摘要:
J. CHEM. SOC. PERKIN TRANS. 2 1992 857 Radical-transfer Catalysis versus Lewis Acid Catalysis by the Copper(1) Chloride/2,2'- Bipyridine Complex: an lllust rat ion of the Synthetic Sign if icance of Ca pt odat ive Rad ica I Sta bi Iizat ion Jan H. Udding, C. (Kees) J. M. Tuijp, Henk Hiemstra" and W. Nico Speckamp" Department of Organic Chemistry, University of Amsterdam, Nieuwe Achtergracht 129, 1018 Amsterdam, The Netherlands The mechanism of the copper( I) chloride/2,2'- biyridine catalysed n-cyclization of N-(ch1oromethyl)alk-3-enylcarbamates changes from a cationic process (leading to piperidines) into a radical-transfer process (leading to pyrrolidines) upon introduction of an ester substituent at the reactive carbon atom, owing to the captodative effect.Recently, we published a novel copper(r) chloride/2,2'-bipy- bipyridine in lY2-dichloroethane at reflux for 18h gave the 5-ex0 ridine catalysed radical-transfer cyclization process of a-chloro-cyclization product 2 in 78% yield, along with a trace amount of glycine derivatives [e.g. 1 (Table 1, entry l)] to substituted the 6-end0 product 3 (2%). Ring closure also took place without pyrrolidine-2-carboxylic esters (proline analogues, e.g. 2).'v2 In catalyst (entry 4), but then produced the same 6-end0 product 3 this communication we wish to report that upon replacing the as the sole cyclization product in 33% yield. The piperidine 3 ester substituent at the reactive carbon atom by hydrogen, the must be the result of a cationic ring closure via iminium ion A mechanism of this ring closure becomes a cationic process, (Scheme 1) as the intermediate.This conclusion is based on leading to piperidines as the main products. We argue that this previous work on the SnC1,-mediated cyclization of 1(entry 3), dual catalytic behaviour of the copper complex is a consequence which gives 3 as the sole product in 80%yield.4 On the other of the presence or absence of captodative radical ~tabilization.~ hand, a genuine radical cyclization of 1 (X = Cl) should only The experimental results are collected in Table 1. Treatment lead to the 5-exo product, as can be inferred from the Bu,SnH- of the N-hex-3-enylcarbamate 1(entry 1) with 0.3 equiv. of the mediated cyclization of 1 (entry 2), which produces only the in-situ formed 1:l complex of copper(1) chloride and 2,2'-proline derivative 4 in 93% yield.' Thus, in the presence of the Table 1 Cyclizationof N-(a-chloroalky1)carbamates to pyrrolidines and piperidines Cyclization precursor Entry Structure X Reaction conditions a Products,% yield (cis/trunscratio) 1 c1 CNbPY )C1 cI 78 (20:80)' --SPh Bu,SnH (22' 93 (35 :65) 3f OAc SnC1, d02Me -4 blank C02Me -GMe rC02Me CO,Me C02Me C02Me 1 2 3 4 ?I 82 (46:54)5 Cl CU(bPY)Cl CI --6 6 SPh Bu,SnH 30 -7' 8 F: OMe SnCl, 95 (5 :95) --C02Me c1 blank -66 (87: 13) I I Ik02Me C02Me C02Me 5 6 7 8 H 9 c1 Cu(bPY)Cl N-C0,Me 30 32 Cl SnCl, --I0 KJ'co2Me11 Cl blank pf (I 37 C02Me 32 9 10 11 Reaction conditions: Cu(bpy)Cl = CuCl(0.3 equiv.), 2,2'-bipyridine (0.3 equiv.), 1,Zdichloroethane (0.3mol dmW3), reflux, 18 h; Bu,SnH = Bu,SnH (1.4 equiv.), AIBN (cat.), toluene (0.04mol dm-3), 8&90 "C, see ref.5; SnCI, = SnCI, (2 equiv.), CH,Cl,, -78 "C-room temp., see ref. 4; blank = 1,Zdichloroethane(0.3 mol drn-,), reflux, 18 h. Isolated yields. All new products were appropriately characterized by their IR, NMR and mass spectra. 'cis/truns ratio pertaining to ring substitution pattern. Both the cis and trans isomer consist of a ca. 1 : 1 mixture of chlorine epimers. See ref. 5. f See ref. 4. 858 C02Me C0,Me R, c---+" " R.~A~-iior iii R.~A~I I 60,Me 6O2Me C0,Me A 1 B U i or ii iii -H R;yAH --5JnX IIC02Me C02Me C02Me C 5of9 D Scheme 1 Reagents: i, SnCl,; ii, Cu(bpy)Cl; iii, Bu,SnH copper catalyst, 1 (X = C1) cyclizes almost exclusively via a radical mechanism.We presume that the radical route is preferred in the presence of copper(I), owing to the formation of the relatively stable captodative glycine radical B (Scheme 1).6 To test this hypothesis, we studied the ring closure of carbamate 5, which lacks the ester function to stabilize the radical. Chloride 5 was readily prepared from the corresponding hydroxymethyl analogue via reaction with PCl, in CCl,. Treatment of 5 with the copper catalyst (entry 5) under the same conditions as for 1 yielded the 6-endo cyclization product 7 in 82% yield as the sole product with 6 being undetectable in the reaction mixture. Without catalyst, cyclization also took place (entry 8) to give the same product 7in 66%yield.Again, the six- membered ring 7must have been formed in a cationic process via C, because the SnC1,-mediated cyclization of 5 (X = OMe) is known to give 7in quantitative yield., The radical reaction of 5 (X = SPh) with Bu,SnH gave only the product of 5-exo cyclization 8 (30%, and 46% of starting material). Thus, in the presence of the copper catalyst, chloride 5 cyclizes in a cationic mechanism via C.The remarkably different stereoselectivities in the formation of the six-membered ring 7 via the three different methods are not readily explained, but do indicate that the copper catalyst functions as a Lewis acid. It is furthermore noteworthy that the radical cyclization of 5 (X = C1 or SPh), which lacks the ester function, is a very slow process.A similar result has been reported for the corresponding wthiosulfon- amides by Padwa et af.* On the other hand, Bachi et aL9 did not observe salient differences between similar tin hydride mediated radical cyclizations, with or without an ester substituent. The Lewis acidic activity of the copper catalyst was confirmed in the next series of reactions. Chloride 9 containing a cyclopentenyl function was prepared in the same way as 5 from methyl cyclopent-2-enylmethylcarbamate. The latter compound arose from the BF,-Et,O-mediated coupling of cyclopent-2- enyltrimethylsilane lo with methyl acetoxymethylcarbamate.4*s When chloride 9 was treated with the copper catalyst, com- pounds 10 (30%) and 11 (32%) were isolated.The bridged bicyclic system 11 is the product of 6-endo cyclization, which is indicative of an ionic mechanism. Without catalyst, 9, sur-prisingly, did not give any trace of cyclization products (entry 11).When treated with SnCl,, chloride 9 cyclized to a mixture of 10 (37%) and 11 (32%). As the regioselectivities of the copper- catalysed cyclization and the SnC1,-mediated cyclization are virtually the same, we presume that the cuprous chloride/2.2'- bipyridine complex now functions as a Lewis acid catalyst. J. CHEM. SOC. PERKIN TRANS. 2 1992 Without copper, the ionic cyclization does not take place, probably because of the difficulty in forming the somewhat strained bicyclic system.In conclusion, the cuprous chloride/2,2'-bipyridine complex may catalyse two different reaction types, namely radical and cationic processes. Depending on the relative ease of radical and cation generation, respectively, the copper complex may act as a radical transfer catalyst or as a Lewis acid catalyst. The a-chloroglycine derivative 1 leads to a relatively stable capto-dative radical B,, whereas the cation is destabilized by the ester function. Therefore, radical reactions prevail with the copper complex. On the other hand, N-chloromethylcarbamates 5 or 9 would lead to radicals D, lacking special stabilization, so that the copper catalyst in this case functions as a Lewis acid catalyst, producing cations C.The copper catalyst thus uniquely illustrates the synthetic relevance of the captodative effect.l1 Our present investigations concentrate on further determining the scope and applications of the copper catalysis as well as the influence of ligand structure. Acknowledgements This work was supported by the Netherlands Foundation for Chemical Research (SON) with financial aid from the Nether- lands Organization for Advancement of Pure Research (NWO). References 1 J. H. Udding, H. Hiemstra, M. N. A, van Zanden and W. N. Speckamp, Tetrahedron Lett., 1991,32,3123. 2 For copper-catalysed atom-transfer cyclization, see also: D. Bellus, Pure Appl. Chem., 1985,57, 1827; H. Nagashima, N. Ozaki, K. Seki, M. Ishii and K. Itoh, J. Urg. Chem., 1989,54,4497; D. P.Curran, in Comprehensive Organic Synthesis, eds. B. M. Trost and I. Fleming, Pergamon Press, Oxford, 1991, vol. 4, p. 7 15. 3 See e.g. H. G. Viehe, Z. Janousek, R. Merknyi and L. Stella, Acc. Chem. Res., 1985,12,148; D. L. Pasto, J.Am. Chem. Soc., 1988,110, 8164; F. G. Bordwell and T.-Y. Lynch, J.Am. Chem. Soc., 1989,111, 7558; 0. Benson, Jr., S. H. Demirdji, R. C. Haltiwanger and T. H. Koch, J.Am. Chem. Soc., 1991,113,8879. 4 P. M. Esch, I. M. Boska, H. Hiemstra, R. F. de Boer and W. N. Speckamp, Tetrahedron, 1991,47,4039. 5 P. M. Esch, H. Hiemstra and W. N. Speckamp, Tetrahedron Lett., 1990, 31, 759; P. M. Escb, H. Hiemstra, R. F. de Boer and W. N. Speckamp, Tetrahedron, in the press. 6 C. J. Easton, C. A. Hutton, G. Rositano and E. W. Tan, J. Urg. Chem., 1991,56, 5614, and references cited. 7 P. M. Esch, R. F. de Boer, H. Hiemstra, I. M. Boska and W. N. Speckamp, Tetrahedron, 1991,47,4063. 8 A. Padwa, H. Nimmesgern and G. S. K. Wong, J. Org. Chem., 1985, 50, 5620. 9 M. D. Bachi, F. Frolow and C. Hoornaert, J. Urg. Chem., 1983,48, 1841. 10 J. M. Reuter, A. Sinha and R. G. Salomon, J. Urg. Chem., 1978,43, 2438. 11 For a comparable change of mechanism in the reaction of u-bromoglycine derivatives with unsaturated stannanes, see D. P. G. Hamon, R. A. Massy-Westropp and P. Razzino, J. Chem. SOC., Chem. Commun., 1991,722. Paper 2/01630C Received 27th March 1992 Accepted 27th March 1992
ISSN:1472-779X
DOI:10.1039/P29920000857
出版商:RSC
年代:1992
数据来源: RSC