7 The Physical Chemistry of Protic Solvents By P. A. H. WYATT Department of Chemistry The University St. Andrews WITH the development of the newer techniques for fast reactions and structure determinations recent effort has naturally been directed principally towards the solution of local problems at the molecular level. Much assorted information has consequently accumulated rather loosely inter-related but all more or less relevant to a proper understanding of the properties of protic solvents. The present survey will be largely confined to acidic systems. Acidic Solvents.-Protic solvents still show extraordinary versatility as vehicles for chemical discovery. Fluorosulphuric acid and its mixtures with SbF and oxides of sulphur have recently yielded a crop of results comparable perhaps to that obtained with sulphuric acid in the early days of physical organic chemistry.Such solvents have figured prominently in Olah's series of papers' on 'stable carbonium ions' which has now passed Part 90 though not all carbonium ions require the new medium.2 The use of fluorosulphuric acid as a solvent was initiated by W0olfy3 who showed that the pure acid has a conductance between those of H,S04 and HF, and that SbF behaves as an acid which can be neutralised with the base KS0,F in a conductance titration. A useful review4 is pravided by Gillespie who has refined and extended the work con~iderably.~ Amongst the advantages of HS0,F over H2S04 as a solvent are its greater acidity its smaller self-dissociation and its much lower freezing point (- 88.98 "C us.10.37 "C for H2S04) and viscosity (1.56 us. 25.54 CP at 25 "C). The pure acid boils at 162.7 "C has a conductance of 1-085 x ohm-' cm-' and a density d425 of 1.726; it can also be handled in conventional glass apparatus. The low freezing point greatly extends the temperature range for liquid phase n.m.r. studies and proton exchange reactions can consequently be decelerated sufficiently for such isomeric structures as those ' G. A. Olah C. L. Jeuell and A. M. White J . Amer. Chem. SOC. 1969 91 3961. * N. C. Deno D. LaVietes J. Mockus and P. C. Scholl J . Amer. Chem. SOC. 1968,90, 6457. A. A. Woolf J . Chem. Soc. 1955 433. R. J. Gillespie Accounts Chem. Res. 1968 1 202. (a) R. J. Gillespie J. B. Milne and M. J. Morton Znorg. Chem. 1968 7 2221; ( 6 ) R.J. Gillespie K. Ouchi and G. P. Pez ibid. 1969 8 63; (c) R. J. Gillespie and G. P. Pez ibid. 1969 8 1229 1233 94 P. A . H . Wyatt of the protonated carboxylic acids to be distinguished : H \ 0 0 RC H (I) RC (11) 0 0 + / \ + / \ \ / / H' H' The two forms are present in approximately equal amounts in the case of formic acid at - 60 "C but acetic acid has only 3 % of form 11. (See Gillespie4 for refs.) The thio-analogues of the carboxylic acids have recently been studied by Olah, Ku and White who now even report a thio-analogue of protonated carbonic acid.6 Among the new cations described5".' are the square 142+ and Se42+ but many simple inorganic gases are either sparingly soluble (N2 02 Ne Xe H2, NF3 CO) or unprotonated (C02 SO2) although HS03F is strong enough as an acid solvent for 1,3,5-trinitrobenzene to be protonated ~ompletely.~' In a detailed study of the effects of several solutes upon the equilibria in HS03F-SbF5 solutions Commeyras and Olah' find evidence that water forms both H30+ and HsO2'.A measurement of the water activity here would help in deciding the constitution of other concentrated aqueous acid systems. In view of the popularity of the HS03F-SbFS system it is interesting that Gillespie Ouchi and Pez5' could find no stronger acid fluoride than SbF5, the dissociation constant of H[SbF,(S03F)] formed in solution from the latter being 3.7 x 10- mol kg- '. This figure makes the acid technically rather weak ; but the concentration of H2S03F+ in concentrated SbF5 solutions must still be very high thus accounting for the extremely high acidity of the medium.The acidity can be increased still further by the addition of SO3 forming eventually the fully ionised H[SbF2(S03F),]. Since the SbF,-HS03F system is therefore not unique Gillespie4 discourages the use of the colourful name 'magic acid' and revives 'superacid' to stand for any system stronger than H2S04 or HS03F themselves a practice now a d ~ p t e d . ~ In the light of all this a report that HS03F-SbF is markedly less acidic than HF-SbF is rather unexpected but Hogeven' proposes a mechanism involving a protonated keto-carbonium ion for a novel conversion of cyclohex-2-enone to 3-methylcyclopent-2-enone and the reaction undoubtedly goes much faster in the HF-SbF5 system. Related kinetic and other studies have appeared re~ently.~ Chlorosulphuric acid is also under investiga-tion" but seems unlikely to produce such spectacular results as the fluoro-compound.G. A. Olah A. T. Ku and A. M. White J . Org. Chem. 1969 30 1827. ' A. Commeyras and G. A. Olah J . Amer. Chem. SOC. 1969,91,2929. * H. Hogeveen Rec. Trao. chim. 1968,87 1295 1303. A. Diaz I. L. Reich and S. Winstein J. Amer. Chem. SOC. 1969,91 5637; P. C. Myhre and E. Evans ibid. 1969 91 5641 ; G. M. Kramer ibid. 1969 91 4819. l o E. A. Robinson and J. A. Ciruna Canad. J . Chem. 1968,46 3197; J. Heubel and M. Wartel Bull. SOC. chim. France 1968 4357 The Physical Chemistry of Protic Solvents 95 Interest has revived in solvent disulphuric acid H2S207. The new work' l a confirms earlier conclusions' ' about the pattern of self-dissociation and the principal anion (HS30 o-) but reports smaller differences in the behaviour of metal cations.The whole H20-S03 system continues to perform general service for kinetic12 and otherI3 applications and it is perhaps rather surprising that there should still be any room for discussion about the constitution of pure sulphuric acid itself. Yet it is suggested14" that the currently accepted values146 of K,, the autoprotolysis constant should be increased by ca. 70 % some points of wider physico-chemical significance being involved. Essentially the argument is that if the forms of dissociation are effectively 2H2S04 = H3S04+ + HS04-2HzS04 = HjO+ + HS207-3H2S04 = HZS207 + HJOf + HS04-there is considerable scope for fitting freezing-point (and other) curves with three disposable equilibrium constants but that precise experimental determinations of the curvatures with composition of three independent freezing-point or say, acidity function curves at zero solute concentration should fix the extents of dissociation in the pure acid.At present the acidity function data are not suffi-ciently reliable near 100% H2S04 and the freezing-point data for the solutes H 2 0 and KHS04 were therefore supplemented by a recent e~fimate'~' of the dissociation constant of solute H2S2O7 to provide the third equation ; but only a drastic revision of the whole dissociation scheme would upset the conclusion that K, should be considerably higher probably near 3 x mo12 kg-2 at 10 "C and 4.6 x mo12 kg-2 at 25 "C.Any significant revision is now likely to come from oleum studies since the interpretation of dilute oleums is still not completely satisfactory e.g. a zero freezing-point depression for KHS04 in fairly dilute oleums seems to require higher polyacid ions than HS3010:, which is nevertheless said to be the principal ion at the H2S207 composition.' There are two interesting consequences to the higher K, values suggested (i) the HS04- concentration in pure H2S04 at 25" now agrees much better with ( a ) R. J. Gillespie and K. C. Malhotra J. Chem. SOC. (A) 1968 1933; (6) B. Dacre and P. A. H. Wyatt Trans. Faraday SOC. 1961 57 1958. l 2 C. W. F. Kort and H. Cerfontain Rec. Trav. chim. 1969 88 860; A. C. Hopkinson, J. Chem. SOC. (B) 1969 203; V. C.Armstrong and R. B. Moodie ibid. 1969 934; D. E. Leyden and J. F. Whidby J . Phys. Chem. 1969 73 3076; I. M. Medvetskaya, M. I. Vinnik and L. R. Andreeva Zhur. j z . Khim. 1969,43 2292; E. S. Mints E. L. Golod and L. I. Bagal Zhur. org. Khim. 1969 5 1203; 0. I. Kachurih and L. P. Mel'nikova Izvest. V. U. Z . Khim.i khim. Tekhnol. 1968 11 102 1 . J. S. W. Carrozza H. A. Garrera and A. J. Arvia Electrochim. Acta 1969 14 205; K. Stopperka and F. Kilz Z . Chem. 1968,8,435; M. Marcantonatos M. I. Bernardo, and D. Monnier Helv. Chim. Acta. 1969 52 291 ; F. S. Dainton and C. Gopinathan, Trans. Faraday SOC. 1969 65 143 1 5 1 . I 4 (a) P. A. H. Wyatt Trans. Faraday SOC. 1969 65 585; (6) R. J. Gillespie and E. A. Robinson 'Nonaqueous Solvent Systems,' ed. T. C. Waddington Academic Press, London and New York 1965 chap.4; (c) G. A. Mountford and P. A. H. Wyatt, Truns. Faraday SOC. 1966,62 3201 96 P. A . H . Wyatt that estimated from Ho data;I5 (ii) the necessity is eliminated for the special “asymmetric dissociation” effect which formerly explained the conductance data so we11,16“ leaving a theory without any illustrative cases either amongst proton-transfer or electron-transfer reactions.’ 6b It is interesting however to notice the family likeness to the recent paper of Onsager and Provencher,” which also describes a fast-reaction effect upon conductance but this time working so as to reduce the asymmetry of the ionic atmosphere and hence the relaxation-time effect by causing greater dissociation (of ion pairs) ahead of the moving ion and less behind it.Papers have also appeared on selenic,’ phosphoric,’ di-n-butylphosphoric,20 formic,2’ and trifluoroacetic22 acids and though no new solvent systems may arise from it the preparation of pure crystalline HMn04 has general chemical interest.23 A report24 on the stability of HC104-perchlorate mixtures may also have direct practical value. Acidity Functions and Related Data.-It is now widely known that indicator acidity scales in concentrated solutions depend very markedly upon the chemical nature of the series of compounds used in setting them up the scales e.g. for amides of azulenes differing from the scale for the substituted anilines largely used by Hammett and Deyrup in constructing the original H o scale. In using a chain of similar indicators of progressively lower basic strength it is assumed that the thermodynamic solvation properties of the base and conjugate acid forms are very similar within the family of compounds and quite independent of the enormous changes in base strength along the series.Where hydrogen-bonding, with its acid-base features is involved in the solvation process it is perhaps not obvious that the chain principle should work satisfactorily over a wide range of acidity. It is therefore reassuring to find that some confirmation is available from thermodynamic measurements. Arnett and Burke found a straight-line relation-ship between the pK values of a series of substituted anilines and their enthalpies of solution in the same concentrated H2S04 solution25 and Boyd’s activity coefficient measurements26a for the different classes of compounds used in indicator work (continued now with other solutes26b) correlated well with the differences in the corresponding acidity functions.l 5 L. P. Hammett and A. J. Deyrup J . Amer. Chem. SOC. 1932 54,2721 ; J. C. D. Brand, A. W. P. Jarvie and W. C. Horning J. Chem. SOC. 1959 3844. l 6 ( a ) P. A. H. Wyatt Trans. Faraduy SOC. 1961 57 773; (6) P. A. H. Wyatt Abstracts, Primera Conferencia Interamericana de Radioquimica Montevideo 1963 p. 355, (C.A. 1966,64 18510b.) L. Onsager and S. W. Provencher J . Amer. Chem. SOC. 1968,90 3134. l 8 M. M. Nour and S. Wasif J . Chem. SOC. ( A ) 1968 3024. l 9 K. Goto and D. Ishii J . Chem. SOC. Japan 1968 89 864. 2 o Z. Kolaiik J. Hejna and H. P. Pankova J .Inorg. Nuclear Chem. 1968 30 2795. 2 1 T. C. Wehman and A. I. Popov J. Phys. Chem. 1968 72 4031 ; J. Inorg. Nuclear 2 2 J. Bessiere Bull. SOC. chim. France 1969 3356. 2 3 N. A. Frigerio J. Amer. Chem. SOC. 1969 91 6200. 2 4 2. I. Grigorovich and V. Ya. Rosolovskii Zhur. neorg. Khim. 1969 14 353. 2 s E. M. Arnett and J. J. Burke J . Amer. Chttn. SOC. 1967 88 4308. 2 6 ( a ) R. H. Boyd J . Amer. Chem. SOC. 1963 85 1555; (6) R. J. Hirko and R. H. Boyd, Chem. 1969 31 2951. J. Phys. Chem. 1969 73 1990 The Physical Chemistry of Protic Solvents 97 Though new scales continue to appear,27 there are signs that the field is settling down to a period of consolidation and refinement. In this respect the two most important papers of 1969 are probably that of Johnson Katritzky and Shapiro28 on the temperature variation of H o in aqueous H2S04 and that of Reagan29 on the acidity function for carbon bases in aqueous H2S04 and HC104.Johnson et al. have paid special attention to checking the consistency of the slopes of their plots of log (indicator ratio) us. acid concentration at the region of overlap between pairs of indicators and they believe their new scale at 25 "C to be even more reliable than that of Jorgenson and Harrter,30 though the differences are not large as the following selected values of - H o show : Wt %H2S04 60 70 80 90 98 J and H scale 4.46 5-80 7-34 8.92 10.41 New scale 4.37 5.82 7.46 9-01 10.43 More significant however is the disagreement with the earlier on the temperature dependence of H . From measurements at 25,40,60,80 and 90 "C they now derive larger values of T - [dHo/d( l/T)] making differences of several kcal to activation energy corrections thus clarifying the interpretation of the kinetics of such reactions as the nitration of heteroaromatics and nitroanilines and acid-catalysed hydrogen exchange in concentrated H2S04 solutions.An equation of the form Ho( T ) = K / T + L was found to fit the Ho data to within 0.2 unit and the older HR data32 to within 0.35 unit HR being the acidity function for triphenylmethanol (secondary) bases. Comparison of the Ho and HR data over the temperature range covered by the latter (up to 45") also reveals a good fit with the equation H = mHo + constant proposed by Yates and M~Clelland.~~ The variation of the value of rn from 2.1 1 at 25 "C to 2-40 at 40 "C complicates present attempts to reassess some kinetic data but it might be worth using partial enthalpy data here to estimate the effect of the temperature coefficient of the extra water activity term contained in '' H.D. Zook W. L. Kelly and I. Y . Posey J . Org. Chem. 1968 33 3477; F. Terrier, Bull. SOC. chim. France 1969 1894; A. Collumeau ibid. 1968 5087; R. Reynaud, Compt. rend. 1968 267 C 989; B. Nahlovskf and V. Chvalovskf Coll. Czech. Chem. Comm. 1968,33 3122; A. P. Kreshkov L. N . Bykova and V. D. Ardashnikova Zhur. analit. Khim. 1969,24 1453; G . N. Novatskii B. I. Ionin L. I . Bagal and E. L. Golod, Zhur. Jiz. Khim. 1968 42 2966. 2 8 C. ID. Johnson A. R. Katritzky and S. A. Shapiro J . Amer. Chem. SOC. 1969 91, 6654.(See also P. Tickle A. G. Briggs and J. M. Wilson J . Chem. SOC. (B) 1970,65.) 2 9 M. T. Reagan J . Amer. Chem. SOC. 1969,91 5506. 30 M. J. Jorgenson and D. R. Harrter J . Amer. Chem. SOC. 1963 85 878. 3 1 A. 1. Gel'bshtein G. C. Shcheglova and M. I . Temkin Zhur. neorg. Khim. 1956 1 , " E. M. Arnett and R. D. Bushick J . Amer. Chem. SOC. 1964,86 1564. 3 3 K . Yates and R. A. McClelland J . Amer. Chem. SOC. 1967 89 2686. 282 506 98 P. A . H . Wyatt HR. In this paper2* Ho values are tabulated at intervals of 2 wt % in H2S04 from 2 to 98 % (plus 99 %) at 25,40,60 80 and 90 "C. Reaganz9 carefully analyses the differences between HR' (= HR - log[water activity]) and Hc his new scale for carbon acids (substituted azulenes 1,l-diarylethylenes and aromatic polyethers) and finds them consistent with Boyd's activity coefficient data.26a From the point of view of the overall picture however, the striking feature is the considerable similarity between the two Hc remains only 0.2 to 0.9 units less negative than Hk over the whole range 1-14~-H,sO,, whereas H o diverges progressively from both and becomes >4 units more positive at 14M.In early attempts34a to devise a successive hydration scheme for the proton which would fit the Ho data the latter scale always seemed some-what cramped in the sense that the equilibrium constant obtained for the first hydration of H 3 0 f at the high concentration end34e9b was too large for the first stage of any statistical scheme set up to fit the more dilute solutions although there was no difficulty in producing a very good fit over a considerable range.If the activity coefficient ratio for the acid and base species were more nearly constant over a wide range of acidity for HR' and Hc indicators than for Ho in-d i c a t o r ~ ~ ~ a simple hydration scheme may be more acceptable and successful if based on Reagan's new scale. It is doubtful however if much purpose would be served by the exercise at this stage in view of the accumulating evidence that changes in general environmental effects may sometimes be just as important in these systems as specific hydration through H - b o n d ~ . ~ ~ ~ ~ In papers on indicator studies for solution in concentrated alkalis however Yagi13'" discusses the application of hydration models to salt effects and gives references to earlier work.He regards the OH- ion as hydrated with three water molecules up to 8~-hydroxide and quotes H - data (for indole derivatives) up to 1 5 ~ for KOH, 1 6 ~ for NaOH and 4-5M for LiOH at 25 "C. (Allowance for ion-pairing will affect the significance of these data however.37b) Good water activity data will always be indispensible for any theoretical treatment of concentrated acid solutions. The extension of the data for aqueous HC104 by Wai and Y a t e ~ ~ ~ is therefore particularly welcome. Using a modified isopiestic technique they have covered the range W 7 5 wt % HClO by com-parison with H2S04 solutions already covered by Giauque et al.39 Up to the highest concentrations studied the rule holds that for a pair of isopiestic solutions, sulphuric acid is much more concentrated than perchloric acid i.e.HC104 is much more efficient for the depression of the water activity e.g. a water activity 3 4 P. A. H. Wyatt (a) Discuss. Faraday Soc. 1957 24 162; (h) Trans. Faraday SOC., 3 5 R. W. Taft jun. J . Amer. Chem. Soc. 1960 82 2965. 36 J. F. Bunnett and F. P. Olsen Canad. J . Chem. 1966 44 1899. " (a) G. Yagil J . Phys. Chem. 1967,71 1034 1045;(h) J. R. Jones Chem. Comm. 1968, 3 8 H. Wai and K. Yates Canad. J . Chem. 1969,47,2326. 39 W. F. Giauque E. W. Hornung J. E. Kunzler and T. R. Rubin J . Amer. Chem. SOC., 1960 56 490. 513. 1960 82 62 The Physical Chemistry of Protic Solvents 99 of 0-00209 found at 83.22% H2S04 (50.56m) is already reached by 74.28% HC104 (29.52m) thus continuing the trend previously observed.34a Whatever the interpretation (thermodynamic) solvation by H 2 0 is evidently larger for the perchlorate than the sulphate system.An interesting experimental fact is that no C1 was found in the isopiestic H2S04 solutions even at 74.28 % HC104, thereby refuting a former suggestion4' that the isopiestic method was restricted to HC104 concentrations < 72% through volatilization of the acid. A short review on the correlation of the composition of acid solutions with activity and acidity functions has also appeared.41 Ground State pK Values.-Aromatic and +unsaturated aliphatic aldehydes, ketones and carboxylic acids must presumably be subject to the hydration complications of the H o indicators; yet they seemed to form a fairly consistent group in H2S04 solution^^^"*^ and to fit the HA (amide) scale better than H o .In terms of the Bunnett and Olsen equation36 all these compounds were assigned 4 values in the amide range 0.474-57. To this group can now be added the sulph~xides,~~" with 4 in the range 0.4-0.6 and pK values ranging from - 1.8 for dimethyl sulphoxide to -2.9 for p-nitrophenylmethyl sulphoxide. It must be emphasized that the pK values assigned depend upon the form of extrapolation technique adopted and differ markedly (by > 2 units in the case of p-N02C6H5SOC6H5)43b from the H o of the solution in which the base is half-protonated. Greig and however, draw attention to the medium effects which cause serious shifts in the absorption spectra of carbonyl compounds and find that these compounds do not after all obey HA but require a function which changes more steeply with acid com-position.A conflict remains to be resolved therefore since Zalewski and Dunn mention the medium effects in their recent paper42b but believe them to be small in the region at which the p K measurements are made. Greig and Johnson44 also review the methods for obtaining pK values ; they favour the Yates-McClel-land method33 which gives similar results to that of Bunnett and 0 1 ~ e n ~ ~ but is easier to apply. 40 Y . L. Haldna I . A. Koppel and K. 1. Kuura Zhurfiz. Khim. 1968,40 1657. 41 C. J. O'Connor J . Chem. Educ. 1969,46 686. 4 2 (a) R. I. Zalewski and G. E. Dunn Canud. J . Chem. 1968,46 2469; (b) ibid. 1969 47, 2263. 43 (a) D. Landini G. Modena G.Scorrano and F. Taddei J . Amer. Chem. SOC. 1969, 91 6703; (6) N. C. Marziano G. Cimino U. Romano and R. C. Passerini Tetra-hedron Letters 1969 2833; P. 0. I . Virtanen and J. Korpela Suomen Kem. 1968, 41 B 326. 44 C. C. Greig and C. D. Johnson J . Amer. Chem. SOC. 1968,90,6453 100 P. A . H. Wyatt Other groups of compounds for which basic strengths have been further studied are alcohols and sulph~namides,~~" sulph~ximines,~~~ hydroxyph-t h a l a n ~ ~ ~ ~ a u l e n e s ~ ~ ~ alkylben~enes,~~' and pr~panediamines.~~~ Aqueous acid strengths are reported for o-substituted benzoic 1,2,3-tria~oles,~~' carboxylic and sulphonic acid amides and hydra~ides,~~' b a r b i t ~ r i c ~ ~ ~ and benzophenone-3,3',4,4-tetracarboxylic acids,46e silanol and 'squaric' acid ( 1,2-dihydroxycyclobutenedione).47 The latter has two dissociation con-stants pK = 1-2 & 0.2 and pK = 3.48 k 0.02 at 25 "C similar to those of oxalic acid.The difference in pK from that of oxalic acid (4.27) is not simply explicable in terms of the extra delocalisation energy of the squarate ion which would produce much too large an effect on its own ; this effect must be compensated by the loss of hydration stabilisation through delocalisation of the negative charge.47 A study has been made48 of the entropies of dissociation of seven moderately strong acids with pK values in the range 0.5-1.4 (i.e. similar to the pK value of squaric acid) mainly to serve as models for ASs from reaction rates. The temperature coefficients of pK were determined so that pK AH" and AS" could be derived at 25" for C12FCC02H F3CC02H C13CC02H +H3NS03-, (CO,H) F2CHC02H and C12CCHC02H.The values of - AS"/cal mol- ' K-' are 7,1,1.4,3.2,9 13 and 12 respectively showing that for the chosen acids in this region of pK the entropy of dissociation (in water) is considerably less negative than that of acetic acid (As0 = - 22 cal mol-' K - ') and may even approach zero. Since the AH" values are mostly small the entropy change is evidently an important factor in determining the strength of this group of acids relative to acetic acid in water at 25°C. (As0 itself is commonly sensitive to temperature however and a different picture could easily emerge at temperatures other than 25°C). The effects upon acid strength of various solvents49 and of adsorption on a micellar surface5' are also reported.4 5 J. Korpela and P. 0. I. Virtanen Acta Chem. Scand. 1968 22 2386; Suomen Kem., 1968 41 B 321 ; ibid. 1969 42 B 142; ( 6 ) S. Oae K. Tsujihara and N. Furukawa, Chem. and Znd. 1968 1569; (c) H. Glinka and A. Fabrycy Roczniki Chem. 1968, 42 1425; (d) H. D . Klotz H. Drost and W. Schulz Z . Naturforsch. 1968 23a 1690; ( e ) T. Rodima U. L. Khaldna and E. E-Yu. Var'end Reakts. spos. org. Soedinenii, 1968 5 466; (f) C. Tissier and P. Barillier Compt. rend. 1969 268 C 1953. 46 ( a ) K . Bowden and G. E. Manser Canad. J. Chem. 1968,46 2941 ; ( 6 ) L. D . Hansen, B. D. West E. J. Baca and C. L. Blank J . Amer. Chem. Soc. 1968 90 6588; (c) S. Kaae and A. Senning Acta Chem. Scand. 1968,22 2400; (d) A. G. Briggs J. E. Saw-bridge P.Tickle and J. M. Wilson J. Chem. SOC. (B) 1969 802; ( e ) G. G. Kryukova, Ya. 1. Tur'yan and A. V. Bondarenko Zhur. obshchei Khim. 1968 38 2177; (f) P. Schindler and H. R. Kamber Helv. Chim. Acta. 1968 51 1781. 4 7 D . J. MacDonald J . Org. Chem. 1968 33 4559. 4 8 J. L. Kunz and J. M. Farrar J . Amer. Chem. Soc. 1969,91,6057. 4 9 R. A. Robinson J . Chem. and Eng. Data 1969 14 247; K. Katoh Japan Analyst, 1968 17 1055; R. Reynaud Bull. SOC. chim. France 1969 699; V. I . Dulova N. V. Lichkova N. V. Arkhipova and Sh. R. Tillyashaikhova Izvest. V.U.Z. Khim. i khim. Tekhnol. 1968,11,867 ; R. Thuaire Compt. rend. 1968,267 C 993 ; V. I. Dulova, N. V. Lichkova and L. P. Ivleva Uspekhi Khim. 1968 37 1893. H. Komara Japan Analyst 1968,17 1147 The Physical Chemistry of Protic Solvents 101 Excited State pK,* Values.-Most investigations on the effects of excitation upon chemical equilibria have been carried out with water as solvent and detailed analysis has often emphasized the important role of the protic solvent in dis-sociative mechanisms both in the excited and in the ground-state proces~es.~ Several papers on excited states have appeared re~ently.~’-~~ Avigal Feitelson and Ottolenghi5’ investigated the quenching effects of a series of carboxylate ions upon the fluorescence of some phenol derivatives as a method for determining singlet pK,* values when the proton transfer from excited ROH* to water is too slow to compete with fluorescence.The quenching rate constants from Stern-Volmer plots satisfy the Brernsted general-base catalysis law and extrapolation to the base strength of water then leads to an estimate of the rate of the reaction between ROH* and H20 from which pK,* is derived by assuming the reverse reaction to have a rate constant of 5 x 10” 1 mo1-l s- ’.Though the authors regard this as a new method for determining pK,* the principle involved does not seem very different from that used by Weller6’ when he found that acridine could not be protonated sufficiently rapidly by H 3 0 + in the required region of pH. He then determined both the forward and the reverse rate constants for protonation with a different acid NH4+ and transposed to the H30+-Hz0 system by means of the pK value of NH,’. The success of the Brernsted general-base test in the latter work is related to the applicability of the Hammett ap treatment to the pK,* values which has been shownS3 to follow when there is a correlation between the shift in absorption spectrum upon protonation of a series of bases and their Hammett a values.Substituted 2- and 4-styrylpyridines show such a correlation ;54 some also reveal large increases in base strength upon excitation unlike pyridine itself (which becomes a slightly weaker base). Evidently the styryl part of the molecule has profound effects upon the changes in electron distribution near the nitrogen in the pyridine ring. The ability to form hydrogen bonds e.g. to the solvent makes an enormous difference to the rate at which proton transfer can occur. In the ground state a A. Weller ‘Progress in Reaction Kinetics,’ ed.G. Porter Pergamon Oxford 1961, vol. 1 p. 189. 5 2 L. Avigal J. Feitelson and M. Ottolenghi J . Chem. Phys. 1969 50 2514. 5 3 H. H. Jaffe H. L. Jones and M. Isaks J . Amer. Chem. SOC. 1964 86 2934. 5 4 J. C. Doty J. L. R. Williams and P. J. Grisdale Canad. J. Chem. 1969 47 2355. ” S. F. Mason J. Philip and B. E. Smith J. Chem. SOC. ( A ) 1968 3051. 5 6 S. F. Mason and B. E. Smith J. Chem. SOC. ( A ) 1969,325. 5 7 B. E. Smith J. Chem. SOC. ( A ) 1969 2673. 5 8 D. L. Horrocks J . Chem. Phys. 1969,50,4151. 5 9 A. C. Hopkinson and P. A. H. Wyatt J. Chem. SOC. (B) 1967 1333. 6 o (a) E. Vander Donckt and G . Porter Trans. Faraday SOC. 1968 64 3215; (6) E. L. Wehry and L. B. Rogers J. Amer. Chem. SOC. 1966,88 351. 6 1 R. C. Dhingra and J. A. Poole J . Phys.Chem. 1968,72 4577. 6 2 A. Gravowska and B. Pakula Photochem. and Photobiol. 1969 9 339. 6 3 K. Nakamaru S. Niizuma and M. Koizumi Bull. Chem. Soc. Japan 1969 42 255. 64 E. Vander Donckt and G . Porter Trans. Furaday SOC. 1968 64 3218. 6 5 J. Bertran 0. Chalvet and R. Daudel Theor. Chim. Acta 1969 14 1. 6 6 N . Tyutyulkov and G. Hiebaum Theor. Chim. Acta 1969 14 39. 6 7 A. Weller Z . Elektrochem. 1957 61 956 102 P. A . H . Wyatt proton attached to carbon ionizes very much more slowly than one attached to oxygen on nitrogen (see e.g. recent papers by Ritchie.68a A related report on negative Brnrnsted coefficients is also interesting.68b) In the excited singlet state, where reactions must occur appreciably within loF8 s if they are to compete with fluorescence the protonation of mono- and bi-cyclic hydrocarbons fails to occur at all even though thermodynamically fa~oured.’~ The absence of hydrogen isotopic exchange in IM-HClO shows that the radiative deactivation rate of an electronically excited aromatic hydrocarbon must exceed the rate of protonation by a factor of >lo’.Oddly enough no hydrogen exchange was detected for methoxybenzene either,’ although quenching occurs in aqueous solutions at acidities where the excited base is expected to protonate. An incipient proton transfer is therefore envisaged which in its early stages accelerates radiationless decay. (For a new self-quenching mechanism involving hydrogen-bonds see Horr~cks.’~) The work of Mason’s s~hool~’-’~ and of others5’ is pushing into regions of acidity where some further insight may be gained from these studies into the differences between the various acidity scales.A concentrated acid form ArC02H2+ which exists at much lower acidities in the excited state, enters into a reassessment of pK,* for the ordinary dissociation to form carboxy-late ion since Vander Donckt and Porter60a believe the presence of the excited acidium species to have invalidated former results.60b Unlike azulene itself the protonated forms of many more complicated azulenoid systems fluoresce normally from the first excited singlet.61 Data are a~cumulating~~ for the comparison of singlet and triplet pK,* values, and Bertan Chalvet and Daude16’ now conclude that the triplet pK,* value for derivatives of naphthalene quinoline etc.is expected to lie between those for the ground state and excited singlet for a P-donor a- or P-acceptor or a hetero-acceptor in the a position but not for an a-donor or a hetero-acceptor in the P position. Kinetic and Structural Studies.-‘Hydrogen-bonded Solvent system^,'^' pro-duced in honour of Professor Wynne-Jones brings the general background up to date and will be required reading for new recruits to the field. More recently on the kinetic side Caldin7O has reviewed the tunnel effect in proton transfer reactions in solution of which there are probably about a dozen instances now. Brickmann and Zimmermann7’ investigate further the theore-tical treatment of this phenomenon calculating the effects of changes in barrier height distance apart and energy difference between the two wells upon the mean ‘lingering time’ of a proton in one of the wells.The results are compared with those from Bell’s quasi-classical method which they resemble qualitatively. Gold72 reviews protolytic processes in H20-D20 mixtures in which connection 6 8 (a) C. D. Ritchie J . Amer. Chern. SOC. 1969,91 6749; (6) F. G. Bordwell W. J. Boyle 6 9 ‘Hydrogen-bonded Solvent Systems,’ ed. A. K. Covington and P. Jones Taylor and ’O E. F. Caldin Chern. Rev. 1969 69 135. 71 J. Brickmann and H. Zimmermann J . Chern. Phys. 1969 50 1608. ’ 2 V. Gold Adv. Phys. Org. Chern. 1969 7 259. jun. J. A. Hautala K. C. Yee ibid. 1969,91 4002. Francis London 1968 The Physical Chemistry of Protic Solvents 103 it is of interest that a revised zero-point energy c a l ~ u l a t i o n ~ ~ for the equilibrium H 2 0 + D20 e 2 H D O brings the theoretical estimate of the equilibrium constant (3-72) into agreement with experiment.A review by Parker,74 comparing protic with dipolar aprotic solvents as media for bimolecular reactions empha-sizes the service done by studies on dipolar aprotic solvents both in a practical way (acceleration of certain reactions by several powers of ten) and in con-tributing towards a better understanding of the dominating influence of local solvation over bulk dielectric constant effects in protic solvents themselves. A survey of the present state of knowledge about activity coefficients of charged and uncharged reactants and of transition states is included (see also ref. 75). It is generally accepted that the ionization of carboxylic and oxyacids proceeds through the initial formation of a hydrogen bond followed by some kind of push-pull mechanism ; which end is pushed and which pulled depends upon the relative acidic and basic strengths at the two ends of the chain (see e.g.Weller”). S~hrnid’~ describes the primary step in the ionisation of a weak acid the formation of the hydrogen-bond to the solvent as exothermic and the secondary step in which the proton transfer is completed as endothermic. He applies his analysis to the mutarotation of a-glucose. Further aspects of pr~ton-transfer~~ and reactions have been covered. and of the hydrated p r o t ~ n . ~ ~ - ~ ~ In further support of his theory of water structure (see ref. 69 p. 9) Walrafen now extends his Raman measurements on H 2 0 solutions in D20 down to 1 mole %79a and still finds a marked asymmetry in the OH stretching contour.An important feature in the argument is that the Several authors deal with various aspects of the structure of 73 J. R. Hulston J. Chem. Phys. 1969 50 1483. 7 4 A. J. Parker Chem. Rev. 1969 69 1 . ’’ P. Haberfield L. Clayman and J. S. Cooper J . Amer. Chem. SOC. 1969 91 787. 7 6 H. Schmid Monarsh. 1968 99 1932. 7 7 D-W. Fong and E. Grunwald J. Amer. Chem. SOC. 1969 91 2413; E. K. Ralph and E. Grunwald ibid. 1969 91 2422; T. H. Marshall and E. Grunwald ibid. 1969 91, 4541 ; D. B. Matthews Austral J. Chem. 1969 22 463; M.-L. Ahrens and G. Maass, Angew. Chem. 1968 80 848; M. C. Rose and J. Stuehr J. Amer. Chem. SOC. 1968, 90 7205; E.K. Ralph and E. Grunwald ibid. 1969,91,2426. ’8 R. P. Bell and P. De Maria J. Chem. SOC. (B) 1969 1057; S. Milstien and L. A. Cohen, J. Amer Chem. SOC. 1969 91 4585; R. C. Fahey and C. A. McPherson ibid. 1969, 91,3865; T. I. Crowell and M. G. Hankins J. Phys. Chem. 1969,73 1380. 79 (a) G. b. Walrafen J. Chem. Phys. 1969,50 560; ( 6 ) ibid. 1969 50 567; (c) J. Schiffer, ibid. 1969,50 566. 8o T. A. Ford and M. Falk Canad. J. Chem. 1968,46 3579. M. Alei jun. and A. E. Florin J. Phys. Chem. 1969 73 863. 8 2 K. Arakawa K. Sasaki and Y. Endo Bull. Chem. SOC. Japan 1969 42 2079; R. W. Bolander J. L. Kassner jun. and J. T. Zung J. Chem. Phys. 1969 50 4402; J. A. Horsley and W. H. Fink ibid. 1969 50 750; K. J. Miller S . R. Mielczarek and M. Krauss ibid. 1969 51 26; L.R. Painter R. D. Birkhoff and E. T. Arakawa ibid., 1969,51,243; D. Lewis and W. H. Hamill ibid. 1969,51,456; H. Taft and B. P. Dailey, ibid. 1969,51 1002; E. R. Lippincott R. R. Stromberg W. H. Grant and G. L. Cessac, Science 1969 164 1482. 83 A. R. Anway J. Chem. Phys. 1969 50 2012. 8 4 N. Salaj Acta Chem. Scand. 1969 23 1534; M. I. Emel’yanov E. A. Nikiforov and N. S. Kucheryavenko Zhur. strukt. Khim. 1968 9 954; I. Olovsson J. Chem. Phys., 1968,49,1063; A. F. Beecham A. C. Hurley M. F. Mackay V. W. Maslen and A. M. Mathieson ibid. 1968 49 3312 104 P. A . H. Wyatt shoulders on the Raman lines reported in recent work can be resolved into two components showing a good isosbestic point. A picture then emerges of a sharp classification of water species into those with a fully H-bonded Cz0 environment of unbroken (linear or nearly linear) hydrogen bonds and those having broken (or distorted) bonds.Schiffer,”‘ however defends the so-called ‘continuum’ model which Walrafen considers to be inconsistent with his Raman data.69*79a*b A further study of the OH and OD stretching vibrations in ice and water has again emphasized the broad distribution of intermolecular energies in the liquid phase.” Evidence for an equilibrium between monomeric and dimeric forms of water is found in water-ammonia solutions at 29.6 “C by ‘H and 1 7 0 n.m.r. experiments, different shifts being assigned to the two species,8 while under mass-spectro-metric conditions water may even form itself into independent chains or whiskers projecting from the field ionisation tip used in the production of hydrated protons in the gas phase.83 Further consideration is also given to the dissociation of water at high pressures and temperature^.^^ In a consideration of the structure of aqueous solutions the study of solid hydrates may be helpful in some cases;86 the variety of structures shown by alkylamine hydrates8’ probably has some relevance to the liquid phase.Specific solvation through hydrogen bonds will depend upon the factors affecting the strength of such bonds as revealed by theoretical,88 n.m.r.,89 i.r.,” and other’’ techniques ; but even a qualitative theoretical model of the structure of hydrogen-bonded solutions will of course require the correct information about the local structures likely to be present.A review by Tuck92 on HX2- and HXY- anions is relevant as also is the unexpected finding by Dewar’s that the forma-tion of an acidium ion from nitric and nitrous acids involves the further proto-nation of the OH group and not one of the free oxygen atoms. It is also of interest to consider whether the inclusion of the environment in the explicit way used by Douglas in the description of silicate systems94 could serve a purpose in hydrogen-bonded solutions a given experimentally recognisable equilibrium being regarded as a type of sum over a family of equations differing in the co-ordination details included. 8 5 W. B. Holzapfel J . Chem. Phys. 1969,50,4424; V. D. Perkovets and P. A. Kryukov, Izvest. sibirsk Otdel. Akad. Nauk Ser. khim. Nauk 1969 No. 3 9. 86 G.Brun Rev. Chim. minerale 1968,5,899; L. W. Reeves Progr. N.M.R. Spectroscopy, 1969 4 193. 8 7 G. A. Jeffrey Accounts. Chem. Res. 1969 2 344. J. N. Murrell Chem. in Britain 1969 5 107. 89 E. Grunwald R. L. Lipnick and E. K. Ralph J . Amer. Chem. SOC. 1969 91 4333; T. A. Wittstruck and J. F. Cronan J . Phys. Chem. 1969,72,4243. 9 0 ( a ) D . M. Mathews and R. W. Sheets J . Chem. Soc. ( A ) 1969,2203; ( b ) A. A. Lipovskii and T. A. Dem’yanova Zhur. priklad. Spektroskapii 1968 9 239. 9 1 H. Wolff and H. E. Hoppel Ber. Bunsengesellschaft Phys. Chem. 1968 72 701 ; J. B. F. N. Engberts Rec. Trav. chim. 1968 87 992; B. B. Bhowmik Indian J . Chem., 1969,7 788; M. Szafran Roczniki Chem. 1968 42 1469. 9 2 D . G. Tuck Progr. Inorg. Chem. 1968 9 161. 93 M. J. S. Dewar M.Schanshal and S. D . Worley J . Amer. Chem. Soc. 1969,91,3590. 9 4 R. W. Douglas Chem. in Britain 1969 8 349 The Physical Chemistry of Protic Solvents 105 Solvation of a more general kind is discussed in a further group of paper^.^'-''^ Erlander proposes a classification of salts into three groups from the point of view of their solubility in water according to the presence of a tightly bound water shell on neither only one or both ions involved and extends the discussion to protein^.^' On a more kinetic basis Safford et a1.96 regard ions as positive or negative 'hydrators' in terms of a lengthening or shortening of the characteristic lifetime exhibited in neutron scattering. On the thermodynamic side Arnett et ~ 1 . ~ ~ report heat capacities of solution of 21 low molecular weight alcohols in water and detect effects due to chain length branching unsaturation and ring closure.Boyd98 has estimated hydration enthalpies for several tetra-alkylam-monium ions from heats of solution and calculated lattice energies and finds them to show a reversal in the usual trend of decreasing hydration heat with increasing size ; the approach to non-electrolyte character in the large ions accounts for their high entropy loss on hydration. In mixtures of water with acetonitrile and ethylenediamine 23Na n.m.r. shows that Naf is preferentially solvated by water in the first case and by ethylenedia-mine in the second."' Dimethyl sulphoxide also seems to be very effective for solvation of cations compared with water."' Such information will help in understanding solute behaviour in mixtures of the two solvents.lo2 Various other aspects of solvation have been studied.lo3 Electrolytic Solutions.-Along classical lines Gardner Mitchell and Cobble1O4 emphasise the value of the 'third-law' approach in an important paper on the thermodynamics of aqueous H2S04 solutions.They point out that the Nernst equation is a necessary but not a sufficient test of the reversibility of a cell and show that the third-law treatment used together with the Nernst equation, provides a much more critical and rigorous test. In the application of the equa-tion 9 5 S. R. Erlander J . Mucromol. Sci. 1968 A2 623. 96 G . J . Safford P. S. Leung A. W. Naumann and P. C. Schaffer J . Chem. Phys. 1969, 9 7 E. M. Arnett W. B.Kover and J. V. Carter J . Amer. Chem. SOC. 1969 91 4028. 9 R R. H. Boyd J . Chem. Phys. 1969 51 1470. 9 9 D. F. C. Morris Structure and Bonding 1969,6 157; G. A. Krestov and V. K. Abrosi-mov Teor. i eksp. Khim. 1969 5 415; K. Uedaira and K. Uedaira Zhur. jiz. Khim., I968,42 3024. 50 4444. l o o E. G . Bloor and R. G. Kidd Canad. J . Chem. 1968,46 3425. ' O r .I. Courtot-Coupez A. Laouenan and M. Le Demezet Compt. rend. 1968,267 C 1475. J.-C. Halle R. Gaboriaud and R. Schaal Bull. SOC. chim. France 1969 1851. l o 3 J. H. Stern and J. M. Nobilione J . Phys. Chem. 1969,73,928; F. Lohmann Chein. Phys. Letters 1968 2 659; F. J. Millero C. Wu and L. G . Hepler J . Phys. Chem. 1969,73, 2453; S. K. Pal U. C. Bhattacharyya S. C. Lahiri and S. Aditya J. Indian Chem. SOC., 1969,46,497; R.N. Butler and M. C. R. Symons Chem. Comm. 1969,71; C. J. Clemett, J . Chem. SOC. ( A ) 1969,761 ; E. M. Verkhovskaya I. N. Shokin and A. G. Kuznetsova, Izvest. V. U.Z. Khim. i khim. Tekhnol. 1968 41 976; A. D'Aprano and R. M. Fuoss, J . Amer. Chem. SOC. 1969 91 21 1 . W. L. Gardner R . E. Mitchell and J. W. Cobble J . Phys. Chem. 1969,73 2021 106 P. A . H . Wyatt the E" scale is anchored at 25 "C with the reliable data of Covington Dobson, and Wynne-Jones,'05 and E" and AC," are tabulated at 5" intervals from 0 to 100 "C for the cell involving the reaction : 2Hg(1) + H2SO4(aq) = Hg2S04(c) + H2(g) and from 0 to 55 "C for the cell involving the reaction : H2S04(aq) + H2(g) + Pb02(c) = PbS04(c) + 2H20 Activity coefficients of H2S04 are listed (cJ also Duisman and Giauque'06).Interest has also revived in aqueous nitric acid'" and in extending the Debye-Hiickel theory,' O 8 and Covington has reviewed ion-selective electrodes.' O9 It is not yet clear what the potentialities will be of photoelectric emission'" in this field. What is clear however as a recent conference at Montpellier shows,' ' ' is that many are now exploring the approach to concentrated salt solutions through the molten state rather than through the properties of protic solvents. A. K. Covington J. V. Dobson and Lord Wynne-Jones Trans. Faraday SOC. 1965, 61 2050. I o 6 J. A. Duisman and W. F. Giauque J. Phys. Chem. 1968 72 562. lo' V. G. Karev G. T. Polovnikova and L. D. Savinskaya Zhur. neorg. Khim. 1969 14, 2281; M. A. Yakimov V. K. Filippov and Ty Ki Zhur. neorg. Khim. 1969 14 551; M. M. Karavaev and A. I. Bessmertnaya Khim. Prom. 1969 515; D. Richter and H. Ullmann Z. phys. Chem. (Leiprig) 1969 236 314; 0. Redlich R. W. Duerst and A. Merbach J. Chem. Phys. 1968 49 2986. E. Glueckauf Proc. Roy. SOC. 1969 A 310 449; A. W. Gardner and E. Glueckauf, ibid. 313 131 ; C. W. Outhwaite J. Chem. Phys. 1969 50 2277. A. K. Covington Chem. in Britain 1969 5 388. J . Chim. phys. Special issue October 1969 pp. 1-214. ' l o B. Baron P. Chartier P. Delahay and R. Lugo J. Chem. Phys. 1969 51 2562