C H - A C I DS A guide to all existing problems of CH-acidity with new experimental methods and data, including indirect electrochemical, kinetic and thermodynamic studies O. A. REUTOV I. P. BELETSKAYA and K. P. BUTIN Chemistry Department, Moscow University, Moscow, USSR Translation Editor: T. R. CROMPTON M.Sc, B.Sc, F.R.I.C., M.A. Chem. PERGAMON PRESS OXFORD · NEW YORK TORONTO · SYDNEY ■ PARIS · FRANKFURT U.K. Pergamon Press Ltd., Headington Hill Hall, Oxford 0X3 OBW, England U.S.A. Pergamon Press Ine -, Maxwell House, Fairview Park, Elmsford, New York 10523, U.S.A. CANADA Pergamon of Canada Ltd., 75 The East Mall, Toronto, Ontario, Canada AUSTRALIA Pergamon Press (Aust.) Pty. Ltd., 19a Boundary Street, RushcuttersBay, N.S.W. 2011, Australia FRANCE Pergamon Press SARL, 24 rue des Ecoles, 75240 Paris, Cedex 05, France FEDERAL REPUBLIC Pergamon Press GmbH, 6242 Kronberg-Taunus, OF GERMANY Pferdstrasse 1, Federal Republic of Germany Copyright © 1978 Pergamon Press Ltd. All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without permis­ sion in writing from the publisher. First edition 1978 British Library Cataloguing in Publication Data Reutov, O A CH-acids. 1. Acids, Organic I. Title II. Beletskaya, IP III. Butin, KP IV. Grib,AV 547 QD305.A2 77-30618 ISBN 0-08-021610-2 In order to make this volume available as economically and as rapidly as possible the author's typescript has been reproduced in its original form. This method unfortu­ nately has its typographical limitations but it is hoped that they in no way distract the reader. Printed in Great Britain by William Clowes & Sons Limited London, Beccles and Colchester Preface Almost all organic compounds are CH-acids because they contain C-H bonds the hydrogens of which are capable of being substituted by metals resulting in carbon-metal bonds. In other words, organometallic compounds are salts of CH-acids and the properties of CH-acids belong in the domain of organo­ metallic chemistry. The authors have been working in this field for many years and this book is indeed addressed to organometallic and organic chemists. This does not necessarily mean that the authors wish to single out a particular audience; in fact, their aim is to explain their special inter­ est in, for instance, proton transfer stereochemistry or structure vs. CH- -acidity patterns. On the other hand, although there is a very direct relation between organometallic reactivity and CH-acidity, this problem is not treated in this book since this aspect of the problem of reactivity will already be well known to organometallic chemists. The present book deals with the acidity proper, not with carbanion chemistry nor organometallic chemistry. Not all aspects of carbanion chemistry or organometallic chemistry pertain to CH-acidity, but both of the problems have much in common. CH-acidity characterises thermodynamic stability of carbanions in a medium containing a proton donor as carbanion acceptor. Many other acceptors may be employed for the purpose, e.g. Hg+, Ag+, RHg+, and other,' cations. How­ ever, proton donors are the most convenient to deal with since, firstly carbanion affinity for the proton was the easiest to determine experiment­ ally and, secondly, the data obtained are comparable with a wealth of data available on OH- and NH-acidities and, thus, the stability of carbanions may be collated with the stability of 0-, H-, and other anions in various solvents and in the gas phase. A problem currently being worked on is the experimental determination of acidity in the gas phase, which would make it possible to construe a quantitative intrinsic acidity scale not influenced by solvent. The first results of this work have already been obtained using the ion cyclotron resonance method (to be discussed in Chapters I and II). The acidity series in the gas phase has been found to be quite different from that obtained in solution. Acidity in solution is of special importance because organic chemistry is, in the main, the chemistry of solutions. Dimethyl sulphoxide, liquid ammonia, cyclohexylamine, and some other solvents are excellent for a study of CH-acidity and it is in these solvents that most data on equilibrium and kinetical acidities has been obtained. However, pKa values found in cyclohexylamine correspond to ion-paired rather than dissociative acidity; 0-bases in dimethylsulphoxide are ion-pairs of alkali cations and this also affects pKa's of CH-acids measured in a dimethylsulphoxide/O-based system. vii Vili Preface On the whole, the effects of ion-pair formation upon equilibrium CH- -acidity have not been exhaustively studied and, in this direction, there is still much to be done, especially on the theoretical level. The first two chapters of the book discuss equilibrium CH-acidity. Chapter I describes methods to study equilibrium acidity and the pKa values of numerous CH-acids are tabulated. The determination of the acidity of weak acids such as CH-acids lead to difficult experimental problems which are concerned with a large pKa scale depending on CH-acid structures. An important task, therefore, is a search for new direct or indirect approaches to determining relative acidities. Recently, a great amount of published work has appeared in this field and this is discussed in Chapter I. It should be noted that for the most part the methods available may be applied only in a rather narrow pKa interval whereas the problem of creating a general-purpose method capable of giving the pKa of any acids with a wider pKa interval still awaits solution. Chapter II discusses equilibrium acidity as a function of CH-acid structures. In Chapter III, the hydrogen isotope exchange in CH-acids, i.e. the so-called kinetic CH-acidity is discussed. Isotope exchange is discussed as a method of finding the relative reactivity of C-H bonds in a series of CH-acids studied in one and the same solvent base system. Chapter IV is devoted to the sterochemistry of proton transfer in CH-acids. Proton transfer from tetrahedral carbon atom allows one to understand better the nature of transition states and the role played by ion-pairing and solvation. Stereochemical problems are specific to CH-acids; they do not arise when studying acids of other types and in this aspect there is a sig­ nificant difference between CH-acids and OH- or NH-acids. The stereo­ chemical aspect is an essentially new aspect contributed by CH-acids in a study of acidity on the whole. Equilibrium vs. kinetic acidity is discussed in Chapter V. The authors believe that the BrBnsted equation is a fundamental law but there are numerous factors obscuring experimental observation of the action of this law. These factors are discussed extensively. Anomalous Brönsted slopes observed in some cases are explained by effects of the medium indicating again the importance of studies of rates and equilibria in the gas phase. We are grateful to A.V.Grib (Cand. Chem. SC, NMR Laboratory, Department of Chemistry, Moscow University) who has translated the Russian manuscript into English. Chapter I Equilibrium Acidity of CH-Acids I. INTRODUCTION Almost any organic compound can, when acted upon by a base (a proton acceptor) of appropriate strength, ionise in solution to give carbanions, that is, negatively charged species whose charge is totally or (more often) partially localised on one of the carbon atoms. This property of CH-bonds gives rise to a great variety of organic reactions in which proton abstraction is a limiting or a pre-equilibrium step. Some examples are carbonyl-methylene condensations, trans-metalation reactions, the allyl rearrangement, cyclisation/recyclisation rearrangements, etc. Shatenshteiir was the first to consider hydrocarbons in terms of the general theory of acids and bases. The recent decade has brought a number of surveys of CH-acidity2»5 , among which the monograph by Cram2 may be mentioned. It is proposed to use the term "acidity" instead of the alternative "acid ionisation constant", in order to emphasise the relation between the equilibrium (thermodynamical) acidity (acid ionisation constant, K ) a RH + B~ -jl^R" + HB pKa · -log kl k-l k-l and the kinetic acidity (k^), the rate of proton abstraction from the acid molecule. The equilibrium and kinetic acidities, i.e. the quantities pK and a logk^, often vary in parallel. To begin with, consider equilibrium CH-acidity. It does not depend on a proton abstraction mechanism and is the best characteristic of the thermodynamical stability of carbanions in a given system. In solution, the energy of any species participating in the acid-base equilibrium should be corrected for the solvation energy. It is advantageous therefore to consider first acid-base equilibria in the gas phase as this reflects the "intrinsic" CH-acidity and is not affected by solvation. Following this, this Chapter will deal with acid-base equilibria in solution. The acidity of any acid in the gas phase may be written as6 X"(gas) + H+(gas) ^=^T HX(gas) + Π where Π is affinity for proton. The Π value is the enthalpy of proton addition to the anion and may be 1 2 CH-acids represented in the following form, Π= "ΔΗΗΧ + ΛΗΧ- + AV where ΔΗ^χ, ΔΗ-, and ΔΗ^+ are the enthalpies of formation of HX, X~ and Hf χ respectively. The heat of proton formation, -ΔΗ^+, is the heat of the following reaction. (1/2)H2 «■ ^ H*(gas) + e~(367 kcal gram-ion1" in vacuo). ΔΗ^χ can be found from experiments directly, while the heat of formation of anions in the gas phase may be obtained by, e.g., the Yatsimirskii? method that assumes that the energy of a salt crystal lattice is the enthalpy of transformation of the solid salt to the ion gas consisting of the same ions UKA - "AHL + ΔΗΚ+ + AHA" where "ΔΗ^, -ΔΗ^+, and ~ΔΗ?~ are the heats of formation of the crystalline salt, gaseous cation and gaseous anion, respectively. The crystal lattice energy is given by the Fayance equation. "UKA * V + V -L S01V· where H and H^ are the heats of ion hydration and L i is the heat of K so v solution of the salt at the ion strength of zero. The data on the heats of formation and on crystal lattice energies allow one to calculate the energy of formation of ions in vacuo and the proton affinities of the anion, the latter not usually being obtainable by direct experiment. The proton affinity may be found through the following thermodynamical sequence 3,8. HX ► X* + H· (DH) X· + e ►X" (-EA) H· + e *-H+ (IP) Sum: HX A" + H+; - Π =DH-EA+IP Equilibrium Acidity of CH-Acids 3 where DH is the X-H bond energy, EA is the affinity for electron of the radical X and IP the ionisation potential of hydrogen. Recently, Brauman^ and other workers9 using the ion cyclotron resonance technique have determined the affinity for the electron for a number of radicals and employed the thermodynamical sequence described above for calculating the values for a number of element hydrides. These data, together with the calculations performed using the method of Yatsimirskii7 are listed in Table 1. TABLE _1 Proton Affinities for some Anions at Medium Vacuum 6"9 Π kcal/mo le Π kcal/mole Anion Anion refs 6, 7 refs 8, 9 refs 6, 7 refs 8, 9 CH - _ _ HS" 343 350 3 NH - 419 407 CH" 325 333 2 OH" 383 390 NO5 320 - F" 363 370 Br" 315 324 PH2' - 364 I 307 314 RCOO" about 350 - HSO4 296 - CN" 348 — CIO4- 285 - The data in Table 1 suggests that the acidity of hydrides in vacuo should rise in going from the left to the right in the Periodic System and downwards in each of the Groups. The data obtained may be summarised to give the series in which Brönsted acids are arranged in the order of the decrease in acidity at medium vacuum at 300°K, 9b . n-C H SH > CH3NO2 > cyclo-C5H > CHCI3 > CH3OOCH3 > CH3CN > CH Cl2, 4 9 6 2 CH3SOCH3 * C H6, tert-C4H90H, ÌSO-C4H9OH > C2H5OH > CH3OH > C3H4, 2 C H CH(CH ) > C H CH > C H > H 0 > C H > H > NH3 > C2H4, C Hi 6 5 3 2 6 5 3 3 6 2 6 6 2 6 2 > (CH)3> CH 2 4 This series is in dramatic contrast with the usual acidity concept based on studying aqueous and alcohol solutions of acids-. Thus the alcohol acidity series in the gas phase ». t-BuOH > i-PrOH > EtOH > MeOH > H 0 2 is the reverse of the series obtained in hydroxyl-containing solvents, H 0 > MeOH > EtOH > i-PrOH > t-BuOH 2 Consequently, the methyl group effect in the gas phase disagrees with the conventional (+I)-pattern. Brauman and Blair introduced an electrostatic model to explain the methyl-induced stabilisation of alkoxide ions. The model includes the interaction of charge with polarisable alkyl groups and predicts that an increase in the size of R in the group R0~ should lead to an increase in the R polarisability and to a decrease in the anion potential 4 CH-Acids energy ("inner solvation"). The R polarisability effect is not observed in hydroxyl-containing solvents because the energy of solvation ("outer solvation") of the negative oxygen with hydrogen bonds is markedly higher than is the energy of interaction of the ion spearhead with the dipole induced by it. The weakest acid in the series discussed above is methane. The acidity rises, however, when one of the hydrogens is replaced by a group that may stabilise the carbanion. Thus the introduction into methane of an electron withdrawing chlorine atom may place the CH-bond value beneath the ammonia - Π value, CH3CI > NH3 > CH4, while toluene, whose phenyl group stabilises the negative charge via a conjugation mechanism, is a stronger acid than is water ^ . Accordingly, eumene in the gas phase is stronger than methane, i.e., alkyl groups do not obey the (+l)-pattern assumed for them conventionally. The energies in Table 1 and the qualitative series given above, when collated, lead to the following acidity series in the gas phase. HBr > RSH, H S> HCN > CHCI > CH3CN > HF, CH CI , CH3SOCH3 > alcohols 2 3 2 2 > C H CH > H 0 > C H > NH3 > CH4 6 5 3 2 6 6 The position of hydrogen fluoride in the series is due to the fact that fluorine ion in Et4NF.2H 0 can split a proton from, e.g., acetonitrilel3. 2 CH3CN + Et4N+F"z^ET N+~CH CN + HF 4 2 The position of hydrogen fluoride agrees with the results published recently by Mclver and Miller158 whose data allow one to arrange alcohols, acetylenes, and hydrofluoric acid in the following series in the gas phase at 298°K. (CH3)3CCH OH > HF > t-BuOH > i-PrOH > t-BuC Ξ CH > PrC Ξ CH > EtOH 2 > CH3C = CH > MeOH The series demonstrates that no distinction exists between CH-acids and acids of other types in the gas phase and all the types are arranged randomly. The same conclusion was reached by Ritchie and Kingl^ who calculated the potential energy surfaces for simpler reactions involving the hydride ion attack along the X-H bond axis. X — H < - - H" O=H,F,0H,NH2,CH3). Water, ammonia, and methane were found to behave similarly, except that the energy differences between the reactants and the proton transfer products were different. Consequently, hydrocarbons and their derivatives are not less efficient at displaying their acid properties than are hydrides and substituted hydrides of other elements. In this respect acetic acid may be regarded as an acetylated oxygen hydride (water). In hydroxyl-containing solvents, however, the acid properties of hydrides and substituted hydrides of electronegative elements(e.g., OH-acids) are Equilibrium Acidity of CH-Acids 5 much more pronounced than are CH-acidities. Hydroxyl-containing solvents favour the ionisation of OH-acids because they can form hydrogen bonds with electronegative elements. The next Section will compare CH- and 0H- acidities in water. II. CH-ACIDITIES IN WATER. A COMPARISON WITH HYDROXY ACIDS Toluene in the gas phase is a markedly stronger acid than is water12. It is profitable to consider whether in a hydroxyl-containing solvent toluene is a stronger or weaker acid than water'5 . The acid-based equilibria in these systems may be written in the following form. H20.solv + B.solv - "" HB+.solv + OH'.solv PhCH .solv + B.solv ~^~*" HB+.solv + PhCHJ.solv 3 To calculate dissociation or ionisation constants in solution, it is necessary to include the solvation of all the equilibrated species, both on the right and on the left hand side of the equations. When the proton acceptor, B, is one and the same and both reactions are made in the same solvent, then the calculation of relative acidities of water and toluene will deal only with the toluene and water solvation energies (on the left) and the hydroxyl and benzyl anion solvation energies (on the right). The difference between solvation of water and toluene in hydroxyl-containing solvents is a result of the fact that the solvation is mainly due to hydrogen bonding. Eigen has grouped the following series as a measure of the ability to form hydrogen bonds16, OH... 0 > OH... N, NH ....0 > NH N > SH X, XH S > PH....X, XH ...P > CH X, XH C where X is any element. It is clear from this tabulation that OH- and CH-acids lie at the opposite ends of the series; the former acids are the strongest, the latter ones the weakest hydrogen bond donors. Consequently, the water energy decreases more than does the toluene energy, on going from the gas phase to a hydroxyl- containing solvent such as water. Similarly, 0-bases are the strongest while C-bases are the weakest hydrogen bond acceptors, so the hydroxyl ion energy decreases more than does the benzyl anion energy on going from the gas phase to aqueous solutions. Probably, the main effect on relative acidities of water and toluene in water is the anion solvation difference. Parker1? who studied reactions and equilibria in which anions participated showed that the solvent-induced increment in the anion solvations provides the most important contribution in the energy. The hydroxide ion, a small ion with a localised charge, is solvated with water very effectively whereas the benzyl anion whose charge is delocalised would have solvated less had it not been protonated instantaneously. This agrees with the hard and soft acids and bases principle introduced by Pearson18. As a result, the difference between the solvation energies of hydroxyl ion and water exceeds noticeably the difference between the solvation energies of toluene and the benzyl anion. In other words, a hydroxyl-containing solvent favours the H2°—* OH"" transformation much more than it does so with PhCH3 -*PhCH2. Therefore, the equilibrium in the water phase as compared with the gas phase is shifted to the right for water much more significantly than it is for toluene.(Table 2). 6 CH-Acids TABLE 2 Water and Toluene Acidities as a Function of Solvation in changing from the Gas to the Liquid Water Phase"""* ^ Extent of solvation of an Total acidity increment ion or a molecule H 0 OH" 2 high high very high PhCH3 (PhCHj small* poor (poor) * That is why the anion is "instantaneously" protonated in water. As a consequence of this, solvation effects in the aqueous phase appreciably increases the acidity of water, a weak acid in the gas phase, whereas toluene does not reveal its acid properties, that is, it remains as weak an acid as it was in the gas phase. It can be concluded that the benzyl anion is a "strong base" although "strong base in water" would be more to the point since it has already been seen that both bases are strong in the gas phase. This example shows why CH-acids in water are markedly weaker than are OH-acids. This is illustrated in Table 3 which compares CH- and OH-acidities for similar structures. In water, the acidities of the CH-acids are, on the average, 20 pKa units lower than those of the respective OH-acids. TABLE 3 Substituent effect on acidities of related OH- and CH- acids; pKa--logKa; ApKa's correspond to the differences between pKa*s of CH- and OH-acids of similar structure; All pKa?s (except for methane) are based on the equilibria H-OR + H2O OR- + H 0+ 3 and H-CH2R + H20 CH2R" + H30+ Acid pKa ApKa Reference H-OH 15.7 >25 H-CH2H >40* 2 H-OCOCH3 4.7 19 ca.16 H-CH2COCH3 ca.21 2 H-OCN 3.7 } ca.20 20 H-CH2CN ca.24 2 H-ONO2 ca -7 21 } ca.17 H-CH2N02 10.2 22 * See section III of this Chapter. However, there do exist CH-acids which are strong even in water. Table 4 lists pKa values for some CH-acids measured in water; much more extensive data on the subject have been published by Ebel23.