4 Reaction Mechanisms Part (iii) Free Radicals By J. A. MURPHY and M. S. SHERBURN Department of Chemistry University of Nottingham University Park Nottingham NG7 2RD The year has seen advances in many aspects of the chemistry of free radicals. The development of alternative reagents as sources of carbon radicals and of modified ways of performing radical reactions has continued apace. Curran has published full details of atom-transfer reactions of hexynyl iodides' and of iodoesters2 (includ- ing iodomalonates). The utility of the atom-transfer approach is seen in the reactions of the iodoester (I) (Scheme 1). Treatment of this molecule with tributyltin hydride and a radical initiator led to the acyclic product (4) but no lactone (3); however reaction with hexabutylditin under photochemical activation gave the iodolactone (2) which was isolated and could then be reduced with tributyltin hydride to the desired target (3).0 0 Reagents i Bu,SnH AIBN; ii Bu3SnSnBul hu Scheme 1 The use of tris(trimethylsily1)silane has been further investigated (see 1988 Report) as an alternative to tributyltin hydride by Gie~e.~ The greater strength of the Si-H bond in this molecule (79 kcal mol-' versus 74 kcal mol-' for Sn-H in tributyltin hydride) makes it a poorer H-atom donor. A range of reactions of carbon radicals including cyclizations rearrangements and addition to activated alkenes have now been shown to proceed well using this silane. Tris( trimethylsily1)silane is a 'special' silane. Roberts4 has examined the possibility of using the less esoteric triethylsilane as a reducing agent for alkyl halides (Scheme ' D.P. Curran M.-H. Chen and D. Kim J. Am. Chem. SOC.,1989 111 6265. * D. P. Curran and C. T. Chang J. Org. Chem. 1989 54 3140. B. Giese B. Kopping and C. Chatgilialoglu Tetrahedron Lett. 1989 30,681. R. P. Allen B. P. Roberts and C. R. Willis J. Chem. Soc. Chem. Commun. 1989 1387. 73 J. A. Murphy and M. S. Sherburn 2). Reactions (1) and (2) are necessary propagation steps in this process. The reaction (2) is normally very sluggish and this renders silanes incapable of sustaining the radical chain reaction. However Roberts has suggested that the sluggishness of the reaction is due to polar factors in the free radicals concerned i.e.that a nucleophilic carbon radical would be slow to abstract an electron-rich hydrogen from a Si-H bond. His analysis has led him to suggest a thiol/silane couple as a means of overcoming this replacing reaction (2) by reactions (3) and (4) and indeed he has demonstrated that the presence of thiol works well in catalysing alkyl halide reduc- tions by radical chain mechanisms. Et,Si' + RHal Et,SiHal + R' (1) R' + Et,SiH aRH + Et,Si' (2) R' + XSH 5RH + XS' (3) XS' + Et,SiH -@+ XSH + Et,Si' (4) Scheme 2 Reactions of radicals which are followed by rapid trapping by an oxidant or metallo-radical have drawn much attention. Minisci' has developed an atom-transfer route for substitution of protonated heterocycles (Scheme 3). The interaction of hydroxyl radicals (generated by reaction of hydrogen peroxide with ferrous salts) with dimethyl sulphoxide is known to produce methanesulphinic acid and methyl radicals.6 The methyl radicals interact exothermically with other alkyl iodides to produce iodomethane and an alkyl radical and as most alkyl radicals are much more nucleophilic than methyl radicals this leads to a rapid addition to a pyridinium salt.The resulting radical cation undergoes proton loss followed by facile oxidation resulting ultimately in an overall substitution of the heterocycle. The selectivity against methyl group substitution is shown by the fact that even when primary alkyl iodides are used it is predominantly the primary alkyl group which is found on the product pyridine; less than 2% of methylated product is detected.0 '0 OH II \/ HO' + MeSMe -MeSMe -MeS0,H + Me' Me' + R-I -Me1 + R' H H H H Scheme 3 F. Minisci E. Vismara and F. Fontana J. Org. Chem. 1989 54 5224. D. Veltwisch E. Janata and K. D. Asmus J. Chem. SOC. Perkin Trans. 2 1980 146. Reaction Mechanisms- Part ( iii) Free-radical Reactions The use of manganese(II1) acetate in the presence of cupric salts has been further investigated. Bertrand' reports on the cyclizations of malonate diesters. The remark- able point about these reactions (Scheme 4) is that the kinetic cyclization products can be trapped with ease by added cupric ion; where the possibility exists for 5-ex0 versus 6-endo cyclization to occur the five-membered ring products are observed.Thus in the example shown the initial cyclized radical (9) is rapidly oxidized to cation (10) and then cyclopropane formation occurs. Snider' reports a detailed evaluation of the structural features required for successful intermolecular addition of radicals derived from malonates and P-ketoesters to alkenes using manganese( 111) acetate. In effect an electron-rich alkene is required; the reaction is also very sensitive to steric effects at the site of cyclization. 0 __. I. 00 0U O M e i I. Scheme 4 Giese9 has used [CpFe(CO),] to generate organic radicals. That this reagent functions as a source of radicals was inferred from the products formed in its reactions (Scheme 5). However the formation of alkenes (11) as well as saturated products (12) is reminiscent of the chemistry of organocobalt reagents where ' H.Ournar-Mahamat C. Moustrou J.-M. Surzcr and M. P. Bertrand J. Org. Chem. 1989 54 5684; Tetrahedron Lett. 1989 30 331. B. B. Snider and B. 0. Buckman Tetrahedron 1989 45 6969. G. Thorna and B. Giese Tetrahedron Lett. 1989 30 2907. J. A. Murphy and M. S. Sherburn $[CpFe(CO),] -[CpFe(CO),]' +R' + CpFe(C0)2X I iii RCH=CYZ + RCH,CHYZ 2RCH,kYZ (11) (12) Reagents i hv;ii RX;iii H,C=CYZ; iv CpFe(C0); Scheme 5 trapping of intermediate radicals by cobalt( 11) species followed by dehydrocobalta- tion leads to unsaturated products. Product analysis has also recently been used to deduce the nature of intermediates in the photolysis or thermolysis of aryl- and alkylcobalt(~r~) The conclusions drawn are that these processes produce free radicals which function essentially independently of the metal.Organocobalt chemistry has developed well since the initial experiments featuring models of the biological chemistry of coenzyme B12.This year has seen (a) further modelling of these biological reactions and (b) the development of synthetic methods some of which are inspired by the B, rearrangement reactions. Pattenden" has developed a synthesis of p- y- and &lactams from acylcobalt salophens. The synthesis of the p-lactams (Scheme 6) is particularly interesting in view of the pharmaceutical importance of these compounds. Trapping of the cyclized radical by Co" is obviously important in efficiently forming these strained products.OR' n = 1,2,3 Scheme 6 Dowd13 Widdo~son,'~ and Golding15 have all made contributions to the B12 modelling experiments. The methylaspartate-glutamate rearrangement (Scheme 7) has been examined by Dowd who has proposed a mechanism for this (Scheme 8) in which methylaspartate forms an imine (13) which then tautomerizes to (14). This is then converted into the radical (15) by coenzyme B1 and rearranges. In the model lo A. J. Clark and K. Jones Tetruhedron Lett. 1989 30,5485. B. Giese J. Hartung J. He 0. Hutter and A. Koch Angew. Chem. Int. Ed. Engl. 1989 28 325. G. B. Gill G. Pattenden and S. Reynolds Tetrahedron Lett. 1989 30,3229. l3 S.-C. Choi and P. Dowd J. Am. Chem. SOC.,1989 111 2313. l4 W.M.Best and D. A. Widdowson Tetrahedron 1989 45 5943. *' S. Ashwell A. G. Davies B. T. Golding R.Hay-Motherwell and S. Mwesigye-Kibende J. Chem. Soc. Chem. Commun. 1989 1483. Reaction Mechanisms- Part (iii) Free-radical Reactions H02CJNH2 H02CqNH2 ~ CO2H CO2H Scheme 7 "02TH2 HO2C /N-R -.JC02H CO2H (17) I COzEt (18) \ iii / (19) Reagents i B,,-dependent enzyme; ii vitamin Biz, room temperature EtOH 5 min; iii Bu3SnH AIBN A Scheme 8 experiments the halide (18) acts as a precursor for a radical analogous to (15); this gives the rearranged product (19) on treatment with either tributyltin hydride or vitamin B,2s. The methylmalonyl coenzyme A mutase reaction continues to inspire discussion of its mechanism. A full report of the modelling experiments conducted by Widdow- son has appeared.14 Although model thioesters do not rearrange well p-ketoesters do undergo a radical-induced rearrangement efficiently.The scope of this rearrange- ment as a synthetic method has now been extended by the synthesis of ring-expanded heterocycles by Dowd16 (Scheme 9). Modelling of biological processes has also been the motivation behind Whiting's recent investigation of reactivity of aryloxymethyl radicals;" it is proposed that the carbon atom of methoxy groups may be used biosynthetically to form carbon-carbon bonds by a radical mechanism and a series of experiments has been carried out 16 P. Dowd and S.-C. Choi Tetrahedron Lett. 1989 30 6129; Tetrahedron 1989 45 77. 17 A.J. Walkington and D. A. Whiting Tetrahedron Lett. 1989 30 4731; S. A. Ahmad-Junan A. J. Walkington and D. A. Whiting J. Chem. SOC.,Chem. Commun. 1989 g13. J. A. Murphy and M.S. Sherburn Bu,SnH AIBN Etozcao L I Ph \ Ph Scheme 9 OMe OMe I S Q OMe OMe Scheme 10 using radical chemistry to mimic this (Scheme 10). One precursor of such a radical was (21). The isolation of the unsaturated compound (23) as the main product is interesting and presumably indicates the intermediacy of the pyridine (22). Keeping with radicals of biological interest leads us to one of the most interesting developments this year namely the study of the related DNA-cleaving agents calicheamicin esperamicins and neocarzinostatin. These molecules are thought to undergo an initial activation by addition of thiol (or possibly thiyl radical2* i,n the case of neocarzinostatin) followed by spontaneous cyclization to form biradicals which in vim abstract two H-atoms from deoxyribose units on complementary Reaction Mechanisms- Part ( iii) Free-radical Reactions strands of duplex DNA.This leads to double-strand cleavage. The reaction is shown here for neocarzinostatin (Scheme 11). The complex structures of the molecules has meant that an enthusiastic search has started for reagents which perform the same function but have simpler structures. Among the contributors this year have been Magnust’ Danishefsky,” Saito,20 Hirama,” Ellestad,22 Snyder,23 Myers,24 and Nic~laou.~~ Simpler model compounds have been shown to mimic the cyclization chemistry.Scheme 12 shows the models (24)24 and (27).18 ArC0,-0-sugar 0-5ugar 1 ArC02-ArC02--W /-OH 0-sugar H 0-sugar Scheme 11 Reports on stereochemical control of free-radical chemistry are prominent. Rajan- babu has continued his studies on the stereochemistry of bicyclization reactions which allow the effects of conformation and substitution pattern to be assessed. Firstly,26 he has identified that for non-flipping cyclohexanes cyclization can occur onto either an axial or an equatorial substituent bearing an appropriate functional in P. Magnus and R. T. Lewis Tetrahedron Lett. 1989 30 1905. 19 N. B. Mantlo and S. J. Danishefsky J. Org. Chem. 1989 54 2781; J. N. Haseltine S. J. Danishefsky and G.Schulte J. Am. Chem. Soc. 1989 111 7638. 20 I. Saito H. Kawabata T. Fujiwara H. Sugiyama and T. Matsuura J. Am. Chem. SOC.,1989 111 8302; R. Nagata H. Yamanaka E. Okazaki and I. Saito Tetrahedron Lett. 1989 30,4995. 21 M. Hirama K. Fujiwara K. Shigematu and Y. Fukuzawa J. Am. Chem. Soc. 1989 111 4120. 22 N. Zein W. J. McGahren G. 0. Morton J. Ashcroft and G. A. Ellestad J. Am. Chem. Soc. 1989 111 6888. 23 J. P. Snyder J. Am. Chem. Soc. 1989 111 7630. 24 A. G. Myers and P. S. Dragovich J. Am. Chem. Soc. 1989 111 9130; A. G. Myers and P. J. Proteau ibid. p. 1146. 25 K. C. Nicolaou G. Skokotas P. Maligres G. Zuccarello E. J. Schweizer K. Toshima and S. Wendeborn Angew. Chem. Int. Ed. Engl. 1989 28 1272. 26 T. V. Rajanbabu and T.Fukunaga J. Am. Chem. Soc. 1989 111 296. J. A. Murphy and M. S. Sherburn 0 \OR H \?R 0\c 0 Reagent i cyclohexa-1,4-diene Scheme 12 1,5 cis trans 22 60 1,5 cis :trans g0 :13 Scheme 13 group (Scheme 13). A full investigation of the stereochemistry of products of cyclization of a series of sugar acetals (Scheme 14) has been published.27 This should prove very useful in predicting the stereoselectivity of related cyclizations of sugar-derived molecules. Although most of the emerging radical chemistry of sugars involves direct attack by tributyltin radicals on a halide or chalcogen Fraser-Reid has used nitrate esters as sources of oxy-radicals which cause intramolecular hydro- gen transfer reactions and so functionalize otherwise unreactive sites on glycals with control of stereochemistry.** On the other hand the ease with which hydrogen Ph’ -LV OBn -Ph’ c;;q OBn OBn OBn Bn = benzyl exclusively trans Scheme 14 21 T.V. Rajanbabu T. Fukunaga and G. S. Reddy J. Am. Chem. SOC.,1989 111 1759. 28 J. C. Lopez R. Alonso and B. Fraser-Reid J. Am. Chem. Soc. 1989 111 6471. Reaction Mechanisms- Part (iii) Free-radical Reactions OBn Bu,SnH b AIBN L Bn = benzyl I OBn BnO -BnO Scheme 15 transfer occurs in sugar radicals has posed problems in De Mesmaeker’s studies on formation of fused-ring (Scheme 15). The stereochemistry of free-radical processes features in a different way in the work of Here attempts at inducing stereochemistry in intramolecular and intermolecular reactions have met with partial success (Scheme 16).Radical additions to activated alkenes of this sort resulted in moderate stereochemical induction. 0 0 14 :1 stereoisomeric ratio Scheme 16 In the fragmentation reactions reported by Bald~in~~ in which medium and large rings are generated the stereochemistry of the final alkene is markedly affected by the orientation of substituents in the starting material (Scheme 17). This is indicative of a concerted radical fragmentation a useful and very rare event in radical chemistry. 29 A. De Mesmaeker P. Hoffmann B. Ernst P. Hug and T. Winkler Tetrahedron Lett. 1989 30 6307. 30 A. De Mesmaeker P. Hoffmann and 9. Ernst Tetrahedron Lett. 1989 30,57. 31 N.A. Porter 9. Lacher V. H.-T. Chang and D. R. Magnin J. Am. Chem. Soc. 1989 111 8309. 32 N. A. Porter D. M. Scott 9. Lacher 9. Giese H. G. Zeitz and H. J. Lindner J. Am. Chem. Soc. 1989 111 8311. 33 J. E. Baldwin R. M. Adlington and J. Robertson Tetrahedron Lett. 1989 30 909 82 J. A. Murphy and M. S. Sherburn 0 eseph 0.1 equiv. Bu,SnH AIBN SnBu 89% 85% Scheme 17 The formation of macrocycles by direct cyclizations has been extended to the synthesis of a natural product -the cembranoid mukulol- by Pattenden34 (Scheme 18). mukulol Scheme 18 Two interesting reactions of macrocycles bearing radicals have appeared. Thus Suarez3’ has studied the transannular functionalization of medium-sized lactams with diacetoxyiodobenzene and has efficiently generated bicyclic products in some cases e.g.the amide (30) produces the isomeric products (31) and (32) in a total yield of 96% (Scheme 19). The cyclization of the radical (33) has been shown36 to Scheme 19 give large amounts of the bridged bicycle (34),which is of interest because of its occurrence in taxane diterpenes (Scheme 20). Unfortunately considerable amounts of non-cyclized reduced product (35) were also seen. The Z-isomer of (33) gave a much higher ratio of (34):(35)than did the E-isomer. 34 N. J. G. Cox G. Pattenden and S. D. Mills Tetrahedron Lett. 1989 30,621. 35 R. L. Dorta C. G. Francisco and E. Suarez J. Chem. Soc. Chem. Commun. 1989 1168. 36 J. D. Winkler V. Sridar and M. G. Siegel Tetrahedron Lett.1989 30.4943. 83 Reaction Mechanisms- Part (iii) Free-radical Reactions Scheme 20 The radical chemistry of small rings continues to feature prominently in the development of synthetic method^.^'-,^ Feldman38 has extended his earlier work to give a synthesis of (*)-yashabushitriol (Scheme 21). The crucial part of this sequence of reactions was the trapping of a peroxyl radical (36) to give specifically the syn stereochemistry in the dioxolane product. Peroxyl radicals have also featured in Bloodworth's studies of cyclization reactions of hydroperoxides?' His experiments have led him to propose that in cyclizations mediated by N-iodosuccinimide a radical chain mechanism prevails (Scheme 22); this contrasts with the case for N-bromosuccinimide which proceeds by a polar mechanism.- O-o* SPh i Ph Ph (34)I 0-0 Ph Ph -Ph Ph yashabushitriol Reagents i O, Ph,S, AIBN hv Scheme 21 0. OH Scheme 22 A review of the chemistry of thiocarbonyl compounds has been written by Cri~h.~~ Several advances have been made involving radical chemistry of sulphur and selenium compounds. Crich has ~ynthesized~~ a model of the A ring of la,25-dihydroxy vitamin D ,developing his earlier work on acyl radical cyclizations from 37 D. L. J. Clive and S. Daigneault J. Chem. Soc. Chem. Commun.,1989 332. K. S. Feldman R. E. Ruckle jun. and A. L. Romanelli Tetrahedron Lett. 1989,30 5845; K. S. Feldman and R. E. Simpson ibid. p. 6985; K. S. Feldman and T. E. Fisher ibid. p. 2969. 39 A.Johns J. A. Murphy and M. S. Sherburn Tetrahedron 1989,45 1625. 40 A. J. Bloodworth and R. J. Curtis J. Chem. Soc. Chem. Commun. 1989 173; A. J. Bloodworth R. J. Curtis and N. Misty ibid. p. 954. 41 D. Crich and L. Quintero Chem. Rev. 1989 89 1413. 42 D. Batty D. Crich and S. Fortt 1.Chem. SOC.,Chem. Commun.,1989 1366. J. A. Murphy and M. S. Sherburn i+ I Scheme 23 acylselenides. Yamamoto has investigated radical cyclizations of homoallylic xan- thate~~~ (Scheme 23) and cyclizations of various derivatives of thionocarbonic acids have been explored44 as routes to thionolactones. During this last study examples of ips0 substitution on aromatic rings were encountered (Scheme 24) e.g.the reaction of (37) gives the expected product (39) in low yield and the anomalous compound (38) as the major product.The proposed route to (38) is shown. This type of substitution has also been seen in Clive’s study of the formation of cis-fused cyclopentanoids by Michael addition and radical cycli~ation~~ (Scheme 25). In this case aromatic sulphones suffered attack and substitution. A novel and very useful means of removing aromatic sulphones in p-ketosulphones using tributyltin hydride has been discovered by Smith;46 these sulphones are frequently difficult to remove by alternative reactions. (37) OKoph -ph% (38) 47% (39) 9% S Ph +pkf (37) Scheme 24 The determination of rate conslants has continued with the publication of values for many important radical processes this year.These figures are extremely helpful in determining the feasibility of synthetic proposals featuring potentially competing reactions. Two examples of cyclopropylalkyl radical ring openings (Scheme 26) serve to show the effect of structure on the rate of reactions in the same series. Thus 43 M. Yamamoto T. Uruma S. Iwasa S. Kohmoto and K. Yamada J. Chem. SOC.,Chem. Commun. 1989 1265. 44 M. D. Bachi and E. Bosch J. Org. Chem. 1989 54 1234. 45 D. L. J. Clive and T. L. Boivin J. Org. Chem. 1989 54 1997. 46 A. B. Smith 111 K. J. Hale and J. P. McCauley jun. Terruhedron Leu. 1989 30 5579. Reaction Mechanisms- Part ( iii) Free-radical Reactions Scheme 25 (41) Scheme 26 the bicyclic compound (40)47 undergoes ring opening with a rate constant of 2.4 x lo9s-l at 37 "C.This has been used to probe the remarkable speed of 'OH rebound' in cytochrome P-450 and hence to help explain how hydroxylation with P-450 can proceed through free radical intermediates and yet feature retention of stereochemistry at the reacting carbon.48 The very much more stabilized radical (41) on the other hand undergoes ring opening with a rate constant of 3.6 x lo5s-' at 22 0c.49 47 V. W. Bowry J. Lusztyk and K. U. Ingold J. Am. Chem. SOC.,1989 111 1927. 48 P. R. Ortiz de Montellano and R. A. Steams J. Am. Chem. SOC.,1987 109 3415. 49 J. Masnovi E. G. Samsel and R. M. Bullock J. Chem. SOC.,Chem. Commun. 1989 1044.