? .778E-04 .153F+0? .134F-01 .32OE+01 .660F-“4 .458F+02 .602E-04 .163F+0? .125F-O? .3hnF+01 0. 0. .6166-04 .I73F+fl? .llRF-0? .180E+OI 0. 0. .664E-04 .IR3F+0? .IOAF-03 .420E+01 0. O. .6??F-04 .193F+n? .976F—04 .46HF+01 0. 0. .SOQE-O4 .701F+0? .906F-04 .500F+01 0. 0. H. 0. .843F-04 .540F+01 0. 0. n. 0. .798F-04 .580F+01 H. u. n. n. o727E’04 .620E+0] no no no no obgaF-04 .660F+01 no 0o no no .696F-04 .700F+01 0. 0. 0. n. .S7nfio04 .800E+01 n. 0. n. n. .498F-04 .90flF+01 ‘0. 0. 0- ”° .QVPF-OQ .100F*02 0. 0o 0. 0. .387F-04 .IIOE+0? 0. O. 0. 0. o3425'04 o1205+02 0. no no no REACTION 0F H20=.fl)?4M AQ- CnNCo o660F-01 .627F-01 .SQQF-O? .5656-01 .SBRE-01 .51?F-03 .485F-03 .464F—03 .441F-01 .4?4€-03 .386F-03 .331F—03 .ZRQF-OW .259F-03 .204F-03 o11RF-03 .155F-03 .131F-03 .114F-01 oggqF-OA .ROPF-OA .671F-04 .SSWF-Oa .469F-04 .414F-04 o36‘F’OA 0. 0. 0. 0. FTUH= 6630;-04 .570F-04 .5]5F-04 .4595*04 .407F-04 .3fiflr—flh .5195-04 .20]c-qa .Pan—Oh .??hF-04 .1995—04 .]ROF-Oh 127 NA +AM- HPU=.0180M TIMF(SEC.) O. .2h7E+00 .533E+00 .800E+00 .lO7E+OI .133E+01 .160590] .213E+01 .267E+01 .320F+01 .373E+OI .427E+0) .4HOE+01 .547E+OI .627E+01 .707E+01 .800E+OI o920E*O) o107E*0? .120E+0? .140E+0? 0. O. 0. 0. O. 00 0. n. 0. + FTOH TN FTOH=.000575M TIMF(SFC.) AD- COMP. 0. .SPCF-nk .?01F+00 .482F-03 .400r+00 .4495-03 .600F+00 .4155-03 .800F+0H .3QIE-n3 .100F+nl .WSIF-n? .12“F+Ol .111E-0? .140F+01 .PQBE-UB .160F+01 .PAOE—01 .180F+01 .294E-03 .??nF+ul .1976-01 .?QOF+OI .1435-0? .3406+Ol .159E-01 .400F+nl .117F-03 .460F+0| .IIHF-03 .6?OF+OI .9706-04 .RROF+01 .790F-04 .640F+0] .fifiHE-OA .7OOE+OI .47HF-04 .760F+0) .190E~04 .R?0F+0) .2905-04 .RROF+0] 0. .940F+01 0. .100F+02 0. .106F+02 n. .II°F+O? 0. n. n. n. n. 0. H. n, n. HFACIION HF K+IFQ- .Onno7su 0. .144E-03 .976F—03 .124F-03 .IQRF-n? .IORE—DB .PQ1F-O? .QQQF-OQ .390F—U? .RAEE-Oa .QRQF-D? .760E-04 ,RRWF-O? .67QE-04 .apnr-n? .AUQF-na .791F-02 .SHHF—fih .874F-0? .606E-04 ,9761-02 .AARE-na .107F-OI .407E-04 fl. .195F-0? .3905-0? .SRSE-O? .781E-0? .976E-O? oll7E'Ol .1375-01 .lSfifi-O) .176E-01 .195F-0) o215€‘01 + H20 IN OMF H20=.0357M AP- .630E-03 .5595-01 o4QfiE-03 o4365'03 oWRSE‘Oj o3406’03 o3006'05 o?66E'03 o23SE‘03 oEOQE‘05 o190F'05 .171E-03 oISQE'03 oquE’03 .125F-03 .116F-03 oln7E-03 .IOEE-03 .QQIE-04 .9015-04 o827F’04 .752E-04 .695E-O4 .642E-04 .589E-04 .530E-04 .536F-04 o491F’04 o444E‘04 o395F’04 THF ETOHz .1675-03 .150E-03 .IBSE-03 o123E'03 ollgE’OB olnlE‘03 .925F-04 .845F-na o7q3E‘04 o7255'04 .683E-04 .616E-04 CUNCo TIMF(SFC.) 0. .133F+00 .667F+OO .100E+01 .133E+01 ol67F*01 .700F+nl .?33F+01 .P67F+nl .300F+0] .333F+01 .567F+fl] .400F+n) o433€*07 .467F+0) oQO0F*01 .513F+OI .567E*OI o600E+0) .633E+nl .667E+01 o700E+01 .733F+nl .767E+fil .ROOF+01 .R?3E+01 .867F101 .900F+01 .933F+Ol .067E+01 .0004191‘5 0. .IQRF-n? .390E-O? .RRSF-n7 .781F-n? .976F-0? .117F-0) .137F-0) .I76F-0) .195F-01 .215F-01 .16?F~04 .143F-04 .IWOF-Oa .117F-Oa .QRBF-OS o970E-0q .RQQF-OG o826E“OS .7R4F-05 .767F-05 0. 0o 0. O. 0. 0. 0. TIHE(§FC.) .)]7F-0) .127E—Ul o137E“0] .léfiF-Ol .156F-01 .1665-01 .176F-01 o18§E“01 oqug'OI .2055~01 0. no Go 0. O. 0. 0. no 128 AP- CONC. .WAHF-Oa .314F-04 o246E‘04 .265E-04 .P3OF-04 o210E’04 .195E-04 .1685-04 .1405-04 .116E-04 0o 0. n. 0. no 0. 0. 0. TIMF(SFC.) o234E‘01 .254F-01 .271E-Ol .295E-0) o312E‘01 o352E-O) o3515‘01 .371F~01 o390E-01 o4]0E’Ol H. H. 0. 0. 0. U. 0. 0. AQ- CONCo .5835-04 .542E-04 o502E’04 oAGEE-OA o432E'04 .QnfiE-OQ .3745-04 o54QE“04 .324E-04 .ZQQF-n4 o266E-04 .2516-04 o231E’04 o213E'04 .202E-04 o190E’04 o187E'04 .1795-04 IIMF(SFC.) .2346-01 .254F-01 o?73F’Ol .793F-01 .1IPF-Ol o51?E-nl .1516-01 .371F-n] .390F-01 .AIOF-nl .aZQF—nl .aaqF-nl .468E-0) .488F-0) .507F-0] .527F—01 .5465-01 .566F-01