摘要:

Chemical Society Reviews Vol 10 No 3 1981 Page Cyanoketenes:Synthesis and Cycloadditions 289By H. W. Moore and M. D. Gheorghiu Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications By J. R. Jones and S. E. Taylor 329 INGOLD LECTURE How does a Reaction ChooseIts Mechanism? By W. P. Jencks 345 Aryliodine (HI) Dicarboxylates By A. Varvoglis 377 The Royal Society of ChemistryLondon Chemical Society Reviews EDITORIAL BOARD Professor K. W. Bagnall, BSc., Ph.D., D.Sc., C.Chem., F.R.S.C. (Chairman) Professor K. R. Jennings, MA., D.Phil., C.Chem., F.R.S.C. Professor G. W. Kirby, M.A., Ph.D., Sc.D., F.R.S.E., C.Chem., F.R.S.C. Professor B. L. Shaw, B.Sc., Ph.D., F.R.S. Chemical Society Reviews appears quarterly and comprises approximately 20 articles (ca.500 pp) per annum. It is intended that each review article shall be of interest to chemists in general, and not merely to those with a specialist interest in the subject under review. The articles range over the whole of chemistry and its interfaces with other disciplines. Although the majority of articles are intended to be specially commissioned, the Society is always prepared to consider offers of articles for publication. In such cases a short synopsis, rather than the completed article, should be sub- mitted to The Managing Editor, Books and Reviews Section, The Royal Society of Chemistry, Burlington House, Piccadilly, London, W 1V OBN. Members of the Royal Society of Chemistry may subscribe to Chemical Society Reviews at E10.50 per annum; they should place their orders on their Annual Subscription renewal forms in the usual way.1981 Annual subscription price, U.K. E31.00, Rest of World E33.00, U.S.A. $78.00 including air speeded delivery. Application to mail at second class postage rate is pending at Jamaica, N.Y. 11431. Change of address and orders with payment in advance to The Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1HN England. Air freight and mailing in the U.S by Publications Expediting Tnc., 200 Meacham Avenue, EImont, New York 11003. All other despatches outside the U.K. by Bulk Airmail, and Accelerated Surface Post outside Europe. Note to subscribers. Regrettably publication of the four issues has still not reverted to the usual quarterly dates. The cause of this is a persisting shortage of articles (the production problems of recent years have been largely overcome) but the setting-up of an Editorial Board should result in an increase in the commissioning of re views. 0Copyright reserved by The Koyal Society of Chemistry 1981 ISSN 0306-4012 Published by The Royal Society of Chemistry, Burlington House, London, WIV OBN Printed in England by Eyre & Spottiswoode Ltd, Thanet Press, Margate.

ISSN:0306-0012    

DOI:10.1039/CS98110FP005

出版商:RSC    

年代:1981

数据来源: RSC

ISSN:0306-0012    

DOI:10.1039/CS98110FX009

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Chemical Society Reviews Vol 10 No 3 1981 Page Cyanoketenes:Synthesis and Cycloadditions 283By H. W. Moore and M. D. Gheorghiu Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications By J. R. Jones and S.E. Taylor 329 INGOLD LECTURE How Does a Reaction Choose Its Mechanism? By W. P. Jencks 345 Aryliodine(LU) Dicarboxylates By A. Varvoglis 377 The Royal Society of Chemistry London

ISSN:0306-0012    

DOI:10.1039/CS98110BX011

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Cyanoketenes:Synthesis and Cycloaddi tions By H. W. Moore DEPARTMENT OF CHEMISTRY, UNIVERSITY OF CALIFORNIA, IRVINE, CALIFORNIA 92717 and M.D.Gheorghiu LABORATORY OF ORGANIC CHEMISTRY, POLYTECHNIC INSTITUTE -TCH BUCHAREST, ROMANIA, 77206 1 Introduction In 1970it was reported that cyanoketenes could be conveniently prepared by the thermolysis of 2,5-diazid0-1,4-quinones.lSince that time agreat deal of work has appeared concerning the synthesis and chemistry of these unusual electron- deficient cumulenes. t-Butylcyanoketene (TBCK) has received the most attention since it provides several experimental advantages; it is unsymmetrical, stable to selfcondensation when kept in solution, and highly reactive towards a large number of keteneophiles. Thus, it is an ideal reagent for investigations of the mechanisms of 2 + 2 cycloaddition reactions as well as a starting material that can be used to prepare a variety of carbocyclic and heterocyclic compounds.To this end, TBCK and some of its analogues have been observed to react with alkenes, alkynes, allenes, ketenes, imidates, imines, aldehydes, isonitriles, amine oxides, azirines, sulphurdi-imides, and a number of heterocyclic substrates. The purpose of this article isto provide a review of cyanoketenes with a focus on their syntheses and cycloaddition reactions. 2 Synthesisof Cyanoketenes Cyanoketenes are but one class of compounds whose synthesis can be viewed as arising from zwitterionic intermediates formed in the thermolysis of appropriately substituted vinyl azides.The general rationale for their formation is defined as outlined in Scheme 1.2 Specifically, vinyl azides of structure (1) cleave to zwit- terions, (2),when X is a substituent capable of cation stabilization and Y and/or Z are anion stabilizing. The zwitterionic intermediate, (2),can then undergo ring closure to (3) or cleave to (4).Thus, a synthetic route to cyanoketenes can be envisaged for cases in which Y is a carbonyl substituent and X is an appropriate leaving group. A particularly suitable class of azides that one would predict to give cyano- ketenes by the above generalized mechanism is 2,5-diazido-l,4-benzoquinones. The thermal chemistry of such compounds has now been explored in moderate depth,2 and it has been found that they do give good yields of alkyl- and aryl- H.W.Moore andW.Weyler,J. Am. Chem. SOC.,1970,92,4132. H. W. Moore, Ace. Chem. Res., 1979,12,125; H. W. Moore, Chem. Soc. Rev., 1974,2,415. 289 Cyanoketenes: Synthesis and Cycloadditions J I z\ C=Y + x: NEC / (4) Scheme 1 cyanoketenes3as well as dicyanoketene.4 An interesting example in this series is 2,5-diazido-3,6-di-t-butyl-l,4-benzoquinonewhich gives a nearly quantitative yield of t-butylcyanoketene (TBCK) when decomposed in refluxing benzene. The mechanism for the generation of this and other cyanoketenes by this method is outlined opposite. Attempts to utilize the diazidoquinone method for the synthesis of halocyano- ketenes met with failure due to the insolubility of the diazidoquinone precursors.However, this problem was circumvented by utilizing 4-azid0-3-ha10-5-methoxy-(SH)-furan-Zones which were found readily to cleave in refluxing benzene to the corresponding halocyanoketene and methylformate.5-' This has now been employed for the synthesis of chloro-, bromo-, and iodo- as well as phenoxy- cyanoketene (see opposite page). With the exception of t-pentyl- and t-butylcyanoketene, the other cyano- ketenes reported readily undergo self-condensation and thus must be generated in situ for their reactions to be studied. This is often an advantage since, under these conditions, the ketene is slowly generated in the presence of the keteneophile. Thus, its concentration is minimized and product yields are often enhanced.The bulky examples mentioned above are stable for days in anhydrous aromatic solvents. However, these, as well as all of the other cyanoketenes studied, are very reactive as electrophilic species in their reactions with a large variety of keteneophiles. W. Weyler, W. G. Duncan, and H. W. Moore, J. Am. Chem. Suc., 1975,97, 6187. R. Neidlein and E. Bernhard, Angew. Chem., Znt. Ed. Engl., 1978, 17, 369. H. W. Moore, L. Hernandez, and A. Sing, J. Am. Chem. Suc., 1976,98, 3728. D. M. Kunert, R. Chambers, F. Mercer, L. Hernandez, and H. W. Moore, Tetrahedron Left. 1978,929. H. W. Moore, R. Czerniak, G. Hughes, C. C. Yu, F. Mercer, D. Goldish, B. Axon, un- published result. Moore and Gheorghiu 0 0 Me IR = Et-C-, Me&-, Me2CH-, Me, Ph-, or -C=NIMe X = C1, Br, I, or OPh A related route to cyanoketenes was reported for the synthesis of phenyl- cyanoketene.Treatment of 3-chloro-4-phenylcyclobutene-1,Zdione with azide ion in acetonitrile at ambient temperature or below gave carbon monoxide, nitrogen, and phenylcyanoketene.8 The synthetic scope of this reaction deserves more attention since it may provide an advantage over the azidoquinone or azidobutenolide methods. That is, the reactive cyanoketenes could conceivably be generated at low temperature. R.C. DeSelms, Tetrahedron Lett., 1969, 1179. Cyanoketenes: Synthesis and Cycloadditions c1 N,-MeCN, -20°C "')=C=O + CO + N2 Ph Ph 3 Cyanoketene-Alkene Cycloadditions Monosubstituted, vicinal disubsti tuted, and trisubsti tuted a1 kenes react with TBCK to give cyclobutanones, in a 2 + 2 cycloaddition.Geminal disub- stituted alkenes, on the other hand, sometimes also give 'ene' reaction products. The cycloaddition of TBCK to styrene has been studied in greatest de~th.~-ll The sole reaction product is the cyclobutanone (5) having a cis stereochemical relationship between the bulky t-butyl and phenyl substituents. Such a result is best rationalized in terms of a concerted 2na + 2ns reaction mode.12 This mechanism further demands preservation of the alkene stereochemistry in the cyclobutanone product. Such was observed for the TBCK cycloadditions to the Z-and E-isomers of monodeuteriostyrenesto give, respectively, thecyclobutanones (6) and (7).The configurations of these cyclobutanones were determined by But Ph NC---TBCK I__j H H P--H But Ph But Ph @ M. D. Gheorghiu, F. Kerek, and M. Avram, Rev. Roum. Chim., 1975, 20, 75. lo M. D. Gheorghiu, L. Piirvulescu, C. Drhghici, and M. Elian, Tetrahedron, 1981, 37, 143. l1 M. D. Gheorghiu, 0. Ciobanu, and M.Elian, J. Mugn. Reson., 1981, 44, 330. l2 R. B. Woodward and R. Hoffmann, 'The Conservation of Orbital Symmetry,' Academic Press, New York,1970. Moore and Gheorghiu theoretical and experimental n.m,r. analyses which included anisotropy effects, as well as solvent and lanthanide induced chemical shift ex~eriments.~-ll 1,2-Disubstituted alkenes react readily with TBCK to give cyclobutanones.Several of these cycloadditions related to the anticipated cis-stereochemical relationship between the bulky t-butyl group and the substitutent at position-3, as well as to the preservation of alkene stereochemistry in the product. For example, treatment of cis-and trans-cyclo-octene with TBCK gave, respectively, (8) and (9).13 Cyclohexene has also been observed to form cycloadducts with TBCK + NJIQ 0 H TBCK.0 But Me TBCK -I(:: TBCK to give (lOj.3 Although stereochemical evidence for (10) is lacking, one can reasonably assume a relationship analogous to that found in (8) and (9), i.e., t-butyl cis to the adjacent CH2. Analogously, cis-but-2-ene gave the cyclo- lS W. Weyler, L. B. Bird, M. C. Caserio, and H. W.Moore, J. Am. Chem. SOC.,1972,94, 1027. 293 Cyanoketenes: Synthesis and Cycloadditions butanone, (I 1).14 Contrathermodynamic cyclobutanones were also observed when the cyclopentene keteneophiles (12)--(20) were treated with TBCK.9J5-18 Me Me It is noteworthy that only cyclobutanones and no rearranged products were observed for the bicyclo[2.2.1 Iheptenes (1 3)-( 18), a result that is again consist- ent with a concerted mechanism for the cycloadditions. The geminal disubstituted alkene, 2-methyipropene9 gave the cyclobutanone (21) and the ‘ene’ product (22) when treated with TBCK.l5 Interestingly, no ‘ene’ l4 P. R. Brook, A. M. Eldeeb, K. Hunt, and W. S. McDonald, personal communication (1978). We thank Professor Brook for providing these results.l6 P. R. Brook and K. Hunt, J. Chem. SOC.,Chem. Commun., 1974,989. M. D. Gheorghiu, P. Filip, C. DrBghici, and L. Plrvulescu, J. Chem. SOC.,Chem. Commun., 1975, 635. M. D. Gheorghiu, C. DrBghici, and L. PPrvulescu, Tetrahedron, 1977, 33, 3295. l8 M. D. Gheorghiu, L. PQrvulescu, and C. Drhghici, Rev. Roum. Chim., 1979, 24, 1005. Moore and Gheorghiu product was observed in the analogous cycloaddition using the trisubstituted alkene, 2-methylbut-2-ene; only the cyclobutanone (23) was obtained.lg TBCK cycloaddition to 1,3,3-trimethylcyclopropene gave three products : the cyclobutanone (24; 28 %), the furan (25; 7%), and the 2:l adduct (26; 40%).2* A similar preference for rearranged products from these strained alkenes was observed when 1-methylcyclopropene was treated with TBCKZ1 Here, the products were (27) and (28), and no cyclobutanone was detected.L OBut CNT But It appears that no examples of the reactions of TBCK with tetrasubstituted alkenes have been reported. Only one brief report has appeared in the literature describing the cyclo- addition of TBCK to an acyclic conjugated diene.3 Treatment of the ketene with trans, trans hexa-2,4-diene resulted in > 80% yield of the cyclobutanone (30). P.R.Brook, A. M. Eldeeb, K. Hunt, and W. S. McDonald, J. Chem. SOC.,Chem. Commun., 1978, 10. ao D. H. Aue, D. F. Shellhamer, and G. S. Helwig, J. Chem. SOC.,Chem. Commun., 1975,603. D. H. Aue and G. S. Helwig, J. Chem. Soc., Chem, Commun., 1975, 604.Cyanoketenes:Synthesis and Cycloadditions Only tentative stereochemical assignments could be made on the basis of n.m.r. data obtained on (30) and the corresponding cyciobutanols formed upon borohydride reduction. However, it is noteworthy that these data are consistent with (30) rather than (31), and the latter would be expected if the cycloaddition TBCK+ were concerted. Thus, a stepwise mechanism involving a zwitterionic inter- mediate such as (29) may be involved. Some cyclic dienes have been reported to give unusual products when treated with TBCK. Cyclopentadiene behaves as expected and gives only the contra- thermodynamic cyclobutanone (32).15322 More interestingly, cyclohexa-l,3- diene resulted in the cyclobutanone (33; 71 %) and the bicyclic enol ether (34; 29%).22 Furthermore, it was observed that (33) can be converted into (34) upon thermolysis.However, it has not been determined if this transformation is a concerted oxy-Cope rearrangement or a stepwise process. Anomalous transformations have also been reported for the cycloaddition of TBCK to certain homoconjugated dienes. For example, bicyclo[2.2.1 Iheptadiene gave the cyclobutanone (35) and the polycyclic ether (36).15J7 The ratio of these products shows little solvent dependence; in benzene the (35): (36) ratio was 2.08 and in acetonitrile it was 1.48. Thus, both products were suggested to arise via a concerted process. Benzobarelene behaved analogously in that the cyclobutanone (38; 33%) and the polycyclic ether (39; 66%) were the observed products.22 Bicyclo[3.2.I]octa-2,5-diene, on the other hand, gave only the cyclobutanone (40).22 Other nonconjugated dienes such as cyclohexa-l,4-diene,15722ci~,cis~~J~ and cis, trans-cyclo-octa-1 ,5-diene15 gave only cyclobutanones when treated with TBCK. 23 M. D. Gheorghiu, L. Plrvulescu, C. Draghici, I. Manolescu, and B. David, unpublished results; presented, in part, at the Chemical Meeting at Timisoara (October, 1979) and at the Meeting of the Chemical Section of the Romanian Academy of Science, Bucharest (March, 1980). Moore and Gheorghiu But H NC ---0 TBCK + OH +hA (34) (33) TBCK NC But TBCK (40) 4 Cyanoketene-Enol Ether Cycloaddition A zwitterion mechanism is most likely for the cycloadditions of TBCK to enol Cyanoketenes: Synthesis and Cycloaditions ethers.23 Ethyl vinyl ether and vinyl acetate gave the corresponding cyclobutanones [(41), (42); 3:1] and [(43), (44); 3.5:1] upon treatment with TBCK.It is significant to note that the stereochemistry of the major isomers [(41) and (43)]is that predicted to arise from a [2 vs+ 2 va]concerted process. However, this was rejected since the products obtained from the related enol ethers, l-ethoxy- propene, 1 -acetoxy- and 1 -benzoxypropene, were (49, (46), and (47), respectively, and these could reasonably arise via zwitterionic intermediates. Thus zwitterionic intermediates were also suggested as the precursors to the cyclobutanones (41 )-(44). The slight selectivity observed for the formation of 0 dOEt TBCK98 % + (0.t + ”$!H NC H NC OEt (41) (42) ,OEt TRCK ‘Me 0 8OtR TBCKNB% But~ ‘Me NC Me NRP Me (46; R = Me) 0 (47; R = Ph) e3 D.Becker and N. C. Brodsky, J. Cliem. SOC.,Chem. Comm., 1978,237. Moore and Gheorghiu the contrathermodynamic products (41) and (43) was viewed as arising by the favoured rotation sequence indicated in the proposed zwitterion, (48), i.e., small cyano-group rotating past small proton. Such a process is consistent with the predictions arising from Orbital Correspondence Analysis in Maximum Symmetry,24125 which proposes specifically oriented zwitterions or diradicals in 2 + 2 cycloadditions. 5 Cyanoketene-Alkyne Cycloadditions Limited reports have appeared concerning the cycloaddition of cyanoketenes to simple alkyne~.~~*27 Here again, the ketene was t-butylcyanoketene, and the cycloadditions gave the cyclobutenones (49a-d) in yield ranging from 40-80 %.R1-C=C-R2 TBCK But (49) U; R1 = Ph, R2 = H h; R1= But, R2 = H C; R1 = But, R2 = Ph d; R1 = R2= Ph For the unsymmetrical alkynes only a single regioisomer was detected. These data are consistent with a concerted cycloaddition. An unusual transformation was observed when the acetylenic transition-metal complexes, (50) were treated with TBCK.28 Rather than cyclobutenones, the cyclopentenones (51) were isolated. Analogous results were obtained with diphenyl ketene. 6 Cyanoketene-Allene Cycloadditions A most interesting series of observations has been recorded for studies of the a4 A.Halevi, Helv. Chim. Acta, 1975, 58, 2136. st. A. Halevi, Angew. Chem., 1976, 88, 664. M.D. Gheorghiu, C. DrBghici, L. StBnescu, and M. Avram, Tetrahedron Lett., 1973, 9. M. D. Gheorghiu, Rev. Roum. Chim., 1977, 22, 1069. L.S. Chen, D. W. Lichtenberg, P.W. Robinson, Y.Yamamoto, and A. Wojcicki, Znorg,Chim. Acra, 1977,25,165. Cyanoketenes: Synthesis and Cycloadditions R (51) M = 775 -CsHsFe(C0)2, R = Me, Ph, or CH2-Fe(C0)2 (q5-Ph) M = ~~-C~H~MO(CO)~,R = Ph cycloaddition of TBCK to allenes. The cyclic and optically enriched allene, cyclonona-1,2-diene, gave the adducts (52) and (53) in a ratio of 3:2, and both showed some optical activity.13 With the optically enriched acyclic allene, penta-2,3-diene, the four possible isomers, (54)-(57), were formed.However, only the minor E-isomers, (56) and (57), showed optical acti~ity.~~.~~ Tetramethylallene gave a 77% yield of (58), and 1,l-dimethylallene gave (59) and (60) in a respective ratio of 65:35. 1-t-Butyl-1-methylallenereacted with TBCK to give four adducts in 27% yield. The major product was (61) and the minor was (62); the other two were (63) and (64).31 Optically enriched 1,3-diphenylalIene gave the cyclobutanones (65) and (66) having only the E-configuration of the benzylidine groups.32 Both were optically active and (65) was shown to have 53 % of the optical purity of that of (66).33 W. G. Duncan, W. Weyler, and H. W. Moore, Tetrahedron Lett., 1973, 4391.30 H. A. Bampfield and P. R. Brook, J. Chem. SOC.,Chem. Commun., 1974, 171. 31 H. A. Bampfield and P. R. Brook, J. Chem. SOC.,Chem. Commun., 1974, 172. 32 H. A. Bampfield, P. R. Brook, and W. S. McDonald, J. Chem. SOC.,Chem. Commun., 1975, 132. 33 H. A. Bampfield, P. R. Brook, and K. Hunt, J. Chem. SOC.,Chem. Commun., 1976, 146. Moore and Gheorghiu TBCK But But I ICN CN (59;657;) (60; 35 %) 27% But + NcjiButTBCK But (61 ;58 %) (62; 3 %) Me But + OM But -f--f-But But +" CN Me CN (63;24 %) (64; 15 %) 0 Ph TBCKLC-------+ 7":"+ Bu:K:Ph But ; CN H CN Ph (65 ;77 %) (66; 23 %) In this last example, any detailed mechanism must account for the following two salient points: (i), the major product is the torsionally strained adduct (65) and it is formed with approximately half the optical purity of (66); (ii), only the E-isomers are formed.These data are consistent with the mechanism outlined in 301 Cyanoketenes: Synthesis and Cycloadditions Scheme 2. The cycloaddition resembles a concerted reaction in that the ketene and allene components approach one another in an orthogonal fashion. However, initial bond formation takes place to give the chiral zwitterions (67) and (68). Ring closure of these gives, respectively, the optically active products (65) and (66). Zwitterions (67) and (68) could assume another chiral conformation, i.e., (69) and (70), respectively, and these could also undergo ring closure to give active (65) and (66).Finally, ion (69) could proceed to the achiral (71) and conrotatory ring closure of this would lead to optically inactive (65). On the other hand, a planar achiral ion arising from (70) would be less likely on the basis of steric effects. Thus, even though the total yield of (65) would be expected to Ph H TBCK + )=C=( H Ph 0 (65; active):>%=? H 'Ph NC;\But 0 Ph H But*:h -(65; inactive) { Ph H (71) Scheme 2 302 Moore and Gheorghiu exceed that of (66), the optical activity of the former should be less than the latter. An analogous mechanism can be envisaged for the conversion of optically enriched cyclonona-l,2-diene into (52) and (53). The mechanism of TBCK addition to penta-2,3-diene is more complex in that both E-and 2-isomers are formed. In addition, only the E-isomers [(56) and (57)] show optical activity; yet the Z-isomers [(54) and (55)] are the major products.These results can be rationalized in somewhat analogous manner to those argu- ments presented for the diphenylallene case and are outlined in Scheme 3. It can Me H TBCK + )=c=( H Me 0 1 0 (76)i Cyanoketenes: Synthesis and Cycloadditions be assumed that (72) and (73) are the initially formed chiral zwitterions and that these lead to optically active E-isomeric products. However, there is a significant driving force for these to proceed to the achiral zwitterion [(76)-(79)] to gain allylic resonance stabilization.Although several planar zwitterions other than those listed [(76)-(79)] are possible, the ones represented do minimise steric interactions. Indeed, one would predict their stabilization order to be (77) > (76) and (78) > (79). Thus, the predominance of the Z-isomers over the corres- ponding E-isomers might be expected. The formation of [(58)-(64)] can also be viewed as arising via dipolar inter- mediates rather than by concerted processes. Clearly additional work is needed before a detailed understanding of the mechanism of cyanoketene-allene cycloadditions can be obtained. However, the results thus far reported suggest these cycloadditions to be non-concerted and to involve zwitterionic intermediates. 7 Cyanoketene-Ketene Cycloadditions In general, the mechanism of ketene to ketene cycloadditions is fraught with ambiguities concerning their concerted versus non-concerted nature.The exception to this concerns the cycloadditions of TBCK to aldo- and ketoketenes since here it is clearly established that these proceed' via a dipolar process.34 The key to this interpretation is that the intermediate zwitterionic specie3 has been independently generated and shown to give the same products as the cyclo- additions themselves. Specifically, it was shown that TBCK reacts with the ketoketenes, dimethyl-,ethylmethyl-, and benzylmethyl-ketene to give, respectively the cyclobutanediones, (80), [(81), (82); 43: 571, and [(83), (84); 46: 541. That these cycloadditions involve a zwitterionic intermediate was established by the observation that the thermolysis of the azidocyclopentenones, (85), (86), and (87) gave the same products and isomer ratios.A zwitterion was established by trapping experiments in the thermolysis of (85). In an analogous set of experi- ments TBCK and methylketene were shown to undergo cycloaddition to give (88), and the same product was formed as the exclusive 1 :1 adduct from the thermo- lysis of (89). The most consistent interpretation of these results is that the two ketenes undergo initial bond formation in a head-to-tail orientation to give the zwitterion represented by conformer (91) (Scheme 4). Such an interpretation is possible since (91) would be exactly the expected conformer to arise when the zwitterion is independently generated from the azidocyclopentenedione precursors, (90).Direct ring closure of (91 ;R = Me) to (80) would involve the orthogonal enolate anion and acyl cation orbitals. The product-forming step apparently experiences very little steric influence from the substituents at positions 2 and 4 of the zwitterion since a nearly equal mixture of cis-and trans-isomers is observed for the ring closure of (91 :R = Et and R = Bn). However, there is a pronounced steric effect on conformational equilibration of the zwitterion as a function of the substituents at position 2. Thus, when R = H rapid rotation to (92) is :jl H. W. Moore and D. Scott Wilbur. J. Org. Chem., 1980, 45, 4483. Moore and Gheorghiu But CN 0 --IlpBu* A Et TBCK Me C Me III + \o 0 N CN CN (81) (43: 57) (82) 305 Cyanoketenes: Synthesis and Cycloadditions I (83) (46: 54) (84) 0 CN But*: N3 0 306 Moore and Gheorghiu Bu&N3 0 R 5--But CN 01 / R = MeI Scheme 4 allowed and subsequent ring closure to the oxetan-Zone (88) occurs.Indeed, oxetan-Zone formation would be expected to be the kinetically favoured route on electronic grounds, but can compete sterically with cyclobutanedione for- mation only when one of the ketene components is an aldoketene. 8 Cyanoketene-Imine Cycloadditions Extensive studies have appeared concerning the cycloadditions of cyanoketenes to formimidates, thioformimidates, and imines. t-Butyl-, methyl-, chloro-, bromo-, and iodocyanoketene were found readily to form cycloadducts with a variety of acyclic formimidates and thioformimidates to give 3-cyano-2-azeti- 307 Cyanoketenes: Synthesis and Cycloadditions dinones (p-lactams) in yields of 46-95 The fact that p-lactams are formed %.6t35 was anticipated since ketene cycloadditions to imidates and imines is one of the oldest synthetic routes to such compounds.However, it was surprising that the cyanoketene cycloadditions proceed in a stereospecific manner to give, in general, those azetidinones having a tvans-relationship between the 3-cyano- and 4-protio-gro~ps.~~~~~ Although over sixty 3-cyano-2-azetidinones have been prepared by this method, the few examples listed below are sufficient to illustrate the transformation. Sufficient data have now been accumulated to establish these cycloadditions ~C6Hll N NC bEtOI' CN OEt (93; 94%) /C6H11 BrN NC EtS b Br CN bEt (94; 63 %) /Ph TBCK bMeS CN SMe (95;40%) dCsHl1 NC +Bun -J' CN S-Bun (96; 50%) 35 H.W. Moore, L. Hernandez, D. M. Kunert, F. Mercer, and A. Sing, J. Am. Chern. Soc., 1981, 103, 1769. 38 R. Chambers, D. Kunert, L. Hernandez, F. Mercer, and H. W. Moore, TefrahedronLet[., 1978, 933. 308 Moore and Gheorghiu to be dipolar in character. The most significant mechanistic finding is of the independent generation of the zwitterion from the thermolysis of 4-azido-2- pyrrolinones. 233 37 For example, chlorocyano ke tene forms cycloadd ucts with U-ethyl-N-cyclohexylformimidateto give (93).The same product was formed when 4-azido-3-chloro-1 -cyclohexyl-5-ethoxy-3-pyrrolin-2-onewas thermally de- composed in refluxing benzene (Scheme 5). In this thermolysis the intermediacy of the zwitterion (97) was established by a series of trapping experiments.38 Thus, it is reasonably assumed that zwitterion (97) is a common intermediate in both the azidopyrrolinone decomposition as well as the ketene cycloaddition. OEt (93) Scheme 5 A variety of cycloadditions of TBCK to azomethines provides additional support for a stepwise mechanism since often 2:l adducts are formed. For example, N-methylbenzylideneamine and TBCK undergo cycloaddition in toluene at room temperature to give a mixture of (98) and (99),3gOn the other hand, if the reaction is accomplished at reflux temperature, or if (98) and (99) are pyrolysed at 180°C,the /3-lactam (100)is formed.40 In addition, if the reaction conditions are toluene-SO2 at -1O"C,then the adducts (101) and (102) are formed in addition to a small amount of thep-lactam (100; 15 %).A particularly unusual result was observed when the reaction was accomplished by adding 37 H. W. Moore, L. Hernandez, and R. Chambers, J. Am. Chem. SOC.,1978, 100, 2245. 38 F. Mercer, L. Hernandez, and H. W. Moore, Heterocycles, 1979, 12, 45. 39 Z. Lysenko, M. M. Joulle, I. Miura, and R. Rodebaugh, Tetrahedron Lett., 1977, 1705. 4o E. Schaumann and H. Mrotzek, Chem. Ber., 1978, 111, 672.Cyanoketenes: Synthesis and Cycloadditions But CN (99; 42%) toluene-so*,-10°C Me + Me But 0 CN (101; 13%) (102; 23 the azomethine to an excess of TBCK containing a catalytic amount of tri- phenylph~sphine.~~Under these conditions a 45% yield of the 3:l adduct, (103), was realized; its proposed mechanism is outlined below, but the effect of tri- phenylphosphine is not understood. Cycloadditions analogous to the formation of (98) and (99) were also observed for TBCK with N-methyl-p-methoxy- and N-methyl-p-nitrobenzylideneamine. However, p-lactam products [(105), (107)] were observed for the cycloaddition to (104)and (106). With N-t-butylbenzylideneaminethe 2:1 adduct, (108) is formed in 45 % yield if the amine is added to the ketene.If the addition is reversed, thq#%lactams (109; 6%) and (110; 31 %) are formed in addition to (108; 10%). Cyclic imines and TBCK give either @lactams or 2:l adducts; the outcome is dependent upon the ring size and substitution pattern of the imine.42 For example, the /%lactams (I 1 l), (1 12), and (1 13) result from the corresponding imines. However, (114) and (117) give the respective 2:l adducts (115),(1 16) and (1 1 8), (1 19). Additional examples resulting in 2 :1 adducts were reported for dihydroiso- quinoline to give (120) and the conversion of (121) into (122) (ketene added to imine sol~tion).~O In this latter case it was initially reported that (123) was the 41 E. Schaumann, H. Mrotzek, and G. Adiwidjaja, J.Chem. SOC.,Chem. Commun., 1978,820. 4a E. Schaumann, H. Mrotzek,and F. Abmann, Jiistus Liebigs Ann. Chem., 1979, 334. 310 Moore and Gheorghiu N/Me TBCKAPh Ph But I f CN Ph N/Me PhASMe TBCK But product when the imine was added to a solution of the ketene.*3 However, the structure has recently been revised to (124)and this arises from the reaction of the ketene dimer with the imine. Finally, one last example of a cyclic imine- D. H.Aue and D. Thomas, J. Org. Chem., 1975,40,2552. Cyanoketenes: Synthesis and Cycloadditions NC But (108 ;45 %) (109; 6%) (1 10; 31 %) TBCK-But CN H (1 1 I ;40:/) 0 TBCK Ph But CN Ph (112; 35%) 0 Ph But , (1 13; 35%) But CN (1 14) (1 15;48%) (1 16; 14%) Moore and Gheorghiu But CN HJ1 U (1 17) (118; 2073 (I 19; 31 ”/,) TBCK aN (120;23 %) 0 .Me 0 But CN Cyanoketenes: Synthesis and Cycloadditions TBCK cycloaddition has been reported.This involves the reaction of 2-(dimethy1amino)thiazole and its 5-methyl analogue with the ketene.4 In both cases the 2 :1 adducts (1 25) and (126) were formed. The observation of 2:l adducts along with the work outlining the independent generation of zwitterionic intermediates, provides solid evidence that Fyanoketene 44 A. Dondoni, A. Medici, C. Venturoli, L. Forlani, and V. Bertolasi, J. Org. Chem., 1980, 45, 621. Moore and Gheorghiu cycloadditions to imines and formimidates is dipolar in character. However, a few additional comments are in order regarding those reactions resulting in the p-lactams (99, (loo), (107), and (1 10)-(113).In general, the stereochemistry of these products is predictable on the basis of zwitterions of structure (127). The 0-major or exclusive, product would arise via a conrotatory ring closure of that zwitterion in which steric interactions (a) and (6) would be minimized. For example, the penultimate precursors to (loo), (107), (110), and (113) would be, respectively, (128), (129), (130), and (131). Particularly noteworthy is the com- 0-parison of (128) and (130), which vary in the steric bulk of the N-substituent. The former leads to the less hindered p-lactam, (loo), and the latter gives the more torsionally strained product, (1 10). The influence of the above-mentioned steric factors also plays a prime role in dictating the products of the cycloadditions of cyanoketenes to a$-unsaturated imine~.~Specifically, steric interactions (u), (b), and (c) (Scheme 6) are of impor- Cyanoketenes: Synthesis and Cycloadditions tance in determining the relative population of zwitterionic intermediates and thus products that arise from such.For example, when using TBCK and when steric interaction (a) is minimized, zwitterion (133) is favoured; this leads to 3-cyano-azetidin-2-ones (1 34) having a trans-relationship between the 3-cyano- and 4-protio-groups. As steric interaction (a) increases, zwitterion (1 32) becomes important, which results in increased amounts of cis-3-cyano-azetidin-2-ones(1 35) and the &lactam (136).Thus, TBCK reacts with the N-p-methoxyphenylimine of cinnamaldehyde to give thep-lactam (137a) in 85% yield. When the steric bulk on nitrogen is increased by utilizing the N-cyclohexylimine, the &lactam (1 396) is formed in 23 % yield and the p-lactams (1 376) and (1 386) in, respectively, 64 % and 10 % yield. The &lactam (1 39b) and /?-lactam (1 386) can be viewed as arising from zwitterion (132; R1= But, R2= C6H11) and the p-lactam (1376) from 0 II cII + R+CN 0-(133)I R1O E : I I CN 1Ph (135) Scheme 6 Moore and Gheorghiu (1 33). Finally, with the N-t-butylimine, the products are thep-lactam (1 37c; 17 %), its torsionally strained diasteriomer (1 38c; 29 %), and the &lactam ( 139c; 52 %).0 / R' 0 R1 LN + + But , i Lph PhdN H R % Yield (a) C6H40Me 85 --(b) C6Hll 64 10 23 (c) But 17 29 52 As expected from the above arguments, the steric bulk of the ketene also has a pronounced influence on the zwitterionic intermediates and thus on product formation. For example, chlorocyanoketene (CCK) undergoes cycloaddition with the N-p-methoxyphenylimine of cinnamaldehyde to give the /3-lactam (141a; 17%), the &lactam (142a; 42 %), and the pyridone (143a; 22 %), whereas, as already mentioned, TBCK gave only thep-lactam (137a; 85 "/,). Six-membered ring formation can be minimized by employing (140; R = Ph) since the additional phenyl group at the /%position would be expected to impart enhanced steric crowding in the ring closure of the zwitterion to a &lactam. Indeed, treatment of (140b) with CCK resulted in a > 90% yield of thep-lactam in (1416), and no &lactam could be detected.These results and mechanistic interpretations are also consistent with data obtained in the study of chlorocyanoketene cycloadditions to trans-cinnam-317 Cyanoketenes: Synthesis and Cycloadditions aldehydes. The aldehydes, unlike their anti-imine 'derivatives, could give zwitterions (144) or (145) in their initial interactions with the ketene. The former should give E-/3-lactones and the latter should give the 2-isomers as well as 8-lactones. Certainly, on the basis of steric arguments, zwitterion (144) would be favoured.It was observed that CCK, but not TBCK, readily undergoes cyclo- addition with (146a-c) to give the dienes (148a-~).~Although undetected, the E-/3-lactones (147a-c) are most likely the precursors to the dienes and give such upon stereospecific decarboxylation under the reaction conditions. Thus, the products of the reaction come exclusively from zwitterion (144). 0 Ph 9 Cyanoketene-Benzaldehyde Cycloadditions Chloro- and bromocyanoketene, but not TBCK, undergo cycloaddition with a variety of substituted benzaldehydes in analogy to the above-mentioned reaction of cinnamaldehyde~.~~ That is, the corresponding p-lactones are formed and suffer stereospecific decarboxylation to give the alkenes (149).It was observed that the relative rates as well as the product yields decreased as the benzaldehyde was substituted with increasingly stronger electron-withdrawing groups. Such observations are consistent with a dipolar mechanism in which the ketene functions as the electrophile and the aldehyde as the nucleophile. This was 45 H. W. Moore, F. Mercer, D. Kunert, and P. Albaugh, J. Am. Chem. Soc., 1979, 101,5435. Moore and Gheorghiu 0 + H XACN R X R % Yield c1c1 2,4-(OMe)2 4-NMe2 92 ‘18 C1 4-OMe 73 Br c1 4-OMe 4-OCOMe 79 61 Br 4-OCOMe 51 Cl H 61 Br H 48 c1 4-Me 54 c1 4-C1 32 Br 4-C1 17 C1 4-NO2 8 Br 4-NO2 <5 further substantiated by the independent generation of the zwitterionic inter- mediate.That is, thermolysis of 4-azido-2-chloro-5-(4-methoxyphenyl)-(5H)-furan-Zone cleaved to zwitterion (1 50)which subsequently gave the same product as the cycloaddition of chlorocyanoketene to 4-methoxybenzaldehyde (Scheme 7). It is particularly noteworthy that dichloroketene gives analogous products, (152), when treated with substituted ben~aldehydes.~~ However, the relative rates and product yields are reversed from those observed with chlorocyanoketene. Thus, whereas chlorocyanoketene functions as the electrophile in these reactions, the dichloro-derivative behaves as a nucleophile. As a result, dipolar character represented by zwitterion (150) controls product formation for the former and (151) for the latter.In another series of studies, dicyanoketene was shown to give (153), (154), and (155) via a cycloaddition-decarboxylation me~hanism.~~*~-*~Products analogous to (153) and (154) were also observed when the ketenes employed were t-butyl- cyano-, cyanomethyl-, and cyanophenyl-ketene. Dicyanoketene was also used to prepare the fulvenes (156),(157), and (158) from the corresponding ketones. 46 H. 0.Krabbenhoft, J. Org. Chem., 1978, 43, 1305. 47 R. Neidlein and K. F. Cepera, Chem. Ber., 1978, 111, 1824. 48 R. Neidlein and G. Humburg, Justus Liebigs Ann. Chem., 1978, 1974. 4g R. Neidlein and E. Bernhard, Justus Liebigs Ann. Chem., 1979, 959. 319 Cyanoketenes: Synthesis and Cycloadditions O+ ___, c1 c1 C1 C1 H c1 ArR Scheme 7 Moore and Gheorghiu CN R=HorMe (1 53) CN CN“AcNNc&N 6CN --\ / Ph Ph Cyanoketenes: Synthesis and Cycloadditions 10 Cyanoket ene-Isoni trile Cycloadditions Little work has appeared concerning the addition of ketenes to isonitriles, but that which has documents the products to be 2:l adducts having the l-imino- cyclopentane-2,4-dione ring system.For example, diphenylketene and benzyl- isonitrile give a 90% yield of (1 59).50 Cyanoketene-isonitrile cycloadditions have 0 Yh Bn-N* C Ph N +Bn been shown to be anomalous to the above. For example, TBCK has been shown to give the 2 :1 adducts (160) when treated with isonitriles at ambient tempera- ture.' Again zwitterions are reasonable intermediates to these products.But 0-CN R-NSC TBC--fN=C N TBCK NC /$ \R But N IK R "/o Yield But 97 Bun 70 C6Hll 82 Bn 73 CH2S02CsH4Me 68 6o I. Ugi and K. Rosendahl, Chem Ber., 1961,94, 2233. Moore and Gheorghiu The adducts, (la), were shown to be the kinetic products of these reactions since thermolysis of the N-t-butyl analogue in refluxing benzene caused its rearrangement to (161). Further thermolysis of (161) at 130°C in o-dichloro-benzene resulted in the butenolide, (1 62). Bu But NqoT:~A NC But-N But NH CN / The cycloaddition of chlorocyanoketene to isonitriles appears to follow a completely different course. For example, when this ketene was generated in the presence of excess t-butylisonitrile, the unusual 3 :1 adduct, (163), was ~btained.~ Ci )cc=o But-N=C NC ?FBUtN N-But I But 11 Cyanoketene-1-Azirine, Oxaziridine, and Thiaziridinimine Cycloadditions TBCK was shown to react with a variety of 1-azirines to give the 2:l adducts CN But Ph b But’yoFyR’ TBCK ,& CN R2 R1 R2 H H H Me Me Me 323 Cyanoketenes:Synthesis and Cycloadditions (164).51 Analogous adducts were observed when diphenylketene was employed.A 1:1 adduct, (165), was observed when TBCK was treated with 2,3-diphenyl-l- azirine. But !? Very limited work has appeared describing the reaction of ketenes with heterocyclic compounds containing two adjacent heteroatoms. Diphenylketene and 2-ethyl-3-phenyloxaziridinecombine to give an oxazolidinone (166; 38 %) and benzaldehyde (50%).52 TBCK reacts with 2’-cyclohexylspiro[fluorene-9,3’-oxaziridine] to give the spiro-oxazolidinone (1 67) and a spiroisoxazolidinone (168) in respective yields of 48 % and 22 %.53 (167; 48 %) (168; 22%) 61 A.Hassner, A. S. Miller, and M. J. Haddadin, Tetrahedron Lett., 1972, 1353. 6s M. Komatsu, Y. Oshhiro, H. Holta, M. Sato, and T. Agawa, J. Org. Chem., 1974, 39, 3198. 6s M. A. Abou-Gharbia and M. M. Joullt, Synrh. Commun., 1979, 9, 871. Moore and Gheorghiu Spiroisoxazolidinones (1 69) were also obtained when TBCK and the corres- ponding N-fluorenylidene alkylamine oxides were subjected to refluxing benzene temperat~re.5~On the other hand, arylamine analogues give (170). Compounds analogous to (170) were also observed when the ketene was cyclopentamethylene ketene or t-butylcarboethoxyketene. TBCK (169; R = H, Me, or Ph) (170; Ar = p-Me-C6H4-, N-Sulphonyliminothiaziridinesgenerated by the thermolysis of 4-alkyl-5-sulphonylimino-l,2,3,4-thiatriazolinesreact with TBCK to give (171).An anal-ogous product was observed when phenylketene was empl0yed.5~ On the last an unusual reaction of TBCK with a heterocyclic compound has been described. Treatment of the ketene with 4-imino-4,5-dihydro-l ,2A6,3- oxathiazol-Zones gave a 70 % yield of (lC12).56 Diphenylketene gave an analogous product. 12 Cyanoketene-Sulphurdi-imideCycloadditions Cycloadditions of ketenes to sulphurdi-imides has received only limited atten- ti~n,~~-60 and no reports have previously appeared where cyanoketenes have 54 M.A. Abou-Gharbia and M. M. Joulle, J. Org. Chem., 1979, 44, 2961. G. L'abbe, G. Verhelst, C. C. Yu, and S. Toppet, J. Org. Chem., 1975, 40,1728. 50 G. L'abbe, C. C. Yu, and S. Toppet, J. Org. Chem., 1979, 44, 3991. 67 H. Grill and G. Kresze, Tetrahedron Lett., 1970, 1427. 58 H. H. Horhold and H. Eibisch, Tetrahedron, 1969, 25, 4277. 69 T. Minami, K. Yamataka, Y. Ohshiro, T. Agawa, N. Yasuoka, and N. Kasai, J. Org.Chem., 1972, 37, 3810. 0o T. Minami and T. Agawa, J. Org. Chem., 1974, 39, 1210. Cyanaketenes: Synthesis and Cyclaadditians been utilized. Such a study has now been accomplished,7 and the results are unusual in that the observed products are generally different from those reported for less electrophilic ketenes such as phenyl-, diphenyl-, and chlorophenylketene. Selected examples employing TBCK and chlorocyanoketene (CCK) are given below. The transformations leading to (174) and (179) are unique in that the thione-S-imides (173) and (178) are formed as intermediates.The former is trapped by an additional molecule of TBCK to give the isothiazolidin-3-one, (174), and the latter undergoes ring closure with loss of HCI to give (179). The formation of (176) and (177) is also without precedent. The formation of (175) is the only example anticipated on the basis of previously reported work employing other ketenes. 13 Conclusion Cyanoketenes are a readily accessible class of electron-deficient cumulenes.They are prepared from the thermolysis of appropriately substituted vinyl azides; the substituent on the cyanoketene moiety can be varied to include t-pentyl, t-butyl, isopropyl, methyl, cyano-, bromo-, chloro-, iodo-, and phenoxy-groups. All of these function as potent electrophiles in their cycloadditions to alkenes, alkynes, Moore and Gheorghiu TBCK N-But+rIqBut~ But-N=S=N-But OC CN But But Ph-N-S-N-Ph TBCK + But But "3.riL + CN -But 327 Cyanoketenes: Syuthesis and Cycloaddirions CCKPh-N-S-N-Ph -O=C=N--Ph+ 1-HCI CN allenes, ketenes, imidates, imines, aldehydes, isonitriles, amine oxides, sulphurdi- imides, and azirines. With the exceptions of the alkene and alkyne cycloadditions the reactions appear to be non-concerted and proceed via zwitterionic inter- mediates.In many cases the reactions are unique in that products result which have not been previously observed to arise from analogous reactions using less electrophilic ketenes. Clearly, a variety of carbocyclic and heterocyclic com-pounds are available from the reactions outlined in this review.

ISSN:0306-0012    

DOI:10.1039/CS9811000289

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications By J. R. Jones CHEMISTRY DEPARTMENT, UNiVERSITY OF SURREY, GUILDFORD GU2 5XH S. E. Taylor* CHEMISTRY DEPARTMENT, QUEEN’S UNIVERSITY, KINGSTON, ONTARIO K7L 3N6, CANADA 1 Introduction The purines [derivatives of (l)] command important positions in both chemistry and biochemistry1 and their versatility is reflected in the number of different roles they play in biological processes, from being one of the fundamental components of nucleic acids to being the basis of energy storage as in adenosine 5’-triphosphate. Furthermore, simple structural modifications of naturally occurring purines have made available a variety of purine analogues which are potent antagonists of many biological systems.IH It is no great surprise, therefore, to find that deuterium- and tritium-labelled purines have found wide application in the biochemical sciences. Consequently, a better appreciation of all the factors that can influence the rates of deuterium or tritium loss will ensure that possible dangers of mjsinterpretation arising from the use of these labelled compounds will be minimized. Conversely, such studies can also lead to optimization of conditions for both labelling and storage of the appropriate compounds. In contrast to other carbon acids that have been the subject of extensive *Present address: Department of Inorganic Chemistry, University of Oxford, South Parks Road, Oxford OX1 3QR. J. H. Lister, ‘Purines’, Vol. 24 (Part 11) of ‘The Chemistry of Heterocyclic Compounds’, ed.D. J. Brown, Wiley-Interscience, New York, 1971. Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications investigation in recent years2 and which undergo ionization as the uncharged species, the purines (like many other heterocyclic compounds) can exist in aqueous solution in a number of ionized forms, and therefore the number of available reaction pathways through which isotopic hydrogen exchange can occur is increased. Consequently, such studies can provide a convenient and often subtle way of obtaining information of both mechanistic and synthetic utility. Hydrogen isotope exchange in heterocyclic compounds3 was first reviewed by one of us in 1974 and more recently Lister,* in dealing with some physicochemical aspects of purines, has considered both hydrogen exchange, radical reactions, and ionic alkylation reactions involving the C-8 position.2 Historical Aspects As long ago as 1952 Eidinoff et aL5reported on a biosynthetic method of incor- porating tritium (from sodium [rnethyl-3H]acetate) into yeast DNA, with sub- sequent hydrolysis leading to the isolation of labelled purine nucleosides and the corresponding bases. Although the position of the label was not identified, it was apparent that since the conditions used in the hydrolysis would lead to the removal of labile N-and O-bound tritium, the most likely sites were the C-8 positions of adenine and guanine. EidinoB and Knoll6 subsequently prepared tritiated and deuteriated adenine and guanine by heating the compounds with HTO or D2O at 100 “C in the presence of a reduced platinum catalyst.In assign- ing the 1H n.m.r. spectrum of purine, Ts’o and co-workers7 found that purine exchanged its H-8 merely by heating in D2O at 105 “Cfor 4 h, thereby dispensing with the need for an expensive platinum ‘catalyst’. The product of this exchange reaction was shown to be identical to that obtained in the desulphurization of 8-mercaptopurine with deuteriated Raney nickel. Further confirmation of the site of exchange was provided by Bullock and Jardetzky* who unambiguously synthesized [8-2H]purine by ring closure of 4,Sdiaminopyrimidine with [~Hz]-formic acid. In addition, hypoxanthine, inosine, adenine, adenosine, and 6- chloropurine were found to exchange H-8 by heating in D2O at 100 “Cfor 10-20 min. Little or no exchange of H-2 occurred under these conditions.FOX,^ using a D20-DMF mixture, found that exchange occurred in adenosine, 6-chloropurine, and 7-and 9-benzyladenine at elevated temperatures. With 3-benzyladenine, however, both C-2 and C-8 deuteriation occurred, as is also the a See forexample (a)J. R. Jones, ‘The Ionisation of Carbon Acids’, Academic Press, London, 1973; (6) R. P. Bell, ‘The Proton in Chemistry’, 2nd edn., Chapman and Hall, London, 1973 ;(c) F. Hibbert in ‘Comprehensive Chemical Kinetics’, Elsevier, Amsterdam, Vol. 16, p. 97, 1976. J. A. Elvidge, J. R. Jones, C. O’Brien, E. A. Evans, and H. C. Sheppard, Adv.Heterocycl. Chem., 1974, 16, 1. * J. H. Lister, Adv. Heterocycl. Chem., 1979, 24, 215. M. L. Eidinoff, H. C. Reilly, J. E. Knoll, and D. H. Marrian,J. Bid. Chem., 1952,199,51 I. * M. L, Eidinoff and J. E. Knoll, J. Am. Chem. Soc., 1953, 75, 1992. M. P. Schweizer, S. I. Chan, G. K. Helmkamp, and P. 0.P. Ts’o,J. Am. Chem. Suc., 1964, 86, 696. F. J. Bullock and 0.Jardetzky, J. Org. Chem., 1964, 29, 1988. J. R. Fox, Ph.D. Thesis, Univ. of Illinois, 1965. Jones and Taylor case in 3-methy1hypo~anthine.l~ More recently, Wong and Keck11 have shown that the presence of an alkyl substituent at the 3-position, with concomitant localization of positive charge in the pyrimidine ring, causes an increase in the ratio of the H-2/H-8 exchange rates.In the detritiation12 of [2,8-3Hz]adenine under neutral conditions, exchange from the C-8 position was found to be ca. 2000 times faster than from the C-2 position. As a preliminary study in the attempted tritiation of nucleic acids, McDonald and Philip@ found that the H-8 of adenosine 5'-monophosphate exchanged in D2O at 92 "C with a half-life of 90 min. Ostermann et al.14 prepared tritiated nucleoside diphosphates by heating the unlabelled compounds in HTO at 100°C, whilst the feasibility of labelling DNA in vitro was demonstrated by Fritzschels who studied the deuteriation of adenosine and guanosine and the corresponding residues in DNA. Shelton and Clark16 incorporated tritium into purine nucleotides by high temperature incubation in HTO and assessed their value as substrates in biochemical research.The tritium label was found to be stable for reasonable periods of time under physiological conditions, although for long-term reactions significant loss of label resulted, as confirmed by Evans et al.17 [3H]DNA prepared in this way has subsequently been used in hybridiza- tion rea~tions,l8~~~ and Wecliter20 has used the same approach to label adenosine specifically in the presence of cytidine. It seems somewhat surprising, therefore, that in view of the many studies that have been carried out, little or no attention has been given to the mechanistic details of these exchange reactions. 3 Experimental Methods Although i.r.,15921 Raman,22-24 and lH n.m.r.7J1p25 spectroscopy have all been used to follow hydrogen isotope exchange in purines, we have found detritiation methods26 to be the most useful.Because of the very low levels of radioactivity lo F. Bergmann and Z. Neiman, Chem. Commun., 1969,992. l1 J. L. Wong and J. H. Keck, jun., J. Chem. Soc., Chem. Commun., 1975, 125. la J. A. Elvidge, J. R. Jones, C. O'Brien, and E. A. Evans, Chem. Commun., 1971, 394. l3 C. C. McDonald and W. D. Philips, Biopolymers, 1965, 3, 609. l4 L. A. Ostermann, V. V. Adler, R. Bibilashvily, and Ya. M. Varshavsky, Biokhimiya, 1966, 31, 398. l5 H. Fritzsche, Biochim. Biophys. Acfa, 1967, 149, 173. l6 K. R. Shelton and J. M. Clark, jun., Biochemisfry, 1967, 6, 2735. l' E. A. Evans, H. C. Sheppard, and J. C. Turner, J.Labelled Compd., 1970,6, 76. K. R. Shelton and J. M. Clark, jud., Biochem. Biophys. Rex. Commii~.,1968, 33, 850. l* D. G. Searcy, Biochim. Biophys. Acfa, 1968, 166, 360. 2o W. Wechter, Collect. Czech. Chem. Commun., 1970, 35, 2003. z1 M. Nakanishi, M. Tsuboi, and I. Nakagawa, Bull. Chem. SOC. Jpn., 1976, 49, 2011. l2 G. J. Thomas jun., in 'Structure and Conformation of Nucleic Acids and Protein-Nucleic Acid Interactions', ed. M. Sundaralingam and S. T. Rao, University Park Press, Baltimore, Md., p. 253, 1975. *s (a) M. J. Lane and G.J. Thomas, jun., Biochemistry, 1979,18, 3839; (b) G. J. Thomas, jun. and M. J. Lane, J. Raman Spectrosc., 1980, 9, 134. 94 (a) J. Livramento and G. J. Thomas, jun., J. Am. Chem. SOC.,1974, 96, 6529; (6) G. J. Thomas jun and J.Livramento, Biochemistry, 1975, 14, 5210. *5 D. Lichtenberg and F. Bergmann, J. Chem. SOC., Perkin Trans. I, 1973, 789. J. R. Jones, Surv. Prog. Chem., 1973, 6, 83. Isotopic Hydrogen Exchange in Purines- Mec han isms and Applications that can be detected by liquid scintillation counting, the tritiated substrate needs to be present in solution only at very low concentration. Consequently, solubility problems rarely arise and measurements can be made in purely aqueous media. Furthermore, by measuring the increase in the radioactivity of the solvent (rather than the decrease in the radioactivity of the substrate) over the first 14% of the reaction, it has been possible to measure the rates of very slow reactions in a relatively short time interval (the so-called initial rate method).In this respect, the ready separation of the substrates from the solvent by freeze-drying is of great assistance. The pseudo-first-order detritiation rate constants (at 85 “C)are frequently of the order of 10-7-10-5 s-l, and being able to use the initial rate method is therefore a great advantage. The recent development of 3H n.m.r. ~pectroscopy~~means that a check on the specificity of labelling in the tritiated substrate can be made. 4 Rate-pH Profiles The detritiation of [8-3H]purineZ8 or [8-3H]adenine29 over a pH range at 85 “C gives rise to a bell-shaped rate-pH profile (Figure 1, curve a).Although there have been one or two reports suggesting that general base catalysis occurs in reactions of this kind,aO in all our studies overwhelming hydroxide-ion catalysis has been the rule.Under the experimental conditions purine and adenine can exist in one of three forms: (i) the neutral, BH; (ii) the protonated (either on N-1 or N-7), BH2+; or (iii) the anionic (ionization of N-g-H), B-. Consequently, it would be reasonable to expect that the observed kinetics would conform to equation (l), Rate = k[BH2+] [OH-] + k’[BHl [OH-] + k”[B-] [OH-] (11 so that a comparison between the relative reactivities of BHz+, BH, and B- could be made. However, in practical terms, this possibility is governed by the values of the respective ionization constants: 27 J. A. Elvidge, J. R. Jones, V. M. A. Chambers, and E. A. Evans, in ‘Isotopes in Organic Chemistry’, ed.E. Buncel and C. C. Lee, 1978, Vol. 4, p. 1. ** J. A. Elvidge, J. R. Jones, C. O’Brien, E. A. Evans, and H. C. Sheppard, J. Chem. SOC., Perkin Trans. 2, 1973, 1889. J. A. Elvidge, J. R. Jones, C. O’Brien, E. A. Evans, and H. C. Sheppard, J. Chem. SOC., Perkin Trans. 2, 1973, 2138. 30 Maeda el claim to have observed catalysis by D,O in (CD,),SO-D,O mixtures for exchange in adenosine, but we suggest that this result can be rationalized in terms of the changing [D,O] affecting K,(D,O), which in turn partly governs the value of the pseudo- first order rate constant according to k@ = k Kw/Ka (the rate of exchange in the pH- independent region of the rate-pH profile for 31a). More recently, Cohen and co-~orkers~~b reported a 4.3-fold increase in the rate of C-2 deuteriation in 1-methylimidazole in the presence of 1 M acetate at pD 4.9.This result was obtained at a very sensitive position on the rate-pD profile for this substrate,3a implying that the effect may, in part, be a consequence of pD or PKa variations at the high salt concentration. In our own studies in which we have utilized several imidazole and purine derivatives, including methylated guanosine and benzimidazole which model the protonated forms of these molecules, we have obtained no evidence for catalysis by bases other than hydroxide- or deuteroxide-ion. Clearly, more work is necessary in order to clear up this point. 31 (a) M. Maeda, M. Saneyoshi, and Y. Kawazoe, Chem. Pharm. Bull., 1971,19, 1641 ;(b) Y.Takeuchi, H. J. C. Yeh, K. L. Kirk, and L. A. Cohen, J. Org. Chem., 1978, 43, 3565. J. L. Wong and J. H. Keck, jun., J. Org. Chem., 1974, 39, 2398. Jones and Taylor C PH Figure 1Rate-pH profiles for isotopic hydrogen exchange from the C-8 position of (a)adenine,29 (6) 9-Pri-purine,2s (c)guan~sine,~~,~~ (e) x~nthosine,~~(d) theobromin~,~' and (f)adenosine 5'-monophosphateq0 Ka (= [BH] [H+]/[BH2+]) and Ka' (= [B-I [H+l/[BHl) The observed rate-pH profile for purine and adenine is of the sameform as that obtained for the detritiation of [2-3H]ben~imidazole;~~by means of similar ex- change studies on some chlorinated benzimidazoles and also on 1,3-dimethyl- benzimidazolium bromide, we were able to show that the observed rate-pH profile is in accordance with equation (2).33Even at pH 12, where very little Rate = k[BH2+][OH-] (2) of the protonated form exists, reaction still proceeds by this mechanism; if the neutral form were reactive at such hydroxide-ion concentrations the rate would increase dramatically.In the rate-pH independent region, the effect of a decreas- ing concentration of protonated substrate is counteracted by the increasing concentration of hydroxide ions. This effect continues until the creation of a negative charge a to the site of exchange makes hydroxide-ion attack unfavour- able on electrostatic grounds (vide infra), leading to a rate reduction. At very low pH, where the substrate is completely protonated, the decreasing hydroxide-ion concentration accounts for the fall-off in rate.34 Sf J.A. Elvidge, J. R. Jones, C. O'Brien, E. A. Evans, and J. C. Turner, J. Chern. Soc., Perkin Trans. 2, 1973, 432. 34 Under these conditions ([H+] >KB)the observed rate of exchange is directly proportional to [OH-], the line, as demonstrated for irnida~ole,~~passing through the origin. 333 Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications If we reduce the number of ionizable forms by, for example, blocking the N-9 position in purine, we see in the results for 9-Pri- and 9-But-purineZ8 (Figure 1, curve 6) how this affects the rate of exchange. At high pH there is a dramatic increase in rate due to the onset of a new mechanism involving hydroxide-ion attack on the neutral form, the appropriate rate equation being: Rate = k[BH2+] [OH-] + k’[BH] [OH-] (3) Adenosine29 gives similar results to the 9-alkylpurines.A logical extension of these studies would be to employ compounds con- taining an ionizable group in the pyrimidine ring, and this has been done for compounds such as guanosine and in0sine3~ (Figure 1, curve c). In such cases the negative charge is developed at a site which is well-removed from that undergoing exchange, and the results suggest that had the studies been extended to still higher pH (> 11.5), hydroxide-ion attack on the anionic form of the substrate would have been observed, in accordance with equation (1). The xanthines provide another example of how the presence of ionizable groups can influence the rates of e~change.~7-3~ Here we can go from the trimethylated species [caffeine; (2), R1 = R2= R3 = Me] which can exist only in protonated (N-9) and neutral forms and which gives a rate-pH profile similar to adeno- sine,37J8 to the three dimethylated species, theophylline [(2), R1= R2 = Me, R3 = HI, which gives a bell-shaped rate-pH profile, and theobromine [(2) R2 = R3 = Me, R1 = H] and paraxanthine [(2), R1 = R3 = Me, R2= HI, which additionally react via the anionic form at high pH (Figure 1, curve d).In general, the less the degree of methylation, the greater the number of potential mecha- nisms, so that in the case of xanthosine (9-/3-~-ribofuranosylxanthine),the pro- tonated, neutral, mono- and di-anionic forms all make a contribution to the overall rate3’ (Figure 1, curve e).In the above examples, the protonation or ionization sites have been confined to either the pyrimidine or imidazole ring systems. There is the possibility that 35 S. E. Taylor, Ph.D. Thesis, University of Surrey, Guildford, 1978. 36 J. A. Elvidge, J. R. Jones, C. O’Brien, E. A. Evans, and H. C. Sheppard, J. Chem. SOC., Perkin Trans. 2, 1974, 174. 37 J. R. Jones and S. E. Taylor, J. Chem. SOC.,Perkin Trans. 2, 1979, 1253. 3* M. Jelinska and J. Sobkowski, Tetrahedron, 1977, 33, 803. 3m M. Jelinska, J. Szydlowski, and J. Sobkowski, Tetruhedron, 1979, 35, 663. Jones and Taylor other groups may also influence the exchange, and this can be seen very clearly when we compare the rate-pH profiles for adenosine 5’-monophosphate40 (Figure 1, curvef) and adenosine.An additional plateau region is observed in the pH range where ionization of the secondary phosphoric acid function occurs. That this is the most likely explanation is supported by the fact that no such second plateau is observed in the case of adenosine 3’,5’-cyclic monopho~phate.~~ 5 Zwitterionic Contributions It is possible that, in certain circumstances (e.g. at high pH), both the neutral form of the substrate and a kinetically equivalent species (a zwitterion) can both undergo exchange. Tomasz et aL41were the first to mention this possibility; in studies of isotopic hydrogen exchange at 37 “Cfrom the C-8 position of guano-sine, they observed a large rate acceleration at pH > 8, and this result, together with their failure to observe a similar acceleration in the case of l-methylguano- sine, was ascribed to the involvement of the guanosine zwitterion (3) rather than the neutral guanosine molecule (4).This idea is an attractive one as the zwitterion 0 1 Ribose Ribose. (3) (4) has a positive charge on N-7, and abstraction of tritium from C-8 would give rise to an ylide intermediate; the process is therefore analogous mechanistically to the reaction pathway operating at low pH. Clearly the involvement of this kinetically equivalent species cannot be discerned from the shape of the rate-pH profiles, but additional evidence is available. Our own studies on adenosinez9 and on 9-Pri- and 9-But-purine,2* which are unable to form zwitterions, showed large rate increases at high pH.More recent results on 1-methylinosine and 1-methylguanosine show the same trend.42 A comparison of these results with those obtained36 for inosine and guanosine serves to show that at low pH, where the effective mechanism is between hydrox- ide ion and the protonated species, methyl substitution at N-1 has but a marginal effect on the rate. In sharp contrast, the results at high pH show that N-l-methylation brings about a large rate retardation. Clearly in both guanosine and inosine the hydroxide ion reacts with the neutral (rate constant k’) and zwitterio- nic (rate constant kh) forms, and the observed second-order rate constant kobs is given by 37942 40 J.R. Jones and S. E. Taylor, J. Chem. SOC.,Perkin Trans. 2, 1980, 441. M. Tomasz, J. Olsen, and C. M. Mercado, Biochemistry, 1972, 11, 1235. 43 J. R. Jones and S. E. Taylor, J. Chem. SOC.,PPrkin 2, 1979, 1587. 335 Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications kobs = k‘ + kh Kzw (4) where Kzw= [BH*]/[BH]. Kinetic evidence in support of the involvement of zwitterions has also been obtained for various ~anthines,~~ andnucle~tides,~O histidine derivative^.^^ X-Ray crystallographic studies44 show that even in the solid state the free acids of nucleotides exist as zwitterionic species. 6 Sites of Protonation In formulating a mechanism for isotopic hydrogen exchange in purines, we assumed that the reactive species is the N-7-protonated form, present in all cases to some extent.However, the extensive non-kinetic investigations which have been carried out in an attempt to identify the site(s) of protonation do not always lead to consistent results. X-Ray crystallographic studies on purine show that the crystalline salt is protonated at N-7,45 whereas N-1 is favoured for adenine46 and adeno~ine.47:~~Alkylation reactions invariably lead to a mixture of isomers49e50 and from a study of the variation in 13C-H coupling constants with pH, it has been concluded that, in solution, purine is partially protonated at N-1, N-3, and N-7.5l On the basis of linear free energy relationships (LFERs), the results of our isotopic hydrogen exchange studies support the idea that protonation does not exclusively involve N-7 in most p~rines.5~ The accumulated mass of detritia-tion rate data in the literature which relates to azolium species can be roughly classified into two main groups, according to their protonation behaviour.52 In the first group,53a data for the compounds imidazole, benzimidazole, 1,2,4- triazole, tetrazole, benzoxazole, benzothiazole, thiazole, and benzoselenazole, which are known to possess a single protonation site (N-3, equivalent to N-7 in purines) adjacent to the position of exchange, exhibit an excellent LFER between log k and pKa (the solid line shown in Figure 2), with a slope of -0.72.52~53a In contrast to this, the data for the purines, in which the protonation site is not as well defined, reveal that only in a few instances do the points lie on the line drawn for the azoles.We have interpreted this in terms of different degrees of N-7 protonation in the purines.52 On this basis, the finding that the data for compounds 1-5 and 8 fall on the line (Figure 2), suggests their preference for exclusive N-7 protonation. Compounds 6, 7, 9-17,23, and 24 appear to fall on a line having approximately the same slope as that derived from the azole data, 4* J. A. Elvidge, J. R. Jones, R. Salih, M. Y.Shandala, and S. E. Taylor, J. Chem. SOC., Perkin Trans. 2, 1980, 447. 44 M. Sundaralingam and P. Prusiner, Nucleic Acids Res., 1978, 5, 4375. 4b D. G. Watson, R. M. Sweet, and R. Marsh, Acta Crystallogr., 1965, 19, 573. 46 W.Cochran, Acta Crystallogr., 1951, 4, 81. “J. Kraut and L. H. Jensen, Acta Crystallogr., 1963, 16, 17. 46 M. Sundaralingam, Acta Crystallogr., 1966, 21, 495. B. C. Pal, Biochemistry, 1962, 1, 558. 6o P. D. Lawley and P. Brookes, Biochem. J., 1964, 92, 19c. 61 J. M. Read, jun. and J. H. Goldstein, J. Am. Chem. SOC.,1965, 87, 3440. I* J. R.Jones and S. E. Taylor, Tetrahedron Lett., 1981, in the press. *’ (a)J. A. Elvidge, J. R. Jones, R. Salih, M. Y. Shandala, and S. E. Taylor, J. Chem. Res. 1980, (S) 172; (M) 2375. (b)D. Lichtenberg, F. Bergrnann, and 2.Neiman,J. Chem. Soc.(C), 1971, 1676. (c) Y. Maki, M. Suzuki, K. Kameyama, and M. Sako, J. Chem. Soc., Chem. Commun., 1981,658. Log k -3 -2 -1 0 1 2 3 1 5 6 P KO Figure 2Plot of log k vs.pKs for the detritiation at the C-2 position of various azolcs (solid line, taken from ref. 53a) and the C-8 position in a number of purines. The numbers refer to the following compounds:1-methylinosine(1 ), inosine(2), 1-methylguanosine(3),guanosine(4), 9-methylhypoxanthine(9,6-mercapto-purine riboside (6), 6-mercaptopurine (7), xanthosine (8), hypoxanthine (9), purine (lo), puromycin w (1 I), guanine (1 2), adenosine (13), adenosine 3’-monophosphate (14), adenosine 5’-monophosphate (1 5) 4 adenosine 3’,5’-cyclic monophosphate (16), adenine (17), paraxanthine (18), theobromine, (1 9), cafeine (20), theophylline (21), xanthine (22), 9-Pri-purine (23), and 9-But-purine (24) (takenfrom ref. 52) Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications but which is displaced by 0.7 units on the log k axis, indicative of a lower (ca.20 %) degree of N-7 protonation. Pointsfor the xanthines [compounds (18)-(22)] are displaced still further, in line with a previous suggestion53b that significant protonation occurs at 0-6 as well as at N-9, making these substrates ca. 30-fold less reactive to exchange. Additional supportive evidence for the participation of varying degrees of N-7(9) protonation in purines is based on further LFER data,53a as well as a very recent paper53C in which pseudo-first-order deuteriation exchange rates were measured for a series of 9-substituted adenines, the highest rates being encountered with those derivatives which showed a greater tendency for N-7 over N-1 protonation (an interpretation not stressed by the authors).7 Metal-ion Effects Recent studies by Kluger54 and Cox55 have drawn attention to the fact that, although the rates of ionization of carbon acids have been extensively investigated both as a function of acid strength and the basicity of the medium, very little attention has been paid to the way in which metal ions can influence the rates, even though in both hydrolysis and hydration reactions quite startling rate accelerations have been reported. Here again, the purines are excellent substrates for investigating such effects, as they are know@ to complex very readily with a number of metal ions; however, little is known about how complex formation affects the reactivity.If, in the absence of zwitterionic contributions, the two reactive forms in isotopic hydrogen exchange from the C-8 position are (5) and (6) it can be seen that in the presence of added metal ions, species such as (7) will be formed, and (5) (7) relative reactivities can be compared. This may be done either in a competitive manner (precipitation of metal ions at high pH rather limits this approach) or by preparing and isolating the appropriately labelled complex and studying its rate of exchange directly. In the one study where both of these approaches have been used, consistent results have been obtained.57 Detritiation of l-methyl[8-3H]inosine under conditions where the operative mechanism is between the protonated substrate and hydroxide ion shows that 64 R.Kluger and P. Wasserstein, J. Am. Chem. SOC.,1973, 95, 1071. bb B. G. Cox, J. Am. Chem. SOC.,1974,96,6823. 66 R. M. Izatt, J. J. Christensen, and J. H. Rytting, Chem. Rev., 1971, 71, 439. b7 J. R. Jones and S. E. Taylor, J. Chem. Soc., Perkin Trans. 2, 1979, 1773. 338 Jones and Taylor metal ions influence the rate in different ways: ZnII ions have virtually no effect, whereas AgI ions retard the rate to a greater extent than do CuII ions. The metal- complexed species, although not as reactive as the protonated substrate, are a.c 104-106 times more reactive than the neutral form, and we have called this the ‘metal activating factor’ (maf), analogous to Stewart and Srinivasan’s ‘proton activating factor’ (~af)~~and our ‘anion deactivating factor’ (adf).59The second- order detritiation rate constant for the complex cation cis-[(l,2-diaminoethane) (g~anosine)z]Pt~~at 25 “C has a value57 of 2.7 x lo5 as compared to 6.1 x lo5 1 mol-ki for 7-methylguanosine (8) which also has a positive charge located at the N-7position.42 These results provide a rationale for the observation that co-ordination of heavy-metal ions to the N-7 position causes a rapid disappearance of the H-8 signal in the lH n.m.r.spectra of inosine and guanosine derivatives in DzO,~Oas well as explaining the absence of rate accelerations when studies are carried oul at low pH.41 This metal activating effect has important synthetic consequences: witness the mercuriation of the C-8 position of purine nucleotides with Hg(0Ac)z under mild conditionsgl and, perhaps more appropriately, the ready formation of C-8-bonded inosine and guanosine methylmercurials,@ reactions which are believed to proceed by the route shown in Scheme 1.The same principles have recently been used to identify the histidine residues that act as metal-ion (ZnII, CuII) binding sites in the metalloenzymes /3-la~tamase-II~~ and superoxide dismutase.64 The latter study, in particular, provides an example of the application of our previous findings5 of decreased rates of detritiation of [2-3H]imidazole in the presence of metal ions, inasmuch as the decreased rate of deuteriation of certain histidine residues is ascribed to metal-ion binding.s4 Similar findings have recently been reporteds6 for some CoIII complexes of imidazole.8 Magnitude of Electrostatic Factors The results in Table 1 show that protonation at the N-7 position of purines which are unable to exist as zwitterions consistently leads to rate accelerations (pafs) of the order of lo7-lo9, ca. 102-103 greater than the effects observed with the few metal ions so far studied.57 Only in the case of tetrazole (9), one of the strongest 68 R. Stewart and R. Srinivasan, Acc. Chem. Res., 1978, 11, 271. I* S. E. Taylor, Can. J. Chem., 1980, 58, 86. 6o (a)S. Mansy and R. S. Tobias, J. Chem. Soc., Chem. Commun., 1974,957; (b)S. Mansy and R. S. Tobias, Biochemistry, 1975, 14, 2952; (c) G. Y. H. Chu and R. S. Tobias, J. Am. Chem. Soc., 1976,98,2641.61 R. M. K. Dale, D. C. Livingstone, and D. C. Ward, Proc. Natl. Acad. Sci. USA, 1973, 70, 2238. 6a (a) E. Buncel, A. R. Norris, W. J. Racz, and S. E. Taylor, J. Chem. Soc., Chem. Commun., 1979, 562; (b) E. Buncel, A. R. Norris, W. J. Racz, and S. E. Taylor, Inorg. Chem., 1981, 20, 98. 6s G. S. Baldwin, S. G. Waley, and E. P. Abraham, Biochem. J., 1979, 179, 459. dl (a) A. E. G. Cass, H. A. 0.Hill, J. V. Bannister, W. H. Bannister, V. Hasemann, and J. T. Johansen, Biochem. J., 1979, 183, 127; (b) J. C. Dunbar, J. T. Johansen, A. E. G. Cass, and H. A. 0. Hill, Carlsberg Res. Commun., 1980, 45, 349. 6s D. H. Buisson, J. R. Jones, and S. E. Taylor, J. Chem. Soc., Chem. Commun., 1975, 856. I6N. S. Rowan, C. B. Storm, and R. Rowan, J.Znorg. Biochem., 1981, 14, 59. 339 Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications MeHgN03I H gMe R R Scheme 1 Table 1 Proton activating factors (pafs), anion deactivating factors (adfs), and metal activating factors (mafs) at 85 "Cfor some purines and azoles Compound Paf adf maf Refs. Adenosine 1.1 x 107 29, 58 9-Pri-purine 1.3 x los 28, 58 9-But-purine 1.2 x 108 28, 58 Caffeine 9.3 x 107 37, 59 Theobromine" 5.9 x 107 7.2 x 10-2 37, 59 Paraxanthine" 5.6 x lo4 1.8 x 37, 59 Guanosinea 5.1 x 105 36,42 1-Methylguanosine 2.1 x 109 6 x 106(Cu11j 42, 57 Inosine" 9.3 x 105 36, 42 1-MethyIinosine 9.5 x 108 1 x loyCu'1) 42, 57 2 x 104(~g9 Thiazole 8.3 x 108 53a Benzoxazole 1.4 x lo8 53a Tetrazole 6.3 x 107 2.3 x 10-7 53a aPotentially zwitterionic molecules.heterocyclic carbon acids,2 has it been possible to compare the effects of protona-tion and ionization at similar sites.53a Indeed, log(adfj for ionization at N-1 (-6.64) is very nearly equal and opposite to log(pafj for protonation at N-4 Jones and Taylor Akibose (7.79).67The latter value is subject to some uncertainty because the pKa used in its calculation was for 1-methyltetrazole.53a These large effects, arising as they do from protonation or ionization at ‘fixed’ sites, contrast markedly with the very small electrostatic effects associated with charged groups in a more ‘flexible’ arrangement.Thus, a comparison of the data for histidine and histamine exchange43 shows that the inhibiting effect of a negatively charged carboxylate group is small, as is also the case for a phosphate group in various nucleo- tides.35940 Similarly, in histidine derivatives when protonation occurs both in the imidazole ring and on the amino-group, the reactivity of the doubly protonated species is similar to that of singly protonated substrates of the same pL43 The main effect of these charged groups in the side-chain is to facilitate proton transfer via zwitterion formation, as shown in equations (5) and (6) for histidine and adenosine 5’-mono p hosp ha te, respectively . In compounds capable of existing as zwitterions, c.g. guanosine, the measured H -0zc NH2 0-I I IRibose-b=O Ribose -p=oI I0-0-It is unlikely that in solution the proton is localized at N-4; however, we are assuming that exchange from the N-4-protonated species will be kinetically dominant.Isotopic Hydrogen Exchange in Purines- Mechanisms and Applications paf values are lower than for compounds unable to form such species (Table 1). This is because the observed second-order rate constant, kobs is (as mentioned previously) a composite of two terms, k’ and k* which refer to the neutral mole- cule and zwitterion, respectively [equation (4)]. In the case of those xanthines that react as the N-1 or N-3 anions (theobromine and paraxanthine, respectively), the corresponding adfs have been evaluated.59 Consequently, as the zwitterions of these species differ from the protonated molecule only inasmuch as ionization has occurred at N-1, equation (7)holds.Using the derived value of k* in equation k+ = k x adf (7) (4) it is now possible to calculate values of Kzw; experimentally, these values cannot be measured dire~tly.5~ The adf values so far. obtained (Table 1) for ionization at N-1 are as expected on electrostatic grounds, and are a good deal smaller than those resulting from ionization at a group immediately adjacent to that undergoing exchange. 9 Ionization Constants In view of the importance of heterocyclic compounds and the fact that they can exist in various ionized forms in aqueous solution, a great deal of effort has been directed to the measurement of their various ionization constants. Isotopic hydrogen exchange studies such as ours provide a convenient route to such information.Thus, for compounds such as purine and adenine, which conform to equation (2) and give bell-shaped rate-pH profiles, it can be shown that the pseudo-first-order rate constant, k+, is given by equation (8). k K, KaKa’k# = Ka + [H+] + -[H+l Ka = [BH][H+]/[BH2+] and Ka’ = [B-] [H+]/[BHl28i29933 In the pH-rate independent region, k,, = k Kw/Ka, so that if the rate at low pH is expressed relative to this value, equation (9) can be derived.33 PH = PKa + logm[R/(l -R)] (9) A plot of pH against logio [R/(l -R)] therefore provides a value of the PKa. At high pH, when Ka 9 Ka’ > [H+], equation (10) holds, such that inspection of the rate-pH profile at R = 0.5 gives the pKa‘ value.tH+lR= [H+] + Ka’ For compounds that give rise to other kinds of rate-pH profiles it is still possible to obtain values of the ionization constants. In such cases, the rates are again expressed in relative terms and the best fit of a theoretical rate-pH profile to the experimental data provides the necessary information. Most of the acidity constant data in the literature refer to a single tempera- ture, usually 25 “C. Frequently the results are required at other temperatures. Jones and Taylor As -d(dG")/dT = AS" and dG" = 2.303RT PKa,, equation (11) holds.68 For equilibria of the kind BHzf + BH + H+, Perrin68 was able to show that AS" values lie in the range -4 f 6 cal K-lmol-l, so that equation (1 1) reduces to -d(pK,)/dT = (pKa -0.9)/T.This equation, named after its author, has met with a good deal of success, and prompted us to look at equilibria of the kind BH + B-+ H+ (ionization constant Ka').ggValues of pKa' at 85 "C for many compounds have been obtained from our isotope exchange experiments and, as equation (1 1)can be rearranged to give equation (12), in which a = -0.218dS0,a plot of pKa' (25°C)against PKa' (85°C) should be linear with slope 1.25. This is pKa'(85 "C)= 1.25 pKa'(25 "C)-0.2% (12) found to be so,G9 and the derived value of a (5.4) converts to AS" = -25 cal K-lmol-l, much more negative than is the case for univalent cations, and consistent with the fact that experimental values of AS"lie in the range -(1 3-22) cal K-lmol-1.56 Although this equilibrium, unlike the first, is not isoelectric, the differences in the solvation requirements of BH2+ and B-are probably approxi- mately constant.Equation (13) therefore describes the temperature dependence of the acidity constant Ka'. 10 Miscellaneous Although the C-8 hydrogen in purines is not sterically hindered in any way, the same is not true in polynucleotides, such as ribonucleic acids; here each nucleo- tide base is in a different environment, governed by the primary structure and secondary factors. The ways in which these considerations can influence rates of exchange have been studied by both Maslova70 and Schimmel71 and their respec- tive co-workers, and it is clear that not only are the rates extremely sensitive to the local microenvironment but that the method offers a powerful way of studying protein-nucleic acid interactions, and should be applicable to a wide variety of systems.(a)D. D. Perrin, Aust. J. Chem., 1964, 17,484; (b)D. D. Perrin, J. Chem. Soc., 1965, 5590. J. R. Jones and S. E. Taylor, J. Chem. Res., 1980, (S) 154. 70 (a)R. N. Maslova, E. A. Lesnik, and Ya. M. Varshavsky, Biochem. Biophys. Res. Commun., 1969, 34, 260; (b) Mol. Biol., 1969, 3, 575; (c) FEBS Lett., 1969, 3, 211; (d) E. A. Lesnik, R. N. Maslova, T. G. Samsonidze, and Ya. M. Varshavsky, FEBS Lett., 1973,33, 7; (e)G. N. Lapiashvili, E. A. Lesnik, R. N. Maslova, and Ya. M. Varshavsky, Nucleic Acids Res., 1977; 4,2181. 'l (a) R.C. Gamble and P. R.Schimmel, Proc. Natl. Acad. Sci. USA, 1974, 71, 1356; (6) R. C. Gamble, H. J. P. Schoemaker, E. Jekowsky, and P. R. Schimmel, Biochemistry, 1976,15, 2791; (c) H. J. P. Schoemaker, R. C. Gamble, G. P. Budzik, and P. R. Schimmel, Bio-chemistry, 1976,15,2800; (d)P. R. Schimmel, Adv. Enzymol. Relat. Areas Mol. Biol., 1979, 50, 187. Isotopic Hydrogen Exchange in Purines-Mechanisms and Applications Rates of detritiation from the C-8 position of purines, as for other exchange reactions, invariably give first-order kinetics. Departure from this behaviour usually signifies the onset of another reaction, and in the case of various purine nucleosides the curvature of the first-order plots can be used to obtain the acid- catalysed hydrolysis rate constant, when analysed according to the sequence shown in Scheme L72 k hydrolysisNucleoside .------++ Base + P-D-ribose detritiation kl detritiation k,1 1Scheme 2 A novel method of assaying the guanine content of DNA has been devised73 that takes advantage of the extreme lability of the C-8 proton in 7-methyl- guanosine.The procedure entailed biosynthetically labelling DNA using [8-3H]guanosine, followed by methylation with dimethylsulphate; the tritium was rapidly released from the 7-methyl-[8-3H]guanosineresidues. 11 Concluding Remarks The biochemists’ frequent preference for 14C-rather than 3H-labelled compounds, despite the fact that their preparation is usually more demanding in time, skill, and expense, is more often than not based on the dangers associated with possible adventitious hydrogen exchange reactions.The chemist, on the other hand, finds that these reactions can be a useful way of investigating details of reaction mechanisms. During such studies one can gain an appreciation of the many factors that can influence the exchange rates, and how these can be applied to good effect. Consequently, when such studies are carried out on important and interesting compounds, such as the purines, one is presented with a fertile area of research, both pure and applied. Various aspects of the work described here have been supported by Amersham International (formerly The Radiochemical Centre), the Science Research Council, and NATO. J. R.Jones and S. E. Taylor, Int. J. Chem. Kinet., 1980, 12, 141. 73 M. Tomasz, Biochim. Biophys. Acra, 1970, 199, 18.

ISSN:0306-0012    

DOI:10.1039/CS9811000329

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

INGOLD LECTURE* How Does a Reaction Choose Its Mechanism? By W. P. Jencks GRADUATE DEPARTMENT OF BIOCHEMISTRY, WALTHAM, MASSACHUSETTS 02254, U.S.A. 1 Introduction Ingold laid the foundation of modern organic chemistry by constructing a classification of reactions and their mechanisms.l Progress in science requires a language and Ingold’s examination of reaction mechanisms led to a classification that provides a rational language for communication, generalization, and prediction in organic chemistry. This language has had an important influence on the development of synthetic organic chemistry and other branches of chemistry, as well as on our present understanding of organic reaction mechanisms. Much of the experimental work on reaction mechanisms has been concerned with fitting reactions into the Ingold scheme or other schemes, such as Winstein’s classification of ion-pair intermediates in solvolysis reactions.2 In comparison there has been surprisingly little inquiry into the question of why a reaction should follow one mechanism rather than another under a particular set of experimental conditions.For example, it is generally agreed that nucleophilic substitution on carbon follows an sN2 mechanism for methyl transfer and an SN~mechanism when a stable carbocation intermediate can be formed easily, but it is not so clear what is responsible for changing mechanisms in the ‘borderline’ region (Figure 1). It is particularly important to have a clearly defined classification of mechanism in this borderline region, which may well be larger than the regions of well established mechanism.In chemistry, as in other areas, lack of agreement upon the position of sharp borderlines invariably leads to conflict. Distinctions between mechanisms of chemical reactions in solution are concerned in large part with the sequence in which reactants are assembled and dispersed in relation to the bond-making and -breaking steps. The purpose of this review is to examine the extent to which the choice of reaction mechanism is dictated by the lifetime of intermediates that may be formed in a reaction. It appears that many reaction sequences are enforced in a simple way by these life- times; a relatively small number have been shown not to be enforced.It has frequently been suggested that a clear-cut distinction between reaction *Delivered at a symposium of The Royal Society of Chemistry Perkin/Faraday Divisions at University College, London, on 12 March 1981. C. K. Ingold, ‘Structure and Mechanism in Organic Chemistry’, 2nd Edn., Cornell Univ. Press, Ithaca, New York,1969. S. Winstein and G. C. Robinson, J. Am. Chem. SOC.,1958, 80, 169. How Does a Reaction Choose Its Mechanism? SN2 ............... ......0%. ..... 3 5-...intimate ................%, .. Ion pai Ts ......... Nu Ph PhSN1 \;I X-L I Ph Figure 1 Mechanisms and borderlines for substitution on carbon mechanisms is impossible because, for example, there is a gradual transformation of an sN2 into an SN~mechanism with no sharp borderline as the transition state develops more carbocation character.3 However, a clear distinction can be made if the classification of mechanism is based upon the lifetime of intermediates rather than the character of the transition state.The lifetime of intermediates permits a fairly sharp qualitative distinction between mechanisms, whereas the character of the transition state or the degree of assistance in a reaction gives only a quantitative description with no sharp boundaries. It is useful to illustrate the distinction with reaction co-ordinate-energy contour diagrams as described by More O'Ferrall in 1970 for elimination reactions4 (Figure 2). A reaction can proceed either through an intermediate in a potential well that provides barriers for both the formation and breakdown of the intermediate (A and B), as in an Elcb elimination mechanism, or through a concerted, one-step mechanism with a single barrier and no intermediate (1,2,or 3), as in an E2 elimination.If an intermediate is said to exist if it has a lifetime longer than a vibration frequency, of the order of 10'3 s-l, there is a sharp border- line between the stepwise and concerted mechanisms.* A concerted mechanism with no intermediate can proceed through transition states with varying degrees *Encounter complexes of the reactants or products are not kinetically significant intermediates in this sense, except in the case of diffusion-controlled reactions. See, for example, S. Winstein, E.Grunwald, and H. W. Jones, J. Am. Chem. Soc., 1951,73, 2700. R. A. More O'Ferrall, J. Chcm. Soc. B, 1970, 274. Jencks Figure 2 Reaction co-ordinate-energy diagram to show how a reaction that requires two processes, A and B,can occur in two steps through an intermediate with a signi$cant lifetime, Int, or through concerted mechanisms, 1,2, or 3, in which the transition states have varying degrees of resemblance to the structure of the intermediate of resemblance to the structure that an ‘intermediate’ might be expected to have and with varying degrees of coupling of the two processes that are involved in the formation and breakdown of the ‘intermediate’, as shown in 1, 2, and 3 (Figure 2). The qualitative distinctions between mechanisms are naive in the sense that they do not take explicit account of the degree of coupling and mechanisms of energy transfer in the activation process.Energy transfer can be relatively slow and energy diagrams certainly do not provide a complete description of the course of a reaction. Nevertheless, these distinctions may be useful as a simple guide for describing and predicting reaction mechanisms. The question remains of what is the meaning of the ‘merging’ of mechanisms? How is one mechanism changed into another as the structure of the reactants or the reaction conditions are changed? One possibility is that one mechanism becomes, or is transformed into, the other. This can happen when the inter- mediate in a stepwise mechanism becomes progressively less stable and event- ually ceases to exist, so that the well corresponding to the intermediate in Figure 2 disappears and the reaction becomes concerted.The other possibility is that the two mechanisms can exist concurrently, so that there is a well for the inter- mediate in the stepwise path but the reaction also proceeds through a concerted path, i.e. 1, 2, or 3. In this case there is a change in the predominant mechanism when there is a reversal of the relative Gibbs energies of the rate-determining transition states for the two coexisting mechanisms. The change in mechanism then may not be enforced by the lifetime of the intermediate, if the reactants are in approximately the same position relative to each other. A change between two coexisting mechanisms will usually give a sharp upward break in structure- reactivity correlations as the second mechanism becomes predominant.However, 347 3 How Does a Reaction Choose Its Mechanism? upward curvature may also occur with a single mechanism as the structure of the transition state changes. There are only a few cases in which this question can be ariswered at the present time. A scheme for distinguishing reaction mechanisms is shown in Figure 3. This PREASSOCI AT I ON ASS I STANCE CONCERTED COUPLED -___------- PC NO ASS I STANCE PREASSOCIATION ASS ISTANCE STEPWISE HYDROGEN BOND1 NG -_---__----NO ASS1 STANCE PS * *?:;'::': Borderline :': ;k 9: ;': ?: $: :t J-J-f LIBERATED c D IFFUS ION CONTROLLED 1NTERMEDI ATE LI ID LI J __-_-___----c ACT1 VAT ION LIM I TED LI *A Figure 3 Classification of reaction mechanisms scheme provides borderlines between mechanisms that depend on the lifetimes of intermediates.We consider here only unstable, steady-state intermediates that do not accumulate during the reaction. There are two primary distinctions or borderlines: (i) between mechanisms that are concerted with no intermediate (except for encounter complexes of reactants and products) and mechanisms that proceed through one or more intermediates in a stepwise process, (ii) between mechanisms in which the intermediate either does or does not have a sufficient lifetime to diffuse through the solvent before reacting with a catalyst or another reactant.When the intermediate does not exist or is too unstable to diffuse through the solvent the reaction must occur through a preassociation mechanism in which the reactants, including the final reactant or catalyst, C, are assembled before the first bond-making or -breaking step occurs. The preassociation mechanism can be either concerted with no intermediate, PC, or stepwise with an intermediate, PS (Figure 3).** If the intermediate lives long enough to diffuse out **The term preassociation has often been applied to stepwise reactions. There is also pre- association of the reactants in concerted reactions and the term is properly applied to both concerted and stepwise reactions in which a final reactant is present at the time of the initial bond cleavage or formation. Jencks of the solvent cage in which it is formed, it becomes a liberated intermediate and can react with a final reactant or catalyst, C, either at a diffusion-controlled rate, LI -D, or in an activation-limited reaction, LI -A (Figure 3, bottom).The mechanism can be described further by the degree of assistance that is provided by a catalyst or reactant in the rate-determining transition state. It is important to separate this quantitative criterion from the qualitative distinction based on lifetimes, because assistance may be either present or absent in both concerted and stepwise mechanisms. A concerted reaction can occur through a coupled mechanism with assistance by the final reactant or catalyst, as in a classical sN2 displacement, or through an uncoupled mechanism in which there also is no intermediate, but the second process has little or no influence on the energy of the rate-determining transition state.These two mechanisms might be described by the solid and dashed lines, respectively, at the top of Figure,3. A stepwise preassociation mechanism can also occur either with assistance, such as hydrogen bonding of an acid catalyst to a basic site in a transition state, or with- out assistance, as in a ‘spectator’ mechanism in which the catalyst is present but does not stabilize the rate-determining transition state of the preassociation mechani~m.~Assistance by the solvent is more difficult to characterize and is not generally useful as a criterion for distinguishing mechanisms. 2 Preassociation Concerted Mechanisms A.Substitution Reactions.-Ingold’s definition of an sN2 substitution is a model of clarity and deserves quotation. The mechanism ‘. . . contains only one stage, in which two molecules simultaneously undergo covalency change.’6 Few will argue today against such a concerted mechanism for nucleophilic displacements on the methyl group, which certainly cannot form a carbocation intermediate with a significant lifetime in the presence of any respectable n~cleophile.~-~ It is surprising, however, that several reactions of methoxymethyl derivatives also appear to proceed by concerted mechanisms, in spite of the potential of these compounds to form the relatively stable oxocarbonium ion (1).Methoxymethyl derivatives, such as methyl chloromethyl ether and fornialdehyde acetals, certainly react through transition states that resemble (1) and have been widely believed to react through a monomolecular mechanism with (1) as an intermediate.1°-12 How-ever, the methoxymethyl derivatives (2) and (3), with dinitrophenolate ion or NN-dimethylanilines as the leaving group, undergo second-order displacement reactions in aqueous solution with various nucleophilic reagents.13J4 The second- s L. D. Kershner and R. L. Schowen, J. Am. Chem. SOC.,1971,93,2014. (I C. K. lngold in ref. 1, p. 423. W. von E. Doering and H. H. Zeiss, J. Am. Chem. SOC.,1953,75,4733. M. H.Abraham and D. J. McLennan, J. Chem. SOC.,Perkin Trans. 2, 1977, 873. W. J. Albery and M. M. Kreevoy, Adv. Phys. Org. Chem., 1978, 16, 87. lo W. Cocker, A. Lapworth, and A. Walton, J. Chem. SOC.,1930, 440. I1 P. Ballinger, P. B. D. de la Mare, G. Kohnstam, and B. M. Presit,J. Chem. SOC.,1955,3641. lP T. C. Jones and E. R. Thornton, J. Am. Chem. SOC.,1967, 89, 4863. 13G.A. Craze, A. J. Kirby, and R. Osborne, J. Chem. SOC.,Perkin Trans. 2, 1978, 357. l4 B. L. Knier and W. P. Jencks, J. Am. Chem. SOC.,1980,102,6789. How Does a Reaction Choose Its Mechanism? order rate constants show a small dependence on the structure of the nucleophilic reagent, which is intermediate between that expected from the Swain-Scott scale for substitution on methyl halides and the Ritchie N+ scale for addition to carbonium ions.The rate constant for solvolysis is accounted for by the rate constant for the second-order displacement reaction with water that is predicted by these correlations, i.e. there is no indication of any solvolysis reaction that proceeds by a different mechanism. There is a large amount of bond-breaking at the leaving group. The reactions of both strong and weak nucleophiles with (3) cannot be accounted for by ion-pait: or ion-dipole intermediates. The reactions occur through an open, ‘exploded’ transition state (4) that closely resembles the oxocarbonium ion (1) but there is significant stabilization of this transition state by the inconling nucleophilic reagent. The concerted reaction mechanism appears to be enforced by the short lifetime of the oxocarbonium ion (1).A lifetime of -s for (1) in water was estimated from an extrapolation of measured lifetimes of oxocarbonium ions derived from acetophenone acetals.l6 Although this estimate is too uncertain to give a definitive conclusion, the estimated ‘lifetime’ of -10-23 s for (1) in the presence of RS-is too short to reconcile with a stepwise mechanism for substitution that proceeds through an intermediate with a significant lifetime.’* The reactions exhibit variable secondary u-deuterium isotope effects ranging up to k~/k~= I. 18 with different nucleophiles. which must reflect differences in the nature of nucleophilic interactions with the central carbon atom in the transition state.The large values for these second-order reactions also show that a-deuterium isotope effects of this magnitude cannot be taken as evidence for a monomolecular reaction mechanism. The transition state (4) may be regarded either as an unusually loose transition state for an SN~reaction or as a carbocation that is stabilized by interactions with both the attacking and leaving groups. Substitution and solvolysis reactions at the anomeric carbon atom of sugars have also been widely believed to proceed through an oxocarbonium ion inter- mediate, but these reactions certainly proceed through a preassociation mech- anism and may well proceed through a concerted mechanism with assistance in l5 P. R. Young and W. P. Jencks, J.Am. Chem. SOC.,1977,99,8238. Jencks nucleophilic solvents. The rate of acid-catalysed hydrolysis of methyl B-D-glucopyranoside is -lo3slower than that of formaldehyde dimethyl aceta1,16J7 so that the oxocarbonium ion derived from sugars is unlikely to be more stable than that derived from formaldehyde derivatives. Solvolysis of a series of 01-and /3-glucosidesin 50 ”/, ethanol-trifluoroethanol gives different product ratios with different leaving groups, which shows that no common intermediate is formed that has a lifetime sufficient to become liberated from the leaving group by diffusion into the bulk solvent.ls These reactions also give product ratios in which substitution by ethanol is favoured over trifluoroethanol by factors of up to 20, which shows that the incoming group can stabilize the transition state by corresponding ratios.Ion-pair intermediates cannot be formed with the uncharged leaving groups, and when phenol is the leaving group there must be bond- breaking in the rate-determining step (phenol is a weaker nucleophile than the solvent,so that ifan intermediate were formed it would give a product rapidly after the bond-breaking step). Thus, the C-1 atom of the sugar is interacting with both the leaving group and the entering group in the transition state (4). It is even more surprising that the relatively small fraction of the reaction that goes with retention of configuration gives similar product ratios that require a similar stabilization of the transition state by the more basic solvent molecule.This is presumably made possible by the open, ‘exploded‘ transition state, which resembles the transition state for diffusion away of the leaving group. Front-side substitution is not usually expected for SK~displacements on carbon but is well known for displacement on silicon, for which there is some theoretical rationale, and for displacement on metals, which also can occur through open, ‘exploded’ transition states that allow a weak interaction with both the entering and leaving groups.19 Lysosyme and related enzymes must provide considerable stabilization to a glycosyknzyme intermediate and to the transition state for its formation, because these enzymes catalyse glycosyl transfer to dilute sugars with retention of configuration, as well as to water.The intermediate cannot be an oxocarbonium ion because an oxocarbonium ion would not have a suficient lifetime to permit diffusion and reaction with a molecule of sugar before it reacts with water. The intermediate is presumably a species that is stabilized by some degree of bonding to the aspartate carboxylate group at the active site of lysosyme.J5 B. Carbanion and Elimination Reactions.-Condensation and elimination reactions that are generally thought to proceed through carbanion intermediates must proceed through a concerted mechanism when the carbanion is not stabilized and has no significant barrier for protonation, condensation, or elimination. The reverse aldol-type cleavage of 1-phenylcyclopropanol to 1-phenylpropanone [equation (l)], for example, would certainly proceed through a l6 P.Salomaa, Suom. Kemistil. By1960, 33, 11. l7 D. Cocker and M. L. Sinnott, J. Chem. SOC.,Perkin Trans. 2, 1975, 1391. M. L. Sinnott and W. P. Jencks, J. Am. Chem. SOC.,1980, 102, 2026. l* N. T. Anh and C. Minot, J. Am. Chem. SOC.,1980, 102, 103. How Does a Reaction Choose Its Mechanism? Ph l* Ph -04+ HA+ carbanion intermediate if that intermediate had a significant lifetime. However, this reaction proceeds through a preassociation mechanism that probably is concerted because the unactivated primary carbanion has a pK 248 and no significant barrier for protonation ;a crude calculation suggests that the carbanion is less stable than the transition state of the observed reaction.20-22 The reaction shows general acid catalysis by buffer acids with a Brsnsted slope of a = 0.25 and a primary deuterium isotope effect of k~/k~= 1.9 & 0.2, consistent with a concerted SE2 reaction mechanism that proceeds through an open, ‘exploded‘ transition state (5).2l The same mechanism must hold for aldol-type condensations through a transition state resembling a homoenolate ion in the reverse direction.21g22 Other elimination reactions of compounds with little or no activation at the p-carbon atom or with leaving groups that are expelled with no activation barrier must proceed by an analogous enforced concerted mech- anism.23 It is conceivable that assistance through a preassociation mechanism of this kind facilitates enzyme-catalysed reactions that would require the rapid formation of unstable carbanions.21 C.Geiieral Acid-Base Catalysis of Complex Reactions.-A concerted mechanism is probable, although not proved rigorously, for general acid catalysis of tri-fluoroethanol addition to formaldehyde, with catalysis at the electrophilic reagent through a class e reaction [equation (2); RX = ROH, >C=Y = HCH01.24 In the reverse direction this mechanism corresponds to concerted general base catalysis of the elimination of ROH from the protonated addition \ +I RX + \C=Y+ HA +[RX.-.C~Y***H...RX-C-YH + A-A]* F= / / i (2) compound to give the carbonyl compound. The rate constant for this elimination is N 105 higher than the calculated rate constant for proton removal to form the 1o C.H. DePuy, Trans. N. Y. Acad. Sci., Ser. 2, 1966, 28, 561. 21 A. Thibblin and W. P. Jencks, J. Am. Chrm. SOC.,1979, 101, 4963. 22 D. H. Hunter, J. B. Stothers, and E. W. Warnhoff in ‘Rearrangements in Ground and Excited States’, ed. P. de Mayo. Academic Press, New York, 1980, Vol. 1, p. 400. 23 W. H. Saunders, jun., Acc. Chem. Res., 1976, 9, 19. 24 L. H. Funderburk, L. Aldwin, and W. P. Jencks, J. Am. Chem. SOC., 1978, 100, 5444. Jencks dipolar intermediate (6) in a stepwise mechanism. The rate constants that would be required for reaction of (6) in order to account for the observed rare have been estimated to be 2 1013 s-l for the expulsion of KOH and > 1014s-l for proto- H I RN--C-OCH2CF3IIHH nation by H30’ in an encounter complex.Kate constants of this magnitude are inconsistent with the existence of two significant barriers for a stepwise reaction through this ‘intermediate’. 1t is even more unlikely that the reaction proceeds through a dipolar intermediate analogous to (6) when the driving force for its breakdown is larger, as in the formation of a more stable carbonyl product such as a ketone or a resonance-stabilized ester or amide. A higher ‘rate constant’ of-10l6s-l would be required for the expulsion of trifluoroethoxide ion from the anion (7; R = p-MeCsH4SO2NH) through a stepwise mechanism, which requires that base-catalysed hydrazone formation from the parent carbinolamine must proceed through a concerted mechanism.25 There is strong evidence supporting a concerted mechanism for catalysis by general bases of the addition of alcohols and water to electrophilic carbon centres, with catalysis at the nucleophilic reagent through a class n mechanism [equation (3)].In the reverse direction this mechanism corresponds to general acid catalysis of the expulsion of RO-. This is a widespread mechanism, which is responsible for hydrolysis and hydration reactions, for example. The concerted mechanism is supported by the occurrence of solvent deuterium isotope effects, usually in the range kROH/kROD = 2-4, and a large body of structure-reactivity data, which provide evidence that both proton transfer and C--0 bond-for-mation or -cleavage are taking place in the transition state.24*2G-28 The simplest J.M. Sayer and W. P. Jencks, J. Am. Chem. Sac., 1977, 99,464. a6 N. Gravitz and W. P. Jencks, J. Am. Chem. SOC.,1974, 96, 507. J. L. Palmer and W. P. Jencks, J. Am. Chem. SOC.,1980, 102, 6466. a8 J. L. Palmer and W. P. Jencks, J. Am. Chem. SOC.,1980, 102, 6472 and references therein. How Does a Reaction Choose Its Mechanism ? such evidence is that there is a smooth transition between the development of net positive or negative charge on the central oxygen atom in the transition state, as shown by the changing dependence of the rate on the pKs of the alcohol (Pnucorpig).This requires that there be a changing balance between the amounts of proton transfer and C-0 bond-formation (or -cleavage) in the transition state with changing substituents on the oxygen atom or the catalyst. The direction and amount of these changes can be explained by a concerted mechanism that corresponds to a diagonal reaction co-ordinate on an energy contour diagram with separate axes for proton transfer and C-0 bond formation. The structure- reactivity data suggest that the reaction is best described as an electrophilic attack on the central oxygen atom by the proton or by the electrophilic carbon centre, which drives the reaction by withdrawing electrons from the bond to the carbon or to the proton, respectively.28 Concerted catalysis of the addition and elimination of amities is less common because the greater stability of protonated ainines favours a stepwise mechanism with a protonated addition intermediate.There is evidence for a concerted class n mechanism for a few reactions that would be expected to give highly unstable intermediates with little or no significant lifetime.29-31 With a bifunctional acid-base catalyst two protons can be transferred in a concerted process through an 8-membered cyclic transition state, because the transfer of each proton increases the basicity of the adjacent basic site and the acidity of the other acidic site by an electrostatic effect. This has been observed for catalysis of the methoxyaminolysis of phenyl acetate by phosphate, arsenate, and similar catalysts [equation (4)].32 Acid catalysts catalyse methoxyamine \/\d /x,Q--$ ‘pHY Hk kl[HA1MeONH2 + MeCOPh F-N ’+ 0--.--fi 0 @+productsk-c \/ k-, ‘ ’‘c/‘OPh kc -1 ‘F, (4)OPh attack by hydrogen bonding to the carbonyl group in a stepwise preassociation mechanism (kl).With monofunctional catalysts there is downward curvature of the Brernsted plot and a sharp maximum in the solvent isotope effect with decreasina acid strength, as the proton-transfer step becomes kinetically sig- nificant near dpK = 0. The same proton-transfer step would be required for bifunctional catalysts if the two proton transfers were stepwise, so that the absence of both downward curvature of the Bronsted plot and an isotope effect gg M. I. Page and W. P. Jencks, J. Am. Chem. SOC.,1972, 94, 8828. so R. Kluger and C.-H.Lam, J. Am. Chem. Soc., 1978, 100, 2191. J. J. Morris and M. 1. Page, J. Chem. Soc., Perkin Trans. 2, 1980, 685. 32 M. M. Cox and W. P. Jencks, J. Am. Clrcm. Soc., 1981, 103, 580. Jencks maximum provides evidence that the two proton transfers with these catalysts occur through a fast, concerted process that never becomes kinetically significant. D. Are Concerted Mechanisms Enforced?-It is true by definition that a reaction is concerted when it proceeds in one step because all ‘intermediate’ species are too unstable to exist. This appears to be thc most common reason that reactions do proceed by a concerted mechanism. The converse question remains: when an intermediate and a stepwise mechanism are known to exist, can a reaction also proceed through a concerted mechanism with the reactants in approximately the same position relative to each other? Jf the answer is no, the merging of mech- anisms represents the transformation of a stepwise into a concerted mechanism as the intermediate ceases to exist whereas, if the answer is yes, the merging of mechanisms represents a change in relative transition-state energies such that the concerted becomes faster than the stepwise niechanism under conditions in which both mechanisms occur concurrently (Figure 2).If a reaction occurs in two steps A and B with an activation barrier for each step, the coexistence of a concerted reaction requires that there be a large advan- tage fr6m coupling the two steps into one so that the barrier for the concerted reaction becomes comparable to or lower than that of both of the steps of the step- wise reaction.In an elimination reaction, for example, this requires that the effective sum of the barriers for proton removal from carbon and for carbon- leaving group cleavage be reduced by coupling between these two processes to give a low-energy transition state for the concerted reaction. This is not so likely when the individual barriers are large, as in many substitution and elimination reactions of carbon compounds; it becomes progressively more likely as the individual barriers for collapse of the intermediate become smaller and is very likely when they disappear, if the geometry of the system is favourable. It is still not clear at what point the concerted pathway appears and whether or not stepwise and concerted mechanisms with a similar geometry can coexist for activation-limited processes of this kind.The notion does not appear to have been disproved that concerted reactions of this kind are concerted simply be-cause intermediates of the corresponding stepwise mechanisms are too unstable to exist, i.e. the reaction will proceed through an intermediate if it can. The barriers for proton transfer between electronegative atoms are generally much smaller than for carbon, so that it is more likely that the advantage from coupling of such a proton transfer with some other step will outweigh the dis- advantage of adding the barriers for the twq steps into a single concerted step.Consequently, stepwise and concerted mechanisms do coexist for complex general acid-base catalysis. For example, dehydration of the carbinolamine of formaldehyde and semicarbazide proceeds by concerted general acid catalysis [kc,equation (5)] in spite of the fact that the leaving oxygen atom is protonated (kl) some 104 faster than the observed dehydration rate under the same con- ditions. The protonated hydroxy-group and a stepwise mechanism of specific acid catalysis must exist, although the equilibrium constant for protonation How Does a Reaction Choose Its Mechanism ? kc \+\NCH20H + HA +N=CH2 + H20 + A-/ /%$ (5) fA-H \NCH2-0+/ / \H and the barrier for C-0 cleavage (klkzlk-1) are too unfavourable for this pathway to make a significant contribution to the observed reaction rate.27 Two kinds of circumstances can favour the coexistence of concerted and step- wise mechanisms.(a) Stepwise and enforced concerted mechanisms can coexist when the individual steps are separated in space and one step occurs by the diffusion-controlled reaction of an intermediate. The reaction then occurs by two separate pathways, in which the first step occurs either in the absence or in the presence of the final reactant. For example, a carbocation may have a significant lifetime in a solvent but no lifetime when it is in contact with azide ion, so that a reaction could occur by a stepwise mechanism, with diffusion- controlled combination of N3- and the intermediate, and by a concurrent nucleo- philic attack of azide through an enforced concerted displacement mechanism.Similar parallel pathways are possible for general acid-base catalysis and other reactions. They may be described by adding one or more ‘wings’ to the diagram of Figure 2 for the diffusional combination steps.33~3~ (b) Different requirements for the concerted and stepwise mechanisms can facilitate their coexistence if the barriers for the two mechanisms are not very different. For example, a concerted E2 elimination may require an antiperiplanar conformation of the reacting atoms that is sterically unfavourable. A concurrent Elcb mechanism that does not require this conformation will then be facilitated, even if the carbanion expels the leaving group with no barrier when it is in the correct conformation.An analo- gous situation is possible for the arrangement of solvent molecules to solvate the leaving group in the concerted reaction. E. Uncoupled Concerted Reactions.-When the coupling between two processes such as nucleophilic attack and leaving-group expulsion is weak because of unfavourable geometry and orbital overlap, a reaction will be concerted only when there is no barrier for one of the steps. All such concerted reactions therefore proceed by an enforced concerted mechanism, as indicated by the dashed line in the top diagram of Figure 3. Although there are few quantitative data available such mechanisms are probable for displacements at sp2carbon, such as acyl-group transfer, nucleophilic aromatic substitution, and nucleophilic vinylic substitution 33 S.Rosenberg, S. M. Silver, J. M. Sayer, and W. P. Jencks, J. Am. Chem. SOC.,1974, 96, 7986. 34 1974, 96,J. M. Sayer, B. Pinsky, A. Schonbrunn, and W. Washtien, J. Am. CIwm. SOC., 7998 356 Jencks with retention, when substrates have good leaving groups and cannot form a stable intermediate. Changes in bond angles are required for expulsion of the leaving group in these reactions [equation (6)].35-37 Nu Disappearance of the barrier for expulsion of a good leaving group can give a concerted mechanism when there is strong stabilization of the carbonyl group by resonance, as in amides, and when the carbanion is unstable in nucleophilic aromatic or vinylic substitution, as in the reaction of equation (7).37 A special problem is posed by certain isomerization and racemization processes that are commonly cited as evidence for reaction intermediates, but may not proceed through intermediates with a significant barrier for collapse to reactants or products.For example, scrambling of labelled oxygen atoms during the solvolysis of esters may proceed through a process in which there is always some degree of electrostatic, if not covalent, bonding between the reacting groups and no significant barrier for the collapse of a carbonium-carboxylate ion pair during the course of the reaction [equation (8); k-1, k-11 > 1013 The * 0 reaction will then proceed by the concerted mechanism shown by the solid line in Figure 4A, rather than by the stepwise mechanism shown by the dashed line.An analogous situation is likely for several other isomerization and racemization reactions. Some of these reactions will proceed with no chemical barrier in the usual sense for collapse to reactants or products in the course of the reaction, but may nevertheless require appreciable the, longer than a vibration frequency, for rotation or other motions within the solvent cage while contact is maintained s5 I. G. Csizmadia, M. R. Peterson, C. Kozmutza, and M. A. Robb in ‘The Chemistry of Acid Derivatives’, Suppl. B, ed. S. Patai, Wiley, New York, 1979, Pt. 1, pp. 1-58. 36 2.Rappoport, Acr. Chem. Res., 1981, 14, 7. 37 G. Modena, Arc. Chem. Res., 1971, 4, 73. How Does a Reaction Choose Its Mechanism ? A 8 C Figure 4 (A) Reaction co-ordinate diagram for scrambling of labelled oxygen atoms during solvolysis through a concerted mechanism when there is no barrier for collapse of the ion pair (solid line) and through a stepwise mechanism when there is a bdrrier and inter- medidtes exist (dashed lines).(B) Reaction co-ordinate diagrams for an uncoupled con-certed reaction that involves two processes, with diflering sensitivities to change in the structure of a leaving group, for example, so that there is a change in the nature of the transition state with changing structure. (C) Structure-reactivity correlation for the reaction in (B), showing a non-linear change in AG+ with changing substituents as the nature of the transition state changes between the reactants. There is no ideal solution to this problem, but it appears most satisfactory to include such reactions in the uncoupled concerted category if there is no intermediate with a significant activation barrier for its collapse.It is important to note that stepwise and uncoupled concerted reactions will often show similar structure-reactivity behaviour, because the unsymmetrical transition state of the uncoupled concerted reaction (Figure 4B)will resemble one or the other transition state of the stepwise reacti~n.l~*~~ A coupled reaction will have a single, more central transition state that represents both processes (Figure 3, upper solid line). Thus, the leaving-group ability of different halide ions will have little effect on the observed rate of nucleophilic viiiylic substitution when the transition state represents primarily nucleophilic addition, even if the reaction is concerted.36~3~A change in the relative leaving ability of the entering and leaving groups, so that the transition state represents predominantly leaving-group expulsion rather than nucleophilic attack, will give the same kind of break in a structure-reactivity correlation for an uncoupled concerted reaction as for a fully stepwise reaction, as shown in Figure 4B and C.14~38 3 Preassociation Stepwise Mechanisms When an intermediate has a short but significant lifetime, a reaction is likely to proceed through a preassociation mechanism in which all of the reactant and catalyst molecules are assembled in an encounter complex before the first covalent change occurs.This is shown in the lower path through the Kasand kl, steps of equation (9) for the general case of a reactant(s), R, that can form an J. F. Kirsch and W. P. Jencks, J. Am. Chem. Soc., 1964, 86, 837. Jencks ka k-a fC (9)Jl intermediate, I, either before or after association with a catalyst or final reactant, C, that is required in order to form products. When the intermediate complex I C breaks down to reactants (k-1.) faster than C diffuses away from it (k-s), the lowest energy pathway for both the breakdown and the formation of I C will be through the lower, preassociation pathway. The reason that the preassociation mechanism must become the favoured pathway when an intermediate becomes sufficiently unstable is shown in Figure 5A.When k-1.becomes large enough that the lowest-energy pathway for reversion A 8 Figure 5 Reaction co-ordinate diagram to illustrate the reason that (A) a preassociation mechanism is preferred when the intermediate I . C reverts to reactants faster than it separates into I and C, k-l# > k-&, and (B) a stepwise mechanism through a free inter- mediate is prejerred when the intermediate is more stable, so that k-a > k-le of the I * C complex to reactants is through the k-1. step, the reverse, kid, step provides the lowest-energy pathway for formation of the complex. This behaviour is expected for condensation reactions that require reaction with C in a final step, as in nucleophilic additions to carbonyl compounds and olefins, and in substi- tution reactions that proceed through a ternary complex containing the elements of all of the reacting molecules.When the intermediate has a longer lifetime, so that k-a > k-l., the I -C complex will break down more rapidly by diffusion away of C, as shown in 359 How Does a Reaction Choose Its Mechanism? Figure 5B.The lowest-energy pathway for the formation of I -C will then be through the free intermediate I, followed by rate-determining diffusion together of I and C with the rate constant ka [equation (9), upper path]. It is important to note that it is the rate of the back reaction, k-l., not the kz step, that determines whether the reaction proceeds by a preassociation or a diffusion-controlled trapping mechanism.When the intermediate becomes still more stable, the kl step will become rate determining. The kinetic requirements for the preassociation mechanism were apparently first described by Sutin, as an explanation for the replacement of water by a ligand on a metal through the dissociative interchange or ‘outer-sphere’ mech- anism.39 This mechanism involves preassociation of the hydrated metal and the incoming ligand before bond cleavage occurs because the lifetime of the ligand- deficient metal, after bond cleavage, is shorter than the time required for diffusion away of the incoming ligand.* Preassociation was described for nitration, nitro- sation, and halogenation reactions by Hartshorn and Ridd,40v41 for general acid- base catalysed reactions by Kershner and Schowen as a ‘spectator’ mechanism, for the case in which the catalyst C does not stabilize the transition state for heavy-atom reorganization,S and by the reviewer for several classes of react ions .42-46 A.Examples and the Question of Assistance.-Preassociation mechanisms were first identified experimentally for reactions in which the final reactant or catalyst C does not directly stabilize the transition state of the bond-making or bond- breaking step. The preassociation mechanism provides a lower-energy pathway than a mechanism that proceeds through a free intermediate because it avoids the higher-energy rate-determining step for diffusion of C to I [ka, equation (9), Figure 5A].For example, a limiting dissociative interchange mechanism of ligand exchange (ID)proceeds through rate-determining dissociation of the metal-ligand bond with the rate constant kl, [equation (lo)] and has a rate constant that is identical for all incoming ligands, except for differences that arise from differences in the equilibrium constant for formation of the initial outer-sphere complex, KoS.39,47 Similarly, the addition of 2-methyl-3-thiosemicarbazideto p-chlorobenzaldehyde is catalysed by general bases through a preassociation mechanism because the base catalyst must be present in the transition state of the kl. [equation (9)] step *A dissociative interchange mechanism could also occur by an uncoupled concerted mech- anism, if there is no barrier for the k-,.step. 3s N. Sutin, Annu. Rev. Phys. Chem., 1966, 17, 119. 40 S. R. Hartshorn and J. H. Ridd, J. Chem. Suc, 3, 1968, 1068. 41 J. H. Ridd, Adv. Phys. Org. Chem., 1978, 16, I. 43 W. P. Jencks and K. Salvesen, J. Am. Chem. SOC.,1971, 93, 1419. 43 W. P. Jencks, Chem. Rev., 1972, 72, 705. 44 J. M. Sayer and W. P. Jencks, J. Am. Chem. SOC.,1973,95, 5637. W. P. Jencks, Arc. Chem. Res., 1980, 13, 161. 46 W. W. Reenstra and W. P. Jencks, J. Am. Chem. Soc., 1979, 101, 5780. 47 C. H. Langford and H. B. Gray, ‘Ligand Substitution Processes’, W. A. Benjamin, New York, 1965. Jencks ka.lrk-a IfrL k8 + M-L 'OH2 in order to remove a proton from the amine immediately after formation of the unstable addition complex (8) [kz,equation (9)]and thereby prevent reversion of the dipolar addition intermediate to reactants.44 The base does not stabilize the transition state for the kla step significantly, so that the Brransted slope is = 0.With weaker bases the Brarnsted plot curves downward because the proton- transfer step and, finally, diffusion away of the protonated base become rate determining. 0-I+ IB * HN-C-II However, if the catalyst or final reactant C is required to be present during the rate-determining step (k~.),because of the short lifetime of the intermediate (large k-l#), it will often stabilize the transition state by facilitating the change in electron density so that there is significant assistance by interaction with the catalyst.This is observed for general acid catalysis of the addition of 2-methyl-3-thiosemicarbazide to p-chlorobenzaldehyde. Hydrogen bonding of buffer acids to the developing negative charge on the carbonyl group (9) stabilizes the transition state and results in a Brarnsted slope of a = 0.2.44 Similar stabilization in a pre- association mechanism has been observed for several carbonyl addition reactions of amines and weakly basic thiol anions, including the attack of methoxyamine on phenyl a~etate.~~-Sl Similarly, the rate constants for ligand exchange on metals by a dissociative interchange mechanism often show small differences that cannot easily be accounted for by differences in Kos.These differences probably represent weak interactions with the incoming ligand that stabilize the transition state for '* M.F. Gilbert and W. P. Jencks, J. Am. Chem. SOC.,1977, 99, 7931. 4* J. J. Ortiz and E. H. Cordes, J. Am. Chem. SOC.,1978, 100, 7080. 5a J. M. Sayer and C. Edman, J. Am. Chem. SOC.,1979, 101, 3010. 51 M. M. Cox and W. P. Jencks, J. Am. Chem. SOC.,1981, 103, 572. 361 How Does n Nenction Choose Its Mechanism? departure of the outgoing ligand.52153 Such a weak interaction is not unexpected for an open, ‘exploded’ transition state (10) of a preassociation mechanism that is almost identical to the transition state for diffusion away of the leaving gro~p.~6 Ar 0 The cleavage of 1-phenyl-2-arylcyclopropanol anions shows general acid catalysis with 01 0 that probably represents stabilization of the developing N benzyl carbanion in the transition state (1 1) by weak hydrogen bonding in a pre-association mechanism.zl Similar electrophilic assistance to carbanion formation is indicated by the primary isotope effect of kH/kD = 2.1-2.5 for the methoxide- induced cleavage of benzyltrimethylstannanesin MeOH-MeOD (I 1 ; A-H = MeOH or MeOD).5* Electron-withdrawing substituents on the benzyl group in this class of reaction cause a sharp increase in the discrimination isotope effect to k~/k~2 lO,55 which suggests that the more stable benzyl anions have a sufficient lifetime to diffuse through the solvent and discriminate between MeOH and MeOD.21 Solvolysis and substitution reactions of mono-substituted phosphates that have been thought to proceed through a metaphosphate monoanion intermediate almost certainly proceed by a preassociation mechanism in hydroxylic solvents, since no free intermediate is formed that can diffuse through the solvent to be trapped or give a constant solvent discrimination.56-S9 These reactions occur through an open, ‘exploded’ transition state with little bond formation and much bond cleavage (12) so that there is only a small amount of assistance by the entering group and &,..It remains uncertain values are small or zer0.5~-~~ whether there is an unstable metaphosphate intermediate with a significant life- time, or whether the reaction occurs by a concerted displacement. A preassociation mechanism is also probable for reactions of mono-substituted 5% C.K. Poon, Coord. Chem. Rev., 1973, 10, 1. 53 J. 0. Edwards, ‘Inorganic Reaction Mechanisms’, W. A. Benjamin, New York, 1964, p. 100. s1 R. Alexander, W. A. Asomaning, C. Eaborn, I. D. Jenkins, and D. R. M. Walton,J. Chem. Soc., Perkin Trans. 2, 1974, 490. C. Eaborn, D. R. M. Walton, and G. Seconi, J. Chem. Soc., Perkin Trans. 2, 1976, 1857; ihid., 1978, 834. 56 S. J. Benkovic and K. J. Schray in ‘Transition States of Biochemical Processes’, ed. R. D. Gandour and R. C. Schowen, Plenum Press, New York, 1978, pp. 493-527. 67 .I.D. Chanley and E. Feageson, J. Am. Chem. Soc., 1963, 85, I18 1. 68 W. P. Jencks and M. Gilchrist, J. Am. Chem. Soc., 1964, 86, 1410.IsA. J. Kirby and A. G. Varvoglis, J. Am. Chem. SOC.,1967, 89, 415. Di Sabato and W. P. Jencks, J. Am. Chem. SOC.,1961, 83,4400. A. J. Kirby and W. P. Jencks, J. Am. Chem. Soc., 1965, 87, 3209. Jencks sulphates with nucleophilic reagents, which show similar characteristics.62 It is also not known for this system whether there is a barrier for reaction of the presumed SO3 ‘intermediate’ with good nucleophiles that makes the reaction stepwise rather than concerted. A preassociation mechanism for acid-catalysed halogenation by hypobromous acid is supported by rate constants that are larger than can be accounted for by diffusion-controlled reactions with Br+ or H2OBr+ and that show a large depen- dence on the structure of the aromatic s~bstrate.4~9~~ This provides an example of a reaction in which the preassociation mechanism is enforced by the rapid dis- sociation of a proton from a protonated reactant, H20Br+ [equation (11); +klH+ + HOBr C HaOBr k-I Kas fArH kaJlk-a fArH\il k + k H+ * HOBr * ArH H20Br -ArH A+productsk-1.k-1. > k-a]. The large dependence of the rate on substrate structure is consistent with either rate-determining halogenation (k2)or a fully concerted preassociation mechanism in which proton transfer is assisted by attack of the substrate. Further work is needed to clarify the relationship between assistance and the lifetime of intermediates in the solvolysis of carbon compounds. When the inter- mediate is unstable the preassociation mechanism, with the incoming nucleophile present in the transition state for bond cleavage, should provide a lower-energy pathway than a Sneen-type mechanism64 involving diffusion and reaction of a nucleophile with an ion pair, for the reason illustrated in Figure 5A.It has been suggested that the solvent plays an important role in assisting the formation of intermediates by nucleophilic participation, as in the sN2 (intermediate) or ‘ion sandwich’ mechanism~,~~.~~a but it has been difficult to prove the existence of intermediates in reactions that show such assistance. a-pNitrophenylethy1 tosylate does not give an intermediate with an appreciable lifetime in hydroxylic solvents and exhibits a healthy second-order reaction with azide, but the small selectivity toward ethanol compared with trifluoroethanol shows that there is little or no solvent assistance in the solvolysis of this compound.66* Olefin-forming elimination reactions are addition reactions in the reverse direction [equation (12)]. The ‘Elcb (ion-pair)’ mechanism67 is a stepwise pre- association mechanism with the catalyst present in the transition state and k2 > k-ld > k-a.The Elcb (irreversible) mechanism67 can be a preassociation 62 J. P. Guthrie, J. Ant. Cliem. SOC.,1980, 102, 5177. 63 H. M. Gilow and J. H. Ridd, J. Chem. SOC.,Perkin Truns. 2, 1973, 1321. 64 R. A. Sneen, Acc. Chem. Res., 1973, 6, 46. 6bT.W. Bentley and P. von R. Schleyer, Adv. Phys. Org. Chem., 1977, 14, 1. 6a (a)F. G. Bordwell, P.F. Wiley, and T. G. Mecca,J. Am. Chem. SOC.,1975,97, 132; (b)J. P. Richard and W. P. Jencks, in preparation. 67 F. G. Bordwell, Acc. Chem. Res., 1970, 3, 281. 363 How Does a Reaction Choose Irs Mechanism ? mechanism with k-1. > kz or a liberated intermediate mechanism if k-a > k-1.. Almost nothing is known about the importance of assistance in these stepwise mechanisms. B. Differential Diagnosis.-It is relatively easy to distinguish between pre-association and liberated intermediate mechanisms, but difficult to distinguish between concerted and stepwise preassociation mechanisms. Criteria for the former distinction include the following: (i) Absolute rate constants for the k-1 and k-1. steps of the reaction [equation (9)] may be estimated by extrapolation from known rate constants of related compounds or calculated from the reverse rate constant and the equilibrium constant of a reaction.4194474830 (ii) The reaction of C with a reactive free intermediate is often diffusion controlled [ka, equation (9)], whereas the rate-determining step for a preassociation mech- anism is not diffusion controlled [kid, equation (9)].Thus, diffusion-controlled trapping by catalysts or final reactants, C, is sensitive to the viscosity of the solvent (although this may be difficult to differentiate from other solvent effects)50~5~>~~-70and will show rate constants that are independent of the basicity, acidity, nucleophilicity, or other chemical properties of C.15971 Such independence does not exclude a preassociation mechanism with no a~sistance,~~ but the observation of different rate constants with different C rules out rate- determining diffusion-controlled trapping of a free intermediate.** (iii) A free intermediate with a given structure, I, must show constant partition- ing, regardless of its source, in its reactions with different solvent components, isotopes, added reagents, or reaction paths [equation (I 3)J.15y72An intermediate complex of a preassociation mechanism that contains the leaving group com- monly shows different partitioning with different leaving groups.17 Unfor- tunately, constant partitioning between different products, such as the products 68 C.Cerjan and R. E. Barnett, J. Phys. Cltem., 1972, 76, 1192.69 M. F. Aldersley, A. J. Kirby, P. W. Lancaster, R. S. McDonald, and C. R. Smith, J. Chem. SOC.,Perkin Trans. 2, 1974, 1487. 'O 1980, 102, 1340.H. Fischer, F. X. DeCandis, and W. P. Jencks. J. Am. Chem. SOC., 71 R. E. Barnett and W. P. Jencks, J. Am. Cliem. SOC.,1969, 91, 2358. 72 D. J. Raber, J. M. Harris, and P. von R. Schleyer, 'Ions and Ion Pairs in Organic Reactions,' Wiley, New York, 1974, Vol. 2, pp. 247-374. Jencks of elimination and solvolysis, does not prove the existence of an intermediate because the same result is expected for late transition states of uncoupled concerted reactions that closely resemble the presumed intermediate. However, only a free intermediate that is formed irreversibly and reacts rapidly with two final reagents, C and D (equation (13); (kc[C]+ kd[D]) > k-l[AJ), will give product ratios that are proportional to the concentration ratio [C]/[D J without affecting the overall rate when the concentrations of C and D are ~aried.7~ Racemization is a special case of constant partitioning, i.e.a free, planar carbanion or carbocation must have the same reactivity on both sides and give complete racemization (in the absence of an asymmetry that hinders reaction on one face). (iv) A reaction that proceeds through a free intermediate frequently shows a change in rate-determining step with changing concentration of the final reactant or catalyst C [equation (14); reaction of A-B with C1.41~71972 There is no such kJC1A-B & B+A --+B-C k-, change in rate-determining step for a preassociation mechanism, because C is present in the transition state of every kinetically significant step [equation (9)1.Thus an increase in the concentration of catalyst or nucleophile can make trapping of a free intermediate so fast that formation of the intermediate (kl) becomes rate-determining and the observed rate becomes independent of the concentration of C. Conversely, if a molecule, A, is released upon formation of the free intermediate [equation (14)], addition of this molecule can make the intermediate revert to reactants by a mass law effect, so that there is a change from rate-determining formation to rate-determining reaction of the intermediate and an inhibition of the observed rate.Such inhibition will be accompanied by incorporation of isotopically labelled A into the reactant. At low [A] the rate of this exchange corresponds to the amount of inhibition. The ‘special salt effect’, a sharp increase in the rate of solvolysis reactions with added salt to a rate that is characteristic of the ‘normal’ salt effect, represents a change in rate-determining step from solvolysis or separation of some ion-pair intermediate to the for- mation of this intermediate.72v73 73 A. Fava in ‘The Chemistry of Organic Sulfur Compounds’, ed. N. Kharasch and C. Y. Meyers, Pergamon Press, New York, 1966, Vol. 2, p. 80. 365 How Does a Reaction Choose Its Mechanism? (v) Changing the structure of C can give a change in rate-determining step that is different for preassociation and liberated intermediate mechanisms.Rate-determining trapping of an intermediate by proton transfer to or from electro- negative atoms commonly gives a Brarnsted plot that follows an Eigen curve for different acids or bases, with limiting slopes of 0 and k 1.0 for strong and weak catalysts that correspond to diffusion-controlled encounter and separation of the catalyst and intermediate, respectively. These lines intersect close to the pKa of the intermediate, at dpK -0. The preassociation mechanism follows the same Brarnsted curve for weak catalysts [kz rate determining in equation (9)] but has a faster rate when kl~is rate determining (Figure 5A). Consequently, the inter- section of the limiting lines of the Brarnsted curve is shifted and, if it is larger than the estimated error for the pKa of the intermediate, this shift can provide evidence for the preassociation mechanism.@*51 (vi) The initial rate of a reaction in which reversible proton removal from carbon gives a free intermediate, such as an Elcb elimination, can exhibit large inverse solvent deuterium isotope effects, such as ~D,o/~H,o= 6.This is the result of a pseudo-equilibrium in the initial step, in which H is removed but D is added back to the carbanion intermediate so that its steady-state concentration is increased in deuterium 0xide.~4 After exchange of deuterium into the starting material is complete such a reaction will not exhibit the primary deuterium isotope effect that is expected for a concerted E2 eliminati~n.~s Criteria for distinguishing stepwise and concerted preassociation mechanisms include the following: (1) Extrapolation of structure-reactivity correlations or calculation of the rate constant that would be required for reaction of an intermediate in order to account for an observed rate constant may give a lifetime of an ‘intermediate’ species that does not correspond to a significant barrier for its breakdown, so that the reaction must proceed by a concerted me~hanism.1~~25 (2) Structure-reactivity correlations and isotope effects can provide evidence that two processes are occurring simultaneously in the transition state to a greater extent than would be expected for a stepwise mechanism with an intermediate.28976 For example, a concerted E2 elimination reaction can show a significant isotope effect and Brarnsted /? value for proton removal and a significant heavy-atom isotope effect and dependence on leaving-group ability (-/hg,‘element effect’) for bond cleavage.A coupled coiicerted mechanism with a single, central transition state should give a linear or smoothly curved structure-reactivity cor-relation, whereas a stepwise or uncoupled concerted mechanism should give a sharp break as one or the other process becomes rate determining, as noted above (Figure 4C).14 (3) A coupled concerted mechanism may be described by a diagonal reaction co- 74 J. Keeffe and w. P. Jencks, J. Am. Chem. SOC.,1981, 103, 2457.75 R. A. More O’Ferrall and S.Slae, J. Chem. SOC.B, 1970, 260. W. H. Saunders, jun. and A. F. Cockerill, ‘Mechanisms of Elimination Reactions’, Wiley,New York, 1973, p. 87. Jencks ordinate on a reaction co-ordinate-energy contour diagram that is defined in terms of structure-reactivity parameters. Changes in structure-reactivity para-meters and in the position of the transition state on such a diagram with changing reactant structure can provide evidence for an interaction between reacting groups in the transition state and for a concerted rnechani~m.~~~~~~~~~~ (4) The demonstration that some time-dependent process occurs faster than collapse of an intermediate through either of two alternative paths shows that there must be a barrier for collapse of the intermediate and that the mechanism is not concerted.An example is the demonstration of equal rate constants for racemization of cis-5-methyl-2-cyclohexenylp-nitrobenzoate and for equi-libration of the carboxyl oxygen atoms in both enantiomers during solvolysis in 80% aqueous acetone.78 This result requires that an intermediate ion pair, (13), must have a sufficient lifetime to allow rotation and complete randomization of the carboxylate oxygen atoms before it collapses to either product or reactant. Most examples of oxygen scrambling and racemization do not prove that there is an intermediate with a significant lifetime, as noted above. (5) Strict stereochemical specificity in the absence of severe steric effects, such as substitution with complete inversion or anti-elimination, provides support for a concerted mechanism.(6) A stepwise preassociation mechanism of general acid-base catalysis involving electronegative atoms, such as the acid-catalysed methoxyaminolysis of phenyl acetate, shows a sharp solvent isotope effect maximum when the proton-transfer step becomes rate determining near dpK = 0, whereas a concerted mechanism shows a solvent isotope effect that does not change with changing pK of the catalyst.51970 However, this criterion needs testing with additional examples. Several of the above criteria can provide evidence for a coupled concerted mechanism but do not distinguish between an uncoupled concerted and a step-?? D. A. Jencks and W. P.Jencks, J. Am. Chem. SOC.,1977,99, 7948. 7n H. L. Goering, J. T. Doi, and K. D. McMichael, J. Am. Chem. SOC.,1964,86, 1951. 367 How Does a Reaction Choose Its Mechanism? wise preassociation mechanism. For example, both stepwise and uncoupled concerted mechanisms for vinylic, acyl, or aromatic substitution and for elimi- nation reactions can show small values of &, element effects, and heavy-atom isotope effects in the leaving group, when attack of the nucleophile or base is the predominant process in the transition state. It is also important to remember that the converse of a criterion need not hold; for example, a reaction that does not show strict stereochemical specificity can be either stepwise or concerted. C. Requirements and Perturbations.-The critical role of the solvent in deter- mining reaction mechanisms by controlling the lifetime of intermediates cannot be emphasized too strongly.The solvent can determine the lifetime of an inter- mediate, I, in two ways that should be distinguished: (a) by its reactivity toward I as the final reactant or catalyst, C, and (b) by altering the stability of I through a solvent effect. A carbocation, for example, is expected to have a shorter lifetime in ethanol than in water because ethanol has both a larger nucleophilicity and a smaller dielectric constant compared to water, whereas acetonitrile will stabilize the cation by its low nucleophilic reactivity and destabilize it by its poor ion- solvating ability. Solvent Reactivity. When the solvent is highly reactive as the final reactant or catalyst, C, an unstable intermediate will simply react non-selectively with the first solvent molecule it sees.A reaction with dilute solute molecules, C ’, will be able to compete significantly with the solvent reaction only if C ’ stabilizes the transition state for theformation of the intermediate, so that a significant amount of I ’ C ’ is formed in the klfstep. In a mixed solvent or a concentrated solution of C ’ the intermediate presumably has the option of reacting with one of several molecules in a surrounding solvent pool, as suggested by Grunwald et al.,79 but little is known about the size or nature of this pool or about the nature of short- range reorientational and translational processes that may influence the relative reactivity of molecules in the pool.Electrostatic, steric, and statistical factors are presumably important. The products that are formed upon the photochemical generation of unstable intermediates may provide more information about these factors. Reaction with the solvent predominates in the dissociative interchange, ID, mechanism for ligand exchange on metals, for example, because the incoming ligand provides little or no stabilization of the rate-determining transition state. Exchange of one ligand for another in aqueous solution almost always proceeds through the aquo complex [equation (10) in reverse], because the solvent is the only significant nucleophile that reacts with the unstable intermediate that is formed upon loss of a ligand.The aquo complex then reacts with the incoming ligand in a second preassociation, outer-sphere step [equation (10) in the forward direction] to form the thermodynamically stable product.80 The solvolysis of R-X occurs by a preassociation mechanism when k-1 and E. Grunwald, A. Heller, and F. S. Klein, J. Chem. Suc., 1957, 2604. R. G. Pearson and J. W. Moore, Inurg. Chem., 1964, 3, 1336. Jencks k-1. are fast and Rf -X-reacts with solvent faster than X-diffuses away [ks > kd, equation (15)]; when R+ X-reacts more slowly, with ks < kd, it undergoes diffusional separation [the solvent is always present and is not shown in equation (I 5)]. When the leaving group is a better nucleophile than the solvent kl R+ -I-X-R-X ck-1 R-Solv.or steric shielding inhibits solvolysis, so that I usually returns to reactants (k-1 > ks), the solvolysis (ks)or diffusional separation (kd) step can be rate determining. If the leaving group stabilizes R+(as in some ion pairs), the pre- association mechanism should become more favourable because this stabilization is completely lost in the transition state for diffusional separation (kd)and is only partly lost in the transition state for solvolysis (ks). Changes in secondary a-deuterium isotope effects are consistent with changes between rate-determining kl, ks,and kd steps for solvolysis reactions with anionic and uncharged leaving groups.8lI82 For example, an increase in the a-secondary deuterium isotope effect for the acid-catalysed hydrolysis of benzaldehyde dimethyl acetal with increasing dioxan concentration can be accounted for by a change to kd as the rate-determining step when k-1 becomes larger than ks.83 Such a reaction will ordinarily show no general acid catalysis, because the proton is completely transferred in the transition ~tate.8~ Molecules other than the solvent become important as the final reactant or catalyst C [equation (9);Nu in equation (15)] when (i) the solvent becomes less reactive or (ii) the molecule C provides assistance to the kl.step. In general acid catalysis, for example, catalysis by a preassociation mechanism becomes increas- ingly important as the pK of I decreases below 16, so that water becomes less reactive as a proton donor to I (the converse holds for general base catalysis).A strong acid or base can give proton transfer every time the intermediate I -C is formed [kz > k-l), equation (9)], so that the importance of the buffer-catalysed relative to the uncatalysed reaction depends inversely on k2 for the water reaction. 85 When bond cleavage occurs in the initial step, as in solvolysis-substitution reactions [equation (131, diffusion apart of the intermediate into its components (kd)competes with other pathways so that there is only a limited region in which V. J. Shiner, jun. in ‘Isotope Effects in Chemical Reactions’, ed. C. J. Collins and N. S. Bowman, Van Nostrand Reinhold, New York, 1970, p. 105. 8z V.P. Vitullo and F. P. Wilgis, J. Am. Chem. SOC.,1981, 103, 880. 8s P. R.Young,R. C. Bogseth, and E. G. Rietz, J. Am. Chem. Soc., 1980, 102,6268. 84 W. P. Jencks, Acc. Chem. Res., 1976, 9, 425. W. P. Jencks and H. F, Gilbert, Pure Appl. Chem., 1977, 49, 1021. How Does a Reaction Choose Its Mechanism? solute molecules, Nu, can play a significant role in a preassociation mechanism without stabilizing the transition state for the k1' step. For example, an added nucleophile can cause a rate increase through a preassociation mechanism in the reaction of equation (15) when k-1, k-11, and k2 are fast. If k2 is large it may be able to compete successfully with ks and kd (when k2 = 10l1s-l, ks = lo9s-l, and kd = 1O1Os-l, for example) and give rise to a significant second-order reaction with Nu, in spite of the unfavourable equilibrium constant for formation of the preassociation complex with the nucleophile.The ratio of the rate constants for reaction with Nu through the preassociation pathway and for reaction through I [when k-1 9 (ks + kd)] is given by k2Kas[Nu]/[(kd+ ks) (k2/k-1' + l)]. This situation has not been identified experimentally for a reaction in which the inter- mediate has been proved to have a significant lifetime in the presence of both solvent and the incoming nucleophile. Reaction through the preassociation mechanism will usually be accompanied by diffusion-controlled trapping of the intermediate by the nucleophile. With sufficiently reactive nucleophiles it will give second-order rate constants that are independent of the reactivity (but not the concentration) of reactive nucleophiles when kl.is rate determining. If kz becomes still larger, there will be no barrier for the k2 step and the reaction will become concerted. It is rare for a preassociation mechanism to be enforced by a fast k-1. step that is second-order, because the first-order diffusional separation step, k-a, will be faster than the k-1. step under most conditions and must always become faster at a low concentration of the reactants for the second-order reaction. However, there are special circumstances in which a fast reaction with another molecule in the k-1 step can give a preassociation mechanism, as in the nitration of p-nitro- aniline in 90%sulphuric acid with k-1.[HzS04] > k2 > k-a [equation (1 6) The reaction involves unprotonated aniline, which is probably reprotonated in this medium (k-1.) faster than N02+ can diffuse away from it (k-a). Although the k-18 step is formally second order, it is effectively first order in 90% sulphuric acid and the mechanism could also be written with Hs0.1-in the preassociation complex. Assistance, Non-enforced Catalysis, and Mixed Mechanisms. For acid-base catalysis involving electronegative atoms a necessary proton-transfer step to or from water will generally be faster than diffusion when the pKof the intermediate I is @ -2 as an acid or 3-16 as a base. The reaction with solvent will then be a pre- association mechanism. The observed reaction will be dominated by the solvent Jencks and its components, i.e.it will be uncatalysed or will show specific acid or base catalysis with no buffer catalysis unless the catalyst can stabilize the transition state of the kli step. An example of such stabilization, by hydrogen bonding, is found in general acid catalysis of the cleavage of carbarnates formed from weakly basic amines [equation (1 7)3.86 Cleavage of the carbamate of p-nitroaniline, which follows a linear Brarnsted plot with a = 0.84, proceeds through a zwitterionic N-protonated intermediate with an estimated pKa of -4.3 and k-1. of 2 x 1O1O s-l for cleavage of the intermediate. This cleavage reaction represents the preassociation mech- anism of equation (9) in the reverse direction, with the k-1.step rate determining and catalysis at the nucleophilic reagent (class n catalysis). Similar assistance is possible for reactions involving carbocation and meta- phosphate intermediates but it remains uncertain whether such assistance is significant for reactions in which the intermediate has been proved to exist, as discussed above. Stabilization of the transition state of the kl, step by C will increase the importance of the preassociation mechanism compared with other mechanisms and, in some cases, can give rise to a non-enforced reaction by a preassociation mechanism. Hydrogen bonding of an acid catalyst to the oxyanion intermediate and the transition state for its formation in a carbonyl addition reaction (9), for example, can stabilize both the intermediate, T-, and the transition state but will not affect the energy of the transition state for diffusional encounter or separation, ka and k-a (Figure 6).This provides an explanation for the non-linear Brnrnsted plot for general acid catalysis of the addition of p-methoxybenzenethiolateanion to acetaldehyde, which is consistent with a trapping mechanism that follows a Brsnsted slope of zero for acids of intermediate strength and a preassociation mechanism with hydrogen bonding and 01 = 0.16 for stronger a~ids.~8 In this and other systems in which the transition state of the kl. step is stabilized by C, the reaction is likely to proceed concurrently by two different pathways, a trapping mechanism and a preassociation mechanism.These are conveniently S. P. Ewing, D. Lockshon, and W. P. Jencks, J. Am. Chem. SOC.,1980, 102, 3072. How Does a Reaction Choose Its Mechanism? Figure 6 Reaction co-ordinate diagram to show how hydrogen bonding of an acid, HA, to an addition intermediate, T-, can stabilize the intermediate and the transition state for its formation so that a preassociation mechanism is favoured over a trapping mechanism. The dashed line shows that the preassociation mechanism is less favourable than the trapping mechanism in the absence of hydrogen bonding described by adding one or more ‘wings’ to the reaction co-ordinate diagram, as shown in Figure 7. The two pathways can have the same form of the rate law and the same composition of the rate-determining transition state, but one transition state represents diffusion together of I and C (ka) and the other represents the chemical step for the formation of I * C (kid).The first chemical step for the trapping mechanism (k1) does not involve C.3334 Figure 7 Diagram with a ‘wing’ to show how a reaction can proceed concurrently by a trapping mechanism, with ke rate determining, and a preassociation mechanism, with k rate determining. A similar diagram can illustrate concurrent trapping and concerted mechanisms, with a direct path for the conversion of R C to products Stabilization of the transition state of the kl#step by hydrogen bonding (9) is also responsible for the weak general acid catalysis of the addition of sulphite to p-methoxybenzaldehyde, with a = 0.06.This represents non-enforced catalysis, because the initial addition intermediate has a sufficient lifetime to diffuse through Jencks the solvent and reach equilibrium with respect to proton transfer.87 Non- enforced general acid catalysis with a = 0.13 is also observed for the addition of MeOOCCH2S-to acetaldehyde at high buffer concentrations, at which the proton-transfer step to the addition intermediate is fast and the addition step becomes rate deterrnini~~g.~~ Such reactions generally exhibit linear Brcansted plots in which the solvent and its components fall on or near the Brarnsted line for buffer catalysts. * ‘Stickiness’ of the I -C complex that arises from hydrogen-bonding, electro- static, dispersion, and hydrophobic interactions will decrease the rate constant for diffusional separation of the complex, k-&,and therefore will favour the pre- association relative to the trapping mechanism (Figure 6).Unreactive Solvents. When a reaction involves bond cleavage before the reaction with C takes place in a second step, it will proceed through a fully dissociative mechanism at sufficiently low concentration of the reactants when the solvent is unreactive. This is because a rate-determining transition state that contains C has a less favourable entropy than one that does not and a reaction will always proceed through a term in the rate law that does not include C when [C] becomes small. Thus, ligand exchange of metal ions is likely to occur by a pure dissociative mechanism (D) in non-liganding solvents, in which the intermediate has an opportunity to diffuse away from the leaving ligand and find a new partner, whereas it is rare in liganding solvents.47 In the cobalamin-cobaloxime series, for example, ligand exchange or addition to the aquo complex appears to follow an IDmechanism in water but occurs through a D mechanism in inert sol~ents.88-~0 Ion Pairs and Weak Complexes.There are few reactions in water and other good ionizing solvents in which an initial bond-breaking step gives an ion pair or other intermediate, I, that has a long enough lifetime to permit a dilute reactant C to encounter and react with it. Equilibrium constants for the formation of ion pairs from singly charged ions in waterg’ are generally c 1 .O M-~so that the first- order rate constant for dissociation of an ion pair is equal to or larger than the second-order, diffusion-controlled rate constant for its formation.The ion pair will then dissociate faster than it can encounter and react with C at a concentra- tion of < 1 M, so that the reaction will proceed largely through the free ions9 or, if k-14 and k2 are fast, through a lower-energy preassociation mechanism. In less ionizing solvents ion pairs have a longer lifetime so that they may diffuse through the solvent before reacting with the solvent or another molecule; *The existence of non-enforced preassociation mechanisms raises a possible ambiguity for the classification of reaction mechanisms in terms of the lifetimes of intermediates.Since there is preassociation of C with the reactants, the intermediate complex I . C does not diffuse through thesolvent, and k, is likely to be fast, it seems preferable to maintain the present nomen-clature rather than to define a new category for this small group of reactions. P. R. Young and W. P. Jencks, J. Am. Chem. SOC.,1977,99, 1206. D. Thusius, f. Am. Chem. SOC.,1971, 93, 2629. R. J. Guschl, R. S. Stewart, and T. L. Brown, Inorg. Chem., 1974, 13, 417. so R. C. Stewart and L. G. Marzilli, J. Am. Chem. Soc., 1978, 100, 817. s1 C. W. Davies, ‘Ion Association’, Butterworths, London, 1962, pp. 77 and 168. 373 How Does a Reaction Choose Its Mechanism? they then provide a major perturbation on the mechanism of solvolysis and substitution reactions.Excellent reviews are available that describe the different types of behaviour that can be explained by different types of ion pairs.’2992 Although many experimental results can be assigned to reactions of these dif- ferent ion pairs, relatively few experiments have been designed to provide a critical test of these assignments. It is not always clear that the behaviour attributed to one species could not be explained by another species or that all ‘ion pairs’ represent intermediates rather than transition states. Some ‘solvent-separated’ ion pairs may really be contact ion pairs and some ‘intimate’ ion pairs may be transition states, which would be consistent with their low reactivity toward nucleophilic atta~k.~3 In fact, the reactivity of an intramolecular carbonium- sulphonate ion pair toward nucleophiles has been shown to be very similar to that of a comparable free carbonium ionq94 The most direct approach to the basic problems of mechanism and reactivity in organic chemistry would appear to be through the examination of reactions in polar solvents in which ion pairs are not formed or play a minimal role.D. Liberated Intermediates.-When an intermediate has a sufficient lifetime to diffuse through the solvent and choose its partner it has crossed a moderately sharp borderline and is free to react with some degree of specificity. This will be the preferred pathway when the complex I * C of an intermediate with a final reactant or catalyst dissociates into I and C faster than it collapses to reactants [k-a > k-18, equation (9) and Figure 5B]or when an intermediate that is formed by bond cleavage dissociates into its components [e.g.kd > k,, equation (15)]. The reaction will then proceed through some fully stepwise or trapping mech- anism and will not contain the final reactant C in the transition state or rate law for the initial step of bond formation or cleavage, unless C is the solvent or accelerates this step. The rate law and rate-determining step of the overall reaction will, of course, include C if the initial step is reversible and trapping by reaction with C is rate determining. An unstable liberated intermediate that encounters C will be likely to react with it faster than C can diffuse away, so that the reaction with C will be diffusion controlled and non-selective (LI-D mechanism, Figure 1).This is the case for a considerable number of reactions that require proton transfer between electro- negative atoms and show diffusion-controlled trapping of an unstable inter- mediate by buffer acids or bases when the proton transfer is strongly favoured thermodynamically. Such reactions follow non-linear Br~rnsted plots that cor- respond to Eigen curves for simple proton-transfer reactions in water."^^^ It is also the case for the reactions of sulphite and hydroxylamine with oxocarbonium ions derived from ketals of substituted acetophenones.l5 The reaction with solvent is activation-limited and selective in these systems, so J.M. Harris, Prog. Phys. Org. Chem., 1974, 11, 89. s3 L. P. Hammett, ‘Physical Organic Chemistry’, McGraw-Hill, New York,2nd Edn., 1970, pp. 163-167. s4 C. D. Ritchie and T. C. Hofelich, J. Am. Chem. SOC.,1980, 102, 7039. Jencks that as the structure of I changes there is a change in the relative reactivity of C and the solvent. For the oxocarbonium ions derived from substituted aceto- phenones the selectivity, l0g(kHzo/kso,z-), has the surprisingly large value of p = 1.6.15This kind of situation, in which one reaction path is diffusion-controlled and non-selective and another is activation-limited, provides one explanation for changes in selectivity with changing reactivity of the reactant, the ‘reactivity- selectivity principle’.95~96 Concerted general acid-base catalysis involving electronegative atoms requires that the pK of a reacting site must change during the reaction so that a proton transfer to or from the catalyst that was initially unfavourable becomes favourable and there is a driving force for the catalysis.97 If the initial proton transfer is thermodynamically favourable it will take place rapidly and, if the immediate product is stable enough to reach equilibrium for the proton transfer step (k-l[A-] > k2 for an acid-catalysed reaction, equation (I 8) 1 it will react in a subsequent rate-determining step in the absence of HA or A-.This represents specific acid or base catalysis and is an example of an activation-limited LI-A mechanism (Figure 1) that is brought about by the long lifetime of the inter- mediate.27 When an intermediate is sufficiently stable to react through an LI-A mech-anism, with an activation-limited process in the final step, this step will show the selectivity and other properties that are expected for such a stable chemical species and we can conclude this description of reaction mechanisms that are enforced by the lifetimes of intermediates. s5 D S. Kemp and M. L. Casey, J. Am. Chem. SOC.,1973,95, 6670. s6 Z. Rappoport, Tetrahedron Lett., 1979, 2559. W. P. Jencks, J. Am. Chem. SOC.,1972,94,4731.

ISSN:0306-0012    

DOI:10.1039/CS9811000345

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Aryliodine(rI1) Dicarboxylates By A. Varvoglis LABORATORY OF ORGANIC CHEMISTRY UNIVERSITY OF THESSALONIKI, THESSALONIKI, GREECE 1 Introduction Since the last general review on polyvalent iodine compounds by Banks1 in 1966, a considerable body of research has been published, especially on aryliodine(II1) dicarboxylates, [ArI(OCOR)2 abbreviated* to AID]. The multifaced interest of these compounds has been the driving force for this review, which is restricted mainly to the title compounds, with incidental citations of related compounds such as aliphatic and cyclic analogues, iodine(Ii1) tricarboxylates, etc. Nomenclature.-A great variety of names have been used for AID (no less than fourteen !). The nomenclature in Chemical Abstracts has been changed several times, e.g.for PID* from (dihydroxyiod0)benzene diacetate to iodoso- benzene diacetate to phenyliodine(Ii1) diacetate to the current phenyliodine bis(acetato-0). The nomenclature used here has been adopted by several journals. 2 Structure and Spectra The crystal structures of PID (1) and PhI(OCOCHC12)2 (2) have been recently determined.2p3 The first study2 was performed at -60 "Cto avoid decomposition by X-rays and revealed a wealth of structural information. Both molecules have in common the T-shaped geometry of dsp3 hybridized trivalent iodine compounds but the overall geometry of iodine can be described as a pentagonal planar arrangement of three strong and two weak secondary bonds (Figures I and 2). In (1) the two 1-0 distances of the covalent bonds are equal and the two secondary I-* -0bonds are intramolecular, forming a four-membered IOCO ring.In (2) the 1-0 distances differ significantly and the molecule is a dimer: one of the 1--0bonds is intramolecular and the other intermolecular, forming a I202 ring. As a suitable bonding model, the overlap of the I-C antibonding orbital with each one of the oxygen atoms lone-pair orbitals is favoured2 (Figure 3). The dipole moments of PID (4.65 D) and other AID have been determined in benzene and compared with theoretical values for various conformation^.^ *Other abbreviations: PID = phenyliodine(ll1) diacetate, PIB = phenyliodine(1ri) bis(tri- fluoroacetate), LTA = lead tetra-acetate. All these compounds are commercially available.D. F. Banks, Chem. Rev., 1966,66, 243. N. W. Alcock, R. M. Countryman, S. Esperas, and J. F. Sawyer, J. Chem. SOC.,Dalton Trans., 1979, 854. C.-K. Lee, T.-C.W. Mak, W.-K. Li, and J. F. Kirner, Actu Cryst., 1977, B33, 1620. 0. Exner and B. PlesniEar, J. Org. Chem., 1974, 39, 2812. Aryliodine(Ir1) Dicarboxylates V Figure 1 ORTEP view of (1) (Reproduced from J. Chem. SOC.,Dalton Trans., 1979, 854) Figure 2 ORTEP view of (2) (Reproduced from J. Chem. SOC.,Dalton Trans., 1979,854) The results clearly showed that all AID are in the 2 conformation, essentially identical to their structure in the crystalline state. No evidence has been found Vurvoglis Figure 3 Pentagonal-plane arrangement around iodide in (I) and (2) (Reproduced from J.Chem. SOC.,Dalton Trans., 1979, 854) in the 1H n.m.r. spectra between 50 and -55°C for the presence of an E conformer or a dissociation of the type shown in equation 1. PhI(OCOR)2)2 + PhI+OCOR + RC02-(1) It seems that in non-polar solvents there is no appreciable dissociation, an observation borne out by freezing point and conductivity measurements.5 In polar solvents, however, there is a small degree of dissociation above room temperature, as shown for PTD by conductivity studies in acetic acid.6 The existence of an ionic equilibrium in Me02H for a series of AID has been claimed on the basis of n.m.r. spectra at -80 to -60 "C, where two acyloxy-groups were observed and assigned to covalent and ionic forms.7 In CH2CIz and CHC13 no such splitting was observed.This interpretation is open to question since PID in CDC13, with two equivalents of MeOH, exchanges its acetoxy-groups slowly enough at -60 "C to be observed by n,m.r.8 [Equation (211. PhI(OCOCH,), + nMeOH + Phl(OCOCH,),-,(OMe), + nCH,CO,H (2) 19F n.m.r. spectra of several metn-and paw-X-substituted fluorobenzenes,g where X = I(OAc)2, I(OCOCF3)2 efc., showed that iodine does not interact with the aromatic ring mesomerically and that only the inductive effect operates in the order i(OCOCF3)z > lCl2 > IF4 > IF2 > I(OAc)2. The 13C n.m.r, spectrum of PID gives a value of 122.2 p.p.m. for the C bpund to 1, far off the value of 94.66 p.p.m. observed for the same C in iodobenzene.10 It seems that the anomalous 'heavy atom' effect of iodine is restricted to its monovalent compounds.The i.r. spectra of several AID have been discussed.lI The carbonyl stretch- ing band is displaced to values cu. 100 cm-' less than those of the corresponding C-compounds, e.g. for PhCH(OAc)2 vC0 = 1755 cm-l and for Phl(OAc)2 KO = 1660 cm-1 (in CC14). This remarkable shift has been attributed to inefficient overlap between the oxygen orbitals: electrons shift towards 0 and W. D. Johnson and N. V. Riggs, Aust. J. Chem., 1955, 8, 457. W. D. Johnson and J. E. Sherwood, Aust. J. Chem., 1971, 24, 2281. S. S. Makarchenko, E. B. Merkushev, M. M. Shakirov, and A. I. Rezvukhin, Izv. Sib. Otd. Akud. Nuuk SSSR, Ser. Khim. Nauk, 1978, 125 (Cfiem.Abstr., 1978, 89, 162876).A. Seveno, G. Morel, A. Foucaud, and E. Marchand, Tetruhedron Lett., 1977, 3349. V. V. Lyalin, G. P. Syrova. V. V. Orda, L. A. Alekseeva, and L. M. Yagupolskii, Zh. Org. Khim., 1970, 6, 1420. (Chem. Abstr., 1970, 73, 87298). lo K. Friedrich, W. Amann, and H. Fritz, Chem. Ber., 1978, 111, 2099. R. Bell and K. J. Morgan, J. Chem. Soc., 1960, 1209. A Aryliodine(Ir1) Dicarboxylates the C=O bond weakens, while the molecule acquires partially ionic character. A cyclic contribution (as that of Figure 3) was rejected. The fact that PID both in nujol and CC14 has low vC0 values suggests that the cyclic form of the crystalline state indeed does not contribute significantly to the weakening of the carbonyl bond. Other characteristic bands near 1290 cm-1 and 670 cm-1 have been tentatively assigned to the stretching of the I-0-(C) system.The mass spectra of a series of aryliodine(r1i) dibenzoates have been reported.12 With the exception of phenyliodine(ir1) dimesitoate, no molecular ions were detected, but several fragment ions arose from thermal decompositions. The main fragmentation pathways include ions of both benzoic acids and iodo- benzene, and a minor fragment has been attributed to the iodole ion in Scheme 1. No molecular ions were seen in the mass spectra of several aryliodine(Ii1) diacet a tes. l3 + Scheme 1 The U.V. spectra of AID have not been published, but those of some diacetates have been found to exhibit a considerable hypsochromic shift14 in comparison to iodoarenes, like other polyvalent iodine compounds.15 3 Synthesis The standard procedure for the preparation of AID with aliphatic acids is the direct oxidation of iodoarenes with peracids, in the presence of the correspond- ing carboxylic acid [Equation (3)].ArI + RCOOOH + RCOOH +ArI(OCOR), + H,O (3) This is especially applicable to the preparation of diacetates, using 30% He02 and acetic anhydride16 or 40% peracetic acid.17 It should be noted that with excess 40 % peracetic acid, iodoxybenzene (PhI02) is f0rmed.1~ This method fails with some ortho substituted iodobenzenes : o-di-iodobenzene forms diace- toxybenzodi-iodoxole18a(3), which was shown by n.m.r. spectroscopy not to be in equilibrium with the isomeric o-iodoso-PID ;o-iodobenzenesulphonamide la E.Malamidou, E. Micromastoras, and A. Varvoglis, Chim. Chron., New Series, 1977, 6, 493. l3 J. A. Gustafsson, L. Rondahl, and J. Bergman, Biochemisrry, 1979, 18, 865. l4 J. Gallos, Ph.D Thesis, to be submitted to the University of Thessaloniki. l5 F. M. Beringer and P. Bodlaender, J. Org. Chem., 1968, 33, 2981. l6 K.H. Pausacker, J. Chem. SOC.,1953, 107. I7 J. G. Sharefkin and H. Saltzman, Org. Synth., 1963, 43,60, 62, 65. la (a) W. Wolf, E. Chalekson, and D. Kobata, J. Org. Chem., 1967, 32, 3239; (b) H.Jaffe and J. E. Leffler, ibid., 1973, 38, 2719; (c), 1975, 40,797. Varvoglis gives the benziodathiazole systemlab (4)and o-iodophenylphosphoric acid the benziodadioxaphosphorinlSc (9,presumably after hydrolysis of its acetate.o-Iodobenzamidelg reacts similarly. OAc I 1 OAc I OAc OH (3) (4) (5) Pentafluoroiodobenzene has been converted into its bis-ttifluoroacetate20 with (CFsC0)20 and HN03, in a reaction where CF3C03H was formed in situ. Aryliodine(rr1) dichlorides react with Ag+ and Pb2+ carboxylatesl and also with aqueous AcOH in pyridine21 to give AID. Iodosoarenes and carboxylic acids1 or their anhydrides20 react to form AID, upon simple stirring in a suitable solvent. Aliphatic iodocompounds do not normally form dicarboxylates with the exception of trans-chlorovinyliodide22 and certain perfluoroiodides; e.g. CF31(0COCF3)2 has been prepared23 from CFJF2 or CF3IO and (CF3C0)20. With peracetic acid, alkyliodides form acetyl hypoiodite, which reacts further with the iodide to give an acetate ester and iodine.24 In the presence of arenes iodination of the aromatic ring takes place.Alkyl iodides with meta-chloroperbenzoic acid are converted mainly into alcohols25 via iodosocornpounds, which rearrange to alkyl hypoiodites; the latter hydrolyse through formation of carbonium ions, so that the yields of alcohols are not always good. Among other products, depending on the nature of alkyl iodide, are ketones and oxiranes. AID exchange their acyloxy-groups with stronger acids.z6 Initially a p-oxo-dicarboxylato-diaryl-di-iodine(6) is formed, which with excess of the acid is transformed to the new AID (Scheme 2). W. Wolf and L. Steinberg, J. Chem. SOC.,Chem Commun., 1965, 449; H.J. Barber and M. A. Henderson, J. Chem. SOC.(C), 1970, 862. ao M. Schmeisser, K. Dahmen, and P. Sartori, Chem. Ber., 1967, 100, 1633, and 1970, 103, 307; L. M. Yagupolskii, I. 1. Maletina, N. V. Kondratenko, and V. Orda, Synthesis, 1977, 574. B. Karele and 0. Neilands, Law. PSR Zinar. Akad. Vestis, Kim. Ser., 1970, 587. (Chem. Absrr., 197 1, 74, 42 033). ap J. Thiele and H. Haakh, Ann. Chem., 1909, 369, 131. 23 D. Naumann and J. Baumans, J Fluorine Chem., 1976, 8, 177. 24 Y. Ogataand K. Aoki, J. Org. Chem., 1969, 34, 3974. a5 H. J. Reich and S. L. Peake, J. Am. Chem. SOC.,1978, 100, 4888; R. C. Cambie, D. Chambers, B. G. Lindsay, P. S. Rutledge, and P. D. Woodgate, J. Chem. SOC.,Perkin Trans. I, 1980, 822; T. L. Macdonald, N. Narasirnhan, and L.T. Burka, J. Am. Chem. SOC.,1980, 102, 7760. E. B. Merkushev, A. N. Novikov, T. I. Kogai, and V. V. Gluskova, Zh. Org. Khim., 1972, 8, 436. Aryliodine(II1)Dicarboxylates Ar Ar ArI(0COMe)z + RCOzH -\1-0-1 / RCOIH, ArI(OCOR)2 / \ -MeColH MeCO2 OzCMe (6) Scheme 2 Using the above method with an isolable p-compound (6), and by careful sequential addition of two different acids, AID with two different acyloxy-groups, e.g. PhI(OCOCH2Cl)(OCOCH2Br),have been obtained.27 Two seven-membered heterocycles were obtained from PID and phthalic acidZ8 (7) and from PID and succinic acid29 (8): The cyclic structures of (7) and (8) have not been rigorously proven and they may well be internal salts. In fact structure (7') has been proposed for the product of the reaction between PID and phthalic anhydride.30 An exchange reaction between PID and acid anhydrides also leads to AID.30 Cyclic AID are also known with five and six-membered rings.l Of special interest are cyclic AID (10) resulting from thermal rearrangement of o-iodobenzoyl aroyl peroxides (9), (Scheme 3).r 1 L i \ OCOAr Scheme 3 27 E. B. Merkushev, T. I. Kogai, L. G. Polyakova, and A. N. Novikov, Zh. Org. Khim., 1973, 9, 1077. 88 E. B. Merkushev, A. N. Novikov, S. S. Makarchenko, A. S. Moskal'chuk, T. I. Kogai, V. V. Glushkova, and L. G. Polyakova, Zh. Org. Khim., 1975,11, 1259. 28 G. P. Baker, F. G. Mann, N. Sheppard, and A. J. Tetlow, J. Chem. SOC.,1965, 3721. 30 A. N. Moskal'chuk, S. S. Makarenko, and A.N. Novikov, Dep. Publ. VINITI 6261-73, 1973 (Chem. Absrr., 1976, 85, 46 116). 382 Varvoglis The above rearrangement was found to occur via a caged radical-pair mech- ani~m.3~It can also take place in the solid state thermally or upon X-irradiation, as revealed during crystallographic studies. Compound (lo), ArCO = o-iodoben-zoyl, appears in two polymorphic forms, whose crystal structures differ signifi- cantly; one of them results in a preferentially oriented single-crystal phase (topotaxy). It should be noted that solid-state chemistry of organic polyvalent iodinecompoundssimilar to( 10)has been shown to be a fruitful area of research.32 A final preparative method for AID is the electrophilic aromatic substitution of various arenes (activated or slightly deactivated) with tris-trifluoroacetoxy- i0dine.3~ Monosubstituted benzenes give exclusively para-isomers, whereas CsH5CF3 affords the meta-isomer. In the presence of strong acids diaryliodon- ium salts are formed prefer en ti all^.^^ Iodobenzene has also been oxidized elec- trolytically35 into PID.The various preparative methods of AID are summarized in Scheme 4. ArlCl2 ArI p-compounds RCOaH,iRCO, -RCOIH RCO,H/ ArH Scheme 4 4 Chemical Properties AID are fairly stable compounds. They decompose only at elevated tempera- tures (Section 4E) and they are not hydrolysed by atmospheric moisture. They are hydrolysed by alkali into iodosoarenes and this is the method of choice for 31 J. E. Leffler, R.D. Faulkner, and C. C. Petropoulos, J. Am. Chem. SOC., 1958,80, 5435; W. Honsberg and J. E. Leffler, J. Org. Chem., 1961,26, 733. 32 J. Z. Gougoutas and J. C. Clardy, Acta Cryst., 1970, B26,1999 and J. SolidState Chem., 1972,4226,230; J. Z. Gougoutas, J. Am. Chem. Soc., 1977,99, 127. 33 1. I. Maletina, V. V. Orda, and L. M. Yagupolskii, Zh. Org. Khim., 1974, 10, 294. 34 F. M. Beringer, R. A. Falk, M. Karniol, I. Lillien, G. Masullo, M. Mausner, and E. Sommer, J. Am. Chem. SOC., 1959, 81, 342. H. Hoffeiner, H. W. Lorch, and H. Wendt, J. Electroanal. Chem., 1975, 66, 183. AryZiodine(1Ii) Dicarboxy lates the preparation of the 1atter.l’ They are sensitive to daylight to a small extent but they easily undergo photochemical decomposition (Section 4E).The reactions of AID are mainly oxidations of various types, where iodine( 111) is reduced to iodine(1). They can be classified as ‘conventional’ oxidations, acetoxylations, and aryliodinations. There are also some substitution reactions at iodine, which is subsequently reduced ; sometimes, however, new stable iodine(iI1) compounds may be formed. A. Oxidations.-Most reactions of AID are oxidations bearing a close analogy with the reactions of LTA, as pointed out by Criegee36 who was the first to explore their chemistry. Although few systematic studies about their structure and reactivity have been performed, acyloxy-moieties from strong acids moderately accelerate reaction rates, whereas substituents in the benzene ring may have a variable role, depending on the reaction mechanism.37 It is possible in some cases for a substituent to exert a powerful influence on the rate.It must be noted that in several oxidations AID react in a unique way and also that PIB reacts with systems inert to PID. (i) Oxidation of N-Compounds. The oxidation of aromatic amines with PID is known to give, in poor yields, azocompounds viu hydrazocompounds in.a reaction involving free radica1s.l Several 2-nitroanilines and heterocyclic vic-nitroamines preferentially undergo an oxidative cyclization into furoxans.lJ8 Dyall et al.39 have studied kinetically the reactions of several ortho-substituted anilines (11) with PID and have proposed a general mechanistic scheme, different from an earlier one suggested for 2-nitroani1ines.l A nucleophilic displacement at iodine leads to the formation of intermediate (12), which by neighbouring-group participation is oxidatively cyclized into (1 3); this is usually a furoxan but it may be also a triazole or an anthranil (Scheme 5).-AcOH N\L 7 Y X’ (12) 0 Ph .t. I (X=Y is N=O,N =NPh, or C=O) Scheme 5 36 R. Criegee, ‘Oxidation in Organic Chemistry’, Academic Press, New York, 1965, p. 365. 37 D. Barbas, J. Gallos, and A. Varvoglis, Chim. Chron., New Series, (in press). 38 A. J. Boulton and D. Middleton, J. Org. Chem., 1974, 39, 2956. 3s L. K. Dyall, Aust. J. Chern., 1973, 26, 2665 and references therein. 384 Varvoglis The intermediate (12) accumulated during the reaction as shown indirectly by comparison of the concentration of 2-benzoylaniline (by i.r.spectroscopy) and total iodine(n1) species (by titration); the apparent aniline concentration was higher in the early stages of the reaction, resulting from a contribution by (12), supposed to absorb at the same region as the symmetric N-H stretch of the aniline. The activation parameters for the cyclization of eleven orrho-substituted anilines with PID fitted an isokinetic relationship, which suggests that a common reaction mechanism is followed in all cases. Amides undergo with PID an oxidative rearrangement40 analogous to the Hofmann rearrangement. With acetic acid as solvent, acetanilides are formed. The reaction has been studied kinetically with various substituted amides; electron donors in the benzene ring accelerate reaction rates, whereas electron acceptors retard them, the Hammett p value being -0.81.The activation para- meters for several benzamides show a good linearity. A mechanism involving the formation of an iodine(IIi)-amide complex, (14) or (15), which rearranges in a concerted manner into the isocyanate has been proposed (Scheme 6). ArCONHz + ArCO-fN';1 Phi(OAC)~ -PhI, -AcOH + ArNH-Ac 4 *'OH ArN=C=O-co, Scheme 6 Amides are converted by PID in MeCN-HzO into amines in high yields,41 provided that the carboxamido-group is not attached to an aromatic ring, in which case the resulting aniline is further oxidized. Formation of N,N-dialkyl ureas, which are by-products in the reaction with PID, is here avoided; in addition, trifluoroacetic acid from equation (4) catalyses the hydrolysis of the -PhI RCONH2 + PhI(OCOCF3)g +RN=C=O + 2CFsCOzH (4) isocyanate [Equation (5)], so that only short reaction times are required with no heating.40 K. Swaminathan and N. Venkatasubramanian,J. Chem. Soc., Perkin Trans. 2, 1975, 1161. 41 A. S. Radhakrishna, M. E. Parham, R. M. Riggs, and G. M. Loudon, J. Org. Chem., 1979, 44, 1746; G. M. Loudon and M. E. Parham, Tetrahedron Lett., 1978, 437. Aryliodine(ii1) Dicarboxylates H+ RN=C=O + H20 4RNH, + C02 (5) Since secondary amides are inert, the above reaction has been applied to the degradation of peptides, serving as a key step in sequence studies.41 The peptide is anchored by the free NH2 of the N-terminal amino-acid to a solid support; the COzH of the C-terminal amino-acid is then converted into CONH2 through a series of reactions and the amide function with PIB is transformed into an amine, which is spontaneously hydrolysed into a new peptide-amine and an aldehyde [Equations (6) and (7)].9 * * *CONH-CH(R)CONH, + PIB -+ * * * .CONH-CH(R)NH3+ (6) .***.CONH-CH(R)NH,+ + RCHO + NH, (7)+ H2O -+ *****CONH, PID oxidizes cyanamide at room temperature, forming in situ cyanonitrene,42 which may be added to thioethers, sulphoxides, phosphines, and olefins. The reagent is mild, with low steric demands, and it is effective in cases where conventional reagents fail. 1 -Amino-2,5-diphenyl-1,3,4-triazolejs also oxidized by PID to the corresponding nitrene, which either decomposes into PhCN and Nz or may be trapped by an 0lefin.~3 The course of the reaction was unaltered when PhI(OCOCHCI2)z or LTA were used.By contrast, oxidation of I-amino-benzotriazole with PID gives not the expected biphenylene, as LTA does, but a mixture of azobenzene, 1-phenylbenzotriazole, and other more complex products.44 Mechanisms where both nitrene and benzyne formation are involved have been proposed. In other similar oxidations45 both PID and LTA give the same products. Nitrene formation occurs also during the PID oxidation of Schiff’s bases.46 The initial addition product (16) is cleaved with rearrangement into a diacetoxy-derivative hydrolysing into an aldehyde and a nitrene, which dimerizes into an azocompound (Scheme 7).A variety of oximes react with PIB in several ways.47 Aromatic syn-aldoximes ArCH = NAr’ + PhI(0Ac)a dArCH-NAr’--+ ArCH(OAc)2 4-[Ar’N : ] .1. J. (16) ArCHO Ar’N = NAr’ Scheme 7 42 J. E. G. Kemp, D. Ellis, and M. D. Closier, Tetrahedron Lett., 1979, 3781. 43 F. Schroppel and J. Sauer, Tetrahedron Lett., 1974, 2945. 44 C. D. Campbell and C. W. Rees,J. Chem. Soc. (C),1969,752;P. G. Houghton and C. W. Rees,J. Chem. RPS. (S), 1980, 303. 45 C. D. Campbell and C. W. Rees,J. Chem. Sac. (C), 1969,742; C. W. Rees and R. C. Storr, ibid., 1969, 760. 46 S. Narasimhabarathi, S. Sundaram, and N. Venkatasubramanian, Indian J. Chem., 1977, 15B,376. 47 S. Spyroudis, Ph.D Thesis, University ofThessaloniki, 1981 ; S.Spyroudis and A.Varvoglis Synthesis, 1975, 445; 1976, 837. Varvuglis give mixtures of aldazine di-ZV-oxides and nitrile oxides, the latter being the major products, provided they are stable. Aliphatic ketoximes form gem-nitroso- trifluoroacetoxy-alkanes, stable only in solution, whereas aromatic ketoximes give mainly the parent ketone. a-Dioximes give furoxans in high yields, whereas p-dioximes ( 17) form mixtures of isoxazoles ( 18) and 4-0x0-4H-pyrazole-di- N-oxides (19) (Scheme 8). R Scheme 8 tram-2-Unsaturated I ,4-dioximes (10)are oxidized by PlB47s4Hinto a mixture of 3a,6a-dihydroisoxazolo[5,4-d]isoxazoles(21) and pyridazine I ,2-dioxides (22) (Scheme 9). R R Scheme 9 No detailed work has been done so far concerning the mechanism of these oxidations, but it seems that at a first stage an unstable substitution product 'C=NOI(OCOCF3)Ph is formed, which is further transformed, mainly /homolytically, with formation of iminoxy-radicals.In some cases, however, a polar or a concerted mechanism may also operate. Other N-compounds oxidized by AID are N-benzylhydroxylamine to give the dimeric a-nitrosotoluene47 and several N,N-benzylhydroxylaniines to give ** A. Ohsawa, H. Arai, H. Igeta, T. Akimoto, A. Tsuji, and Y. litaka, J. Org. Ch(Jm.,1979, 44, 3254. 387 Ary liodine( 11I) Dicarboxylates nitrones.49 Again, homolytic pathways are assumed to operate. In the last case various oxidants were shown to produce essentially the same results.(ii) Oxidation of Alcohols and Ethers. Perhaps the best known reaction of PID is the cleavage of glycols into carbonyl corn pound^.^^^^ The reaction is similar to the Nal04 or LTA oxidation and an analogous mechanism involving a cyclic trivalent iodine intermediate has been proposed.’ PID has been found to cleave several steroidal glyc0ls5~ at a rate about 100 times slower than LTA. Although again a cyclic intermediate is favoured, the fact that PID reacts with trans-decalin-9,lO-diol suggests that an alternative mechanism is also available (Nal04 does not react with this diol). It is pertinent that d,l-tartaric acid is oxidized by PID slightly faster than its meso-isomer.51 PIB is a superior reagent to PTD and no heating is required for glycol cleavage, which is effected very quickly.47 In an analogous manner N,N-dialkyl-1 ,2-aminoalcohols (23) are cleaved by PJD to aldehydes and an immonium salt (25), which hydrolyses to an amine and formaldehyde; the latter being eventually oxidized to COz.A cyclic intermediate (24), as with glycols, is likely to be formed52 [Scheme (lo)]. RCH-OH RCHO + CHzNRz’ %[ RC\H +/I”] -HAR2 [CH2=&R2 AcO--!%CH20 + R2NHJ (23) (24) (25) Scheme 10 When the reaction was run in the presence of acidic N-H compounds, such as isatin, derivatives of 1,l -diaminomethane (Mannich’s bases) were formed from (25).The reaction of PTD with phenols leads usually to resinous products1 but upon reaction with 4-X-phenols (X = electron acceptors) p-benzoquinones or, mainly, o-iodophenyl ethers are formed53 (Section 4C).Substituted pyro- catechols54 and hydroquinones47 are easily oxidized by PID and PIB to the corresponding quinones. A special case is the oxidation of certain bisnaphthols (26) to the spirocompounds (28) (Scheme 11). Although the reaction proceeds with a great variety of oxidants, only PID gives exclusively the above (R*R*) diastereomer. Its stereospecificity has been 49 P. A. Smith and S. E. Gloyer, J. Org. Chem., 1975, 40,2508. 6o S. J. Angyal and R. J. Young, J. Am. Chem. SOC.,1959,81,5251. 61 K. Vaidyanathan and N. Venkatasubramanian, Indian J. Chem., 1973, 11, 1146. sz H. Mohrle and S. Dornbrack, Pharmazie, 1974, 29, 573, 757. 5s A. R. Fox and K. H. Pausacker, J.Chem. Soc., 1957, 295. 54 A. T. Balaban, Rev. Roum. Chim., 1969, 14, 1281 ; A. Suzuki and K. Sato, Japan Kokai, 156 863 (1977), (Chem. Abstr. 1978,88, 152 42 1). Varvoglis (26) Ph J w'0 0 Scheme 11 attributed to the intermediate formation of (27) from which the less hindered (28) results. Several bisnaphthols and other related di01s~~ react in a similar fashion. Another special case, from alkaloid chemistry, is the intramolecular oxidative coupling of reticuline (29) to salutaridine (30) and other analogous reactions56 (Scheme 12). Many other oxidants failed to effect this oxidation, but several AID were successful. Me0 H AID _______, Me0 I1HoBxMeQoHOMe 0 Scheme 12 55 D. J. Bennett, F. M. Dean, G.A. Herbin, D. A. Matkin, and A. W. Price,J. Chem. Soc., Perkin Trans. f, 1980, 1978; F. M. Dean, G. A. Herbin, D. A. Matkin, A. W. Price, and M. L. Robinson, ibid., 1980, 1986. j6 C. Szantay, G. Blasko, M. Barczai-Beke, P. Pechy, and G. Dornyei, Tetrahedron Lett., 1980,21, 3509. Aryliodine(I I I) Dicurboxylates Alcohols do not react appreciably with AID at ambient temperature, but they are not suitable as solvents for measurements of U.V. spectra.14 At 80°C PIB oxidizes both primary and secondary alcohols into carbonyl compounds; although the presence of pyridine is beneficial, yields are still moderate.47 Alkoxymagnesium salts, ROMgBr, are oxidized by PID and other oxidants into carbonyl compounds.57 Although the double bond of a heptenol derivative remained intact, it is doubtful whether the double bond would not react in other cases (Section 4B).Mandelic acid undergoes oxidative decarboxylation with PID to give ben~aldehyde.5~A kinetic study of several a-hydroxy-acids has shown a similar behaviour for most of them; an iodine(Ir1) ‘ester’, RCH(COzH)OI(OAc)Ph, is formed initially, and then is cleaved either directly or through a cyclic inter- mediate.51 Esters of a-hydroxy-acids are also oxidized by PID but more slowly than the corresponding acids and evidently to a-keto-esters. Examples of glycol cleavages can be found also in the chemistry of carbo- hydrates. 9, O An oxidative cleavage of dibenzyl ethers (31) by PIB has been observed with formation of benzaldehyde and benzyl trifluoroacetate.61 The mechanism of the reaction involves an oxonium ion intermediate (32) and an intermediate (33) with iodonium-like character (Scheme 13).-CF&OaHPhCH2-0-CHzPh ’IB PhCH -O+yCHzPh Id I d -PhCHO Ph-I-CH20PhI OCOCF3 (3 1) CF3CO2- OCOCF3 J. (32) CF3C02CH2Ph Scheme 13 Benzyl alkyl (or benzhydryl, trityl) ethersreact analogously to give benzaldehyde and trifluoroacetates; the latter are usually hydrolysed during work up. Thus the reaction may be used for the removal of the protective benzyl group from alcohols containing reducible functions. Benzyl aryl and alkyl aryl ethers give with PIB iodonium salts.47 (iii) Oxidation at C. Apart from the above reactions and acetoxylations and aryliodinations of various substrates (Sections 4B and 4E), oxidations at C are .57 K.Narasaka, A. Morikawa, K. Saigo, and T. Mukaiyama, Bull. Chem. SOC.Jpn., 1977, 50, 2773. 68 R. Criegee and H. Beucker, Ann. Chem., 1939,511, 218. 6* S. Ukai, H. Idemitsu, and T. Takaoka, Japan. Patent, 1974, 10933 (Chem. Abstr., 1974, 81, P63924). 6o S. C. Pati and R. C. Mahapatro, Natl. Acad. Sci. Lett. (India), 1978, I, 325 (Chem.Abstr., 1979, 90, 23 444), and Proc. Indian Acad. Sci., Sect. A, 1979, 88, 203 (Chem. Abstr., 1979, 91, 158004). 61 S. Spyroudis and A. Varvoglis, J. Chem. SOC.,Chem. Commun., 1979. 615. 390 Varvoglis not common ; e.g. 9,lO-dihydroanthracene is oxidized by PIB under drastic conditions and in poor yield47 to anthracene, although PID oxidized * --CH2CH2COOR to -.CH=CHCOOR, when this group was attached to a triazolium ring, in high yield.62 Ethyl 4-aryl-2,4-dioxobutanoates(34)are cleaved by PID in AcOH containing a catalytic amount of water into benzoic acids,63 whereas under similar conditions the corresponding acids give 5-hydroxy-2-ary1-4-aroyl-furan~,6~ present in their tautomeric form (36).This unusual reaction is thought to pro- ceed through an initial oxidative coupling to (33, which under loss of C02 and H2O is transformed into (36),(Scheme 14). ArCOzH + C02 R =y ArCOCHzCOCOzR (34) \ Ar IR=H c-0 ArCOCHCOCOzH I ArCOCHCOCO2H (35) Scheme 14 Related oxidative cleavages have been reported for fluorenyI a-ketoesters and 2-acyl-l,3-indanediones.65 Another oxidative dimerization has been observed with a-cyanocarboxylatess (37).Initial attack from N to I leads to the intermediates (38) or (39), which viu free radicals (40),dimerize into (41)or (42).The reaction is carried out in methanol and an enol ether (43)may also be formed (Scheme 15). 62 G. Doleschall and G. Toth, Tetrahedron, 1980, 36, 1649. 63 B. D. Podolesov, God. Zb., Prir.-Mar. Fak. Univ. Skopje, Mat. Fiz. Henr., 1974, 24, 51 (Chem. Absfr., 1975, 82, 124993). 64 N. Bregant, J. Matijevic, 1. Sirola, and K. Balenovic, Bull. Sci. Cons. Acad. sci. Arts RSF Yougosl.,Sect. A, 1972, 17, 148 (Chrm. Abstr., 1973, 78, 4047). 65 M. R. Korunoski and B. D. Podolesov, God. Zb., Prir.-Mat. Fak. Univ. Skopje, Mat. Fiz. Hem., 1974, 24, 55, and 1978, 28, 87. 391 Aryliodine(u1) Dicarboxylates RCH -CN + P~I(OAC)~--+ RC=C=N-14Ac or (RC=C=N)d-Ph CO2R bh k02R1 (37) (38) (39) CN .1IRC-N =C=CR or [RC- CN]2 MeOH [RC=~Z--CCN]'I CO2R CO2R LO&*I (42) + (411 CN OMe /RI I/RC-NH-C= C ICO2R YC02R (43) Scheme 15 3-Cyanosuccinimides with PID in methanol are converted into acyl wethan@ by a mechanism analogous to that of Scheme (15); ketimines like (42) have been detected in the reaction mixture from their i.r.absorption (2030 cm-l).. An interesting case of regiospecificity has been observed in the oxidation of tetraketones67 (44a), which with PID and LTA give the isomeric pyrones (45) and (46), respectively (Scheme 16), probably through their ring-chain tautomers (44b).RCOCHzCOCOCHzCOR Scheme 16 (iv)Oxidationat S, Se, and P.Thiols are oxidized with AID into dis~lphides~~~~* 6* G. Morel, E. Marchand, A. Seveno, and A. Foucaund, Tetrahedron Lett., 1977, 3353. 67 M. Poje, Tetrahedron Lett., 1980, 21, 1575. 68 T. Mukaiyama and T. Endo, Bull. Chem. SOC.Jpn., 1967, 40, 2388. Varvoglis exceedingly easily. Thioethers give with PID sulphoxides in moderate yields,69 but sometimes the conditions may be drastic. A similar oxygenation occurs in phenothiazines and pheno~elenazines.~O After treatment with PID, followed by hydrolysis, trityl phenyl sulphide and trityl benzyl sulphide give triphenyl carbinol as the only isolable product. Triphenyl phosphine with PIB gives phosphine oxide, whereas trialkyl phosphites give mixtures of complex products,14 as with LTA.Phosphorus ylides of the type Ph3P=C(OMe)COR are oxidized by PID, LTA, and other oxidants into Ph3PO and RCOCOOMe.71 Several mechanistic schemes may be envisaged for these oxidations but no systematic studies have been reported yet. B. Acetoxy1ations.-One of the earliest reactions of PID is the addition of twc acetoxy-groups to an ethylenic double bond.58 Cyclopentadiene reacts exotherm- ically with formation of both 1,2- and 1’4-addition products. Anethole (4-Me0- C6H4-CH-CHMe) reacts faster with PID’s bearing electron donors on the benzene ring than with PID’s bearing electron acceptors. PID and catalytic amounts of oso4 have been used for effective hydroxylation of the double bonds in several steroids.72 Alkenes also react with PIB47 to give mixtures of cis-and trans-l,2-bis-trifluoroacetoxy-derivatives;when they bear aryl groups other products are formed as well, mainly carbonyl compounds resulting either from scission of the double bond or from rearrangements.With tetraphenylethy- lene the following products were obtained, besides the normal addition product: PhzCO, Ph3CCOPh, 9,1O-diphenylanthracene,and tetraphenyloxirane. The reaction appears to have a close analogy with the reaction of iodine(u1) tris- (trifluoroacetate), (47)and alkene~,~~ where 1,2-bis-trifluoroacetoxyalkanes(50) are also the main products. The mechanism of this reaction involves initial addition of I(OCOCF3)2 and OCOCF3 to the double bond.Thus, with ethylene the aliphatic AID (48) is formed, which expels CF3C02I and is transformed to the isolable dioxolane (49). Under the experimental conditions (49) rearranges to (50) but an acylal(51) may also result as a by-product after a 1,2-shift (Scheme 17). The addition is mainly cis for 1,2-disubstituted alkenes but cis, trans-for tri- and tetrasubstituted alkenes. PIB adds also its trifluoroacetoxy-groups to alkynes.74 When the triple bond is internal two equivalents of PIB form a tetra-trifluoroacetoxy-derivative,which on hydrolysis affords an a-diketone. Diynes may give a diketone with the second triple bond intact. When the triple bond is terminal, one equivalent of PIB forms the relatively stable aryl alkynyl iodonium salt (52), detected by i.r.H. H. Szmant and G. Suld, J. Am. Chem. SOC.,1956, 78, 3400, K. C. Schreiber and V. Fernandez, J. Org. Chem., 1961, 26, 2478, 2910, J. P. A. Castrillon and H. H. Szmant, J. Org. Chem., 1967, 32, 976. 70 B. D. Podolesov, Prikl. Maked. Akad. Nauk Umet., Od. Prir.-Mar. Nauk, 1978, 10, 49; B. D. Podolesov and V. B. Jordanovska, Croat. Chem. Acta, 1972,44,411. 71 E. Zbiral and E. Werner, Monatsh. Chem. 1966, 97, 1797. 72 J. A. Hogg et al., J. Am. Chem. SOC.,1955,77,4436, 4438, 6401. 73 J. Buddrus and H. Plettenberg, Chem. Ber., 1980, 113, 1494.’* E. B. Merkushev, L. G. Karpitskaya, and G. I. Novosel’tseva, Dokl. Akad. Nauk. SSSR, 1979,245, 607. AryZiodine(m) Dicarboxylates AciO CH2-0, ,CF3 CHzOAci \+ I ‘C’ -ICH-CH3 CHz-0 ’\OAcf CHzOAcf (Acr0)zCH-CH3 Scheme 17 spectroscopy; (52) is hydrated to (53) and then hydrolysed to an a-hydroxy- acetophenone (54), (Scheme 18).&CzCH% Arc-C- ;-Ph AciO- Hzo ArC=C-;-Ph 1 1OH H Hzo F ArC-CH201 -PhI 11 0 (52) (53) (54) Scheme 18 Acetoxylations also take place at sp3 carbon atoms. Thus several P-diketone~7~ and acetophenones76 react with PID to give acetoxy-derivatives. The reactions proceed in AczO-AcOH or in aqueous AcOH with HzS04 as a catalyst through the enolic form of the ketones, with intermediate formation of an O-phenylio- dinated species (59, (Scheme 19). Aliphatic and cyclic ketones react similarly.77 75 0. Neilands and G. Vanags, Dokl. Akwd. Nauk SSSR, 1960, 131, 1351 (Chem.Abstr., 1960, 54, 21 080). F. Mizukami, M. Ando, T. Tanaka, and J. Imamura, Bull. Chem. Soc. Jpn., 1978, 51, 335. 77 S. C. Pati and B. R. Dev, Indian J. Chem., Sect. A, 1979, 17, 92, and 1979, 18A, 262; V. Mahalingan and N. Venkatasubramanian, Indian J. Chem., Sect. B, 1979, 18, 94, 95; M. Higuchi, and R. Suzuki, Japan Kokai, 1973,68575 (Chem. Abstr., 1974,80,36997). 394 Varvoglis ArC-CH3 II 0I1 r ArC=CH:! 1 L J (55) Scheme 19 It must be noted that most /3-diketones react with PID but they yield iodine y1ides rat her than acetoxy-derivatives (Section 4C). PIB reacts with aryldiazomethanes and 1 -aryl-l -diazoethanes in a complex way.78 The main products are esters of trifluoroacetic acid, ArCHZOCOCF3 or ArCH(CH3)0COCF3, resulting from an intermolecular hydrogen transfer, as deuterium labelling has shown.The mechanism of the reaction remains unknown. The reaction of 4-R-CsH4NHAc (R = electron donor), (56),with PID affords 3-acetoxy-derivatives1 (62) and it was initially thought to proceed via free radicals. It has subsequently been suggested that it is an electrophilic displace- ment involving direct transfer of acetoxy-cation (from Ph-I+-OAc) to the ace- tanilide.79 The mechanism of this interesting reaction has now been elucidated and the acetoxy-group actually enters the aromatic ring as a nucleophile.80 Initially (56) attacks PID at iodine with formation of an iodonium salt (57). This expels iodobenzene and gives a nitrenium ion (58), which in the presence of MeOH (solvent) or AcOH forms the dienoneimine (59), the protonated form of which (60a) rearranges, possibly intermolecularly through (61), to the final product (62).In MeOH (60b) gives the addition product (63), which eventually solvolyses to the stable cyclohexa-2,5-dien-3-one(64).In the presence of MeNH2 (60a) forms 3-methylamino-acetanilide.The same dienone (64) is also formed from PID and N-t-butyl-p-toluidine (Scheme 20). The reaction between PID or PIB and iodine leads initially to acetoxylation of iodine,sl with formation of an unstable hypoidite, [Equation (S)]. PhI(OCOR)2 4-12 __I, PhI + 2RCO2I (8) The hypoiodite may add to alkenes with formation of 2-iodoalkyl acetates or may react with several aromatic compounds to give mono-, di-, or tri-iodo 78 B.Axiotis, S. Spyroudis, and A. Varvoglis, Chim. Chron., New Series, in press. 79 W. D. Johnson and N. V. Riggs, Aust. J. Chem., 1964, 17,787; W. D. Johnson and J. E. Sherwood, ibid., 1972, 25, 1213. P. Kokil, S. Partil, T. Ravidranathan, and P. Madhavan Nair, Terruhedrun Lett., 1979, 989. Y. Ogata and K. Aoki, J. Am. Chem. SOC., Jpn.,1968, 90, 6187, and Bull. Chem. SOC. 1968,41, 1976; E. B. Merkushev, N. D. Sirnakhina, and G. M. Koveshnikova, Synthesis, 1980,486. Aryliodine(II1) Dicarboxylates R (56) NHAc NHAc--H+ $ y ! E L IH+ NHAc0 MeCOO R YO R Me (ma), Y = Ac (61) (60b),Y = Me 0 R OMe R OMe Scheme 20 derivatives; halobenzenes are also iodinated, exclusively at the para-position. Good yields, mild conditions, and simple isolation procedures make these reactions synthetically attractive.PID may acetoxylate the iodine anion also, forming acetyl hypoiodite, which equilibrates with AcO- to give a diacetoxy- iodate(1) anion, [Equations (9) and (10)3. Varvoglis PhI(0Ac)a + I-+ PhI + IOAC + AcO-(9) IOAc + AcO-+ I(OAc),-(10) Upon mixing of tetraethylammonium iodide and PID in CHC13 the stable salt EtaN+I(Ac0)2- (65) is isolated in high yield.62 Compound (65) acetoxylates the triazolium salts (66) into (67); the same acetoxylation takes place with PID and iodides or, better, tri-iodides of (66). Kinetics established that I(Ac0)2- is the active species in both cases rather than IOAc.The salts (67) on hydrolysis are converted into an aldehyde (68) and the salt (69). Since (66) are easily prepared from carboxylic acids and S-methyl-l,4-diphenylth iosemicarbazide, their acetoxylation and hydrolysis has served as a convenient method for the conversion of acids into aldehydes with one carbon atom less62 (Scheme 21). Scheme 21 Compound (65) acetoxylates several active CHZcompounds, introducing one (diethyl malonate) or two (acetophenone) acetoxy-groups. It also reacts with cyclohexene to give trans-l-acetoxy-2-iodocyclohexane.In all these cases I(0Ac)z- is the active species. The bulkiness of I(0Ac)z- is probably the reason why with diethyl n-butylmalonate it does not act as an acetoxylating but as a halogenating agent, affording n-butyliodomalonate.In this reaction, and also in the reaction of (65) with 4-hydroxybenzaldehyde to give its 3,5-di-iodo derivative, IOAc is probably the iodinating agent. Another reaction of (65) is the oxidation of secondary alcohols into ketones; curiously, primary alcohols do not react. Trialkylboranes react with PID and boron is acetoxylated into RzBOAc with formation of an alkyl acetate.82 A polar mechanism has been proposed, [Equations (11) and (12)]. PhI(OAc), + R3B 2PhfOAc + R,BOAc (11) PhIbAc + R3BOAc --f R,BOAc + [Ph-I(R)OAc --f ROAc + PhI] (12) LTA reacts similarly but prefentially with boranes bearing secondary alkyl Y.Masuda and A. Arase, Bull. Cliem. Sac. Jpn., 1978, 51, 901. Aryliodine(1ir) Dicarboxylates groups, whereas PID converts only the primary ones. Thus the two reagents have a complementary action.It is of interest to note that the reaction with LTA follows a homolytic pathway. With 1-alkenyl dialkylboranes, both PID and LTA react similarly affording alkeness3 (Scheme 22). R X R PID \ /x\f= c/\ -F= CkH BRz H X = H, C1, or Br Scheme 22 The reaction is stereoselective, E-isomers being formed in a high proportion (E:Z-9 :1). Similarly, 1-bromo-and 1-chloro-1-hexenyl dialkylboranes react to give haloalkenes. Bulky alkyl groups and low temperatures favour Z-isomers for the bromo-compounds, while 2-isomers are formed from the chloro-compounds almost exclusively. The mechanism of these reactions is not yet known. C. Transfer of the Aryliodine Group.-The reaction of AID with activated aromatic compounds is one of the various methods developed by Beringer for the preparation of iodonium salts (70), [Equation (13)].H+ArH + ArI(OCOR)2-+ Ar-I+-Ar RC0,-+ RCOzH (13)(70) This reaction has been studied in detail, mainly with benzene, toluene, and polymethylben~enes~*-~~and it is a typical electrophilic aromatic substitution. Sulphuric acid is necessary for the reaction to proceed and it exerts a marked catalytic effect. Electron donors in the benzene ring of AID accelerate slightly the reaction rate, which is considerably greater with toluene than with benzene. The product distribution with toluene is approximately 90% for the para-and 10% for the ortho-isomer, in contrast to earlier observations where exclusive formation of the para-isomer was claimed.There is a primary hydrogen iso- tope effect in the reactions of PID with C6H6 and KH/K~H= 1.4. A decrease of basicity for AID (when there are electron acceptors in the ring) and an increase of basicity for the arene both lead to a decrease of the acti- vation energy, which varies between 40 and 80 kJ mol-1. These findings have been explained by assuming that sulphuric acid protonates PTD rather than pro- moting its dissociation [Equation (14)]. 83 Y. Masuda and A. Arase, Bull. Chem. SOC.Jpn., 1980,53, 1652. 84 D. J. LeCount and J. A. W. Reid, J. Ghem. SOC.(C),1967, 1298. 85 J. M. Briody, J. Chem. SOC.(B), 1968, 93. H. Hoffelner, H. Schneider, and H. Wendt, Chem.-Ztg., 1978, 102, 53.87 N. I. Nogina and V. A. Koptyng, Zh. Org. Khim.. 1972 8, 1495. 398 Varvoglis H+ H+ PhI+OAc + AcOH + PhI(OAc), + [PhI(OAc),H]+ (1 4) The protonated PID6 rather than PhI+OAc is the active electrophilic species which gives with the arene first a wcomplex and then a o-complex. The latter is transformed irreversibly into the iodonium salt and this is the rate determining step of the reaction,96 [Equations (15) and (16)]. [PhI(OAc),H]+ + ArH --+ [PhI(OAc)ArH]+ + AcOH (1 5)--f [PhI(OAc)ArH]+ + PhI+Ar + AcOH (16) Numerous iodonium salts have been prepared by the above reaction, especially with thiophene,a8 because thienyl iodonium salts are potential antimicrobial agents. Bifunctional AID, such as p-(AcO)zIC6H4I(OAc)z, may give not only bis-iodonium salts, but also polymeric iodonium salts.89 Of special interest are various heterocyclic iodonium salts, resulting from in situ I-acetoxylation of suitable iodoarenes, which subsequently cyclize with HzS04, e.g.(71),1 (72),9* (73),91 and (74).92 CH (74) (75) Such iodonium salts have been used in the preparation of 5-aryl-5H-diben-ziodoles93 and other compounds such as carbazole from (71), 2,2’,6,6’-tetra- 88 Y. Yamada and M. Okawara, Bull. Chem. SOC.Jpn., 1972, 45, 1860, 2515. as Y. Yamada, IS.Kashima, and M. Okawara, J. Polym. Sci.,Polym. Lett. Ed., 1976, 14, 65. so D. Hellwinkel, W. Lindner, and W. Schmidt, Chem. Ber., 1979, 112, 281. . s1 R.B. Sandin, J. Org. Chem., 1969, 34, 456. 92 F.M. Beringer, L. L. Chang, A. N. Fenster, and R. R. Rossi, Tetrahedron, 1969, 25, 4339. s3 F. M. Beringer and L. L. Chang, J. Org. Chem., 1971, 36, 4055; H. J. Reich and C. S. Cooperman, J. Am. Chem. SOC.,1973,95, 5077. Aryliodine(mr) Dicarboxylates iodobiphenyl from (73),91 and the chelate complex (75) of dibenziodolium cation.94 Perfluoroalkyliodine(IIr) bis(trifluoroacetates) react with tolueneg5 to give the stable iodonium salts (76), where there is an I-C bond to the sp3 carbon atom. These salts react under mild conditions with various nucle~philes,~~ which attack exclusively the sp3carbon atom. An interesting case of phenyliodination has been reported97 in the reaction of the phosphorane Ph3P=-CHCOOR with PID in the presence of HBF4, leading to the double salt (77).Me-d-/>-i-& ‘I+-Ph BF4-In certain favourable cases the iodonium salt of a heterocycle may be converted by alkali into an inner salt, i.e. an ylide of iodine (iodinane). Pyrazole’(78) gives, for example, the salt (79) with PID and p-toluenesulphonic acid and this is converted into the ylide (80),98 Scheme 23. Ph-I+ R TsOH N 1 I IH H H Scheme 23 With activated heterocycles iodine ylides are formed directly in an alkaline medkm, e.g. (82) from indole (81),99 in Scheme 24. O4 R. E. Lee and R. H. Soderberg, J. Am. Chem. SOC.,1972,94,4132. Db V. V. Lyalin, V. V. Orda, L. A. Alekseyeva, and L. M. Yagupolskii, Zh. Org. Khim., 1971, 7, 1473. 9(1 L. M. Yagupolskii, I. I. Maletina, N. V.Kondratenko, and V. V. Orda, Synthesis, 1978, 835. 97 0. Neilands and G. Vanags, Dokl. Akad. Nauk SSSR, 1964, 159, 373 (Chem. Abstr., 1965, 62, 6510). 98 B. Karele, S. V. Kalnin, I. Grinberga, and 0. Neilands, Khim. Geterosikl. Soedin., 1973, 245. Varvoglis IH (81) Scheme 24 It is of interest that indole and PID in MeOH without alkali give intractable tars, while 2,3-dimethylindole (83) reacts to give initially a N-I intermediate (84), which reacts further with alcohols to give a 3-alkoxyindolenine100 (Scheme 25). Reserpine reacts similarly. Iodine ylides may also come from substrates with sp3 carbon atoms i.e. 1 I5H Ph’ COAc Scheme 25 compounds with active methylene groups. A comprehensive list of iodine ylides prepared up to 1978 mostly from PID has been published.101 Among the various substrates used, the following are mentioned : barbituric acid, Meldrum’s acid, dimedone, Chydroxycoumarins, uracil, and acetoacetic and malonic esters.Several ylides have also been prepared from cyclopentadiene salts and PID or CICH=--CHI(IOAC)~.~~ A new type of I-N ylide is formed from PID and p-toluenesulphonamide,lo2[Equation (17)]. KOH PhI(OAc), + TsNH, +PhI+N-Ts (17)MeOH An interesting reaction leading to an iodine ylide is that between 4-nitrophenol and P1D.l An ester (86), is initially formed which rearranges to the iodonium OP B. Karele, L. Treigute, S. Kalnina, I. Grinberga, and 0.Neilands, Khim. Geterosikl. Soedin, 1974, 214. loo D. V. C. Awang and A.Vincent, Can. J. Chem., 1980,58, 1589. lol T. Kappe, G. Korbuly, and W. Stadlbauer, Chem. Ber., 1978, 111, 3857. lU2Y. Yamada, T. Yamamoto, and M. Okawara, Chem. Left., 1975, 361. Aryliudine(w) Dicarboxylates salt (87); this can be isolated but on storage over KOH in a desiccator it is transformed into the stable ylide (88). Crystallographic studies favour the betainic form (88b)l03 rather than the ylidic (88a),*04 (Scheme 26). P?I, ,OAc __3 +-+ NO2 NO2 NO2 AcO-NO2 NO2 Like other iodine ylides, (88) and related compounds with an electron acceptor X instead of NO2 rearrange thermally into 2-iodo-4-NOz (X)-diphenyl ethers, which may be used for the preparation of dibenzofurans.lo3 D. Substitutions at Iodine.-A fruitful area of PID chemistry deals with its reactions with alkenes in combination with MesSiN3.In a series of papers Zbiral and co-workers105 have shown that the double bond of both nucleophilic and electophilic alkenes reacts with PID-Me3SiN3, which are in equilibrium, as shown by i.r. spectroscopy, according to equation (18). PhT(OAc), + nMe,SiN, + PhI(OAc),-n(N,)n + nMe,SiOAc (18) The above system is very reactive; it reacts at -15 to -55 "C in two main ways: either the double bond is cleaved and a ketonitrile is formed (Scheme 27) PID-Me,SiN, HsO&-CN Scheme 27 or only the 7~ bond is cleaved with formation of an a-azidoketone (Scheme 28). lo3S. W. Page, E. P. Mazzola, A. D. Mighell, V. L. Himes, and C. R. Hubbard,J. Am. Chem. Suc., 1979, 101, 5858.lo* P. B. Kokil and P. M. Nair, Tetrahedron Lett., 1977, 4113. lo5 E. Zbiral and G. Nestler, Tetrahedron, 1970, 26, 2945; E. Zbiral and J. Ehrenfreund, ibid., 1971, 27, 4125; J. Ehrenfreund and E. Zbiral, ibid., 1972, 28, 1967; J. Ehrenfreund and E. Zbiral, Ann. Chem., 1973, 290; E. Cech and E. Zbiral, Tetrahedron, 1975, 31, 605. Varvoglis ArCH =CH2 PID + Me,SiN, H,O F ArCOCH2N3 Scheme 28 Various mechanistic schemes are possible for these reactions, the addition of PhIOAc and N3 being always the initial step. With 1,3-dienes 1,4-addition prevails but the final products are often the result of further transformation, e.g. from 2,3-dimethylbutadiene a mixture of 3,4-diazido-3-methylbutan-2-one and 2,3,5-trimethyl-3,6-diazidomethylohepta-l,5-dienehas resulted.The ternary system PID + MesSiNa + AcCl forms in situ PhI(N3)CI which adds to alkenes, with eventual formation of 1,2-chloro-azidoalkanes.~~5 isIt noted that PID with acyl chlorides alone105 as well as PhIO and acyl chlorides106 react to give phenyliodine(~~~) dichloride and acid anhydrides, probably through a common intermediate (Scheme 29). PhI(0Ac)z + AcCl Ad1PhI(0Ac)Cl PhIC12-Ac,O PhIO + AcCl Scheme 29 Apart from its exchange with carboxylic acids (Section 3), PID may form, with certain dicarboxylic acids, polymeric AID with unusual properties,l07 Equation (19). nPhI(OAc)z + nHO2C. * * * * *CO2H-+ [-OI(Ph)OCO* * *CO-]n (19) An interesting exchange of PID with substituted benzenesulphonic acids leads curiously to the hydroxy-derivatives (*89)10*(Scheme 30).Attempts to ob- tain the bis-tosyloxy-derivative were unsuccessful. The crystal structure of (89), where X = Me, has been determined and some of its reactions examined, main- ly with ArSiMe3 to iodonium salts. With aryl iodides metathetical redox reactions take place. PID also exchanges with hydroperoxides.lOg The reaction at room temperature is highly exothermic and CH4, C2H6, MezCO, and C02 are among the products when ButOOH is used. At -80 "C,however, in CHzCl2 a peroxy-derivative of J. Wicha, A. Zarecki, and M. Kocor, Tetrahedron Lett., 1973, 3635. lo' H. K. Livingston, J. W. Sullivan, and J. I. Musher, J. Polym. Sci. (C), 1968, 195; I. Kresta and H. K. Livingston, Macromolecules, 1972, 5, 25.lo8 0.Neilands and B. Karele, Zh. Org. Khim., 1970,6, 885; G. F. Koser and R. H. Wettach, J. Org. Chem., 1977, 42, 1476, 1980, 45, 1542, 4988; G. F. Koser, R. H. Wettach, J. M. Troup, and B. A. Frenz, ibid., 1976, 41, 3609; G. F. Koser, R. H. Wettach, and C. S. Smith, ibid., 1980, 45, 1543. loo N. A. Milas and B. Plesniear, J. Am. Chem. Soc., 1968, 90, 4450. 403 AryZiodine(111) Dicarboxylates /002s PhI(OAc)2 + H03SeX -HI0 PhI \OH X = H, Me, NOz, or Cl Scheme 30 1111 is formed, PhI(OOBut)z, which is unstable and by expelling ButOO. is reduced to PhI. The radicals combine initially to a tetroxide and then after a multistep process ButOOBut is eventually formed with 02 evolution. PID and 3-nitro- PJD form, with dilute HN03, stable p-oxo-dinitrato-diaryl-di-iodine(II1) derivat ives,llO Ar(ONO2)I-O-I(ONO~)Ar, while PID in fuming HN03 has been claimed89 to give PhT(ON02)2.The crystal structure of the p-phenyl com- pound has been rep0rted.l" E.Thermal and Photochemical Decomposition.-Early work on AID established that on heating they produce alkyl or aryl radica1s.l The thermal decomposition of AID poses interesting mechanistic problems and has been studied in detail by Leffler and his co-workers.ll2 Two major pathways have been recognized, involving heterolysis and homolysis (Scheme 3 1). + RCOz PhI + [2RC02] / PhI(OCOR)% + RC02-PhOCOR + [RCOJ] Scheme 31 The homolytic reaction is the minor one. The RC02. radicals (aroyloxy or acetoxy) lose C02 and dimerize into R-R, or RC02.and R- may combine to RC02R. Another possibility is the reaction of Re with the solvent. The hypo- 110W. E. Dasent and T. C. Waddington, J. Chem. SOC.,1960, 3350; W. E. Dasent and L. E. Sharman, ibid., 1964, 3492. ll1 N. W. Alcock and R M. Countryman, J. Chem. SOC.,Dalton Trans., 1979, 851. *laJ. E. Leffler, W. J. M. Mitchell, and B. C. Menon, J. Org. Chem., 1966, 31, 1153; J. E. Leffler and L. J. Story, J. Am. Chem. SOC.,1967, 89, 2333; T. T. Wang and J. E. Leffler, J. Org. Chem., 1971,36, 1531 ;J. E. Leffler, D. C. Ward, and A. Burduroglu, J. Am. Chem. SOC.,1972, 94, 5339. Varvoglis iodites formed in the heterolytic reaction decompose homolytically to COz and RI or they can be trapped with added alkene.In chlorobenzene, phenyl- iodine(u1) dibenzoate gives chlorobiphenyls (all isomers) in fair yield. The reaction is catalysed by benzoyl peroxide, which with chlorobenzene forms re- active phenylchlorocyclohexadienyl radicals serving as chain-transfer agents. When substituted PID’s are used the reaction is accelerated with electron-with- drawing groups (p Hammett value +0.8). By introducing aroyloxy- instead of acetoxy-groups the decomposition is accelerated and the reaction shifts towards the radical pathway; both Cnitrophenyl and Cmethoxyphenyl groups accelerate the reaction rate. The thermolysis of PhI(0COPh)z was also studied in bromobenzene at relatively high temperat~res.”~ In this system PhCOzI reacts with the solvent, furnishing 2- and 4-(but not 3-) bromoiodobenzenes.Also, the homolytic path produces PhCOzH, bromobiphenyls, and I2 as additional products. Iodine results from homolysis of PhCOJ, while PhCOzH comes from both the re- action of PhC02I with PhBr and the reaction of PhI-OCOPh with PhCeH4Br. Mechanistic details are beyond the scope of this article but it is of interest to note that at least twelve individual steps are necessary to account for the experimental data. In AcOH the decomposition of PID is very slow, even at 100°C, but it is greatly enhanced in the presence of perchloric acid, which effectively protonates PID, as has been shown by U.V. spectroscopy and conductivity studies.6 The protonated PID is equilibrated with phenyl acetoxyiodonium ion by a series of ion-pair equilibria and through a concerted reaction with AcOH gives the observed products, [Equations (20) -(22)].PhI(OAc), + H+ + [PhI(OAc),H]+ (20) [PhI(OAc),H]+ + PhI+OAc + AcOH (21) PhI+OAc + AcOH + PhI + CO, + MeC0,Me (22) The thermolysis of ArI(OAc)z, where Ar = mesityl, duryl, and pentamethyl- phenyl, was found to proceed in the absence of solvents at 130-170°C via the polar me~hanism.11~ The aryl acetates initially formed do not always survive the drastic conditions and in the case of pentamethyl PID, for example, the isolated products were 2,3,5,6-tetramethyl-4-iodophenylacetate and 1,4-diacetoxy-2,3,5 ,&tetramethyl benzene. In connection with the thermolysis of ATD113 it was found that PhI(OCOPh)2 when irradiated with U.V.light in the presence of 2,3-dimethylbutane forms almost exclusively the 2-benzoyloxy-derivative as in Equation (23). hv Me,CH.CHMe, IF PhI(OCOPh), --f Me,CH.CMe,OCOPh (23) Since benzoyl peroxide under similar conditions gives MezCHCMezOCOPh and Me2CHCHMeCH20COPh in a ratio of about 2: 1, it was concluded 113 C,J. Grill, G. B. Gribb, and H. A. R. El-Jamali, J. Chem. Suc., Perkin Trans. 2, 1977, 860. 114 E. B. Merkushev, A. N. Novikov, and L. F. Kharitonova, Zh. Org. Khim., 1971, 7, 519. AryZiodine(rI1)Dicarboxylates that both in photolysis and thermolysis the active species is PhIOCOPh rather than PhCOO.. PID is decomposed photochemically6 in AcOH into CsH51, C02, MeCOzMe, and probably ethane, with a quantum yield of 0.62.Kinetic studies established that the rate determining step of the reaction is the activation of PID. 5 Analytical Applications AID normally oxidize iodide to iodine (see however the reaction with Et4N+I-, p. 397). This reaction can serve as an assay of their purity and also in kinetic measurements where unreacted AID liberates iodine, which is titrated with thiosulphate. Quantitative determination of AID may be effected also by automatic potentiometric titration.l15 PID has been proposed as a redox titrant116 in oxidationsof Ass+, Fe2+, [Fe(CN)6I4-, T1+, Sb3+, and N2H4. Organic reductants undergoing quantitative oxidation include hydroquinone and ascorbic acid; phenol, aniline, and oxine (8-hydroxyquinoline) were also oxidized in the presence of bromide, which is first oxidized to bromine and then forms a tri- or dibromo-derivative.The use of PID as a redox titrant has been extended to the determination of thiols, dithiocarbamates, and ~anthates.11~ Volumetric determinations of thiols, thioureas, and other sulphur compounds have also been reported.118-120 Determinations of AID and iodonium salts in mixtures have been performed by polarography.86 6 Miscellaneous An interesting property of AID is their action as hydroxylating agents in bio- chemical reactions.13J21 It has been found that, in addition to NADPH, several AID can effect steroid and fatty acids hydroxylations in rat-liver microsomes, by acting as oxygen donors to cytochrome P-450. The most efficient was 2-nitro-PID, (several hundred times as active as NADPH), whereas 2-cyano- PID was inactive.PPD can induce polymerization of 2,6-dimethylphenol.l22 Poly-p-styryl- iodine(rI1) diacetate may be used as a polymer reagent.123 PID has been proposed as an oxidant for leuco anthraquinone dyes in colour photography,12* for the dissolution of cellulose ethers in water,125 and as a component of a thermal 115 S. S. Makarchenko, A. N. Novikov, and G. M. Kropacheva, Dep. Doc. VZNZTZ, 1975, 239-75 (Chem. Abstr., 1977, 87,95045). 116 V. N. Pilai and C. G. R. Nair, Talanta, 1975, 22,57. 117 K. K. Verma, J. Ahmed, M. P. Sahasrabuddhey, and S. Bose, J.Indian Chem. SOC.,1977, 54,699. 118 K.K. Verma, Fresenius’ Z. Anal. Chem., 1975, 275,287. llS K.K. Verma and S.Bose, J. Indian Chem. SOC.,1973, 50, 542. la0 K. K. Verma and S. Bose, Philipp. J. Sci.,1974, 103, 187 (Chem. Abstr., 1976, 84, 25 586). 121 J. A. Gustafsson and J. Bergman, FEBS Lett., 1976, 70,276. lar J. Kresta and H. K. Livingston, J. Macromol. Sci,, Chem., 1970, 4, 1719 (Chem. Abstr., 1970, 73, 99 300). lZ3 Y. Yamada and M. Okawara, Makromol. Chem., 1972, 153; M. L. Hallensleben, Angew. Makromol. Chem., 1972, 27,233. S. J. Ciurca, jun.,Fr. Demande 2251 841 (Chem. Abstr., 1976, 84, 128747). Ips C. D. Callihan and J. R. Boudreaux, U.S.P.337628 (Chem. Abstr., 1968,68, 106207). 406 Varvoglis reaction battery.1z6 The most unexpected entry in Chemical Abstracts was the ‘effect of PID on the pacemaker action-potential and contraction in isolated guinea-pig atria’.lZ7 Acknowledgment.I wish to thank Mr. E. Kritsis for his assistance in the prepara- tion of the manuscript. lZ6 A. A. Benderly and D. R. Hartler, U.S.P. 3819415 (Chrm. Absfr., 1974, 81, 108466). 12’ M. Tani, Acfa Med. Nagasaki, 1968, 12, 99 (Chrm. Abstr. 1969, 70, 95 115).

ISSN:0306-0012    

DOI:10.1039/CS9811000377

出版商:RSC    

年代:1981

数据来源: RSC