Stereochemistry of the aldol reaction between 4-t-butyldimethylsilyloxy-5-(3-t-butyldimethylsilyloxyoct-1-enyl)cyclopent-2-enylethanal and cyclopentanone enolates; a key step in the total synthesis of prostacyclin

摘要

J. CHEM. SOC. PERKM TRANS. I 1983 Stereochemistry of the Aldol Reaction between 4-t-Butyldimethylsilyloxy-5-(3-t- butyld imethylsilyloxyoct-1 -enyl) -cyclopent-2-enylethanal and Cyclopentanone Enolates IA Key Step in the Total Synthesis of Prostacyclin Anthony D. Baxter, Stanley M. Roberts, Basil J. Wakefield, and Geoffrey T. Woolley The Ramage Laboratories, Department of Chemistry and Applied Chemistry, University of Salford, Salford M5 4WT Roger F. Newton * Chemica ?search Department, Glaxo Group Research, Ware, Herts. SG12 ODJ Reactions of lithium, boron, zirconium, and tin enolates derived from cyclopentanone with 4-t- butyl- dimet hylsilyloxy-5- (3-t- butyldimethylsilyloxyoct-1 -enyl)cyclopent-2-enyl ethanal (6) gave mainly the 5R, 6R, and 5S,6s (prostaglandin numbering), threo-adducts, regardless of the nature of the enolate; the highest yield of threo-products (83 of a total yield of 93) was obtained with lithium enolate in pentane at -78 "C.In contrast, the ratio of 5R,6R to 5S,6S products was significantly affected by the nature of the enolate, ranging from ca.1 :1.3with a boron enolate to ca. 1 :3 with a tin enolate, and ca. 1 :4 with a lithium enolate. 5-(3-t-Butyldimethylsilyloxyoct-1-eny1)-4hydroxycyclopent-2-enyllethanal (1) l*'* is a valuable intermediate for the synthesis of prostanoids, since it has the required stereo- chemistry and functionality about the cyclopentane ring, and the aldehyde function is available for the elaboration of the upper side-chain. For example, it has been converted into 6P-prostaglandin II2 and 9-deoxa-9,lO-didehydroprostaglan-din D2.4For the present work we have used the route to the aldehyde (1) described in ref.1;it has proved to be capable of furnishing the compound in gram quantities. The least reliable stage was the reaction of the tricyclic lactone (2)with the side-chain cuprate reagent in the presence of hexa-methylphosphorous triamide (HMPT), which occasionally gave the amide (3) as well as the required acid (4). However, the amide (3)could be converted directly into the aldehyde (5) by reduction with di-isobutylaluminium hydride. Our route for the synthesis of prostacyclin methyl ester from the silylated derivative (6) of the hydroxyaldehyde (1) is outlined in Scheme 1.5 In this synthesis, the stereochemistry of the aldol reaction is crucial. Most importantly, as a con-sequence of the trans-elimination in the last step leading to prostacyclin methyl ester, only the threo-isomers (7) and (8) can give the required E-geometry at the A5 double bond; the erythro-isomers (9) and (10) will give the 2-isomer.Further- more, the ease of halocyclisation is greatly dependent on the stereochemistry of the aldol products. In a similar case, reported earlier, the 6(S)-isomer (12)cyclised much more slowly than its 6(R)-epimer.' Our initial experiments were made with the lithium enolate of cyclopentanone, and gave four aldol products, as expected, together with small amounts of the self-condensation product of cyclopentanone. T.1.c.on silica (5 ethyl acetate: 95 chloroform) showed the aldol products grouped as two pairs (amp; 0.76 and 0.72;and 0.42 and 0.32).On the evidence of n.m.r. spectroscopy, the first pair were judged to be the threo- isomers and the second pair the erythro-isomers. Thus, the signals for the CHOH proton occurred at 6 3.84 and 3.76, for the threo-isomers and S 4.27and 4.20,for the erythro- t The analogous tetrahydropyranyl-protected aldehyde has also been prepared by a multi-step synthesis from 7-benzyloxymethyl-norborn-5-en-2-0ne.~ CH0 '.J But Me2Si0 C 5Hll (1) R=H ( 3 1 R = CONMe2 ( 6 1 R = ButMe2Si (4) R =C02H (5) R =CHO isomers. The chemical shifts for this proton in the known lo threo- and erythro-adducts of cyclopentanone enolate with benzaldehyde occurred at 6 4.75and 5.25 respectively. These assignments were later confirmed by the conversion of both the threo-isomers (7) and (8) into prostacyclin methyl In our initial experiments we discovered that only one of the threo-isomers underwent iodocyclisation, to the iodo-compound (11; X = I).5 By analogy with the earlier case, we suggest that it was compound (7) (stereochemistry 5R,6R using prostaglandin numbering and designated f.3 for convenience), which cyclised.In this paper we describe and discuss the stereochemistry of the reaction of the aldehyde (6) with a variety of cyclopent-anone enolates. Our results are summarised in the Table. These results display two striking features : the comparative lack of sensitivity of the threolerythro ratio to the nature of the enolate, but the large variation in the ratio of components within each diastereoisomeric pair.In various studies of the reactions of enolates with simple aldehydes, very high degrees of threo- or erythro-selectivi ty have been observed in certain cases.' Various hypotheses have been advanced to account for this stereo~electivity,'~** but it seems clear that energy differences between the transi- tion states must be small, so that while they are significant in simple cases, they may be outweighed by other steric con- straints for more complex systems. In our case, the reaction is apparently inherently threo-selective (compare our entry 1, J.CHEM. SOC. PERKIN TRANS. I 1983 + -9(6) ?o i,ii (11) X = Br or I , then H iii, iv, v ,viI v iic Scheme. Reagents: i, X+; ii, BuJSnH; iii, m-CIC6H4C03H;iv, K,C03-MeOH; v, MeSO2C1-Et3N; vi, MeC02H-THF-H20 ; vii, DBU Table, with the corresponding reaction with benzaldehyde, which gives little or no stereoselection9). This selectivity could be explained in terms of a Zimmerman-Heathcock transition state for the formation of the threo-aldol such as (13) in which all the bulky groups lie at equatorial positions to the pseudo-chair ring. However, in the absence of a detailed knowledge of the preferred conformations of the aldehyde (6) or of the mechanism of the reaction we prefer simply to accept the threo-selectivity as an empirical fact.Some vari- ation from the 'natural ' ratio is observed with enolates that have been reported to be erythro-selective (Table, entries 4, 7, 8), notably the zirconium and tin enolates (cf. refs. 10, 6Si MeZBut (12) OBR, ii (yC80' 11). As a variation on the simple boron enolate * (entry 4) we tried the 'ate complexes' (14), in which the negative charge is presumably largely localised on boron. These gave somewhat less erythro-products than the dialkylboron enolate but such reagents are much more easily prepared and may be useful in simpler cases.* The large variation in the ratios of the two threo diastereo-isomers and the ratios of the two erythro diastereoisomers is perhaps even more surprising. Since the formyl group in (6) is not bonded to a chiral centre, generalisations such as Cram's rule are not strictly applicable, and little stereo- selectivity would be predicted.For example, in recent experi- ments on the addition of Grignard reagents to S-(+)-3-methylpentanal, the maximum ratio of the products was 1 : 1.3.12 However, although a similar ratio was observed between the two threo-isomers in the reaction of the aldehyde (6) with the dibutylboron enolate (Table, entry 4), with the lithium enolate in ether at -120 "C it was as high as 1 :4.1 (entry 3). Once again, the variation in this ratio must be due to subtle steric factors. For the purpose of the synthesis of prostacyclin, since the problem of the halocyclisation of the epimer (8) has now been overcome,6 the optimum conditions are those of the Table, entry 2, i.e.lithium enolate in pentane at -78 "C, which give a 93 yield of aldol products, 83 of which have the required threo-configuration. Work is now in progress on epimeris- ation of the erythro-isomers, to enable all the aldol products to be utilised. Experimental Unless otherwise stated, 'H n.m.r. data were recorded for solutions in deuteriochloroform with tetramethylsilane as external standard; i.r. spectra were recorded for neat films; mass spectra were determined after ionisation by electron impact at 70eV (e.i.m.s.) or chemical ionisation using ammo- nia (c.i.m.s.); t.1.c. was carried out with Camlab Polygram pre-coated silica gel plates ; column chromatography was carried out using medium pressure over silica (Merck Kiesel- gel, Art.7 736; ca. 20 x weight of substrate); light petroleum refers to the fraction, b.p. 40-60 "C. All reactions involving * Since the completion of our work, it has been noted that in the presence of an excess of trialkylboron, the reaction between benzaldehyde and cyclopentanone enolate becomes highly threo-selective: Y. Yamamoto, H. Yatagai, and K. Maruyama, Tetra-hedron Lett., 1982, 23, 2387. J. CHEM. soc. PERKIN TRANS. I 1983 1811 Table,Reaction of the aldehyde (6) with enolates of cyclopentanone Metal counterion of enolate Reaction temp. (ldquo;C) Solvent Total yield() i; : Product hratio () (8) : (9) : (10; Total three"/, Li + -78 EtzO 80 22 53 6 19 75 Li+ -78 Pentane 93 26 57 4 13 83 Li + -120 EtzO 92 15 62 3 19 77 BBU~ -78 EtzO 70 28 36 17 19 64 BEt3 -Li+ -78 Etz0-THF 87 27 40 13 20 67 BBuSSLi+ -78 Etz0-THF 72 23 49 6 22 72 Zr(cp)ClSnPh3 -78 -78 THF THF 61 78 24 14 38 48 9 10 28 28 62 62 organometallic compounds were carried out under an atmosphere of dry, oxygen-free nitrogen, using dried appa- ratus and solvents. 5-(3-t-Buty ldimethy lsily loxy oct- 1-enyl)-4- hydroxypen t-2-enyllethnal (l).-(a) The aldehyde (1) was prepared essenti- ally as described in ref.lb. On one occasion the reaction of 2-oxabicyclo3.3.0.04~6oct-7-en-3-onewith the cuprate re-agent (1.5 mol equiv.) derived from 1-iodo-3-t-butyldimethyl-silyloxyoct-l-ene in the presence of HMPT (3.5 equiv.) gave, besides the acid (4) (1273, NN-dimethyl-4-exo-(3-t-butyl-dimethylsilyloxyoct-1 -enyl)bicyclo 3.1 .Ohex-2-ene-6-carboxamide (3) (42); vmX.1 640 crn-rsquo;; 6, 0.0 (6 H, S, SiMe2),0.9 (12 H, m, 4CH3), 1.25 (8 H, my4CH2), 1.7-2.2 (2 H, m, 1-, 5-H), 2.4 (1 H, my 6-H), 2.95 (3 H, s, NMe), 3.1 (3 H, s, NMe), 3.65 (1 H, my4-H), 4.0 (1 H, m; 3-Hrsquo;), and 5.3-5.9 (4H, my2-, 3-, 1rsquo;-,and 2rsquo;-H) Found :(e.i.m.s) M+, 391.2903. C23H41N02Sirequires M, 391.29041. (b)A solution of di-isobutylaluminium hydride (M in hexane, 0.52 ml) was added dropwise during 30 min to a solution of the amide (3) (0.20 g, 0.55 mmol) described in (a) in light petroleum (7 ml) at -78 ldquo;C. The mixture was stirred at -78 ldquo;C for 1 h.Methanol (2 ml), water (20 ml), and 2~-hydrochloric acid (6 ml) were added. The organic layer was separated, and the aqueous phase was extracted with ether (3 x 50 ml). The combined organic layers were neutralised with 2u-sodium hydroxide and then re-extracted with ether (3 x 50 ml) Conventional work-up of the com- bined organic layers, followed by chromatography on silica, gave the bicyclic aldehyde (5) (116 mg,ardquo;/,),identical (i.r., n.m.r.) with the material previously described.rsquo;* 2-4-t-Butyldimethylsilyloxy-5-( 3-t-butyldimethylsilyloxyoct-1-enyl)cyclopent-2-enylethanal(6).-To a solution of the aldehyde (1) (2.8 g) in DMF (100 ml) was added t-butyl- chlorodimethylsilane (2.34 g) and imidazole (2.34 g).The homogeneous mixture was set aside at room temperature for 15 h. Water (100 ml) was added, and the solution was ex- tracted with ether (4 x 100 ml). Conventional work-up and chromatography (silica, 2 ethyl acetate in light petroleum) gave the protected aldehyde (6) (2.94 g, 800/,),vmx. 1 720 cm-l; 6,O.O (12 H, s, 2 SiMez),0.7-1.0 (21 H, m, 7 Me), 1.05-1.6 (8 H, m, 4 x CHt), 2.0-2.85 (4 H, m, 2-, 1rsquo;-, and 5rsquo;-H), 4.05 (1 H, m, 3rdquo;rsquo;-H), 4.50 (1 H, m, 4rsquo;-H), 5.35-5.6 (2 H, m, 1rdquo;-and 2rdquo;-H), 5.6-5.85 (2 H, s, 2rsquo;-and 3rsquo;-H), and 9.8 (1 H, t, J 1.5 Hz, CHO). of the self-condensation product of cyclopentanone was observed but was not recovered quantitatively. (a)To a stirred solution of di-isopropylamine (39.5 mg, 0.39 mmol) in diethyl ether at 0 ldquo;C was added dropwise n-butyl-lithium (1.44~in hexane; 0.27 ml), the temperature being kept below 3 ldquo;C.The solution was stirred at 0 ldquo;Cfor 10 min, then cooled to -78 ldquo;C.Cyclopentanone (32 mg, 0.39 mmol) was added dropwise during 2 min, and the mixture was stirred at -78 ldquo;C for a further 10 min. A solution of the aldehyde (6) (185 mg, 0.376 mmol) in dry ether (3 ml) was added rapidly, and the mixture was stirred at -78 ldquo;C for 10 min. Saturated aqueous ammonium chloride (20 ml) was added, and the mixture was allowed to warm to room temper- ature. The mixture was extracted with diethyl ether (4 x 25 ml), and the combined extracts were washed with water (2 x 20 ml), dried (MgS04), filtered, and evaporated, to leave a yellow gum (0.21 g).The gum was subjected to repeated short-path column chromatography (silica, eluant 1 ethyl-acetate, 19 dichloromethane, 80 light petroleum, to afford the following. (i) threo-amp;l-(2-Oxocyclopentyl)-2-4-endo-t-butyldimethylsilyloxy-5-exo-(3-t-butyldimethyl-silyloxyoct-l-enyl)cyclopent-2-enylethanol (7) (30.4 mg), t .l.c. RF 0.76 (5 ethyl acetate5 dichloromethane); vmx, 3 450 and 1 720 cm-rsquo;; 0.0 (12 H, s, 2 x SiMe2),0.7-1.0 (21 H, my 7 x Me), 1.0-2.5 (18 H), m, 2.62 (1 H, m, 1rdquo;rsquo;-H), 3.84 (1 H, tm, largest coupling 7 Hz, 1-H), 4.08 (1 H, m, 3rdquo;-H), 4.55 (1 H, t, J 7 Hz, 4rsquo;-H), 5.54 (2 H, my 1rdquo;-or 2rdquo;-H), 5.70 (1 H, dm, largest coupling 5 Hz, 2rsquo;- or 3rsquo;-H), 5.95 (1 H, dm, largest coupling 5 Hz, 2rsquo;-H or 3rsquo;-H) Found (e.i.m.s.) M+ -141, 423.2750.C32amp;04Si2 -CsHsO -C4H9 requires mlz 423.27481. (ii) The threo-a-isomer (8) (74.5 mg); t.1.c. RF0.72; vmX. 3 450, and 1 720 ern-'; 6, 0.0 (12 H, s), 0.7-1.0 (21 H, m), 1.0-2.4 (18 H, m), 2.55 (1 H, mylrdquo;rsquo;-H), 3.76 (1 H, tm, largest coupling 9 Hz, 1-H), 4.06 (1 euro;3, q, 3rdquo;-H), 4.15 (1 H, m, OH), 4.50 (1 H, dd, J 10 Hz and 4.5 Hz, 4rsquo;-H), 5.56 (2 H, m, 1rdquo;-and 2rsquo;-H), 5.65 (1 H, my2rsquo;-or 3rsquo;-H), and 5.88 (1 H, d, J 5 Hz, 2rsquo;-or 3lsquo;-H) Found (e.i.m.s.) M+ -141, 423.2750. C32Hb004Si2-C5H80-C4H9requires m/z 423.27481. (iii) The erythro-P-isomer (9) (8.5 mg); t.1.c. RF 0.42; vmx. 3 450 and 1720 cm-rsquo;; 6, 0.034, 0.040, 0.053, 0.056, 0.062, 0.071, 0.073 (12 H total, 2 SiMe2),0.888, 0.890, 0.895, 0.898 (18 H total; 2 x But), 1.0-2.4 (21 H, m), 2.55 (1 H, m, lrdquo;rsquo;-H), 4.07 (1 H, m, 3rdquo;-H), 4.27 (1 H, m, 1-H), 4.53 (1 H, m, 4lsquo;-H), 5.50 (1 H, m, 1rdquo;-or 2rdquo;-H), 5.53 (1 H, m, 1rdquo;-or 2-H), 5.69 (1 H, dq, J 6 Hz and 2 Hz, 2rsquo;- or 3rsquo;-H), 5.90 (1 H, my 2rsquo;-or 3rsquo;-H); mass spectrum m/z 564 (M+),507 (M -C4H9), 423 (M -C4H9-C5HsO)Found: (e.i.m.s.): M+,564.4053.C32H6004Si2Reactions of 4-t-Butyldimethy~sily~oxy-5-(3-t-butyldi~ethy~-requires M, 564.40271. with Enolates of (iv) The erythro-a-isomer (10) (26.3 mg); t.l.c., RF 0.32silyloxyoct-1-enyl)cyclopent-2-eny~ethanal(6) Cyc1opentanone.-The yields of the products of the reactions vmx. 3 450 and 1 720 cm-rsquo;; 6, 0.031, 0.054, 0.062, 0.067, described below are recorded in the Table. The yields take 0.073 (12 H total, 2 Mez), 0.883, 0.892, 0.895 (18 H, total, into account recovered starting material.In each case a little 2 But), 1.00-2.35 (21 H, m), 2.42 (1 H, m, 1rdquo;rsquo;-H), 4.07 (1 H, 1812 9,J 5 Hz,3"-H), 4.2 (1 H, dq, J9.5, 3.5 Hz, 1-H), 4.50 (1 H, dm, largest coupling 11 Hz, 4'-H), 5.6-5.45 (2 H, m, 1"-and 2'-H), 5.70 (1 H, tt, J 6.5, 2 Hz, 2'- or 3'-H), 5.85 (1 H, tm, largest coupling 6.5 Hz, 2'-or 3'-H); mass spectrum as for the erythro-p-isomer Found : (e.i.m.s.); M+, 564.4037. C32H6004Si2requires M, 564.40271. Starting aldehyde (6) (36 mg) was also recovered. The dimer of cyclopentanone was detected by t.l.c., but not isolated. (b) Reactions with the lithium enolate in pentane at -78 "C and in ether at -120 "Cwere carried out similarly. (c) To a stirred solution of ethyl di-isopropylamine (0.163 g, 1.26 mmol) in pentane (2 ml) at -78 "C was added with care a solution of di-n-butylboron trifluoromethanesul-phonate (0.32 g, 1.17 mmol) in pentane (1 ml).Cyclopentanone (91 mg, 1.08 mmol) was added dropwise, and the resulting suspension was stirred at -78 "C for 30 min, allowed to warm to -20 "C during 30 min, and then re-cooled to -78 "C. A solution of the aldehyde (6) (433 mg, 0.902 mmol) in pen- tane (1 ml) was added dropwise, and the mixture was stirred at -78 "C for 45 min; then as it warmed to 0 "C during 95 min. The reaction mixture was poured into water buffered to pH 7 (phosphate), and worked up via extraction and chro- matography as described above. (d) To a solution of the lithium enolate of cyclopentanone (0.625 mmol) prepared as described in (a), was added at -78 "C triethylboron (M in THF; 0.625 ml).The solution was stirred at -78 "C for 30 min, and a solution of the alde- hyde (6) (246 mg, 0.512 mmol) in ether (2.0 ml) was added dropwise. After a further 20 rnin at -78 "C, aqueous ammo- nium chloride was added, and the products were isolated as described in (a). A similar experiment was carried out with tri-s-butylboron in place of triethylboron. (e) To a stirred solution of the lithium enolate of cyclo-pentanone (51 mg, 0.615 mmol), prepared as described in (a), but in THF instead of diethyl ether, was added dropwise at -78 "C a solution of dicyclopentadienylzirconium di-chloride (181 mg, 0.615 mmol) in THF (3.0 ml).The resulting suspension was stirred for 45 rnin as it warmed to -50 "C, and was then re-cooled to -78 "C.A solution of the aldehyde (242 mg, 0.504 mmol) in THF (2.0 ml) was added dropwise, and the mixture was stirred at -78 "C for 30 min and then for a further 15 min as it warmed to -50 "C.Methanol (6 ml) and water (6 ml) were added, and the mixture was extracted with ether (4 x 30 ml). The combined extracts were washed with water (2 x 25 ml), and the combined aqueous layers were further extracted with ether (30 ml). The products were isolated from the combined organic layers as described in (a). (f)To a stirred solution of the lithium enolate of cyclo-J. CHEM. SOC. PERKIN TRANS. I 1983 pentanone (46 mg, 0.555 mmol), prepared as described in (a), but in THF, was added dropwise at -78 "C a solution of bromotriphenyltin (0.239 g, 0.555 mmol) in THF (2 ml) and the resulting suspension was stirred at -78 "C for 30 min.A solution of the aldehyde (6) (219 mg, 0.456 mmol) in THF was added in one portion, to give a clear solution, which was stirred at -78 "C for 30 min. Methanol (5 ml) and water (7 ml) were added, and the products were isolated as des- cribed in (e) and (a). Acknowledgements We thank the S.E.R.C. for CASE awards (to A. D. B., G. T. W.); Dr. B. E. Mann (University of Sheffield) and Dr. D. Moorcroft (University of Manchester) for high-resolution n.m.r. spectra; Dr. M. D. Lynch for assistance with large-scale syntheses of compound (6); and Dr. F. Scheinmann for useful discussions.References 1 (a)S. M. Ali, M. A. W. Finch, S. M. Roberts, and R. F. Newton, J. Chem. SOC., Chem. Commun., 1979, 679; ibid., 1980, 74; (6) S. M. Ali, C. B. Chapleo, M. A. W. Finch, S.M. Roberts, G. T. Woolley, R. J. Cave, and R. F. Newton, J. Chem. Soc., Perkin Trans. I, 1980, 2093. 2 M. A. W. Finch, S. M. Roberts, and R. F. Newton, J. Chem. Soc., Perkin Trans. I, 1981, 1312. 3 C. Gandolfi and G. Doria, Farmaco, Ed. Sci., 1974, 29, 405. 4 M. A. W. Finch, S. M. Roberts, G. T. Woolley, and R.F. New-ton, J. Chem. Soc., Perkin Trans. I, 1981, 1725. 5 R. F. Newton, S. M. Roberts, B. J. Wakefield, and G. T. Woolley, J. Chem. SOC., Chem. Commun., 1981,922. 6 A. D. Baxter, F. Binns, M. D. Lynch, R. F. Newton, S. M. Roberts, P. Sadler, F. Scheinmann, B. J. Wakefield, and G. T. Woolley, International Conference on Prostaglandins, Florence, 1982; Abstracts p. 131. 7 (a)C. H. Heathcock, in ' Comprehensive Carbanion Chemistry,' ed. T. Durst and E. Buncel, Elsevier, Amsterdam, 1982, vol. 11; (6) C. H. Heathcock, Science, 1981, 214, 395. 8 Y. Yamamoto and K. Maruyama, J. Am. Chem. SOC.,1982,104, 2323. 9 D. A. Evans and L. R. McGee, Tetrahedron Lett., 1980, 3975; Y.Yamamoto and K. Maruyama, Tetrahedron Lett., 1980,4607. 10 Y. Yamamoto, H. Yatagai, and K. Maruyama, J. Chem. SOC., Chem. Commun., 1981, 162. 11 D. A. Evans, E. Vogel, and J. V. Nelson, J.Am. Chem. SOC.,1979, 101, 6120; S. Masamune, S. Mori, D. Van Horn, and D. W. Brooks, Tetrahedron Lett., 1978, 1665; but see S. Shenvi and J. K. Stille, Tetrahedron Lett., 1982, 23, 624. 12 I. G. Vasi and K. R.Desai, J. Inst. Chem., Calcutta, 1982,54, 19. Received 18th January 1983; Paper 3/074