MeO MeO O NMe MeO MeO NO2 O MeO MeO NH2 O O NCOR O MeO MeO NHCOR O MeO NCOR O O OH MeO MeO NCOR O 19 2 OH 19 1 O R¢O O 7 5 NMe 8 9 O R¢O O 17 NMe 1 2 3 9 14 10 4 NH NH CH2R 4 7 i R O ii iii O MeO viii vii NMe iv 5 vi O N 8 6 O R iv v,iv 10 Me Ph PhCH2CH2 PhCH2O PhCH2 PhCH2CH2CH2 CH3[CH2]4 a b c d e f g R HO ix O NH O MeO O N O R O R Me 11 + 8 4 MeO O N O R H OH MeO O N O OH O O N OH 18 R O R Me 12 + 7 + 152 J. CHEM. RESEARCH (S), 1997 J. Chem. Research (S), 1997, 152–153 J. Chem.Research (M), 1997, 1001–1020 Short Synthesis of 14b-Acylaminocodeinones from the Cycloadducts of Thebaine and Acylnitroso Compounds Ross I. Gourlay and Gordon W. Kirby* Department of Chemistry, University of Glasgow, Glasgow G12 8QQ, UK Thebaine 1 has been converted in four steps, via its cycloadducts 7 with acylnitroso compounds and the derived dimethyl ketals 8 or ethylene ketals 13, into 14b-acylaminocodeinones 5 (Rp=Me), analgesics formerly prepared from 14b-aminocodeinone dimethyl ketal 3. 14b-Aminocodeinone dimethyl ketal 3, originally prepared2 from thebaine 1 via the nitro ketal 2, was a key intermediate in the synthesis of the acylamino- 5 (Rp=Me) and alkylamino- codeinones 6 (Rp=Me) (Scheme 1).1 These codeinones and the corresponding morphinones (Rp=H) have shown promise as clinically useful analgesics. For example, 14b-pentylaminomorphinone, pentamorphone 6 (R=Bu, Rp=H), has been evaluated3 in man and identified as an effective analgesic with clinically tolerable side-effects in the dose range 0.12–0.24 mg kgµ1.In the mouse hot-plate test, pentamorphone showed4 1872 times the potency of morphine and 4 times that of fentanyl. Since the compounds having substantial analgesic potency are all N-acyl (5) or N-alkyl (6) derivatives of the parent aminocodeinone, we have now devised a synthetic route from thebaine 1 in which the required acyl groups are introduced directly in the first step and which avoids nitration with tetranitromethane and the consequent formation of potentially hazardous salts of trinitromethane. Thebaine 1 reacts with transient acylnitroso compounds, RCONO, generated in situ by oxidation of hydroxamic acids, RCONHOH, with periodate, to give the cycloadducts 7 in high yield.7 In principle, a general synthesis of the acylamino ketals 4 might then be completed in two, unremarkable steps, viz.methanolysis of the cyclic ketals 7 and deoxygenation of the resulting hydroxamic acids 8.The 3-phenylpropanoyl cycloadduct 7c was selected initially for detailed study since the corresponding codeinones 5c (Rp=Me) and 6c (Rp=Me) and morphinones (Rp=H) were especially potent analgesics. Brief treatment of the cycloadduct 7c with dry, methanolic hydrogen chloride at 0 °C gave the dimethyl ketal 8c in good yield. However, attempted deoxygenation of the hydroxamic acid 8c with standard reagents was uniformally unsatisfactory. For example, zinc in acetic acid, or acetic acid alone, simply regenerated the cycloadduct 7c, while zinc and ammonium acetate caused no significant change.Other, more powerful reducing agents caused reductive removal of the entire 14-acylamino group. However, phosphorus trichloride in pyridine10 at 10 °C rapidly gave, in high yield, the required amide 4c, which was hydrolysed with methanolic hydrochloric acid to afford the codeinone 5c (Rp=Me). Other transformations of the cycloadduct 7c gave successively the hydroxamic acid 9c and the bridged phenol 10c (Scheme 1)7 and the oxazolidine 12c (Scheme 3).Unexpectedly, a serious limitation of this route to the dimethyl ketals 4 became apparent when other cycloadducts 7 were investigated. Although methanolysis of the 3-phenylpropanoyl derivative 7c was essentially quantitative, incomplete conversion was observed for all the other cycloadducts tested. The approximate positions of the equilibria 7M8, determined from the 1H NMR spectra of the reaction mixtures, are expressed, as follows, as % conversions into the ketals 8: 8a, 85%; 8b, 63; 8c, a95; 8d, 50; 8e, 90; 8f, 60; 8g, 70.Since methanol was already employed in a large excess as the solvent, there was no means of displacing the ketal equilibria substantially in the forward direction. Attention was there- Scheme 1 Reagents and conditions: i, C(NO2)4 in MeOH–NH3; ii, Zn–NH4Cl in MeOH; iii, RCOCl–pyridine; iv, HCl–H2O; v, B2H6–THF or LiAlH4; vi, RCONHOH–NaIO4; vii, dry MeOH–HCl; viii, PCl3–pyridine at 10 °C; ix, NaOEt–EtOH at 20 °C Scheme 2 Reagents and conditions: SO2 in pyridine at 115 °C Scheme 3 Reagent: NaBH3CN–HCl in dry THF *To receive any correspondence.O O O NCOR OH O 14 O O NHCOR 13 7 5 i ii iii J.CHEM. RESEARCH (S), 1997 153 fore turned to the thermodynamically more stable, cyclic, ethylene ketals 13 (Scheme 4).6 Treatment of the cycloadduct 7c in dichloromethane with an excess of anhydrous, glycolic hydrogen chloride at room temperature effected essentially quantitative formation of the ethylene ketal 13c.Significantly, the benzoyl cycloadduct 7b, which gave only ca. 63% of the dimethyl ketal 8b, also underwent essentially complete conversion into the ethylene ketal 13b. Although the ketals 13 were deoxygenated effectively with phosphorus trichloride in pyridine, an alternative method,13 which gave cleaner products, was devised. Thus, solutions of the ethylene ketals 13 in pyridine were saturated at room temperature with sulfur dioxide then heated under reflux, to afford the amides 14 in good yield.When this method was applied to the dimethyl ketal 8c, concomitant deoxygenation and cyclisation gave the oxazoline 11c (Scheme 2). Finally, hydrolysis of the ethylene ketals 14 with hydrochloric acid gave the acylaminocodeinones 5 (Rp=Me). In conclusion, the route (Scheme 4) involving deoxygenation of the ethylene ketals 13 with sulfur dioxide in pyridine is recommended generally for the conversion of thebaine, in four steps, into 14b-acylaminocodeinones.No chromatographic purifications are necessary and yields of 70–80% per step are usual. The route (Scheme 1) employing deoxygenation of the dimethyl ketals 8 with phosphorus trichloride in pyridine was satisfactory only for the cycloadducts 7a, c and e; with the other derivatives the equilibration 7M8 reduced the overall yield. We thank the SERC for financial support, Reckitt and Colman Pharmaceutical Division for pharmacological tests and their interest in this work and Dr J.W. Lewis for helpful discussions. Techniques used: IR, 1H NMR, mass spectrometry References: 14 Schemes: 4 Received, 23rd December 1996; Accepted, 22nd January 1997 Paper E/6/08557A References cited in this synopsis 1 R. J. Kobylecki, I. G. Guest, J. W. Lewis and G. W. Kirby, DTOLS 2,812,580, 1978 (Chem. Abstr., 1979, 90, 87709t); DTOLS 2,812,581, 1978 (Chem. Abstr., 1979, 90, 39100r); see also J. W. Lewis, C. F. C. Smith, P. S. McCarthy, D. S. Walter, R. J. Kobylecki, M. Myers, A. S. Haynes, C. J. Lewis and K. Waltham, NIDA Res. Monogr. Ser., 1988, 90, 136. 2 R. M. Allen, G. W. Kirby and D. J. McDougall, J. Chem. Soc., Perkin Trans. 1, 1981, 1143; see also C. F. Henderson, G. W. Kirby and J. Edmiston, J. Chem. Soc., Perkin Trans. 1, 1994, 295. 3 P. S. A. Glass, E. M. Camporesi, D. Shafron, T. Quill and J. G. Reves, Anesth. Analg. (N.Y.), 1989, 68, 302. 4 F. G. Rudo, R. L. Wynn, M. Ossipov, R. D. Ford, B. A. Kutcher, A. Carter and T. C. Spaulding, Anesth. Analg. (N.Y.), 1989, 69, 450. 6 G. W. Kirby and D. McLean, J. Chem. Soc., Perkin Trans. 1, 1985, 1443. 7 G. W. Kirby and J. G. Sweeny, J. Chem. Soc., Perkin Trans. 1, 1981, 3250. 10 E. Ochiai, J. Org. Chem., 1953, 18, 534. 13 Cf. G. W. Kirby, H. McGuigan and D. McLean, J. Chem. Soc., Perkin Trans. 1, 1985, 1961. Scheme 4 Reagents and conditions: i, dry (CH2OH)2–HCl at 20 °C; ii, SO2 in pyridine at 115 °C; iii, HCl–H2O–MeOH