OBun O O I ButO N O Ph OBun + O Ph OBut (55%) (28%) TEMPO K2CO3 Scheme 1 3 Reaction mechanisms Part (iii) Free-radical reactions By STEPHEN CADDICK VERN M. DELISSER and CRAIG L. SHERING The Chemistry Laboratory University of Sussex Falmer Brighton UK BN1 9QJ 1 Introduction Progress continues unabated in the area of free radical chemistry and numerous exciting developments have been reported in the last year with an increasing number of review articles highlighting the speed with which the state of the art changes.1 Most of the important synthetic developments have been a consequence of the increased availability of quantitative data and more important additions have been made to this data.2 Once again there has been considerable debate on the intermediacy of free-radicals in iron(III)-mediated alkane oxidation reactions and there appears in the literature to be a general consensus that radicals are involved in some of these reactions.3 2 Initiators promoters reagents and precursors The search for alternative methods for generating and utilising radicals in synthesis is a high priority.Recently Jang has shown that hypophosphorous acid can mediate iodide and certain bromide reductions in aqueous solution.4 Recent work has demonstrated that hypervalent iodine reagents can be used to generate iodine radicals which can then undergo oxidation reactions. The intermediacy of free radicals has been demonstrated by trapping with TEMPO (tetramethylpiperidineN-oxyl) as shown in Scheme 1.5 Magnus et al. have developed oxidation procedures using iodosylbenzene and Royal Society of Chemistry–Annual Reports–Book B 55 Si O 1 Si O Si O N3 N3 SiO N3 N3 2 91% PhIO TMSN3 TEMPO –45 °C PhIO TMSN3 TEMPO –45 °C 90% Scheme 2 trimethylsilyl azide to mediate some interesting transformations.It has been found that treatment of 1 with TMSN 3 and PhIO in the presence of TEMPO leads to the diazide 2 and further experiments support the intermediacy of azido radical formation (Scheme 2).6 TEMPO has also been used as a trap for a range of organometallic reagents as illustrated by Nagashima and Curran (Scheme 3).7 The work provides a simple procedure for the transformation of a halide to an alcohol although the success of the transformation is dependent on the organometallic reagent used. The advent of combinatorial synthesis has demanded new reactions and technologies and although radical chemistry has much to o§er this area little work has been reported so far.However recently Curran and Hadida have reported a new reagent tris[(2-perfluorohexyl)ethyl]tin hydride which is readily prepared and will promote standard radical reductions under stoichiometric and catalytic conditions (using NaCNBH 3 ). The reactions are carried out under biphasic conditions using tert-butyl alcohol and (trifluoromethyl)benzene and the products are readily separated without using chromatography. 19F NMRcan be used to detect the presence of any impurities (which were not observed) and the reagent’s use was extended to some simple addition processes.8 3 Intramolecular reactions Radical cyclizations continue to find considerable utility in organic synthesis and in a very nice piece of methodology Crich et al.have examined the use of polarity reversal catalysis to improve the ratios of five- versus six-membered ring formation in vinyl radical cyclizations.9 It is well known that the ratio of endo to exo products is directly 56 S. Caddick et al. n–C6H13M TEMPO THF –78 °C – room temp. N n–C6H13O M Yield (%) CuCN•ZnI CuCN•Li MgBr Li 26 68 70 78 Scheme 3 I OH OH H H OH TBTH 3 1 PhSeSePh/TBTH 5 1 3 Scheme 4 related to concentration; the use of a high stannane concentration is required to obtain the 5-exo products. The problem is that these conditions may lead alternatively to direct reduction and therefore it is important to be able to modify the concentration but still minimize rearrangement; this requires the use of a more e¶cient hydrogen atom donor.Benzeneselenol is an excellent hydrogen atom donor for alkyl radicals [k 25 \2.1]109 s~1 cf. tributyltin hydride (TBTH) k 25 \2.4]106 s~1] and can be generated by reaction of PhSeSePh with TBTH. Treatment of 3 with stoichiometric TBTH and catalytic PhSeSePh led to improved ratios of 5-exo products as shown in Scheme 4. Presumably benzeneselenol is a poor hydrogen atom donor for vinyl radicals; this is consistent with earlier work from the Crich group which had shown that benzeneselenol is a poor hydrogen atom donor for aryl radicals. Some other notable recent examples of intramolecular cyclization reactions include addition of alkyl radicals generated from 4 to enynes using TBTH or SmI 2 to give dienes 5.10 Fu and Hays report addition of a-alkoxyl radicals to alkenes; the generation of the a-alkoxyl radical uses the addition of TBTH to aldehydes and is perhaps unsurprising.However the reactions are carried out using only 5–15% (Bu 3 Sn) 2 Oand 0.5 equiv. of PhSiH 3 ; the reaction proceeds by generation of TBTH using the known transformation of tin alkoxides with silicon hydrides. The procedure is simple to carry out and proceeds to give good yields of cyclization products (Scheme 5).11 Russell and Li have described cyclization of N-phenylacrylamide with ButHgI–KI followed by photolysis with Ph 2 S 2 as shown in Scheme 6.12 The reactions succeed because the initially formed a-carbonyl radicals are relatively unreactive toward PhSSPh but the product nucleophilic radicals trap out rapidly. Booker-Milburn et al. have extended some of their results on iron(III)-mediated radical cyclization reactions.13 They have found that cyclopropanone acetal undergoes oxidative cyclization with Fe(NO 3 ) 3 and can be trapped with external radical 57 Reaction mechanisms Part (iii) Free-radical reactions X O Br X O 5 4 SmI2 Bu tOH DMPU 79–84% TBTH AIBN 80 °C 32–74% O Ph OH Ph Ph OH + (Bu3Sn)2O PhSiH3 EtOH AIBN 80 °C 85% 1.2 1 or Scheme 5 Scheme 6 Me3SiO EtO SPh EtO O 66% Fe(NO3)3 (PhS)2 DMF Scheme 7 traps to give the desired product in good yields.The use of ferric nitrate appears to be advantageous even when preparing the chlorine atom transfer products and has promise as a non-tin method for the production of carbon-centred radicals (Scheme 7). The use of Lewis acids has become extremely popular in radical chemistry usually as a method for attempting to control the stereoselectivity of a particular process.However Bowman et al. have elegantly shown the di§erential reactivity of a complexed and thus electrophilic aminyl radical to great e§ect in some tandem cyclization reactions as exemplified in Scheme 8.14 The addition of aryl radicals to cyclic vinylsilanes has been reported by Mignani and co-workers who have used this to prepare novel bicyclic adducts (Scheme 9).15 The radical variant of the Brook rearrangement described by Tsai and co-workers has been used to develop some nice cyclization sequences and an extremely clever extension has been recently reported.16 The incorporation of the radical acceptor onto the silane to allow migration and thus allow ultimate trapping by addition is an interesting concept which may find some use in target synthesis (Scheme 10).58 S. Caddick et al. PhSe N Bn TBTH MgBr2•OEt2 27% N Bn Scheme 8 Si Ph Ph N H Br Pr nSnH AIBN 40% Si NH Ph Ph 3 Scheme 9 Si Br O TBTH AIBN HO OH + HO OH 21% 40% Scheme 10 Sibi and Ji have described an investigation into the radical cyclization of Nenoyloxazolidinones. The use of oxazolidinones in mediating diastereoselective radical reactions has been fairly extensively explored; one of the problems which may arise is the predominant conformation at low temperatures. For example treatment of 6 with TBTH under photochemical conditions gave no desired cyclization but by carrying out the reaction at elevated temperature the desired cyclization product 7 was isolated (Curtin–Hammett principle). Most interesting was the behaviour of the substrate toward tris(trimethylsilyl)silane (TTMSS) which under analogous conditions gave only a 13% yield of the desired product.The authors propose that the presence of tri-n-butyltin halide a§ects conformation and consistent with this hypothesis the addition of tri-n-butyltin chloride to the TTMSS-mediated reaction led to good yields of cyclized product. Indeed even the TBTH-mediated reaction showed an improvement when carried out in the presence of the added tin chloride reagent (Scheme 11). Although tri-n-butyltin chloride is likely to be a§ecting the conformation it is worth noting the poor yields of cyclized products that are isolated when reactions are carried out in the presence of an additive at room temperature.17 Togo et al. have proposed that nitrogen-centred radicals can be generated by irradiation of the parent amine and diacyloxyiodoarenes using a tungsten lamp to give bicyclic products 8 in moderate to good yields.18 The presence of the sulfonyl group was essential for cyclization; when other electron-withdrawing groups were used (trifluoroacetyl benzoyl etc.) and if a suitably positioned hydrogen was available an alternative Ho§man–Loßer–Freytag pathway could be observed.The same workers have described related work on alkoxylation (Scheme 12).19 Some intramolecular substitution reactions of sulfonyl-substituted indoles and related systems have been reported. Thus fused indoline systems 9 were synthesised in 59 Reaction mechanisms Part (iii) Free-radical reactions Scheme 11 NH SO2CF3 I2 hn 40–80% N SO2CF3 8 OH O I2 hn ArI(O2CMe)2 60% Scheme 12 good yield by ipso-substitution of the sulfonyl residue and vinyl and aryl radical additions were demonstrated (Scheme 13).20 Radical based carbonylations developed by Ryu and others have gained increasing utility in synthesis and Fallis and Brinza have described a tandem reaction involving the cyclization of acyl radicals onto hydrazones as shown by the transformation of 10 into 11 (Scheme 14).21 In the natural product area Murphy and co-workers have extended their versatile tetrathiafulvalene(TTF)-mediated radical polar crossover technology to model studies of Aspidosperma alkaloids as shown in the transformations in Scheme 15.The incorporation of the amide functionality by trapping the intermediate sulfonium species is a very elegant method for the introduction of such a useful functional group.22 60 S.Caddick et al. TBTH AIBN N O 9 57% N Br SO2Tol TBTH AIBN N 31% N TolSO2 Br O Scheme 13 H Br N Ph2N 10 AIBN C6H6 TBTH 80 °C 1100 psi 5 h CO NHNPh2 O 11 69% Scheme 14 N N2 + SO2CH3 TTF CH3CN 48 h 41% N HN CH3 O SO2CH3 Scheme 15 4 Intermolecular reactions From a synthetic viewpoint intermolecular radical reactions have been less extensively used than intramolecular variants; however in the last few years there has been an increase in activity in this area. Of course intermolecular radical reactions are wellestablished in the polymer area. In attempts to develop methods for the synthesis of polymers containing multiple stereogenic centres Porter et al. have examined the e§ect of a range of chiral auxiliaries on simple radical telomerizations of acrylamides and 61 Reaction mechanisms Part (iii) Free-radical reactions I B O O B O O 71% OBu OBu TBTH ln•(initiator) Scheme 16 Br Br TBTH AIBN CN H H CN Br 59% H H CN Br H H CO2Me CN 30% TBTH AIBN CO2Me Scheme 17 conclude that the penultimate chiral centre controls the stereochemistry of the ultimate chiral centre.The nature of the e§ect is dependent on the auxiliary used; thus with oxazolidines the erythro product is formed but with sultams the threo isomers are formed.23 The versatility of the carbon–boron bond in organic synthesis makes methods for its introduction into organic molecules of particular interest. Batey et al. have described the generation of a-boryl radicals which undergo addition and substitution reactions as shown in Scheme 16. The report showed using competition experiments that such radicals are ambiphilic and yields of adducts are optimal when telomerization is avoided.24 The related use of a-bromo radicals has also been explored by Tanabe et al.who have shown that gem-dibromocyclopropanes undergo highly stereocontrolled sequential radical addition reactions. The first addition leads to a monobromocyclopropanone which of course could have considerable synthetic potential in non-radical reactions; however the authors show its ability to undergo further inter- or intramolecular reactions which proceed with inversion due to steric hindrance (Scheme 17).25 Perhaps one of the most innovative developments in this area comes from Dang and Roberts who have described a superb new homolytic aldol reaction.26 It has been shown that acyl radicals will undergo radical addition to alkenes although it is often ine¶cient due to poor hydrogen atom abstraction of the adduct radical from the donor.These workers showed that thiols can be used as catalysts to abstract the acyl 62 S. Caddick et al. R H O + OAc R O OAc RSH In 60 °C 55–80% CATALYTIC CYCLE R O • OAc R H O R O OAc RS• RSH R O OAc RSH • Scheme 18 hydrogen e¶ciently and thence to act as a hydrogen atom donor to the adduct radical as shown in the catalytic cycle illustrated in Scheme 18. The reactions proceed in moderate to good yields and there is now a clear and exciting opportunity for asymmetric catalysis using a chiral non-racemic hydrogen atom donor. Substitutions Peukert and Giese have reported an important radical-induced S N 1 substitution reaction.27 They showed that radicals 12 and 13 produce 15–18 via a common intermediate proposed to be 14 (Scheme 19).The understanding of such mechanistic pathways is essential for the design of therapies which involve DNA modification. Barton and Fontana have extended their work on radical based transformations of carboxylic acids to the synthesis of dialkyl selenides. As with the related transformations which lead to phosphonic acids and thiols the photolysis of pyridinethione oxycarbonyl (PTOC) esters with elemental selenium gave dialkyl selenides.28 The addition of aminyl radicals to allylstannanes provides a basis for a reductive allylation of sulfonyl azides by allylstannane as shown in Scheme 20.29 The generation of the aminyl radicals takes place by addition of the stannyl radical to the azido 63 Reaction mechanisms Part (iii) Free-radical reactions O Ph OR OPO(OEt)2 12 O Ph OR • • OPO(OEt)2 13 O Ph OR • + 14 O Ph OR OMe O Ph OR O Ph OR O Ph OR OMe OMe OMe 15 16 17 18 MeOH TBTH Scheme 19 SO2N3 + SnPh3 AlBN Benzene SO2NH 91% Scheme 20 R N2 O + SnBu3 i AlBN Benzene R O 60–80% ii KF–H2O Scheme 21 functional group.The reactions appear to work well using a range of substrates and this is an important addition to the growing use of azides in radical transformations. The addition of stannyl radicals to carbenoids has been described by the same group giving a-stannyl ketone radicals which can be quenched using a hydrogen atom donor or by addition (Scheme 21).30 Allyl sulfones have been shown by Chatgilialoglu et al. to be useful partners in simple radical allylation reactions.31 Quiclet-Sire and Zard have described a very simple and elegant radical allylation sequence which has been used to give a range of interesting products.The reactions proceed because the allyl aryl sulfone is used in excess (3–6 equiv.) and inevitably is the species which performs the allylation; the ArSO 2 radical does not undergo a-elimination and therefore undergoes addition to 64 S. Caddick et al. Scheme 22 CF3 Ph 20 49–61% Ph SO2CF3 (19) AlBN Scheme 23 N Boc SO2Tol N Boc SnBu3 60% TBTH AlBN Scheme 24 the alkly allyl sulfone which after a-elimination then propagates the chain. Clearly it is generally possible to intercept the intermediate radical with another addition process providing it is faster than the allylation (Scheme 22).32 Sulfones have also gained utility in a novel alkynylation procedure described by Gong and Fuchs.They reported alkynylation of activated and unactivated hydrocarbons using alkynyl trifluoromethyl sulfone 19 with appropriate initiation. The mechanistic details are still unclear and may involve either direct ipso substitution or carbene formation followed by rearrangement. Either case leads to the formation of the trifluoromethylsulfonyl radical which can undergo a-elimination and in a very elegant experiment these workers treated the sulfone 19 with an alkene and isolated addition product 20 in good yields (Scheme 23).33 Another example of ipso substitution but using an aryl sulfone is shown in Scheme 24. Thus treatment of the indoline-7-sulfone with TBTH gives the desired stannane presumably via radical ipso substitution.34 Minisci and co-workers have also recently described an ipso substitution reaction of vinyl and aryl chlorides.35 65 Reaction mechanisms Part (iii) Free-radical reactions O N Ph O O Lewis acid R–I TBTH Et3B O2 CH2Cl2 78 °C O N Ph O O R 0.5 MgI2 0.5 Ligand 86% Yield 79% ee O N N O Ligand = Scheme 25 Sn CH3 H 21 Stereoselectivity More reports on studies relating to asymmetric radical reactions have appeared and such is the interest that there is a new book on the subject.36 The use of substrate37 or induced-substrate (auxiliary) control38 as a mechanism for achieving stereocontrol has been described.An alternative to these strategies is the use of a reagent based approach which may employ Lewis acids and can pave the way for catalysis.39 The major benefit of using a reagent based strategy is really only founded if a catalytic variant can be developed.In order for catalysis to be viable there are several classical problems which need to be addressed; of particular importance is the dissociation of the catalyst –substrate to enable turnover. Sibi et al. have described the first example of a catalyst which promotes enantioselective radical additions to electron deficient alkenes as shown in Scheme 25. The choice of such a transformation is not only synthetically important but also provides a basis for turnover as the substrate and products di§er in their Lewis basicity. The yields were good and the enantioselectivities are very promising. It will be interesting to see how widely these principles can be used; perhaps such asymmetric radical reactions will become powerful tools in synthesis.40 A conceptually di§erent approach has been described by Curran41 and Roberts.42 In the Curran approach a C 2 -symmetric chiral non-racemic tin hydride 21 has been prepared and used in the reduction of 22 which proceeds to give modest but meaningful levels of enantioselectivity (Scheme 26).The Roberts procedure uses an optically active thiol thus avoiding the use of tin reagents. The group had previously demonstrated that hydrosilylation of prochiral alkenes can be improved by using a thiol catalyst to accelerate the hydrogen atom transfer step. In a beautifully conceived approach the feasibility of enantioselective 66 S. Caddick et al. Ph Ph O Br 12–70% Yield 11–40% ee In Ph Ph O H 22 21 Scheme 26 O O O O Ph3Si * Ph3SiH In R*SH 23 72% (50% ee) Scheme 27 hydrogen atom transfer in the enantioselective hydrosilylation of methylenelactone 23 was demonstrated.Although the levels of enantioselectivity are presently modest it should be possible to improve the method (Scheme 27). References 1 P. J. Parsons C. S. Penkett and A. J Shell Chem. Rev. 1996 96 195; C. H. Schiesser and L. M. Wild Tetrahedron 1996 52 13 265. 2 S. J. Garden D. V. Avila A. L. J. Beckwith V. W. Bowry K. U. Ingold and J. Lustyk J. Org. Chem. 1996 61 805; W.R. Dolbier Jr. X. X. Rong B. E. Smart and Z.-Y. Yang J. Org. Chem. 1996 61 4824; C. Tronche F. N. Martinez J. H. Horner M. Newcomb M. Senn and B. Giese Tetrahedron Lett. 1996 37 5845. 3 D.H.R. Barton Chem. Soc. Rev. 1996 237; F. Minsici F. Fontana S. Araneo F.Recupero and L. Zhao Synlett 1996 119; D. W. Snelgrove P. A. Macfaul K. U. Ingold and D. D. M. Wayner Tetrahedron Lett. 1996 37 823; M. Newcomb P. A. Simakov and S.-U. Park Tetrahedron Lett. 1996 37 827. 4 D.O. Jang Tetrahedron Lett. 1996 37 5367. 5 M. Ochiai T. Ito H. Takahashi A. Nakanishi M. Toyonari T. Sueda S. Goto and M. Shiro J. Am. Chem. Soc. 1996 118 7716. 6 P. Magnus J. Lacour P. A. Evans M. B. Roe and C. Hulme J. Am. Chem. Soc. 1996 118 3406. 7 T. Nagashima and D. P. Curran Synlett. 1996 330. 8 D.P. Curran and S. Hadida J. Am. Chem. Soc. 1996 118 2531. 9 D. Crich J.-T. Hwang and H. Liu Tetrahedron Lett. 1996 37 3105. 10 J.-P. Dulcere E. Dumez and R. Faure Synlett 1996 391. 11 D. S. Hays and G. C. Fu J. Org. Chem. 1996 61 4. 12 G. A. Russell and C. Li Synlett 1996 699.13 K. I. Booker-Milburn B. Cox and T. E. Mansley Chem. Commun. 1996 2577. 14 W.R. Bowman P. T. Stephenson and A. R. Young Tetrahedron 1996 35 11 445. 15 D. Damour M. Barreau F. Dhaleine G. Doerflinger M. Vuilhorgne and S. Mignani Synlett 1996 890. 16 Y.M. Tsai K. H. Tangard W. T. Jiaang Tetrahedron Lett. 1996 52 7767. 17 M.P. Sibi and J. Ji J. Am. Chem. Soc. 1996 118 3063. 18 H. Togo Y. Hoshina and M. Tokoyama Tetrahedron Lett. 1996 37 6129. 19 T. Muraki H. Togo and M. Tokoyama Tetrahedron Lett. 1996 37 2441. 20 S. Caddick K. Aboutayab K. Jenkins and R. I. West J. Chem. Soc. Perkin Trans. 1. 1996 675. 21 I. M. Brinza and A. G. Fallis J. Org. Chem. 1996 61 3580. 22 M. Kizil C. Lampard and J. A. Murphy Tetrahedron Lett. 1996 37 2511. 23 N. A. Porter R. L. Carter C. L. Mero M. G. Roepel and D.P. Curran Tetrahedron 1996 52 4181. 24 R. A. Batey B. Pedram K. Yong and G. Baquer Tetrahedron Lett. 1996 37 6847. 25 Y. Tanabe K. Wakimura and Y. Nishii Tetrahedron Lett. 1996 37 1837. 26 H.-S. Dang and B. P. Roberts Chem. Commun. 1996 2201. 27 S. Peukert and B. Giese Tetrahedron Lett. 1996 37 4635. 28 D. H. R. Barton and G. Fontana Tetrahedron 1996 52 11 163. 67 Reaction mechanisms Part (iii) Free-radical reactions 29 H.-S. Dang J. Chem. Soc. Perkin Trans. 1 1996 1493. 30 H.-S. Dang J. Chem. Soc. Perkin Trans. 1 1996 769. 31 C. Chatgilialoglu A. Alberti M. Ballestri D. Macciantelli and D. P. Curran Tetrahedron Lett. 1996 37 6391. 32 B. Quiclet-Sire and S. Z. Zard J. Am. Chem. Soc. 1996 118 1209. 33 J. Gong and P. L. Fuchs J. Am. Chem. Soc. 1996 118 4486. 34 K. Aboutayab S. Caddick K. Jenkins S. Joshi and S. Khan Tetrahedron 1996 52 11 329. 35 S. Araneo R. Arrigoni H.-R. Bjorsvik F. Fnatana L. Liguori F. Minisci and F. Recupero Tetrahedron Lett. 1996 37 6897. 36 Stereochemistry of Radical Reactions. D. P. Curran N. A. Porter and B. Giese VCH Weinheim 1996. 37 M. Roth W. Damm and B. Giese Tetrahedron Lett. 1996 37 351. 38 R. A. Ewin K. Jones and C. G. Newton J. Chem. Soc. Perkin Trans. 1. 1996 1107. 39 M.P. Sibi and J. Ji J. Org. Chem. 1996 61 6090. 40 M.P. Sibi J. Ji J. H. Wu S. Gurtlet and N. A. Porter J. Am. Chem. Soc. 1996 118 9200. 41 D. Nanni and D. P. Curran Tetrahedron Asymm. 1996 7 2417. 42 M.B. Haque and B. P. Roberts Tetrahedron Lett. 1996 37 9123. 68 S. Caddick et al.