2 Physical Methods Part (iii) High-pressure Chemistry By N. S. ISAACS Department of Chemistry University of Reading Whiteknights Reading Berkshire RG6 2AH With some 60 groups carrying out aspects of chemistry at pressures in the kbar range this topic must be regarded as one of major importance. Two principal objectives may be sought in carrying out reactions under pressure in solution. Rates may be measured as a function of pressure and the volume of activation AV* obtained according to equation (1) (AV* is defined as the difference in partial molar volumes of the transition state and reagents).* This is most usefully combined with a measurement of the overall volume change for the reaction to give a ‘volume profile’ which is proving to be a unique probe into the nature of the transition state.A similar relationship holds for equilibria equation (2) where AV is the volume change for the reaction. Pressure will therefore favour a process that involves a diminution in volume and conversely. The other aspect of interest is to employ this principle for preparative chemistry in order to bring about reactions that have a large negative value of AV* but which will not occur at 1 bar. Much activity around the world is evident on both fronts. -RT[d In k/dp]p+o = AV* (1) -RT[d In K/dp]p+O= AV (2) The subject has been thoroughly reviewed recently in two general review articles and data compilation^^.^ in which some 1500 volumes of activation and volumes of reaction of both organic and inorganic reactions are collected.Pressure effects on enzymic reactions,’ on displacements from co-ordination com- pound~,~*’ and upon isotope effects’ are among more specialized articles. The 50th Commemorative Volume of Reviews of Physical Chemistry of Japan is devoted to review articles on Modern Aspects of Physical Chemistry at High Pressure; topics * The units of molar volume throughout this article are cm’ mot-’. * ‘High Pressure Chemistry’ ed. H. Kelm Reidel Amsterdam 1978. ’ N. S.Isaacs ‘Liquid Phase High Pressure Chemistry’ Wiley Chichester 1981. T. Asano and W. J. le Noble Chem. Rev. 1978,78,407. W.J . le Noble and H. Kelm Angew. Chem. Znt. Ed. Engl. 1980 19,841. E. Morild ‘Theory of Pressure Effects on Enzymes’ Adv. Protein Chem. 1981,34,93.G. A.Lawrance and D. R. Stranks Acc. Chem. Res. 1979,12,403. T.W.Swaddle Inorg. Chem. 1980,14 3203. N.S.Isaacs ‘Effect of Pressure on Kinetic Isotope Effects’ in ‘Isotopes in Organic Chemistry’ Vol. 6 in press. 29 N. S. Isaacs of interest to organic chemists treated here include pressure effects on conforma-ti~n,'~ micelle f~rmation,~' volume profiles," kinetic and thermodynamic parameter^,'^ electro~triction,~' inorganic reactions," radical and biophysical chemistry.'" The proceedings of the 7th AIRAPT conference" contains much information concerning current interests. N.m.r. techniques and results at high pressure have also been reviewed." There are no physical measurements which it seems cannot be carried out under high pressure.However the problem remains of interpreting high-pressure data and extracting the mechanistic information which is latent in values of AV' or AV and their dependence on temperature solvent or other variable. Values of AV* and A are often partitioned into two components A V,,the intrinsic volume change i.e. change in van der Waals volumes of reagents in conversion into products or transition states and A V, the electrostrictive volume change comprising the change in volume of the solvation sphere. Ionogenic reactions are expected to show a large negative value of A V that may mask A Vi,while the latter may be the quantity most easily interpreted in terms of mechanism. Asano has derived an expression [equation (3)] whereby AV; may be obtained as the intercept of a linear plot,'* where k, k are rates at P and 1 bar respectively K is the compressibility of the solvent and B its constant in the Tait equation.As an instance the Menshutkin reaction between Et3N and EtI has AV* = -54 whereas the calculated AVi = -8.7 it follows that AV; = -45.3 consistent with an SN2displacement accom- panied by large charge separation. Analysis of actual molar volumes of alkanes in terms of three contributions from the molecular van der Waals volume V, the void volume V (space between molecules) and expansion volume V, has led to the conclusion that it is the latter which accounts for the increase in volume on cyclization rather than void volume at the centre of the ring which is sometimes assumed to be the case.13 Association of two molecules with no change in charge is accompanied by a reduction in volume.The formation of Meisenheimer complexes between nitroaryl ethers and methoxide in methanol is no exception but the overall volume changes are much less negative as a result of delocalization of charge and relaxation of the tight solvation sphere around the methoxide ion.l4 In water the attachment of OH- is accompanied by a much larger release of electrostricted solvent. Hydrogen bonding causes only a small volume change for example association of phenol in carbon tetrachloride is accompanied by A e = -2.4,15 whereas charge-transfer interactions between tetracyanoethylene and arenes is much more variable -3 to (a)E. Whalley Review of Physical Chemistry of Japan 50th Commemorative Edition 1980 119; (b) H..W.Offen ibid. p. 97; (c)W. J. le Noble ibid. p. 207; (d)B. S. El'yanov and E. M. Vasilvitskaya ibid. p. 169; (e) S. D. Hamann ibid. p. 147; (f) R. Van Eldik and H. Kelm ibid. p. 185; (g) M. Zhulin ibid. p. 217; (h)K. Heremans ibid. p. 259. lo B. Vodar and P. Marteau Proc. AIRAPT Conf. Le Creusot 1979. A. E. Merbach and H. Vanni Helv. Chim. Acra 1977.60.1124. T.Asano Rev. Phys. Chem. Jpn. 1979 49 109. l3 T. Asano Rev. Phys. Chem. Jpn. 1979,49 56. M. Sasaki N. Takisawa F. Amita and J. Osugi J. Am. Chem. SOC.,1980 102,7268. Is C. Josefiak and G. M. Schneider J. Phys. Chem. 1980,84 3004. Physical Methods and Techniques -Part (iii) High Pressure Chemistry 31 -15 if reliance is placed upon the Benesi-Hildebrand method for evaluating equilibrium constants for association.l6 Ion pairing of small highly charged ions is well known to bring about a large increase in volume owing to a reduction in solvation but the association of large delocalized ions such as (1)and (2) in water at least shows a small reduction indicating little change in the solvation pattern.17 Release of the solvation shell around Group IA cations is evident when they are complexed by cryptands and related polydentate ligands. The better the fit between ion and complexing agent the more positive the value of AV." [2,2,2]-Cryptand for example shows a maximum value of A V for complexation of K' and Rb' (+16) and a minimum for Li' (+3). There is much interest in complexation by cyclodextrins as the pressure effect of the equilibrium formation of inclusion complexes with some fluorescent indicators has been considered to model the behaviour of ligand-protein complexes1g thereby permitting a deeper understanding of these processes.Many proton-transfer reactions have been examined under high pressure both as equilibria and as rate processes. The latter are frequently fast enough to require the use of stopped-flow P-or T-jump techniques. Further work on reaction (4) NMe, / 02N~CH,N02 (4) -+ NH=C \ NMe, / 02N0CHN02 H2&C \NMe, -originally studied by Caldin,20 has been reported. Volumes of activation have been determined for both proton- and deuteron-transfer by stopped-flow analysis in CH2C1 as AV*(H) = AV*(D) = -15 but in toluene AV*(H) = -18 AV*(D) = -24.21*22 The isotopic effect has been shown' to be associated with reactions in which hydrogen transfer partakes of quantum mechanical tunnelling and is equivalent to a kinetic isotope effect which is diminished at high pressure.In this l6 T. Nakayama Rev. Phys. Chem. Jpn. 1979,49,25. R.K.Williams J. Phys. Chem. 1981,85 1795. N. Morel-Desrosiers and J.-P. Morel J. Am. Chem. SOC. 1981 103 4743. l9 P. M. Torgerson H. G. Drickamer and G. Weber Biochemistry 1979.18 3079. 'O C.D. Hubbard C. J. Wilson and E. F. Caldin J. Am. Chem. SOC. 1976,98 1870. M. Sasaki N. Sugimoto and J. Osugi Chem Lett. 1980 887. 22 N. Sugimoto M. Sasaki and J. Osugi submitted to Bull. Chem. SOC. Jpn.; personal communication. 32 N.S. Isaacs case the effect amounts to kH/kD=11.9 (1 bar) 9.3 (1 kbar) and is similar to effects previously Pressure can affect conformational equilibria favouring the conformer of lowest volume.9a Recently a Raman investigation on conformational equilibria in 1-bromoalkanes has shown that the gauche conformer at C-1-C-2 is lower in volume than the trans by 0.6-1.3°h.24 This is true also of 1,1,2-trichloroethane,2s The conformational preference about the N-CO bond of a cyclic amide depends upon the ring size above 15-membered rings the trans-geometry is favoured by pressure whereas below the cis-isomer is preferred.26 The volume of activation for conformational rotation in dimethylacetamide has been obtained from the pressure dependence of the ‘H n.m.r.signals of the CH groups. In many aprotic solvents AV* == 10 whereas in water the value becomes ~0.~’ The explanation tendered seems to imply that specific solvation of the acetamide occurs in the non-protic solvents requiring an increase in volume during rotation. Water however forms an open hydration shell within which the acetamide may rotate and undergo conformational change with little or no increase in volume. It seems possible that some sort of compensating effect is also operating. However in water A V*rapidly increases upon the addition of small quantities of ionic solutes or urea known to have drastic effects upon solvent structuring which are held to confirm this analysis. N.m.r has also been used to probe the rotational volume requirements of the ribose-base bond of a nucleoside using (3) as a model.’’ The small value of AV* in water is taken as evidence that rotation occurs in a number of small discrete steps.H OVNV0 (3) cis-Azocompounds can be generated by photolysis under pressure and their thermal isomerization to the trans- compounds studied kinetically. Mechanisms suggested are either by inversion (both aromatic rings remaining parallel) or by rotation (in which they become orthogonal). Isomerization of the conjugated and dipolar 4-amino-4’-nitroazobenzenes has been used as a probe into the mechanism (Scheme 1). Inversion should require only a small change in volume as the dipole remains intact throughout but rotation by decoupling the aryl rings would result in a loss of the dipole in the transition state and should result in a larger positive value of AV*.The experimental value obtained is very solvent dependent 0 in hexane but -22 in benzene suggesting a change of mechanism from inversion to 23 N. S. Isaacs K. Javaid and E. Rannala J. Chem. SOC.,Perkin Trans. 2 1978.709. 24 D. J. Gardiner R. W. Jackson and B. P. Straughan J. Chem. SOC.,Chem. Commun. 1981 159. ’’Ref. 2,p. 174. 26 A. Sera H. Yariada and H. Masaki Chem. Lett. 1980 1533. 27 G. Volkel J. H.iuer and H.-D. Ludemann,Angew. Chem. In?. Ed. Engf. 1980 19,945. 28 G.Klimke J. Hauer H.-D. Ludemann and W. Pfleiderer,J. Chem. Res. (S),1981,80. Physical Methods and Techniques -Part (iii) High Pressure Chemistry NR2 transition state Scheme 1 rotation respecti~ely.*~*~~ A 3,3'-bridged azobenzene which cannot undergo inversion for steric reasons has AV' = -22 and the rotational mechanism is inferred.30 Cycloadditions are invariably associated with large negative volumes of reaction and activation and have been extensively in~estigated.~ Pressure effects on several new types have been recently reported.Addition to a pyridinium betaine (4) resembles in its volume profile a Diels-Alder reaction (AV' and AV both -30 to -40),despite the loss of the dipolar character of the betai~~e.~' Ar / 00-7ZOOEt Et02CJ-J EtOH N I AV' = -36 Ar AV = -37 (4) Scheme 2 Intramolecular Diels-Alder reaction of (5) shown in Scheme 3 is found to have A V*= A which is typical of many intramolecular examples although slightly less 29 T.Asano. J. Am. Chem. SOC.,1980,102,1205 T. Asano T. Okada S. Shinkai K. Shigematsu Y.Kusano and 0.Manabe J. Am. Chem. Soc. 1981 103,5161. 31 N. S. Isaacs and P. Van der Beeke J. Chem. SOC.Perkin Trans. 2 in press. N. S. Isaacs negative (as is also the entropy of a~tivation).~~ This may be held to support the view that the association of two molecules brings about a volume reduction of ca. -10 and the formation of a covalent bond a further reduction of the same magnitude. Only the former quantity is absent from the intramolecular reaction. Ph' (5) AV' = -25 AV = -25 Scheme 3 The ene reaction has also received attention and the reaction of (6) with (7) exemplifies the existence of a generally tight product-like transition An exception to this pattern appears to be the dimerization of 2,3-dimethylbuta-1,3-diene (8).34In this case the volume of activation is considerably less negative than the volume of reaction suggesting a transition state which does not lie close to products.The entropy of activation is correspondingly less negative than usual ( -28 cal K-' mol-' compared to -41 for isoprene dimerization). It is suggested that in this case a 2-step diradical mechanism may be operating or at least is in competition with the normal concerted process. This seems the more likely since this diene is known to have an energy barrier to the cis-coplanar conformation required for reaction. Considerable interest is being shown in the use of high pressures for synthetic purposes taking advantage of the large negative activation volumes of cycloaddi-tion reactions.In such work the highest pressures attainable are used frequently in the 10-15 kbar range. Some examples of recent successes include the single-step syntheses of thia- and selena-fulvalenes (Scheme 4),35*36 synthesis of cantharadin (9) (Scheme 5),37 and synthesis of cyclobutanones (Scheme 6).38 32 N. S. Isaacs and P. Van der Beeke Tetrahedron Lett. in press. 33 M. Papadopoulos and G. Jenner Tetrahedron Lett. 1981,22,2773. 34 G. Jenner and J. Rirnrnelin Tetrahedron Lett. 1980,21,3039. 35 Y.Okarnoto and P. S. Wojciechowski J. Chem. SOC.,Chem. Commun. 1981,669. 36 J. E.Rice and Y. Okarnoto J. Org. Chem. 1981,46,446.37 W. G.Dauben C. R. Kessel and K. H. Takernura J. Am. Chem. SOC., 1980,102,6893. 38 N.S.Isaacs and P. Van der Beeke unpublished work. Physical Methods and Techniques -Part (iii) High Pressure Chemistry 35 -RCECR' + 2CX R = COOEt,R' = H R X X R (9) cis and rrans Scheme 4 Scheme 5 Ph,C=C=O 30 "C + 0-ph-m 0 Scheme 6 The use of high pressure techniques in the synthesis of heterocycles has been reviewed. 39 The complex addition-elimination reaction between dibutylamine and benzo- quinone (Scheme 7) is extremely pressure-dependent and evidently proceeds via a dipolar intermediate which may be reached by an electron tran~fer.~' .yBu2 '-# H -'.-.* 0 0 I ;HBu, Bu,N bNBU2 0 0-AV*(M~CN) = -67 Reagents i HNBu,; ii p-benzoquinone Scheme 7 Pressure studies of additions of methanol to nitriles forming iminoethers have also been reported and are moderately favoured by pressure.41 Furan will undergo '' K.Matsumoto T. Uchida and M. Acheson Heterocycles 1981 16 1367. M. Sasaki M. Bando Y. Inagaki F. Amita and J. Osugi J. Chem. SOC. Chem. Commun. 1981 725. 41 H. Inoue Rev. Phys. Chem. Jpn. 1978,48 105; ibid. 1979,49,95. 36 N. S. Isaacs a Friedel-Crafts reaction with a,&unsaturated carbonyl compounds at 100 “Cand atmospheric pressure with very poor yields. At 3 kbar 75% of 3-(2-furyl)pro- panoate may be obtained from the addition of a~rylate.~~ On the other hand dissociative processes will often be in evidence by their positive volumes of activa- tion for example the retro-aldol reaction shown in Scheme 8.43When catalysed by OH- AVS = +6.9,whereas catalysis by ethylamine is accelerated by pressure.These figures point to rate-determining steps which are different in type. Similarly the loss of nitrogen from a diazonium ion is accompanied by an increase in volume of +8 to +10 cm3 mol-’ a value in accordance with the formation of an aryl cation and dinitrogen as initial prod~cts.~~*~~ 2CH3COCH3 CH2COCH2 I ,C-CH3 OH H3C ‘CH2-c-/ \ CH3 NR2 Scheme 8 Solvolytic reactions have received as might be expected a great deal of attention by workers studying high-pressure reactions and dozens of activation volumes have been recorded. Unfortunately the results are not always straightforward since in most cases both Vi and V make major contributions to the experimental value and little attempt has been made to disentangle these effects.Seldom can one use an experimental volume of activation as a criterion of mechanism in this field at the present although undoubtedly valuable information is there. Values of A V* are usually negative but systems undergoing sN1 and sN2 mechanisms tend to be indistinguishable e.g. solvolyses of Me-OTos and But-Br in aqueous acetone have AV* = -17 and -24 respectively. Some additional examples do little to clarify the It is shown however that such data may be extremely temperature dependent; the acetolysis of 2-aryl-2-propyl tosylate (presumably by an ionization mechanism) has a maximum negative volume of activation at 74 0C.26The difficulties experienced in this field seem on the whole to point to the inability to interpret complex solvation effects.Line broadening at high pressure of the ‘H n.m.r. spectrum of the cation (lo) which undergoes rapid degenerate rearrangement (Scheme 9) yields an activation volume A V* = 8.7 in agreement with the charge dispersal in the transition 42 G. Jenner J. Rimmelin F. Antoni S. Libs and E. Schleiffer Bull. SOC.Chem. Fr. 11 1981,65. 43 F.Gr~nland and B. Anderson in ref. 10 807. 44 T.Kuokkanen Finn. Chem. Lert. 1980 189. 45 T.Kuokkanen Finn. Chem. Lett. 1980 192. 46 0.C. Kwun J. R. Kim and J. C. Ryu Taehan Hwahakhoe Chi 1981,25,152. 47 V.Nunes and M. Calado Rev. Port. Quim. 1979 21 139. 48 J. Hwang S.Yoh and J. Jee Taehan Hwahakhoe Chi 1980,24,150. 49 W.J. le Noble S. Bitterman P. Staub F. R. Meyer and A. E. Merbach J. Org. Chem. 1979,44,3263. Physical Methods and Techniques -Part (iii) High Pressure Chemistry 37 Me0 OMe Me0 4-(10) Scheme 9 Valuable information concerning 1-and 2-bond fission of peroxides azo-compounds and other radical precursors has been published by Neuman," in which the solvent-cage effects are considered particularly important. An examin- ation of pressure effects on rates and product yields from the decomposition of azocumene (11)gives evidence that the unstable 'semibenzene' products (12)result from cage recombination of geminate radical pairs rather than by diffusion apart followed by attack on a molecule of the azo-compo~nd.~~ AV' = +5 ArCMe,N=NCMe,Ar -ArdMe + N, 11 AVt = +6 ArCMe H E.s.r.of radicals has been successfully adapted to high-pressure work and the volumes of activation for radical reactions can now be studied directly. The radical dissociations in equations (5) and (6) have been studied in this way.52 The less- positive value for the phosphorus radical is taken as indicating an earlier transition state for bond fission than that of iminyl radical since the latter reaction is much more endothermic. The kinetic isotope effect of the internal hydrogen-atom transfer of the radical (13) has been measured by e.s.r. at high pressure. This reaction has a very curved Arrhenius plot and shows characteristics of a tunnelling reaction.It is of interest therefore that this reaction is a further example of one whose isotope effect diminishes with pressure or volumes of activation are i~otope-dependent.~~ BU'*C=& + BU" + BU'CN AV' = +3 (5) Bu'O+(OEt) + But' + O=P(OEt) AV* = +0.2 (6) CL3 L "P2Me *"'VC\ I ,Me B~tQ~'Me AV'(H) = +5.3 Me + \ AV*(D) = -1.2 \ Bu' But (13) L = H or D " R. C. Neuman Acc. Chem. Res. 1972,5,381. '' R. C. Neuman and M. J. Amrich J. Org. Chem. 1980,454629. 52 P. R. Marriott and D. Griller J. Am. Chem. Soc. 1981,103 1521. 38 N. S. Isaacs Radical recombination reactions are normally associated with a decrease in volume. It is surprising to learn therefore that several radicals generated by a laser flash apparently undergo recombination as judged by spectrophotometric observations at rates which diminished with pressure.53 In some instances both dissociation and recombination processes were studied and both assigned values of A V’ of the same sign.The meaning of these results is far from clear and indeed have been queried.54 There is considerable current interest in pressure effects upon reactions of biological although often their complexity renders interpretation somewhat difficult. Detailed study of reactions catalysed by isolated enzymes with simple substrates offers a better chance of elucidating some details of the mechan- isms. For example hydrolysis of pentyl and heptyl benzoates by a-chymotrypsin has been studied under pressure.55 Michaelis-Menten kinetics were observed and the Michaelis constant Kappand the catalytic constants k,, were evaluated giving volumes of dissociation of the enzyme-substrate complex of -7 and volumes of activation for the rate-determining step of -20.The latter was consistent with that expected for formation of two hydrogen bonds from the substrate and the co- ordication of serine-195 to give the tetrahedral intermediate together with the breaking of hydrophobic interactions inferred to be involved in the deacylation step. A similar type of reaction is the hydrolysis-aminolysis of long-chain p-nitrophenyl esters which serves as a model system for enzymic proteolytic processes.56 Activation volumes for displacement of p-nitrophenoxide by four competing processes were evaluated separately [by buffer OH- RNH2 and (RHN2)2] i.e.nucleophilically assisted aminolysis. Also the effect of chain length of both amine and ester on AV’ values led to the conclusion that there is a contribution to the activation volume of about +1 cm3mol-’ for each pair of methylenes on the two chains in contact owing to hydrophobic interactions. Finally mention must be made of current interest in the pressure effects upon micelle formation of detergent molecules. The solubilization effects of these micro- emulsions stabilized by a combination of hydrophobic and polar forces make such systems of more than scientific interest. The critical micellar concentration (c.m.c) is an easily measured property representing the saturation concentration of the free solute and increases to a maximum with pressure usually in the 2-5 kbar range depending upon the compo~nd.~’ The method of measurement i.e.electrical conductivity or addition of naphthalene as a fluorescent probe may however lead to conflicting values.’’ Micellar stability is assessed from the volume and compressi- bility of the solution both above and below the c.m.c. It appears that micelles are more compressible than solutions containing dispersed molecule^.^^ Micellar size however is relatively unaffected by pressure. ” A. I. Yasmenko I. V. Khudyakov A. P. Darmanjan V. A. Kuzmin and S. Claesson Chem. Scr. 1981,49. 54 W. J. le Noble personal communication. ’’ Y. Taniguchi and K. Susuki Bull. Chem. SOC.Jpn. 1980,53 1709. E. Morild and G. Aksnes Acta. Chem. Scand Ser. A 1981,35 169. 57 Y. Taniguchi and K. Susuki Rev. Phys. Chem. Jpn. 1980,49,91. S. D. Hamann Rev. Phys. Chem. Jpn. 1978,48,60.