Synthesis from thebaine of 10-oxothebaine, potentially a precursor for κ-selective opiates

摘要

J. CHEM. soc. PERKIN TRANS. 1 1994 Synthesis from Thebaine of 10-Oxothebaine, Potentially a Precursor for K-Selective Opiates Calum F. Henderson, Gordon W. Kirby" and (in part) John Edmiston Department of Chemistry, University of Glasgow, Glasgow G12 800,UK Thebaine 3 has been converted in an improved yield (72). with tetranitromethane, oxygen and ammonia, into the 14P-nitro 8a,lO~-epidioxide 4, which reacted with base to give the known 14~-nitro-lO-oxo compound 7. Reduction of this nitro compound with tributyltin hydride gave the 8p-hydroxy-1 0-oxodihydrothebaine 8, which was dehydrated with phosphorus oxychloride to give 10-oxothebaine 9. As expected, this conjugated diene reacted with but-3-en-2-one to form a Diels-Alder cycloadduct 10, the 10-0x0 derivative of thevinone 11, the precursor of a well known series of opiate analgesics.Unexpectedly, the epidioxide 4 was found to isomerise cleanly on activated alumina to give the 8a,lOa-epoxy acetal 5, in which a peroxide oxygen has been inserted into the 7,8-bond of the bridged peroxide 4. Opiate analgesics act at several, well defined sub-populations of receptors in the central nervous system. For example, morphine 1acts predominantly at p receptors while ethylketocyclazocine 2 is a prototypical ligand for K receptors. Several clinically use- ful opiates have been synthesis4 from the opium alkaloid 10 1 2 thebaine 3 and, consequently, like morphine, they have C-10 methylene groups. 10-Oxothebaine 9 therefore is an attractive starting material for the synthesis of potentially K-selective opiates.Direct conversion of the 10-methylene group of several morphinan derivatives into a 10-keto group has been achieved with chromium(w) reagents, although yields are generally modest.' However, the methoxy diene system of thebaine 3 is attacked by powerful oxidants. We report here a 4-step synthesis of 10-oxothebaine 9 from thebaine 3 via the known epidioxide 4 and 10-0x0 derivative 7(Scheme 1). We reported2 that the thebaine 3 reacts in benzene with tetranitromethane and oxygen to give, unexpectedly, the bridged peroxide 4 as a major product. However, about half the thebaine is concurrently converted into its trinitromethane salt. For the present study, we increased the yield of the epidioxide 4 to ca.70 by conducting the nitration in benzene with slow passage of streams of oxygen and ammonia2C through the mixture. Treatment of the epidioxide with sodium hydroxide in ethanol, as before,2b gave the keto alcohol 7 in high yield. Removal of the 14P-nitro group was effected with an excess of tributyltin h~dride,~ with heating under reflux in toluene. The radical generator azoisobutyronitrile was added slowly to promote the reduction, which proceeded in cu. 70 yield. Finally, the codeine derivative 8 was dehydrated with phosphorus oxychloride in pyridine to afford 10-oxothebaine 9 (73). The structure 9 was readily discerned by comparison of the product's spectra with those of thebaine 3, and by the following transformation.Thevinone 11, the cycloadduct of thebaine and but-3-en-2- one, is the precursor of an important series of potent analgesic^.^ When 10-oxothebaine 9 was heated with an excess of the MeO' 3 6 Ii 4 7 I Iii iv Me0 MeO MeO' OH 5 8 IV Meo-NMe MeO' MeO' OMe 1o;x=o 9 11; X = H2 Scheme 1 Reagents and conditions: i, C(NO2),4,-NH, in PhH at 2OOC; ii, Al,O,; iii, NaOH in EtOH at 20OC; iv, Bu,SnH and MezC(CN)Nlz in PhMe at 11 1 "C; v, POCl3-C5H,N in PhMe at 11 1 "C;vi, MeCOCH=€H, in PhH at 80 OC Me0 H 12 butenone, the corresponding cycloadduct 10 was formed in 86 yield. Again, the structure and stereochemistry of 10- oxothevinone 10 was deduced from the close similarity, mutatis mutandis, of its 'H NMR spectrum with that of thevinone 11.A by-product of the cycloaddition was tentatively identified ('H NMR) as the 7p epimer of 10-oxothevinone 10. The epidioxide 4 was routinely purified2' by chromatog- raphy on neutral, grade I11 alumina, prior to recrystallisation. When grade I alumina was inadvertently employed, efficient conversion into the isomeric 'isodioxide' 5 took place. The structure 5 was deduced from the following chemical and spectroscopic properties. Unlike the epidioxide 4,2b the iso- dioxide 5 did not react with triphenylphosphine, or sodium iodide in acetic acid. Also, the mass spectrum showed an intense molecular ion peak, unlike the spectrum of the epidioxide. Clearly then, the isodioxide was not a peroxide.The IR and NMR spectra showed the absence of hydroxy and carbonyl groups and of any newly formed C=Cdouble bond. Therefore, the isodioxide 5 was hexacyclic, like its precursor 4. The 'H and I3C NMR spectra indicated the presence of a new, acetal, methine group 8-H, 6, 105.6 and 5.37 (s). Significantly, one other methine group 7-H, 6, 120.0 and 6, 5.73 (d, J 1.4 Hz) had also become attached to oxygen during the isomerisation. Clearly, rearrangement had occurred with cleavage of the peroxide link in the epidioxide 4, perhaps catalysed by alumina acting as a Lewis acid, and migration of an adjacent methine carbon, C-7 or C-9, on to oxygen. Accordingly, the alternative structures 5 and 6 were considered for the isodioxide.The former better accounted for the 'H and I3C chemical shifts and the vicinal coupling constant J9.104.4 Hz. Structure 5 was con- firmed when I3C-'H NMR correlations showed (see Experi- mental section) that the acetal carbon, 6 105.6, was C-8 rather than C-10 (6 76.6). Migration of C-7, rather than C-9, on to oxygen (0-8) in the epidioxide 4, concerted with peroxide cleavage, would be favoured by the anti arrangement of the 7,8- bond and the peroxide link, as shown in the diagram 12, based upon the published2" X-ray structure. Attack of the other oxygen (0-10) on the cation generated at C-8 would then form the 5-membered, acetal ring. A similar, 8a,lOa-epoxy link is present in the epoxide formed 2b by reduction of the epidioxide 4 with triphenylphosphine.Experimental General.-M.p.s. were determined on a Kofler, hot-stage microscope. IR spectra were recorded on either a Perkin-Elmer 580 or 953 spectrometer for chloroform solutions. 'H NMR spectra at 90 MHz were obtained with a Perkin-Elmer R34 spectrometer. I3C NMR spectra at 90.6 and 50.3 MHz, and 'H spectra at 360 and 200 MHz were obtained with Bruker spectrometers. The 'H spectrum at 100 MHz was obtained with a Varian XL 100 spectrometer. All NMR spectra were recorded for deuteriochloroform solutions. J Values are in Hz. Mass spectra were produced by EI at 70 eV with an AEI MS9 instrument. TLC employed Merck GF254 silica gel, the flow J. CHEM. SOC. PERKIN TRANS. 1 1994 was assisted by a water pump.' Woelm alumina was used for column chromatography.Organic solvents were dried, unless otherwise stated, over magnesium sulphate and evaporated with a Buchi Rotavapor. Improved Preparation of the Epidioxide 4.-Tetranitro- methane was prepared from fuming nitric acid and acetic anhydride.6 For best results, the fuming nitric acid was freshly redistilled from a mixture with one third of its volume of concentrated sulphuric acid. The tetranitromethane was steam distilled, washed successively with aq. sodium carbonate and water and then dried (Na2S04); 6J50.3 MHz; CDCl,-C,D, (4: l) 119.8 (nonet, JC,N9.4). The 'H spectrum likewise showed no signals from any significant impurities. Dry ammonia gas and dry oxygen were passed slowly through a solution of thebaine (18.73 g, 60.2 mmol) in benzene (500 cm3) at room temperature after which tetranitromethane (1 8.87 g, 96.3 mmol) in benzene (100 cm3) was added dropwise during 20 min with continuous passage of both gases. The passage of both gases was continued for a further 2 h during which time a red oil separated from the reaction mixture and collected in the bottom of the flask.The supernatant, benzene solution was decanted off the oil, which was then washed with benzene (2 x 150 cm3). CAUTlON: the red oil contains ammonium trinitromethide, which may decompose violently when heated. The combined benzene solutions were washed successively with aq. sodium carbonate and water and then were dried and evaporated. The residual oil was chromatographed on a column of either neutral grade 111 alumina or TLC grade silica to give the epidioxide 4 (16.76 g, 72), m.p.160 "C (from ethyl acetate) (lit.,2b 160- 161 "C). The Isodioxide 5.-The epidioxide (400 mg) 4 was adsorbed onto a column of neutral, grade I alumina (1 5 g) which after ca. 2 h was eluted with chloroform-hexane to give the isodioxide 5 (380 mg). Recrystallisation of the product 5 from ethanol gave material (295 mg, 74), m.p. 270-271 "C (Found: C, 59.0; H, 5.0; N, 7.1; M', 388.1276. C19H2*N20, requires C, 58.8; H, 5.2; N, 7.2; M, 388.1270); v,,,/cm-' 1549, 1450, 1225 and 1285; 6,(360 MHz) 6.86 and 6.75 (ABq, J 8.1, 1-and 2-H), 5.73 (d, J 1.4, 7-H), 5.37 (s, 8-H), 5.21 (d, J 1.4, 5-H), 5.12 (d, J4.4, 10-H), 4.47 (d, J4.4, 9-H), 3.88 (s, 3-OMe), 3.44 (s, 6-OMe), 2.54 (s, NMe), 2.53 (m, 16-H,), 2.39 (ddd, J 13.0, 10.3 and 6.9, 15,,-H) and 1.78 (dt, J 13.0 and 2.7, 15eq.-H); 6,(90.6 MHz) 154.8 (C-6), 146.1 (C-3), 145.1 (C-4), 130.3 (C-12), 125.9 (C-1 l), 120.0 (C-7), 119.3 (C-1), 113.9 (C-2), 105.6 (C-8), 95.5 (C-14), 92.7 (C-5), 76.6 (C-lo), 71.3 (C-9), 56.3 (OMe), 56.1 (OMe), 46.6 (C-l3), 45.8 (C-l6), 43.1 (NMe) and 31.3 (C-15).I3C-'H Correlations (50.3 and 200 MHz) showed, inter aka, that C-8 in structure 5 rather than C-10 in structure 6 was the acetal carbon (6, 105.6). Thus, the following 1-and 3-bond 13C-lH couplings allowed signals for 10-H, C-10, 8-H and C-8 to be successively identified: C-12 (8 130.3) coupled with 1-H (6.87), 15,,-H (2.38) and 10-H (5.12); C-10 (76.6) with 10-H and 8-H (5.37); and C-8 (105.6) with 8-H and 10-H. 8,14-Dihydro-8/3-hydroxy-14P-nitro- 1 0-oxothebaine 7.-The epidioxide 4 (600 mg) was suspended in ethanol (75 cm3) at room temperature and treated with aq.sodium hydroxide (4 mol dm-3; ca. 0.2 an3).The mixture was stirred for 18 h, by which time a clear solution had formed. Work-up2' gave the nitro ketone 7 (560 mg, 9379, m.p. 220°C (from methanol) (lit.,2b 219 "C) (Found: C, 58.8; H, 5.3; N, 7.1. Calc. for C19H20N20,: C, 58.8; H, 5.2; N, 7.2). The UV and 'H NMR spectra agreed well with those reported.2b 8,14-Dihydro-8/3-hydroxy-10-oxothebaine -The nitro ketone 7 (2.00 g, 5.15 mmol) and tributyltin hydride (9.00 g, 3 1 mmol) were heated under reflux in dry toluene (250 cm3) under J.CHEM. SOC. PERKIN TRANS. 1 1994 nitrogen and azoisobutyronitrile was added slowly in toluene to the mixture, the course of reaction being monitored by TLC. Addition was continued until reduction of the nitro compound was almost complete. The time (ca. 2 h) and the quantity of azo compound (a large excess) required depended upon the rate of addition. The mixture was heated for a further 2 h and then was evaporated to give an oily residue. The residue was shaken with acetonitrile (100 cm3) and hexane (100 cm3). The acetonitrile layer was washed with hexane (4 x 100 cm3) and then evaporated to give a yellow-green solid. Chromatography on a silica gel (TLC grade) column eluted with ethyl acetate gave successively the nitro compound 7 (0.27 g) and the hydroxy ketone 8 (1.28 g, 72), m.p. 158-160 "C (from methanol) for the first batch prepared; later batches had m.p. ca.260 "C (decomp.) with sintering from 235 "C (from ethanol) (Found: C, 66.3; H, 6.15; N, 4.0; M+, 343.1413. C,,H,,NO, requires C, 66.5; H, 6.2; N, 4.1; M, 343.1419); v,,,/cm-' 3600, 3400, 1672, 1620 and 1600; v,,,(KBr)/cm-l 3528, 1665, 1624 and 1605; 6,(200 MHz)7.38and6.80(ABq, J8.4, l-and2-H),4.95(d, J1.3,5-H), 4.82 (d, J 1.8, 7-H), 3.92 (s, 3-OMe), 3.78 (dt, J 9.3 and ca. 1.5, 8-H), ca. 3.56 (signal partly obscured, 9-H), 3.55 (s, 6-OMe), 2.72 (ddd, J 12.1, 5.0 and 1.8, 16eq.-H), 2.44 (s, NMe), 2.37 (td, partly obscured, J 12.1 and 4.5, 1 6ax.-H), 2.27 (dd, J 9.3 and 2.6, 14-H), 2.10 (td, J 12.3 and 5.0, 150x.-H), 1.95 (ddd, J 12.5, 4.3 and 1.8, 15eq,-H) and 1.72 (br s, OH, exch.with D,O); 6,(50.3 MHz) 34.8 (C-15), 42.3 (C-13), 43.5 (NMe), 47.3 (C-16), 51.5 (C-14), 54.9 (OMe), 56.3 (OMe), 65.9 (C-8 or -9), 66.0 (C-9 or -8), 87.5 (C-5), 104.7 (C-7), 113.5 (C-2), 118.7 (C-1), 125.2 (C-1 l), 136.7 (C-12), 143.8 (C-4), 150.2 (C-3), 151.6 (C-6) and 192.4 (C-10). 10-Oxothebaine 9.-The hydroxy ketone 8 (1.00 g), dry pyridine (4 cm3) and freshly distilled phosphorus oxychloride (1 cm3) were heated under reflux in dry toluene (1 50 cm3) under nitrogen for 3 h, by which time a black oil had separated out. The mixture was evaporated to low volume and the residue was treated with sufficient aq. sodium hydrogen carbonate to render the mixture distinctly alkaline.The toluene and aq. layers were separated and the aq. layer was extracted with toluene (4 x 40 cm3). The combined toluene solutions were washed with aq. sodium hydrogen carbonate, dried and evaporated. Pyridine was removed from the residue by repeated addition and evaporation of toluene. The resulting yellow- green solid was chromatographed on a column of silica gel (TLC grade). Elution with chloroform then chlorofom-methanol (10: 1) gave a green oil (0.77 g), which crystallised slowly. Recrystallisation from ethanol gave 10-oxothebaine 9 (0.69g, 7373, m.p. 195-198 "C(Found: C, 70.3; H, 5.8; N,4.3; M+ 325.1305. Cl9H1,NO4 requires C, 70.1; H, 5.9; N, 4.3; M, 325.13 14); v,,,/cm- ' 1680, 1666 and 1607; A,,,(MeOH)/nm 263 (E 12 900 dm3 mol-' cm-'), 285 (10 450) and 330 (8470); 6,(200 MHz) 7.35 and 6.76 (ABq, J 8.4, 1-and 2-H), 5.64 (d, J 6.5,8-H), 5.38 (s, 5-H), 5.09 (d, J 6.5,7-H), 3.92 (s, 3-OMe), 3.62 (s, 6-OMe and 9-H), 2.77 (ddd, J 12.6, 6.5 and 2, 16eq-H), 2.69 297 (td, J 12.4 and 3.5, 16ax-H), 2.52 (s, NMe), 2.28 (td, J 12 and 6.4, 15,,-H) and 1.87 (ddd, J 12.6, 3 and 2, 15eq-H); 6,(50.3 MHz) 36.8 (C-l5), 43.4 (NMe), 47.4 (C-16), 47.7 (C-l3), 55.1 (OMe), 56.3 (OMe), 70.5 (C-9), 88.5 (C-5), 96.4 (C-7), 1 12.3 (C-8), 113.0 (C-2), 119.7 (C-1) 125.4 (C-11), 130.7 (C-12), 138.7 (C-14), 143.9 (C-4), 149.6 (C-3), 152.4 (C-6) and 193.8 (C-10).10-Oxothevinone lO.-lO-Oxothebaine 9 (120 mg) and freshly distilled but-3-en-2-one (1.2 cm3) were heated under reflux with benzene (1.2 cm3) for 5 h.The solution was evaporated and the residue was chromatographed first in diethyl ether on a short column of silica gel (TLC grade) and then on silica plates developed with diethyl ether. The band having R,ca. 0.4 was extracted with ethyl acetate to give 7a-acetyl-6,7,8,14-tetra-hydro-1O-OXO-~~,14a-ethenothebaine (1 0-oxotheuinone) 10 (1 26 mg, 8679, m.p. 129 to 131 "C (from methanol) (Found: C, 69.8; H, 6.4; N, 3.5. C23H,,N0, requires C, 69.9; H, 6.3; N, 3.5); vmax/cm-' 1719 and 1675; 6,(200 MHz) 7.28 and 6.76 (ABq, J 8.4, 1- and 2-H), 5.98 (dd, J8.8 and 1.2, 18-H), 5.55 (d, J8.8, 17-H),4.68(d,J1.2,5-H),3.92(s,3-OMe),3.61(s,6-OMe),3.18 (s, 9-H), 2.48 (s, NMe), 2.16 (s, Ac) and 1.39 (m, 8a-H); m/z 395 (M'), 352 and 325.The product 10 before TLC showed weak 'H NMR signals at 6 6.13 (dd, J ca. 9 and 1, 18-H), 5.44 (d, J 8.8, 17-H) and 5.05 (d, J ca. 1, 5-H) possibly arising from the protons indicated in the corresponding 7p-acetyl derivative. Acknowledgements We thank the SERC and Reckitt and Colman Ltd (Hull) for financial support, Drs. C. P. Chapleo and J. W. Lewis (Reckitt and Colman) for their interest in this research and Drs I. H. Sadler (Edinburgh) and D. S. Rycroft (Glasgow) for NMR spectra. References 1 S. Archer, A. Seyed-Mozaffari, S. Ward, H. W. Kosterlitz, S. J. Paterson, A. T. McKnight and A. D. Corbett, J. Med. Chem., 1985, 28, 974, and references cited therein. See also A. B. McElroy, D. E. Bays, D. 1. C. Scopes, A. G. Hayes and M. J. Sheehan, J. Chem. Soc., Perkin Trans. I, 1990, 1563. 2 (a) R. M. Allen, C. J. Gilmore, G. W. Kirby and D. J. McDougall, J. Chem. Soc., Chem. Commun., 1980, 22; (b) R. M. Allen, G. W. Kirby and D. J. McDougall, J. Chem. Soc., Perkin Trans. I, 1981, 1143; (c) R. J. Kobylecki, I. G. Guest, J. W. Lewis and G. W. Kirby, DTOLS 2,812,580/1978 (Chem. Abstr., 1979,90, 87709t). 3 N. Ono, H. Miyake, R. Tamura and A. Kaji, Tetrahedron Lett., 1981, 1705. 4 K. W. Bentley and D. G. Hardy, J. Am. Chem. Soc., 1967,89,3267. 5 L. M. Harwood, Aldrichimica Acta, 1985, 18,25. 6 P. Liang, Organic Synth., Coll. Vol. 3, John Wiley, New York, 1955, p. 803. Paper 3/05418G Received 9th September 1993 Accepted 11th October 1993