摘要:

1976 1037Intramolecular Catalysis. Part Il1.l Effect of a Neighbouring Hydroxy-group on the Opening of Steroidal Aziridines with Azide AnionsBy Yoram Houminet, Department of Organic Chemistry, The Hebrew University, Jerusalem, Israel5a,6a-lminocholestan-3~-ol (IV) and its 3a-hydroxy-isomer (V) have been prepared and their structures estab-lished. Their reactions with sodium azide in acetone-water (2 : 1 ) produce the corresponding trans-diaxialamino-azides. The ratio of the reaction rates is ca. 1 : 2, respectively. The mechanisms of the reactions of thesteroidal aziridines are discussed and comparison is made between these compounds and the related epoxides.WE have recently reported that 5a,6a-epoxycholestan-3a-01 reacts with sodium azide in acetone-water fasterthan its 3p-hydroxy-is0mer.l A similar phenomenonhas also been observed with the 7p- and 7a-hydroxy-4aJ-5a-epo~ycholestanes.~~~ It has been suggested that theneighbouring hydroxy-group reacts as an internalelectrophile and facilitates the opening of the epoxidering by delocalization of the negative charge whichdevelops on the ring oxygen atom in the transitionstate.l The present investigation extends this study to5a,6a-iminocholestan-3a-ol and its 3p-hydroxy-isomerJsteroidal aziridines possessing similar stereochemistry.We were also interested in comparing the reactivities ofaziridines and epoxides of similar structures.Theaziricrlines were synthesized from the corresponding5a-azido-6~-chlorocholestane derivatives. 5a-Azido-6P-chlorocholestan-3~-o1 (I) was prepared by basic hydro-lysis of its known a ~ e t a t e .~ Oxidation of (I) with sodiumdichromate in acetic acid gave the ketone (11) in almostquantitative yield. Reduction of (11) with an excess ofsodium borohydride at room temperature gave amixture of two main products, of which the major one(50%) isolated by t.l.c., was 5a-azido-6~-chlorocholestan-3~-01 (111). The niinor product (33%) was the 3p-01(I). The n.m.r. spectrum of (111) indicated its stereo-chemistry. The CN-OH signal in (111) was a multiplet,of W,,, (7.2 Hz (the signal overlaps with that of the6~-proton, and both give a multiplet of WIlz 7.2 Hz),characteristic of an equatorial proton [the multipletcorresponding to the 3a-proton in (I) has U1;,, ca.24Hz]. Chemical evidence for the structure (111) wasobtained by its smooth oxidation to regenerate theketone (11). I t is well known that reduction of a 3-0x0-group in the cholestane skeleton with sodium boro-hydride gives the 3p-hydroxy-isomer as the majorp r o d ~ c t . ~ However, it has been shown that thepresence of a substituent at the 5a-position affects thestereochemical course of the reduction so that therelative amount of the 3a-hydroxy-isomer is increased,and in some cases it even becomes the major product.The reduction of (11) to (111) and (I) further demon-strates this phenomenon.Treatment of the 3P-01 (I) with a large excess of lithiumaluminium hydride in bis-(2-methoxyethyl) etherPart XI, Y . Houminer, J.C.S.Perkin I, 1976, 1663.I>. H. R. Barton and 1'. Houminer, J.C.S. Chem. Comm.,G. Snatzke and A. Veithen, A m d e n , 1967, 703, 169. * L. M. Jackman and S. Sternhell, 'Application of NuclearMagnetic Resonance Spectroscopy in Organic Chemistry,'Pergamon, London, 1969, p. 288.1973, 839.afforded 5a,6a-iminocholestan-3p-o1 (IV) in good yield.Similarly, the 3a-01 (111) was almost quantitativelyconverted into 5a,6a-iminocholestan-3a-ol (V) . Theconfiguration of the aziridine ring in both (IV) and (V)has to be 5a,6aJ since it is well established that the ringretains the original configuration of the a~ido-group.~~~In the n.m.r. spectrum of (IV) the CH*OH signal occursat 6 3.75, as a multiplet, W1lz 23 Hz. The correspondingH NH OY NH NH(1Y) (Y) (Yi)proton in the spectrum of (V) resonates at 6 4.09 withWIl2 6.8 Hz.This then establishes the configurationsof the hydroxy-group in (IV) and (V) as equatorial andaxial, respectively. The C-18 protons of both (IV) and(V) appear as singlets at 6 0.61, but the correspondingprotons of cholestan-3p-01 resonate at 6 0.65. Thisdifference is due to a long-range shielding effect and canbe ascribed to the 5a,6a-aziridine ring current. Asimilar phenomenon is also observed in the n.m.r.spectra of the corresponding 5a ,6a-epoxycholest anes,in which the singlets corresponding to the 18-protonsappear at 6 0.62.It has been shown1 that reduction of 5aJ6a-epoxy-cholestan-3-one with sodium borohydride gives a t .l.c.-separable mixture of the 3a- and 3p-hydroxy-epoxides.The same reaction was attempted as an alternativeroute for the preparation of (V). Oxidation of (IV)by Jones, method afforded 5a,6a-iminocholestan-3-one(IV), treatment of which with sodium borohydride inmethanol gave a mixture of (IV) and (V).However, the0. R. Vail and D. M. S. Wheeler, J. Org. Chem., 1962, 27,3803; 0. H. Wheeler and J. L. Mateos, Can&. J . Chem., 1958,36,1049.Y . Houminer, J . Org. Chem.., 1975, 40, 1361, and referencescited therein.G. J. Matthews and A. Hassner, Tetrahedroiz Letters, 1969,1833; A. Hassner and L. A. Levy, J . Amer. Chem. SOC.. 1965,87,42031038 J.C.S. Perkin Itwo isomers could not be separated by t.1.c. The n.m.r.spectrum of the mixture indicated that (IV) was itsmajor component.The reaction of either (IV) or (V) with sodium azidein acetone-water (2 : 1) gave the corresponding amino-azides, (VII) and (VIII) respectively.In the n.m.r.HO & c@ I ’ OH NH2wn I cvm,spectra of these the CHN, signals appear as multipletsat 6 3.26 with W1l2 5.0 Hz and at 6 3.28 with Wll2 5.6Hz, respectively. These values are characteristic ofequatorial protons,4 thus establishing that in bothcompounds the 6-azido-group is axial and in the p-configuration. The formation of both (VII) and (VIII)indicate that the ring opening of the 5~,6a-imines withsodium azide produces the trans-diaxial amino-azides.This result fits the general pattern of the uncatalysed andacid-catalysed ring opening of aziridines, which usuallyproceeds stereospecifically t~ans.~*Qand in the light of this it has been emphasized that, foraziridines to undergo ring opening with a nucleophileunder ordinary conditions, the nitrogen atom must firstbe protonated.Thus, reaction via Scheme l(a) is almostimpossible and the route shown in Scheme l(b) isalways preferred. Indeed, most known reactions ofthis type are carried out under acidic cataly~is.~ How-ever, there are several examples,l0J1 including our ownwork, in which aziridines react with nucleophiles inneutral aqueous solutions. These observations do notnecessarily imply that under these conditions thereactions proceed via route (a) in Scheme 1, since wateris sufficiently acidic to provide a very low concentrationof aziridinium ions.9 Our results support this con-clusion (see preceding discussion).SCHEME 2Nu The tram-diaxial structure of both amino-azides(VII) and (VIII) suggests an S N ~ mechanism for thereaction of (IV) and (V) with azide anions, as shown inScheme 2 for the case of (IV).An SN1 mechanism(Scheme 3) would afford the corresponding 5-azido-6a-amino-compound resulting from the more stable tertiarycation. This compound is not formed, thus indicatingthat an SN1 mechanism is not important in our case. Asimilar stereochemical argument was taken as evidencefor an SN2 mechanism in the reaction of ZP,SP-imino-Based-catalysed ring opening of aziridines, unless they cholestane with nucleophiles.8have electron-withdrawing N-substituents, is rare.8 The kinetics of the reactions of both (IV) and (V)The imino-group, being a poor leaving group can react with sodium azide in aqueous acetone were followed by- Nu- 1 I & INH2 - NHNu -* H+NH, NH +NH2SCHEME 1SCHEME 3only with very strong bases such as carbanions andamide ions.9 There is a significant body of data9 con-cerning the opening of aziridines by various nucleophilesA.Hassner and C. Heathcock, J . Org. Chem., 1966, 80, 1748and references cited therein.0. C. Dermer and G. E. Ham, * Ethylenimine and OtherAziridines, Chemistry and Applications,’ Academic Press, NewYork, 1969, pp. 203-300, and references cited therein.n.m.r. The 18-protons in both (IV) and (V) resonateat 6 0.61, whereas those in the corresponding amino-azides (VII) and (VIII) both resonate at 6 0.69.Inte-gration of the two separate signals made possible theRichardson, Proc. Chem. Soc., 1963, 84.79, 734.10 R. D. Guthrie, D. Murphy, D. H. Buss, L. Hough, and A. C.11 R. Ghirardelli and H. J. Lucas, J . Amev. Chem. SOC., 19571976 1039determination of the amounts of the starting materialand the product in the reaction mixture. For com-parison, the kinetics of the reactions of the epoxides(IX) and (X) with sodium azide to give the correspondinghydroxy-azides (XI) and (XII)1 were followed by aH 0'H OH(XI 1OH OH(Xn)similar method. The 18-protons in the epoxidesresonate at 6 0.62 and in the hydroxy-azides at 6 0.68.Since the kinetics of the reactions of the epoxides aremore simple we shall first discuss these results.Thereactions of both (IX) and (X) with a 10-120 foldexcess of sodium azide in acetone-water (2 : 1 v/v) werefound in each case to be first order in the epoxide blotsof ln(lOO/lOO - x ) against t , where x is the reaction per-centage, gave straight lines with correlation coefficients>0.9995]. The results with various concentrations ofsodium azide (Table 1) show that the reaction is also firstTABLE 1Rate coefficients for the reactions of 6a,6a-epoxycholestaneswith sodium azide in acetone-water (2 : 1 v/v) a t 50 O C10[NaN3]/ los x k,' b / lo6 x ka/Steroid a moll-' S-1 1 mol-1 s-10.31 0.16 j, 0.005 6.161.54 0.78 -J= 0.02 5.063.08 1.60 j, 0.02 6.193.08 a 1.62 f 0.03 6.266.16 2.48 & 0.04 4.030.31 1.16 4 0.03 37.421.64 6.63 f 0.10 36.563.08 11.64 -J= 0.15 37.80 I 6.16 18.66 f 0.23 30.115a,6a-EyoxYcho1estan- 3.08 11.40 f 0.13 37.015a, 6a-Epoxycholestan-3p-013a-01Steroid concentration 2.49 x 10" mol 1-'.Pseudo-first-In the presence of order rate coefficient (k2' = kgtNaNs]).NaOH (4.84 x 10-3 mol 1-l).order in the sodium azide. Therefore, the reactionkinetics fit the equation: rate = K,[St][N,-] (St =steroid). The results in Table 1 suggest that at concen-trations of sodium azide up to 3.08 x 10-1 mol 1-1 thesodium azide is fully dissociated. At higher concen-trations, the observed rate coefficient is less, and thismay indicate that some of the sodium azide exists as ionpairs. Table 1 also shows that the reaction rates are notdepressed by additions of sodium hydroxide.The Figure shows pseudo-first-order plots for thereactions of the aziridines (IV) and (V) with sodiumazide.For both aziridines the rate decreases with theprogress of the reaction. Table 2 summarizes theTABLE 2Initial rates for the reactions of 5a, 6cc-iminocholestaneswith sodium azide in acetone-water (2 : 1 v/v) at 50 OCSteroid' lO[NaN,]/m011-~6a, 6a-Iminocholes tan- 3 p-01 3.083.081.640.315a, 6a-Iminocholestan-3a-01 4.623.083.081.54109 x initialrate b/l mol-l s-l4.63<0.504.003.108.268.070.656.81* Steroid concentration 2.49 x mol 1-l. &4%. a Inthe presence of NaOH (4.84 x rnol 1-l).initial rates for the reactions of the aziridines withsodium azide at various concentrations; it is clear thatthe reaction for both (IV) and (V) is less than first-order in sodium azide.Table 2 shows also that thereaction rates are markedly lowered in the presence ofsodium hydroxide.I cdReactions of (A) 6a,6a-iminocholestan-3@-01 and (B) the 3a-01moll-1) with sodium azide (3.08 x 10-1 moll-1) in (2.49 xacetone-water (2 : 1 v/v) at 50 "CIn view of the above results, we propose a mechanismfor the reaction of both epoxides and aziridines asdemonstrated in Scheme 4 for the 3p-hydroxy-isomersof these compounds. Under our conditions, water orhydroxide anions are not involved as nucleophiles, sincethe only products are the azides. The reaction rate isgiven by equation (1).rate = K,[St][N,-] + &[StH+][N,-] (1)Aziridines and epoxides of similar structures differstrongly in their basicities.There are no reported pKvalues for the epoxides (IX) and (X) or the aziridines(IV) and (V), but it can be estimated, on the basis of acomparison between ethers and secondary amines,12that (IV) and (IX) as well as (V) and (X) differ in theirpK values at least by ten orders of magnitude. There-fore, in the case of the epoxides we can conclude that,under our conditions, which are slightly basic owing tola J. March, 'Advanced Organic Chemistry, Reaction Mech-anisms and Structure,' McGraw-Hill, New York, 1968, pp. 219-2211040 J.C.S. Perkin Ithe solvolysis of sodium azide in aqueous solutions,[StH+] is negligible, and although it is clear thatk, < k3 (the protonated epoxide must be more reactivetowards nucleophiles) the fact that [StH+] N 0, simpli-fies equation (1) into rate = K,[St][N,-].This con-clusion is supported both by the kinetics of the reactionsand by the observation that the rates are not depressedby the presence of added sodium hydroxide, whichshould decrease [StH+].The reaction kinetics in the case of the aziridines aremuch more complex. Although both [N3-] and P,O]remain practically constant during one kinetic run, thereactions have no pseudo-first-order characteristic(Figure). Therefore the decrease in rate during onekinetic run, both with (IV) and (V), is evidently due to thebuild-up of [OH-]. Indeed, added sodium hydroxidestrongly lowered the reaction rates (Table 2).TheseHO H,O & HO&+OH- \k - 1 t* $4-i XHHzO fast 1*o& % -I- OH'1XHSCHEME 4 X = 0 for epoxides, NH for aziridinesresults support our previous arguments on the mechan-ism of the opening of aziridines which we assumed toproceed exclusively through the protonated aziridines.The results (Table 1) verify our previous observationson the existence of electrophilic intramolecular catalysisin the opening of the epoxides (IX) and (X) with azideanions. We expected that such internal catalysiswould also be shown in the related aziridines, and sincethe epoxy-group in the epoxides is a better leavinggroup than the imino group in the aziridines, we evenexpected to find a larger catalytic effect in the lattercase. However, our results show that (V) is onlyslightly more reactive than (IV), and it is clear that(V) does not exhibit intramolecular catalysis.Thisphenomenon can be explained as follows. Electro-philic assistance by the hydroxy-group in (V) [but notin (IV)] can occur only on the unprotonated aziridine,in the transition state of which, (XIII), a negativecharge is developed on the nitrogen atom. Our resultssuggest that the reaction proceeds exclusively via theprotonated aziridine and therefore no such internalcatalytic effect can be observed.Both Table 2 and the Figure indicate that (V) isabout twice as reactive as (IV). The proximity of the & .N$b Hrxm)3a-hydroxy-group to the imino-group in (V) suggeststhe possibility of internal hydrogen bonding betweenthe two groups (mainly O-H***NH and not H O * * *H-N ; compare for example the case of ethanolamine 13),thus increasing slightly the concentration of a quasi-protonated aziridine in the case of (V) as compared with(IV).However, the i.r. spectra of both (IV) and (V)do not indicate any intramolecular hydrogen bonding in(V), similar to the case of the related ep0xides.l There-fore, internal hydrogen bonding in (V) is not a likelyexplanation for its higher reactivity. We believe thatthe greater reactivity results from the higher basicityof (V), i.e. the 3~-hydroxy-group can stabilize the positivecharge on the protonated nitrogen by internal solvation,thus increasing the basicity of the amino-group.It is clear from the results that any attempt to com-pare the rcactivities of epoxides and aziridines of re-lated structures towards nucleophiles is meaningless,since the two series react by two different mechanisms.If we compare (IV) and (IX), in both of which no intra-molecular catalysis is possible, the reactivities aresimilar in the absence of sodium hydroxide but differmarkedly in the presence of the base.Thus, in general,the relative reactivities of epoxides and aziridinesshould be strongly pHdependent.EXPERIMENTALR4.p.s were determined with a Fisher- Johns apparatus.1.r. spectra were recorded with a Perkin-Elmer 254 spectro-photometer. Optical rotations were determined for solu-tions in chloroform with a Perkin-Elmer 141 polarimeter.N.ni.r. spectra were taken for solutions in deuteriochloro-form with a Varian T 60 spectrometer with tetramethylsilaneas internal standard.Mass spectra were recorded on aVarian MAT 311 spectrometer. T.1.c. was carried out onsilica gel G. Plates were eluted with light petroleum (b.p.60-80') containing 20-4074 acetone.5u-Rzido-6~-chZo~ochoZestan-3~-oZ (I) .-A suspension of3~-acetoxy-5~-azido-6~-chlorocholestane (1.50, g) m.p.121-122", [a], -448" (lit.,3 m.p. 120-122", [a], -50"),in methanol (250 ml) containing potassium hydroxide(2.0 g) was refluxed for 30 min (the solid had dissolved after5 min). Water was added and the product was filtered off.Recrystallization from methanol afforded needles (1.30 g,95Oj,), 1n.p. 130-142", [a], -22" (C 0.78), vmx (eel,)3 620, 2 105, 1 165, 1040, and 665 cni-l, 6 0.70 (s, lS-H,),1.28 (s, 19-H,), 3.98 (m, W',~z ca.24 Hz, H-Sa), and 4.1213 P. J . Krueger and H. D. hlcttee, Canad. J. Chcm., 1965, 43,29701976 1041(m, W1ls 7.0 Hz, H-64, m/e 4651463 (0.1/0.2%, M ) , 442(4),and 382(15) (Found: C, 69.5; H, 9.7; C1, 7.55; N, 9.05.C2,H4,C1N,0 requires C, 69.85; H, 10.0; C1, 7.65; N,9.05%).5a-Azido-6~-chlmochoZest-3-one (11) .-TO a suspension of(I) (0.60 g) in acetic acid (20 ml) a t 50°C was added asolution of sodium dichromate dihydrate (0.4 g) in aceticacid (5 ml) . The mixture was stirred at 50 "C for 30 min(the solid had dissolved after 5 min). Water was added andthe solution was cooled in an ice-bath. The product wasfiltered off and the solid was washed with water.Re-crystallization from methanol afforded plates (0.53 g, 88%),m.p. 143-145", [a], -25' (c 0.39), vmBx (CCl,) 2 100,1725, 910, 665, and 650 cm-1, 6 0.72 (s, 18-H,), 1.45 (s,19-H,), and 4.10 (m, IY1/2 6.8 Hz, H-6a), mfe 442(26%),420(15), 418(39), 400(21), 398(50), 384(70), and 382(100).5%-A zido-6~-chlurocholestan-3a-ol (111) . T h e ketone (11)(500 mg) in propan-2-01 (100 nil) was treated with a largeexcess of sodium borohydride, and the solution was left atroom temperature for 2 h. Work-up as usual1 and re-peated t .l .c. separations afforded 5a-azido-6~-chZorocholestan-3a-ol(III) (250 mg, 50%). Recrystallization from methanolgave amorphous powder, m.p. 49-54', [a], -12.5" (c1.56), vmaX. (CCl,) 3 625, 3 600sh, 2 105, 1 170, 930, 918, and655 cm-1, 8 0.72 (s, lS-H,), 1.23 (s, 19-H,), and 4.12 (m,Wl,2 7.2 Hz, H-3P and H-Ga), m/e 437/435 (8/21%, M- N2), 422(2), 420(5), 400(100), 384(31), and 382(26)(Found: C, 69.65; H, 10.1; C1, 7.15; N, 8.95.C2,H4&lN,0requires C, 69.85; H, 10.0; C1, 7.65; N, 9.05%). Alsoseparated was 5a-azido-6~-chlorocholestan-3~-ol (I) (165mg, 330/,), m.p. 140-142" (from methanol), identical withauthentic (I) (mixed m.p., [a],, i.r., and t.1.c.).Oxidation of the 3%-Alcohol (111) .-The steroid (111)(30 mg) in acetic acid (3 ml) was treated with sodiumdichromate dihydrate (30 mg) and the solution was leftat 50 "C for 30 min. Work-up as usual and recrystallizationfrom methanol afforded plates (21 mg, 70%), m.p. 142-144",[u], - 25" (G 0.45), identical with authentic 5a-azido-6P-chlorocholest-3-one (11) (mixed m.p., ix., and t.1.c.).5a, 6a-IminochoZestan-3~-ol (IV) .-5a-Azido-6P-chloro-cholestan-3P-01 (I) (500 mg) in anhydrous bis-(Z-methoxy-ethyl) ether (60 ml) was treated with a large excess oflithium aluminium hydride and the mixture was stirred for2 h at 100°C.The excess of the reducing agent wasdestroyed according to the procedure of Steinhardt .14The granular precipitate was filtered off and dried invacuo. It was washed with acetone (20 ml) and then withmethylene chloride (20 ml) . Washings were repeated, asabove, until no more organic material was eluted. Thecombined organic solutions were evaporated under reducedpressure. Almost pure (IV) was obtained (390 mg, 9Oy0),m.p. 202-205".Recrystallization from acetone affordedneedles, m.p. 210-212", [a], -50" (c 0.41) (lit.,, m.p.210-215", [a], -49.9"), v,, (CCl,) 3 630, 3 260 w, vbr, and1032 cni-l, vms. (Nujol) 3 340 and 3 265 cm-l, 6 0.61 (s,18-H3), 1.06 (s, 19-H,), and 3.75 (m, W112 23 Hz, H-34,mfe 401 (53y0, M ) , 386(100), 372(26), 368(19), 358(16),344(16), 330(11), and 316(19).5a, 6a-Iminocholestan-3a-01 (V) .-A solution of 5a-azido-6~-chlorocholestan-3a-ol (111) (500 mg) in anhydrousbis-(2-methoxyethyl) ether (60 ml) was treated with a largeexcess of lithium aluminium hydride and the mixture wasstirred for 2.5 h at 85 "C. Work-up as above gave almostpure (V) (345 mg, 80%), m.p. 135-138". Recrystallization437/435 (2/7, M - N,), 422(2), 420(6), 400(100), 384(18),from acetone afforded needles, m.p.14%-144", [a],, -70"(c 0.43), vmx. (CCl,) 3 610, 3460, 3 320sh, and 1010 cm-1,vmax. (Nujol) 3 510 and 3 305 cm-l, 6 0.61 (s, 18-H,), 1.05(s, 19-H,), and4.09 (m, W,~z6.8Hz,H-3P),m/e401 (52%,lM),386(100), 384(31), 383(35), 372(11), 368(39), 358(11), 355(14),344(14), 330(39), and 316(11) (Found: C, 80.5; H, 11.85;N, 3.8. C,,H,,NO requires C, 80.75; H, 11.8; N, 3.5%).Attempt to Prepare the 3a-Alcohol (V) by Reduction ofthe Ketone (VI) .-To a solution of 50, 6a-iminocholestan-3p-01 (IV) (100 mg) in acetone (70 ml) at 0 "C was addedJones reagent (0.2 ml), and the solution was stirred a t0 "C for 10 min. Water was added and the product wasextracted with chloroform. The solution was washed withsodium hydrogen carbonate solution, dried (Na,SO,) , andevaporated under reduced pressure.A yellow oil wasobtained which showed vmX (CCl,) at 1720 cm-l and noabsorption at 3 630 cm-1. T.1.c. indicated the presence ofone major product. However, the pure ketone could notbe isolated either by recrystallization or by t.1.c. There-fore, the crude mixture in methanol (40 ml) was treated witha large excess of sodium borohydride, and the solution wasstirred at room temperature for 15 min. Work-up as usualgave a white solid. Its n.m.r. and i.r. spectra showed thepresence of both (IV) and (V). However, the two isomerscould not be separated by t.1.c. under various conditions.The n.m.r. spectrum of the mixture indicated that (IV) wasits major component.5a-Amino-6P-azidocholestan-3P-ol (VII) .-A solution of theaziridine (IV) (200 mg) in acetone (70 ml) and water (35ml) containing sodium azide (2.0 g) was refluxed for 192 h.Water was added and the product was extracted withchloroform.The solution was washed with water, dried(MgSO,) , and evaporated under reduced pressure. T.1.c.gave pure 5a-amim-6~-azidocholestan-3~-ol (VII) [ 155 mg,70%; the rest of the material was unchanged (IV)], m.p.146145" (needles from methanol-water), [a], - 46"(c 1.22), vwx. (CCl,) 3 620, 3400, 3 320, 2 095, and 1 040 cm-l,6 0.69 (s, lS-H,), 1.18 (s, 19-H,), 3.26 (m, W1/, 5.0 Hz,H-Ga), and 4.10 (m, W1/2 21.5 Hz, H-34, m / e 416 (12%,374(100), 368(20), 344(27), 330(31), and 316(32) (Found:C, 72.6; H, 11.0; N, 12.15.C,,H,,N,O requires C, 72.9;H, 10.9; N, 12.6%).5a-Arn~no-6~-az~do~holestan-3a-ol (VIII) .-A solution ofthe aziridine (V) (120 mg) in acetone (60 ml) and water(30 ml) containing sodium azide (1.2 g) was refluxed for96 h. Work-up as for the case of (VII) and t.1.c. gave pure5u-amino-6~-azidocholestan-3a-ol (VIII) [ 105 mg, 80% ; therest of the material was unchanged (V)], m.p. 187-189"(needles from acetone-water) , [a], -41°, vmx. (CCl,)3 360 vbr and 2 105 cm-l, 6 0.69 (s, 18-H,), 1.07 ( s , 19-H,),3.28 (m, Wl12 5.6 Hz, H-6a), and 4.00 (m, W112 8.8 Hz,H-SP), mfe 416 (18%, M - N,), 401 (69, M - HN,), 386(29), 384(20), 383(20), 374(100), and 344(20) (Found:C, 73.05; H, 10.85; N, 12.3%).Rate Measurements.-A sample (30 mg) of the 5a,6a-iminocholestane derivative was weighed into each of five orsix measuring flasks. A solution of sodium azide inacetone (B.D.H. AnalaR) and water (conductivity grade)(2 : 1 v/v) was added, and the stoppered measuring flaskswere introduced into a bath a t 50 f 0.1 "C, shaken tocomplete dissolution, and withdrawn, each after a pre-determined time. The acetone was evaporated off at rooml4 L. F. Fieser and M. Fieser, ' Reagents for Organic Synthesis,'Wiley, New York, 1967, p. 584.M - N,), 401 (31, M - HNJ, 386(38), 384(20), 383(19)1042 J.C.S. Perkin Itemperature under reduced pressure. Water (100 ml) was the ratio of the signal a t 6 0.61 (18-H, of the 5a,6cc-imino-added, and the products were extracted with methylene compound) to that a t 6 0.69 (18-H, of the 5a-amino-@-chloride (3 x 50 ml). The combined extracts were washed azido-compound) . A similar procedure was used forwith water (2 x 100 ml) and evaporated under reduced studying the reaction rates of the corresponding 5a,6a-pressure. An n.m.r. spectrum of the dry reaction mixture epoxycholestanes.was taken, and the reaction percentage was calculated from [5/1823 Received, 22nd September, 1975

ISSN:1472-7781    

DOI:10.1039/P19760001037

出版商:RSC    

年代:1976

数据来源: RSC