PART 1 AN ATTEMPTED WOLFF-KISHNER INITIATED FRAGMENTATION REACTION PART 11 LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT Thesis for TIN Degree of DILD. MICHIGAN STATE UNIVERSITY RonaId H. Starkey 1968 “I LIBRARY Michigan State University This is to certify that the thesis entitled Part I: AN ATTEMPTED WOLFF-KISHNER INITIATED FRAGMENTATION REACTION Part II: LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT presented by Ronald H. Starkey has been accepted towards fulfillment of the requirements for Ph. D. degree in Chemistry ”/4254; H, )CLWA Major professor Date September 4, 1968 0-169 ABSTRACT PART I AN ATTEMPTED WOLFF-KISHNER INITIATED FRAGMENTATION REACTION PART II LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT BY Ronald H. Starkey Attempts to effect a fragmentation reaction1 by means of a number of modifications of the Wolff-Kishner reaction on trans-5—pftoluenesulfonoxy-l-decalone (I) have failed. Reasons for expecting the fragmentation to give trans,trans- 1,6-cyclodecadiene are presented. The failure of the fragmentation to occur is due to a facile competing 1,3-elimination yielding tricyclic cyclo- propyl compounds (II). An explanation for the facility of the elimination compared to the desired fragmentation is discussed. I II Ronald H. Starkey Lithium aluminum hydride reduction of BS-pftoluenesul- fonoxycholest-S-en-4B-ol (III) has been reported to give three products; A, B, and C.2 Product A was subsequently identified correctly by another research group3 as cholest- 5-en—4a—ol (IV), B was not identified, and C was shown to be cholest-4-ene (VI). The present investigation confirms the formation of IV and VI; and identifies product B as 35—hydroxymethy1-A- norcholest-S-ene (V). A possible mechanism for the observed rearrangements is proposed and substantiated by the reaction of III with lithium aluminum deuteride, and preparation of a proposed intermediate cholest-S-en-4-one. ¢/[:;:j::E:I;\ llllhlller~ :!I[illll;\ lllllflllll)\ OH III OH v VI H IV TSO 0... REFERENCES Grob and P. W. Schiess, Angew. Chem. (Int. Ed.L, .A. 1 (1967). C Q, 2. P. Karrer, H. Asmis, K. Sareen, and R. Schwyzer, Helv. Chim. Acta., Q2, 1022 (1951). 3. D. N. Jones, J. R. Lewis, c. w. Shoppee, and G. Summers, J. Chem. Soc., 2876 (1955), PART I AN ATTEMPTED WOLFF-KISHNER INITIATED FRAGMENTATION REACTION PART II LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT By . K’q Ronald HI Starkey A THESIS Submitted to Michigan State University in partial fulfillment of the requirments for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1968 To Sandy ii ACKNOWLEDGMENTS The author expresses his appreciation to Professor William Reusch for his guidance and assistance during the course of this investigation. Appreciation is also extended to the National Institutes of Health for a pre—doctoral fellowship from September 1966 to September 1968, and to the Dow Chemical Company for a fellowship during the summer of 1966. iii TABLE OF CONTENTS PART I AN ATTEMPTED WOLFF-KISHNER INITIATED FRAGMENTATION REACTION HISTORICAL AND INTRODUCTION . RESULTS AND DISCUSSION . . . . EXPERIMENTAL .. . . . . . . . . General . . . . . . . . . trans-Decalin-1,5-dione . Reduction of trans-decalin—1,5—dione Tosylation of trans-5-syn-hydroxy-1-decalone (gas) Wolff-Kishner Reduction (Huang-Minlon Modifica- tion) of Ketotosylate £1 Reaction of Ketotosylate g1 with Base Wolff-Kishner Reduction (Lock Modification) of Ketotosylate QZ’ . . . . Wolff-Kishner Reduction (Cram Modification) of Ketotosylate S1, . . . . Reaction of trans-5-anti-hydroxy—1-decalone (35) with Phosphorous Tribromide . . . . . . . iv Page 15 26 26 27 27 28 29 30 3O 31 32 TABLE OF CONTENTS — Continued PART II LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT Page HISTORICAL AND INTRODUCTION . . . . . . . . . . . . 34 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 37 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 53 General . . . . . . . . . . . . . . . . . . . . 53 3B-proluenesulfonoxycholest-5—en—4B-ol (52) . . 54 Lithium Aluminum Hydride Reduction of 3fi§p¢ Toluenesulfonoxycholest-5-en-4fi-ol (22) . . . 54 Ascending Dry Column (ADC) Chromatography of the LAH Reduction Product . . . . . . . . . . 55 Compound A . . . . . . . . . . . . . . . . 56 Compound B . . . . . . . . . . . . . . . . 56 Compound C . . . . . . . . . . . . . . . . 57 Lithium Aluminum Deuteride Reduction of 36527 Toluenesulfonoxycholest-S-en-4B-ol (5g). . . . 57 Jones Oxidation of Cholest-5-en-4a-Ol (62) . . . 58 Reaction of Sodium Hydride with 3ijrToluene- sulfonoxycholest-5-en-4B-ol (52) . . . . . . . 59 Lithium Aluminum Hydride Reduction of Cholest-5— en-4-one (61) . . . . . . . . . . . . . . . . 59 FIGURES . . . . . . . . . . . . . . . . . . . . . . 61 REFERENCES . . . . . . . . . . . . . . . . . . . . . 7O LIST OF TABLES TABLE Page 1. Mass spectrum of tricyclo[4.4.0.01:5]decane (SQ)....................18 2. ADC chromatography of the LAH reduction product 55 vi LIST OF FIGURES FIGURE Page 1. Infrared spectrum of trans-5—anti-hydroxy- 1-decalone (66) . . . . . . . . . . . . . . 61 2. Infrared spectrum of trans-S-syn-hydroxy-l- decalone (66) . . . . . . . . . . . . . . . 61 3. Infrared spectrum of trans-S-synapftoluene— sulfonoxy-l-decalone (62) . . . . . . . . . 61 4. Infrared spectrum of tricyclo[4. 4. 0. O1 5]- decane (38) . . . . . . . . . . . . . . . 62 5. Infrared spectrum of tricyclo[4.4.0.01:5]- decan-Z—one (66) . . . . . . . . . . . . . 62 6. Nuclear magnetic resonance (nmr) spectrum of trans-5-anti-hydroxy1-1-decalone (66) . . . 63 7. Nmr Spectrum of trans-5-syn-hydroxy-1- decalone (66) . . . . . . . . . . . . . . . 63 8. Nmr spectrum of tricyclo[4.4.0.01'5]- decane (22) . . . . . . . . . . . . . . . . 64 9. Mass spectrum of tricyclo[4.4.0.01I5]- decane (22) . . . . . . . . . . . . . . . . 64 10. Infrared spectrum of cholest-5-en-4a-Ol (62). 65 11. Infrared spectrum of 36—hydroxymethyl-A- norcholest—5-ene (62) . . . . . . . . . . . 65 12. Infrared spectrum of cholest-4—ene (61) . . . 65 13. Nmr spectrum of cholest—5-en-4a-ol (62) . . . 66 14. Nmr spectrum of BB-hydroxymethyl-A-norcholest- 5-ene (62) . . . . . . . . . . . . . . . . 66 vii LIST OF FIGURES - Continued FIGURE Page 15. Nmr spectrum of cholest-4-ene (61) . . . . . 66 16. Nmr spectrum of 4B-deuterocholest-5-en-4a-ol (66) . . . . . . . . . . . . . . . . . . . 67 17. Nmr spectrum of 3S-hydroxydeuteromethyl—A- norcholest-S-ene (66) . . . . . . . . . . . 67 18. Nmr spectrum of 4,66—dideuterocholest-4-ene (66) 67 19. Nmr spectrum of 3B-hydroxymethyl-A-norcholest— 5-ene (66) . . . . . . . . . . . . . . . . 68 20. Nmr spectrum of 36-hydroxydeuteromethyl-A- norcholest—S-ene (66) . . . . . . . . . . . 68 21. Mass spectrum of cholest—5-en-4a-Ol (62) . . 69 22. Mass spectrum of 35-hydroxymethyl-A-norcholest- 5-ene (66) . . . . . . . . . . . . . . . . 69 23. Mass spectrum of cholest-4-ene (61) . . . . . 69 24. Mass spectrum of 46-deuterocholest-5-en-4a-ol (66) . . . . . . . . . . . . . . . . . . . 69 25. Mass spectrum of 36—hydroxydeuteromethyl-A- norcholest-S-ene (66) . . . . . . . . . . . 69 26. Mass spectrum of 4,66-dideuterocholest-4-ene (66) . . . . . . . . . . . . . . . . . . . 69 viii PART I AN ATTEMPTED WOLFFfKISHNER INITIATED FRAGMENTATION REACTION HISTORICAL AND INTRODUCTION Recently there has been widespread interest in medium- size ring compounds.1 Many new methods for the synthesis of these interesting and synthetically challenging struc- tures have been developed. Among the most promising syn- thetic approaches to medium-ring compounds are fragmentation (reactions. Fragmentation reactions have been known for quite some time, but only recently have been recognized as a distinct class of reactions.2 Recent reviews by Grobzl3 illustrate and classify the various types of fragmentations and show the great diversity of the reaction. In this context fragmentation refers to the cleavage of a molecule, symbolized by gfgfgf§:§, in the manner shown by equatiOn 1. e-E-s-Q-I >e=e+s=s+f ’ (1) In a heterolytic fragmentation reaction, which is the most common type, EJQJEJ and g_represent atoms, such as carbon, oxygen, and nitrogen. which are capable of forming multiple bonds.. i represents a leaving grouP such as haloe gen or tosylate, gfig iSfthe portion«offthe~m01ecule Which is 2 3 able to supply electrons, and in most systems 9 and Q are carbon atoms. Fragmentation reactions occur under a variety of con- ditions on a multitude of substrates. The fragmentation of 3-bromoalcohol 1 (equation 2), where the leaving group §.is bromide and the §f§_portions are the hydroxyl and carbinol methylene, takes place in 15% potassium hydroxide solution.4 Equation 3 illustrates an acid catalyzed fragmentation5 of HOWE? ——-9 OZCHZ + >:CH2+ Br (2) 1 ——’0=< +>=<+H20 (3) HO OH 2. Cl __.. CH2 : + Cl (4) {1' + N=CH2 CH3 1 l CH3 a 1,3-diol. Grob6 has shown that y-halo amines, such as 6 undergo thermal fragmentations. A commonly encountered fragmentation reaction, which is not generally thought of - - 7 as such, is the retro-aldol reaction (equation 5). OMs OMs . m . o- O 4 _ OH (6) WCOZH Fragmentation reactions have been used to open the rings of cyclic compounds8 (equation 6). When the ring Opening is performed on a properly substituted bicyclic system the fragmentation effects a ring expansion. This technique has been applied to substituted hydrindane systems by Corey9 in the synthesis of Carophyllene (6) (equation 7), and by Tanabe10 for the preparation of a 13,14-secosteroid (equation 8). Decalin derivatives have been used by 2CD Weston11 (equation 9) and Wharton12 (equation 10) to prepare unsaturated ten—membered rings. OMS Cb ——» OH 9 ~ 0 SS Br Br (10) Zn N 0.. Br 11 nBr ‘32 £6 The preferred geometry for effecting a fragmentation reaction is dependent on stereoelectronic factors3. The bonds participating in the fragmentation should be arranged so that overlap of the developing prorbitals will be at a This geometry is attained maximum in the transition state. in the anti conformation13 as shown in structure 16, Thus Grob14 asserts that the reason the 5-equatorial tosylate of N-methyl decahydroquinoline (16) fragments readily to give 16) while the corresponding axial tosylate 11'only undergoes elimination (equations 12 and 13), is due to the preferred 66£1_geometry of 16; This argument is weakened since the different mode of reaction of 11 may be due only to the fact that a more rapid Eggggfdiaxial B- elimination of the tosylate is now competing with the frag- mentation process. H C! 3 |CH3 :N + N / I s (12) // OTs 16 12, "“ 1m" CH3 .N H l O \______ N s (13) OTs ii 17 7 Wharton15 found that hydroxy tosylate 16'gave the fragmentation product 26’in greater than 90% yield, while under the same conditions the isomer 26) which cannot at- tain the preferred 22E; geometry, could be recovered un- changed. Even erythro-dibromohydrocinnamic acid (22), which H OTS H H HO ~ A (X) 12. OH 2:? 0 21 W does not have a fixed conformation, undergoes fragmentation in acetone exclusively via the anti transition state to give cis-bromostyrene (26).16 Br H C6H5 C6H5 23 Support for the maximum overlap principle comes from a study of the ultraviolet spectra of piperidones by Cookson.17 Piperidone 26) which probably has the nitrogen n-electron pairs oriented in the axial (endo) position, shows no intense absorption in the accessible ultraviolet region. A similar structure (26), in which the electron pairs are forced to be equatorial and therefore anti to 8 the 1,2-bond, has a strong absorption band at 262 mu (6 = 3600) which disappears in acid solution. 0 ‘ \x ‘\ CH —N \N-—CH3 \ 3 C) . C) @NVN® 24 62 m The 5~substituted~1~decalone 26 was chosen as a model for the synthesis of ten membered diene rings via a frag- mentation reaction. 0 Of) X 26 (I: In this rigid decalin ring system the leaving group X occupies an equatorial position and is fixed in the prefer- red gagi-orientation to the 9-10 bond. The equatorial ori- entation is also advantageous since the preferred trans- diaxial geometry for B-elimination of H—X cannot be attained; thus, any competitive elimination reactions would be mini- mized. A nucleophilic displacement of X- would also be quite unfavorable since the axial hydrogens on carbons 7 and 9 would make back—side attack on carbon 5 quite hindered. 9 The 1,5 arrangement requires that an anion be generated on carbon 1 to effect a fragmentation. This electron pair preferably should be in an equatorial orientation, as this would also make it EEEA to the 9-10 bond. The 1-ketone was chosen because ketones are usually easily prepared and a method for generating an anion on the carbonyl carbon appeared to be at hand. Wolff-Kishner re- ductions18 of ketones, according to the accepted mechanism,19 proceed Iii a hydrazone (21) which rearranges to a diimide anion structure (26) and then loses nitrogen to leave a carbanion (26) on the original carbonyl carbon (equation 15). R2C=O NHZNHZa R2C=N-NH2 fie 21 (15) .. .._ -N2 .. _ RZCH-N=N > RZCH 3.3.. 32. Under normal Wolff-Kishner conditions this intermediate anion would pick up a proton from the reaction medium and form the hydrocarbon. The anion formed from a Wolff—Kishner reaction on a ketone such as 26 has another alternative available to it—— it can initiate a fragmentation reaction (equation 16). The hydrogen transfer (21 -—¢ 26) would probably yield 22 since this places the diimide anion in the sterically less 10 8 X X 9 ' I _\ " 30 N :N- N—NHz rw H ‘N2 (16) 55“ H _ o 'X h H “©- 31 gg hindered equatorial position. Regardless of whether loss of nitrogen and fragmentation is concerted, the equatorial diimide anion would assure that the electron pair would be REE; to the 9-10 bond, and thus all steric requirements for fragmentation would be met. A fragmentation reaction in the course Of a Wolff- Kishner reduction of w-bromo camphor (66) has been observed20 (equation 17). F Br Br I re- "——’ (17) N=:N 33 /’ m In effect the Wolff-Kishner initiated fragmentation reaction is an "ethylagous" analog of the well known Kishner elimination21 (equation 18). 11 NHZNH2 6 -N2 | ( 18 ) \ O N-NHZ Leonard and Galfand22 have found that the greater the ability of the leaving group Y to accommodate a negative charge, the better the yield of Kishner elimination product. In a study of a-halo ketones Wharton, Bunny and Kreb523 found that the Kishner elimination product yields were in the order F > C1 > Br > I > OTs——the reverse of the normal order based on leaving group ability. This was attributed to a more facile competing B-elimination in these non-rigid systems. Consequently the a—fluoro—compounds did not under— go B-elimination, but remained intact until they could be eliminated by the Kishner route. In addition, the poorer leaving fluoro-compounds required higher temperatures to effect Kishner elimination. Generalizing from the Kishner elimination to its ethylog, the fragmentation reaction, it appears the best leaving group (X) would be a sulfonate derivative, 3,3, tosylate, mesylate, or a reactive halide such as iodide or bromide. This is borne out by a comparison of the rates of frag— mentation of a steroidal tosylate (66) and a chloride (66).24 The chloride reacted much slower in the fragmentation reac- tion with potassium Efbutoxide than did the tosylate. Thus 12 T30 OH 34 leaving group ability for fragmentation reactions seems to parallel that found for nucleophilic dISplacement reactions. The geometry of the double bonds formed in fragmenta— tion reactions can be deduced by mechanistic considerations. In the trans,trans-1,5-diequatorial substituted system (66), the geometry of the ring system demands that a concerted fragmentation lead to ££§g§,££ggg-1,6-cyclodecadiene (62). This geometrical stereospecificity is evident from the fragmentation of 16,and 66,15 Compound'16) in which the hydrogens on carbons 5 and 6 are 55223, gives the EEEBET Olefin, whereas when the same hydrogens are £6§_(66), the gigfolefin results. In our model system both sets of hydrogens which would become the vinyl hydrogens (hydrogens on carbons 1, 9, 10 and 5 of structure 66) are ££6£§_and therefore the expected product would have two ££66§_double bonds in a ten membered ring. Models indicated that 1,6—trans,trans double bonds can be accommOdated in a cyclodecane ring and that it can attain an essentially strain free conformation, and there- fore should be quite stable. In fact a trans,trans-1,6- cyclodecadiene is postulated25 as an intermediate in the 13 H OTs H0 .132. o H OH TsO H 36 biogenesis of eremophilone (66) from trans-farnesol (61). /’ OH 37 38 A substituted trans,trans-1,5-cyclodecadiene structure 1a,c occurs in nature and has recently been synthesized by Corey26 and Marshall.27 Qggdgig-l,6-cyclodecadiene has been prepared and is quite stable.28o29:3° During the course of the present investigation trans, trans-1,6—cyclodecadiene itself was prepared by Heinback31 using a photolytic isomerization (equation 19) of cis,trans- 1,5-cyclodecadiene (66), and Marshall and Bundy30 prepared 14 a methyl substituted trans,trans-l,6-cyclodecadiene (61) by means of a fragmentation reaction (equation 20). by 19 Fe(co)5 ( ) 39 62 OMs OMs J ____., 35—. (:3 (20) 40 E] H 41 ’W \-—' B\b\ m \/ 0—H RESULTS AND DISCUS S ION Reduction Of Eggggfdecalin—l,5-dione32 with one equi- valent of lithium tri(§;butoxy)aluminum hydride33 in tetra- hydrofuran (THF), gave a mixture of ketoalcohols 66 and 66, Preparative thin layer chromatography34 (prep-tlc) afforded pure 66 in 23% yield and pure 66 in 52% yield. Li[(CH3)3c013AIHb THF . . (a) fi (9) : H HO -' Ed 36 These two epimeric alcohols were differentiated by their infrared and nuclear magnetic resonance (nmr) spectra. The hydroxyl group of 66 displayed an oxygen-hydrogen (0-H) stretch at 3620 cm.1 and a carbon-oxygen (C-O) stretch at 985 cm_1, while epimer 66 showed 0-H and C-0 stretches at 3600 and 1045 cm_1 respectively. The hydroxyl group of 66' was assigned an axial configuration and that of 66 an equatorial, since the 0-H stretch of axial alcohols in six- membered rings is known to occur at higher frequencies than that of corresponding equatorial alcohols, and C-0 stretching 15 16 frequencies are observed at 1040 cm”1 for equatorial alcohols and around 1000 cm.1 for their axial epimers.35 The nmr spectrum of 66’shows a fairly sharp singlet for the carbinol hydrogen (line width36 6 cps) at 6 3.97. while the carbinol hydrogen of 66 is a broad unresolved multiplet (line width 22 cps) at O 3.40. This supports the assignment of figggg-Sjgggif-hydroxy-1-decalone to 66’ and Egggg-Sfigygf-hydroxy-1-decalone to 66/ because a car- binol hydrogen that is axial is known36 to absorb at higher field than the correSponding epimeric equatorial hydrogen. It is also well established38 that vicinal axial-equatorial OH 36 hydrogen Splitting is much less (J = 1-5 cps) than the split— ting between axial-axial hydrogens (J = 8-14 cps). Thus the line width of the axial carbinol hydrogen signal of 66) which is subject to vicinal axial-axial splitting, is broader than the signal from the corresponding equatorial hydrogen of 66. * The configuration of the hydroxyl is described as syn or anti with respect to the nearest bridgehead hydrogen.37 17 Tosylation of ketol 3Q with p-toluenesulfonyl chloride in pyridine gave the equatorial tosylate §Z'(trans-5-syn*- pftoluenesulfonoxy—l-decalone) in 83% yield. The nmr spec— trum confirmed the equatorial orientation of the tosylate m 0 H O ___, W (22) ; OTs H0' 36 37 rw by the presence of a one hydrogen multiplet at 5 4.4 (line width 22 cps) due to the axial carbinol hydrogen. Mild conditions were employed in the first attempt to effect a fragmentation reaction. The Wharton reaction,39 a special type of Kishner elimination, where the leaving group is epoxide oxygen, proceeds using hydrazine hydrate and potassium acetate in boiling ethanol with no added strong base. Under the conditions of the Wharton reaction39 ketotosylate 31 was recovered unchanged. The next attempt to produce fragmentation was by the more vigorous conditions of the Huang—Minlon modification of the Wolff-Kishner reduction. A solution of tosylate 31' in diethylene glycol (DEG) containing hydrazine hydrate and potassium hydroxide was heated to a temperature of 200°. A volatile material isolated from the reaction in 30% yield and purified by glc was identified as tricyclo[4.4.0.01'5]- .decane (3g). 18 Huang-Minlon_y (23) Modification )‘ 37 38 O --- '8 U} The structure proof of 38 is based on chemical and spectroscopic evidence. A positive tetranitromethane test40 indicated 38’contained either olefinic unsaturation or a cyclopropane ring. The infrared spectrum showed a cyclo- propyl C-H stretch at 3000 cm-1 and cyclopropyl skeletal vibration41 at 1020 cm_1. Nmr revealed two cyclopropyl hydrogens at high field (6 0.69-0.77), a methylene envelope (6 1.0-2.0) containing fourteen hydrogens, but no vinyl hydrogen signals. The mass spectrum confirmed the molecular weight as 136, and examination of the intensity of the M + 1 (molecu— lar ion + 1) and M + 2 peaks relative to the molecular ion peak (Table 1) indicated42 a molecular formula of C10H16. Table 1. Mass spectrum of tricyclo[4.4.0.01I5]decane (38) m/e Experimental Calculated43 %ofM %ofM M. 136 100.0 100.0 M + 1 137 11.4 11.06 M + 2 138 1.0 0.55 19 Apparently under the basic conditions used, formation of a carbanion on carbon 9 and displacement of the tosylate by this carbanion occurs much faster than the Wolff—Kishner reduction of the ketone (equation 24). Because of its prox— imity to carbon 5, an electron pair on C-9 is able to dis- place the tosylate group and give the cyclopropyl-ketone 42’ (24) “:5 Mt ” 3~9v I'w O W lff- 0 ”12‘2“ Kishner :22. , ketone 49 could sub- Sm' Since hydrazine is present in excess sequently undergo reduction by way of the normal Wolff- Kishner reaction (equation 25) to give the hydrocarbon (38) Abstraction of an a—hydrogen under Huang-Minlon condi- tions has been documented by Djerassi, Grossnickle and High44 in the reduction of epimeric digitonins Both 41' and 42 gave a single reduction product 43; 20 Support for the intermediacy of tricyclo[4.4.0.01'5]— decane—2-one (42) during the reduction of 31 was provided by its synthesis from 31 on treatment with potassium hydrox- ide in refluxing DEG (equation 24). Compound 42 showed infrared cyclopropyl absorption at 3000 cm-1, a cyclopropyl conjugated carbonyl at 1660 cm_1, and an ultraviolet absorp- tion maximum at 219 mu (E = 103). This maximum is typical of cyclOpropyl ketones45 and is reasonably close to the value of 212 mu calculated for 42 by the method of Dauben and Berezin.46 The mass spectrum of gg'indicated a molecular weight of 150 and the molecular formula42 C10H14O. The Lock modification of the Wolff-Kishner reduction is reported to be useful for base sensitive carbonyl com- pounds.47 In this method the hydrazone is first prepared in the absence of strong base. The Wolff-Kishner reduction is then effected by heating with base. This method appeared promising for the system under investigation, since it should retard 1,3-elimination of tosylate. This conclusion is based on the assumption that hydrogens a to a hydrazone function are less acidic than those a to a carbonyl group, 21 since nitrogen is less electronegative than oxygen. Also, the two N—hydrogens of the hydrazone should be quite acidic and would probably be abstracted in preference to the a- hydrogens, decreasing even further the hydrazone function's ability to stabilize an q—negative charge. The N—hydrogen abstraction is also necessary in the first step of the Wolff— Kishner reduction.19 In contrast to these arguments, there is a report of epimerization during Wolff-Kishner reduction of a similar system.48 When semicarbazone 44 was reduced under Huang- Minlon conditions, it gave a mixture of gé’and 46 in a 45:55 ratio. However, the possibility that 44 was a '———-> + (26) H 111 u. :1: N \NH-CONHZ 45 46 $4. molecular compound composed of both the gg§_and trans epi— mers cannot be ruled out, since the only criterion given for its purity was a melting point which did not change on recrystallization. The equilibrium composition of the ketone from which gg’is derived is 40% gig and 60% trans, The Lock modification was performed by warming §Z’with' 95% hydrazine in DEG, followed by addition of potassium hydroxide in DEG and slow heating to 2400 (equation 27). The tricyclic hydrocarbon 38 was the only material which 22 could be isolated from the distillate. The product was isolated by means of gas chromatography and identified by a comparison of its infrared spectrum with tricyclo- [4.4.0.01'5]decane from the Huang—Minlon modification of the Wolfanishner reduction. Lock . . .t’ (27) Modification st 37 38 NV Again a—hydrogen abstraction and 1,3-elimination is the mode of reaction. Thus it appears that the rate of abstraction and subsequent displacement, even in the hydra- zone, is much greater than the rate of Wolff-Kishner re- duction. The low temperature Cram modification49 of the Wolff- Kishner reduction was attempted next. The dimethyl sulfox- ide (DMSO) used as the solvent in this modification in- creases the reactivity of the base and therefore should facilitate both the Wolff—Kishner and hydrogen abstraction reactions. However, the use of Efbutoxide, since it is a bulkier base than hydroxide, should make abstraction of the tertiary a-hydrogen more difficult. A fragmentation initi— ated by the Cram modification has been reported20 (equation 17). 23 The product from slow addition of the crude hydrazone (41) in DMSO, to a DMSO solution of potassium tfbutoxide was isolated and purified by distillation and gas chroma- tography and identified as tricyclo[4.4.0.01r5]decan-2-one (42) by comparison of its infrared and ultraviolet spectra with that of 42 obtained from base treatment of ketotosylate 31. The reaction apparently proceeds via 1,3—elimination N-NHz N-NH2 0 (28) Cram Modification) I J J}. 2.9. of tosylate from hydrazone 41 to give the tricyclic hydra- zone 48/ which is then transformed to 49 during the work-up. Finally, an attempt was made to prepare trans-5-gyn— bromo-l—decalone, since bromide is a poorer leaving group than tosylate and might therefore be less reactive in the 1,3-elimination, but still adequate as a leaving group in the fragmentation. The example of Gustafson and Erman20 shows that bromide can serve as the leaving group in a Wolff-Kishner initiated fragmentation reaction (equation 17). Preparation of alkyl bromides by treatment of alcohols with phosphorous tribromide in pyridine proceeds mainly with inversion of configuration at the carbinol carbon.50 On this basis a bromide prepared from trans-S-EQEifhydroxy- l-decalone (35) should yield a decalone with a bromine in 24 the equatorial position (42). Since the intermediate phos- m--. CH ‘ git-2’ Br phite ester (22) is in the axial position and is well ori— ented for an anti elimination, attempts to prepare the hmomide have led only to 1,2-elimination. P/‘ o H HI 50 51 Treatment of 22 with phosphorous tribromide in pyridine gave A9-octalone-1' (24) in good yield. The assignment of structure 2; to this product is based on its infrared spectrum.51 A 1,2—diaxial elimination of the phosphite ester and isomerization of the double bond to a position conjugated with the carbonyl would give rise to the observed product. The tendency of axial alcohols to undergo elimination rather than substitution is nicely illustrated by an at- tempt to prepare the 7—chlorides from the corresponding sterols.52 The reaction of phosphorus oxychloride with the axial alcohol (52) gave only elimination, whereas the 25 equatorial alcohol (22) gave predominantly substitution. ——————> ‘ ~ \ ACO 0H 52 N \ ~‘——. AcO OH ~\ AcO Cl 53 NV 1 1 Attempts to effect a fragmentation reaction by means of a number of modifications of the Wolff-Kishner reaction on Egagg-5-pftoluenesulphonoxy-l-decalone (22) have failed. The failure of the fragmentation to occur is due to a facile competing 1,3—elimination yielding tricyclic cyclopropyl compounds. EXPERIMENTAL General. Infrared spectra were recorded on a Perkin— Elmer 237 B grating spectrophotometer, using sodium chloride cells. A Unicam Sp.800 spectrophotometer was used for ultraviolet spectra. Nuclear magnetic resonance spectra were taken on Varian A-60 and HA-100 high resolution spec— trometers, using tetramethylsilane as an internal standard. A Consolidated Electrodynamics Corp. Type 21-103c and an LKB Type 9000 mass spectrometer were used to obtain the mass spectra. Gas chromatography was performed on an Aerograph A-90-P3 with a thermal conductivity detector, using helium as the carrier gas. A 6% carbowax 20M on 80/100 Aeropack 30 column was employed. Preparative thin layer chroma- tography was done on 1.5 mm layers of silica gel PF254 on 20 x 20 cm glass plates. Melting points were determined on a Koefler hot stage, and are uncorrected. Analyses were performed by Spang Microanalytical Labs, Ann Arbor, Michigan. ed sodium chloride solution. Brine refers to a saturat All materials were checked for purity by analytical thin layer chromatography on silica gel G. 26 27 In all runs of the Wolff—Kishner reduction no isolable product could be obtained by extraction of the residue from the distillation of the crude products. trans-Decalin—1,5-dione. A 20 g sample of decalin-1,5— —‘__ diol53 was oxidized according to the procedure of Johnson, Gutsche and Banerjee32 yielding 12 g (60%) of trans—decalin- 1,5—dione: mp 165—1660 [lit32 mp 166°]; ir (cnc13) 1715 cm"1 (c=o). Reduction of trans-Decalin-1,5—dion§. To a stirred solution of 12.0 g (72.5 mM) of Erangfdecalin—1,5—dione in 1.2 l of tetrahydrofuran (distilled from LiAlH4), which 'was maintained at 0° under a nitrogen atmosphere, 23.0 g (90 mM) of lithium tri(£7butoxy) aluminum hydride (Alfa Inorganics) in dry tetrahydrofuran (THF) was added dropwise. After the addition was completed, the solution was allowed to stand at room temperature for 12 hours. Addition of one equivalent of water and evaporation of the THF gave a light yellow solid, which was dissolved in a mixture of ether and 5% aqueous hydrochloric acid. The separated ether layer was washed with water and brine, and then dried over sodium sulfate. Evaporation of the ether gave 9.0 g (75%) of tan crystalline crude product. Separation of the two alcohols from a 2.0 g sample of the crude product was accomplished by preparative thin layer chromatography34 (prep-tlc) on two 20 x 20 cm glass plates coated with 1.5 mm layer of silica gel PF254 (E. Merck). 28 Development with anhydrous ether and ultraviolet visualiza- tion showed the two alcohols as bands of Rf 0.5 and Rf 0.4. Elution of the Rf 0.5 band with ether gave 45.6 mg (23% of crude) of the axial ketol 22; Crystallization from hexane-benzene gave white crystals: mp 94-950; ir (CHClg) 3620 (o-H), 1710 (c=o), and 985 cm’1 (C-OH). [Figure 1]; nmr (CDC13) 5 3.97 (s, 1H, line width36 6 cps, carbinol hydrogen), 2.7-1.2 (m, 15H, methylene and methine hydrogens with hydroxyl at 6 2.2), [Figure 6]. Elution of the R 0.4 band with ether gave 103.8 mg f (52% of crude) of the equatorial ketol 22; Crystallization from benzene gave white crystals: mp 118-1190; ir (CHCla) 3600 (O-H), 1710 (c=o) and 1045 cm—1 (C—OH) [Figure 2]; nmr (CDC13) 5 3.4 (m, 1H, line width 22 cps, carbinol hydro- gen), 2.8 (s, 1H, O-H), 2.6-1.0 (m, 14, methylene and methine hydrogens) [Figure 7]. Angl, Calcd for C10H1602: C, 71.39; H, 9.59. Found: C, 71.39; H, 9.44. Tosylation of trans-5-syn-hydroxy-1—decalone (22). To a stirred and cooled (0°) solution of 1.0 g (6.0 mm) of 22 in 5 ml of pyridine, 1.6 g (8.4 mM) of solid pftoluene— sulfonyl chloride was added in small portions. After the solution was stirred for 12 hrs at 25°, the pyridine was evaporated on a rotary evaporator and the resulting amor- phous solid was dissolved in ether. The ether solution was washed twice with water, twice with 5% hydrochloric acid, 29 and then with brine, followed by drying over sodium sulfate. Evaporation of the ether and crystallization of the crude solid from ether, gave 1.6 g (83%) of the equatorial keto— tosylate 22; mp 94-950; ir (CHC13) 1715 (c=o), and 1175 cm-1 (tosylate) [Figure 3]; nmr (CDCl3) 5 2.45 (s, 3H, CH3Ar), 4.4 (m, 1H, line width 22 cps, CHfOTs), 7.6 (AB— quartet, 4H, g—Ar) . 522$: Calcd for C17H2204S: C, 63.33; H, 6.88; S, 9.95. Found: C, 63.58; H, 6.93; S, 9.90. Wolff—Kishner Reduction (Huanquinlon Modification) of Ketotosylate 21, To a solution of 400 mg (7.1 mM) of potas— sium hydroxide dissolved54 in 5 ml of diethylene glycol (DEG), 0.3 ml (6.2 mM) of hydrazine hydrate (99%) and 388 mg (1.2 mM) of ketotosylate g1 were added. The distillate, collected while the mixture was heated from 25° to 220° (over a 2 hr period) and then maintained at 2200 for 4 hrs, was dissolved in ether and washed with water and with brine. Evaporation of the ether gave 50 mg (30%) of tri- cyclo[4.4.0.01'5]decane (22), which was collected by and shown to be nearly pure (> 98%) by glc (6% carbowax at 125°). The glc-collected product gave a positive tetranitromethane reaction4° (bright yellow) and was characterized spectro— scopically: ir (neat) 3000 (cycloprOpyl C-H), 2975, 2925, 2850, 1445 and 1020 cm_1 (cyclopropyl) [Figure 4]; nmr, IOOMC (CDClQ 50.69 (m, J = 2 cps), 0.77 (5) combined rela- tive area of 0.69 and 0.77 signals = 1, 1.0-2.0 (many 30 signals) combined relative area - 7 [Figure 8]; mass spec- trum (70 eV) gig (rel. intensity) 138 (1.0), 137 (11.4), 136 (100) [Figure 9]. Reaction of Ketotosylate 21 with Base. Ketotosylate (21) (200 mg, 0.62 mM) was added slowly to a solution of 200 mg (3.5 mM) of potassium hydroxide in 3 ml of DEG at 25°. The resulting mixture was warmed slowly to reflux temperature and then allowed to reflux for 2 hrs. The solu— tion was cooled and poured into an equal volume of water and extracted twice with ether. The ether extract was washed with water, brine and then dried. Evaporation of the ether gave 77 mg of a yellow oil, from which could be isolated by prep—tlc (silica gel PF254, developed with ether— chloroform [3:2], Rf 0.67) pure tricyclo[4.4.0.01'5]decan— 2-one4°: ir (CHC13) 3000, 2950, 2875, 1660, 1450 and 1370 cm-1 [Figure 5]; uv max (cyclohexane) 219 mu (6 - 103); nmr (CDC13) 5 1.0—2.5 (m); mass spectrum (70 eV) gig (rel. intensity) 152 (0.78), 151 (11.8), 150 (100). Wolff-Kishner Reduction (Lock Modification)47 of Ketotosylate 37. Ketotosylate §Z_(322 mg, 1.0 mM) was dis- solved in 6 ml of DEG. After 1.0 ml (30 mM) of 95% hydra- zine was added the solution was warmed slowly to 170°. The solution was cooled and 500 mg (9 mM) of potassium hydroxide in 4 ml of DEG was added. The volatile material was dis- tilled as the solution was heated slowly to 240° (required 31 2 hrs). The distillate was poured into water and the water was extracted with pentane. The pentane solution was washed with water, brine, dried over sodium sulfate and the pentane evaporated. A substance which had the same glc retention time and an almost identical infrared spectrum as that of tricyclo[4.4.0.01I5]decane (22) was the only material which could be isolated from the residue by gas chromatography. Wolff-Kishner Reduction_(Cram Modificatiog)59 of Keto- tosylate 37. Preparation of the hydrazone of 21 was accomp- lished by the procedure of Gustafson and Erman5° Its infra- red spectrum confirmed the crude product as a hydrazone:55 ir (CHC13) 3300 and 1665 cm—1. Tlc showed that no 21 was present in the product. The crude hydrazone of 21 (500 mg, 1.5 mM) in 4 ml of dimethyl sulfoxide (DMSO), was added dr0pwise over a 6 hr period to a solution of 336 mg (3 mM) of sublimed potassium Efbutoxide in 3 ml of dry DMSO. The solutions were protected from moisture and the reaction mixture was continuously swept with nitrogen. After being stirred overnight at room temperature and then heated at 90° for 4 hrs the mixture was cooled, poured into 75 ml of ice water, acidified with 5% hydrochloric acid, water, brine and dried over sodium sulfate. Evaporation of the ether and distillation of the resulting brown oil up to a temperature of 130° (1.5 mm Hg) gave 80 mg of a yellow liquid (38%). Gas chromatography on 6% carbowax at 180° showed a nearly pure major product, 32 which was identified as tricyclo[4.4.0.01'5]decane-2—one (42), by comparison of its infrared and ultraviolet Spec- trum with that of 42 from base treatment of the ketotosylate 21, Reaction of grgng—5—antithydroxy71-decalone (35) with PhOSphorous Tribromide. To a flask fitted with a drying tube and containing 0.2 ml (2 mM) of pohsphorous tribromide in 7 ml of dry pyridine, 500 mg (3 mM) of 22 was added. The solution was stirred for 4 hrs at 0°, 1 hr at 25° and then maintained at 100° for 0.5 hrs. After standing for 12 hrs at room temperature, water was added and the reaction mixture extracted with ether. The ether extract was washed twice with water, twice with 5% hydrochloric acid, once with brine and dried with sodium sulfate. Glc of the residue from evaporation of the ether, on 6% carbowax at 165° af- forded 220 mg (52%) of A9—octal-1-one51 (24): ir (CC14) 1665 and 1625 cm‘l. PART II LITHIUM ALUMINUM HYDRIDE REDUCTION OF A STEROID TOSYLATE. A NOVEL REARRANGEMENT 33 HISTORICAL AND INTRODUCTION In 1951 Karrer, et al5.6 reported the results of a lithium aluminum hydride reduction of tosylate 24. Three compounds, denoted by A, B, and C, were obtained after numerous column chromatographic separations. C8H17 +B+C (29) Compound A, obtained in about 20% yield, was assigned the structure cholest-5-en-4 B—ol (22) on the basis of catalytic hydrogenation to a saturated alcohol, melting point 189°, which was thought to be 5a-cholestan-4B-ol (22). x \ Oij/\ 57 OH .5555 OH 851 - Compound C, obtained in only 2% yield, was identified as cholest-4-ene (57) by a mixture melting point with authen- tic material. 34 35 Compound B (ca. 4% yield) was not identified, but was characterized by a melting point of 98° and a specific ro— tation ([a]$3) of -27.1°. Karrer apparently assumed that simple displacement of the tosylate group by hydride was responsible for the major product, 22’. In an earlier study Schmidt and Karrer57 found simple displacement accompanied by i—steroid (22) formation in the reaction of lithium aluminum hydride with cholesterol tosylate (22) (equation 30). \ \ r v T50 58 22’ 22' NV In 1951 Barton and Rosenfelder58 reported the prepara- tion of both cholestan-4B-ol (22) and cholestan—4a-ol (22). They found that the physical properties of the compound as- sumed to be cholestan—4B-ol by Karrer, §E_§l.55 actually corresponded to the 4a—hydroxy epimer 22. 36 Shoppee and co-workers59 in 1955 prepared both the a(22) and 2(22) epimers of 4-hydroxycholest—5-ene, and demon- strated that the a epimer (22) was identical with the pro— duct of the lithium aluminum hydride reduction of tosylate 24. The structural assignments of 22 and 22 were later con- firmed by Becker and Wallis.°° Thus, it appears that the 4Q-hydroxy group of tosylate 24 has been epimerized during the reaction with the metal hydride. A structure for Karrer's compound B has never been pro— posed and it is difficult, on mechanistic grounds, to see how olefin 21 (Karrer's compound C) can be formed in the reaction. A reinvestigation of the reaction of 3fifipftoluene— sulfonoxycholest-5-en-4 B-ol (24) with lithium aluminum hydride was undertaken because: a) the structure of Karrer's compound B remained unknown; b) the epimerization of an alcohol by lithium aluminum hydride is unprecedented and, therefore the mechanism of the formation of A (22) is not obvious; c) the formation of compound C (21) is unusual and should be confirmed. RESULTS AND DISCUSSION Reduction of BB—Eftoluenesulfonoxycholest-5- en-4fi—ol (24) by lithium aluminum hydride (LAH) in refluxing ether gave three products which could be separated very effectively by ascending dry column (ADC) chromatography on neutral alumina. C8H17 TSO 0 0H 54 0H 9.3. ~ (31) X '\ .Q + Q. @1on £53 2?. The major product, isolated in 48% yield, was cholest- 5-en-r4a-ol (22). The assignment of structure 22 to this product is consistent with the infrared and nmr spectra and r the physical properties reported for the authentic material. Compound 22) mp (hexane) 141-1420, exhibited a Specific rotation of [a];5 -53°, and appears to be identical to Karrer's compound A (mp 1440: [0157.5 '550) and authentic 37 38 SEE, prepared by Sheppee. gt_al_.59 (mp 144°, [0‘15 -500) and Becker and Wallis°° (mp 143°, [a]D -54°). The nmr spectrum of 22 displayed a carbinol hydrogen signal of line width 21 cps36 at 5 4.25. This hydrogen should be axial and thus a fairly broad signal is expected, since the dihedral angle favors a large Spin-Spin coupling.38 The infrared spectrum showed hydroxyl O-H stretching at 3600 cm_1 and C-0 stretch— ing at 1070 cm_1. This C—O stretching frequency is char- acteristic of an equatorial hydroxyl group.35 The mass spectrum of 22’displayed a molecular ion at m/e 386 and was consistent with a molecular formula42 of C27H460. AS ex— pected, abundant fragments were observed at m/e 371 (M-CH3) and 368 (M-HZO). ’ Karrer's compound C, obtained in 3% yield by prepara- tive thin layer chromatography of the least polar fraction from ADC chromatography, was identified as cholest-4—ene (21) by comparison of its physical properties, melting point 80—81°, [a]D +67° with reported values. (Broome and co- workers61 report a melting point of 80-81° and [a]D +67O for cholest-4—ene). The nmr Spectrum shows a vinyl hydrogen signal at 5 5.25 and the C-19 methyl at 5 1.00. The value calculated62 for the C-19 methyl of a A4—steroid iS 591-025- The mass Spectrum exhibited42 a molecular ion at m/e 370. and indicated a molecular formula C27H46- The third product, obtained directly from the ADC Chromatography in 33%3yield, was crystallized from methanol, mp 99°, [a];5 -24.4°. These physical constants are 39 essentially identical to those reported by Karrer56 for compound B. The structure 35-hydroxymethyl-A—norcholest—5— ene (22) was assigned to this compound on the basis of its spectral properties and a comparison with authentic 22, The ir spectrum of 22 diSplayed hydroxyl O-H stretching at 3600 cm_1 and C-0 stretching at 1020 cm-1, as expected for a primary alcohol. The mass Spectrum, molecular ion at m/e 386 and prominent fragments at m/e 371 (M-CH3) and 355 (M-CHZOH), supported the molecular formula C27H460. The nmr spectrum [Figures 17 and 19] showed a vinyl hydrogen of rela— tive area 1 at 5 5.5, a multiplet of area 1 at 5 2.7, and a "doublet" with an area corresponding to two hydrogens cen- tered at 5 3.62 (J = 7.2 cps). This is a deceptively Simple pattern for the carbinol hydrogens, which should be the AB portion of an ABX system, Since they are adjacent to an asymmetric center (carbon—3, 22a) and are therefore magnetically non—equivalent. Evi- dently the chemical shifts of HA and HB are nearly equal and the qstem is essentially of the AA'X type. An AA'X system63 would account for the observed carbinol hydrogen 40 doublet and the broad multiplet for the hydrogen on C-3 at 5 2.7. An example of an AA'X system is found in the nmr spectrum of 2—furfurol (24), which displays a doublet for hydrogens HA and HA, and a triplet for HX.64 Finally Whitham and Wickramasinghe65 have prepared 22) by an independent route and report a melting point of 99- 101° and [a]D -26° (9 1.56, CHC13). Comparison of an infrared Spectrum of 22 obtained from the reduction of 24 with that of a sample of 35—hydroxymethyl—A-norcholest-5— ene, kindly supplied by Prof. G. Whitham, shows the two to be identical. It should also be mentioned that the 3g- hydroxymethyl epimer of 22 has been prepared by Dauben and Ross°°; this epimer is reported to have a melting point of 102-1030 and [5135 -45°. Direct crystallization of the crude lithium aluminum hydride reduction product from benzene affords a crystal- line molecular-compound of 22 and 22 with a Sharp melting point of 123-124°, which does not change on recrystalliza- tion from acetone-hexane. This molecular-compound exhibits an [a];5 -38.6°, indicating it is a 1:1 complex of 22 and 22; Interestingly, this molecular compound showed only one spot by tlc on silica gel, however, tlc on alumina gave two SPots of R 0.43 and 0.32. f A possible mechanism for the transformation of tosylate 24 to alcohols 22land 22 is depicted in Scheme 1. The com- mon intermediate for formation of both products is the metal TsO OH 54 41 Scheme 1 ‘\ j x ’ 2 TSO ‘0 , 65 Path A Path B 1 \ \ o .92, LiAiH4 H 1 \ 1 HO 42 alkoxide 22 (M represents either aluminum or lithium). In the chair conformation of this intermediate (22;) the Egg; orientation of the tosylate and the 4—5 bond is favorable for a semi—benzilic type rearrangement (path A) giving alde— hyde 22, This aldehyde would be reduced immediately by the excess hydride present to give the observed product (22). In a twist-boat conformation (22b) the tosylate can attain an orientation 22££ to the C-4 hydrogen, and a semi—benzylic rearrangement of hydride (path B) would lead to ketone 21. This ketone would then have to be reduced by lithium alum- inum hydride from the B side of the molecule to give the equatorial alcohol 22. Exampless".68 of this type rearrangement, that ilus— trate a geometrical preference in bond migration, are Shown in equations 32 and 33. Barton has pointed out that the \ t-BuOK , — t . (32) -§—BuOH 69 O m OH ‘ (r bond which migrates is always the one anti to the leaving group.69 43 1,2-Hydride shifts are less prevalent than 1,2-alkyl shifts. The elimination of HBr from compound 12 can be viewed either as an E2 elimination or as a hydride shift, as shown in equation 34.70 The proposal of a hydride shift 1V 0 W t 00 is not unreasonable since it has been demonstrated in the benzylic acid type rearrangement of keto-aldehyde 22'4equa- tion 35).71 O _ OH HO - (35) ’ C6H5 D C6H5 O 75 0 22. M. NI ( x M+— 5 77 M+—- 0 12 The only reasonable alternative to the hydride shift of path B (Scheme 1) is a simple E2 elimination of HOTs to give enolate anion lg. This alternative is unlikely since there is evidence that enolate anions are not readily reduced by lithium aluminum hydride. Thus LAH reduction of enol 44 lactone lg affords ketone 22 in 65% yield. Presumably the enolate anion shown in equation 36 is the product of the hydride reduction and upon work-up gives the ketone.72 Dauben and Eastham report that ketone Q} can be obtained in 34% yield from the LAH reduction of enol acetate §2/ and that the pre-formed lithium enolate of cholestan-4-one (21, M = Li) is quite resistant to hydride reduction.73 If the enolate anion IQ is an intermediate in the reduction of tosylate Q4 one would expect to recover some cholest—S-en- 4—one (QZ), but no ketonic product is observed. ' (36) /& LiAlH4 fl H20 ' \fl/\L . ; / M 0 0 OH 731 ' 22 1' . I W .. (37> AcO 80 O §l. m Since in the first step of the proposed mechanism LAH is functioning only as a base, reaction of the starting tosylate 54 with a base that is not a carbonyl reducing agent, should afford the proposed intermediates fig and El; Treatment of tosylate lewith sodium hydride in THF afforded cholest-S-enFW4-one (g1) in 48% yield (equation 38). 7 was identified by a comparison of its physical (W Enone EtOH . ted properties (mp 111-1120, Amax 241 mu) With repor v 45 values74, and is identical with the product obtained by X X. T80 OH 0 67 22. Jones oxidation75 of cholest- -5-en-4a— ol (62) (equation 39) \ Jones Ox. Cr+g 22. It is not surprising that the intermediate aldehyde Qfilcould not be isolated, since under the strongly basic conditions employed the aldehyde could easily undergo fur- ther reactions. A similar aldehyde (g2) has been prepared76 and is found to be quite unstable. 1 ’ . I H' OHC If enone 67 is an intermediate in the formation of fig from 54 it must lead to the a alcohol (62) when reduced with LAH under similar conditions. This reduction of enone 67 afforded 62 in 95% yield, thus confirming reports that enone 67 is reduced by LAH almost exclusively from the fi side of the molecule.59t60 46 ‘\ LiAlH 4 ., IIIII (40) The results of lithium aluminum deuteride (LAD) reduc- tion of tosylate lealso support the proposed mechanism. The products, obtained in yields comparable to those of the LAH reduction, are shown in equation 41. N T50 s OH 54 OH 83 \ \ .9 + 4. 6 CHDOH 84 D D 85 m m The major product 4B-deuterocholest-5—en-4a—ol (Q3) was identified by its melting point (141°), and infrared spectrum (2110 cm-1, v C-D). The incorporation of one deuterium was shown by mass Spectrometry. The molecular ion appeared at m/e 387, the largest fragment was at m/e 369 (M-Hzo) and a prominent fragment was apparent at m/e 372 (M—CH3) [Figure 24]. The absence of the carbinol hydrogen signal in the nmr spectrum of §§iindicated the deuterium was in the 46 position. This product would re- sult from the LAD reduction of the proposed enone inter- mediate 61, 47 The second product, 35-hydroxydeuteromethyl—A-nor- cholest-S—ene (§4), mp 99°, showed C-D stretching in the infrared spectrum at 2140 cm-1. Mass spectrometry showed a molecular ion at m/e 387 indicating a deuterium content of one atom per molecule [Figure 25]. In addition the mass spectrum showed fragments at m/e 372 (M-CH3) and 355 (M-CHDOH). The m/e 355 fragment, which is also present in the corresponding undeuterated alcohol 63/ indicates the deuterium is on the carbinol carbon, since it is lost with the hydroxyhiethylene fragment (CHDOH). The nmr spectrum of §4 shows two doublets at 6 3.56 and 3.62 (J - 7.2 and 7.2 cps) with a combined relative area corresponding to one hydrogen [Figure 20]. Since LAD reduction of the proposed intermediate aldehyde figican occur from either side of the carbonyl group, two epimeric alcohols (§2§ and §4b) should result. The presence of the deuterium changes the magnetic environment of the D (42) carbinol hydrogen and causes the difference in chemical shifts of hydrogens HA and HB to be slightly greater in an and §4b than in the corresponding undeuterated alcohol (fifi). Since the product of the reduction is a mixture 48 consisting of two epimeric alcohols having carbinol hydro— gens with different chemical shifts and each being of the AX type, the resulting signal is two doublets with g_= 7.2 cps, one due to epimer §4a (JA ) and the other from ggb X (JBX). The doublets are separated by the difference in chemical shifts (3.75 cps) between HA and HB' This product is consistent with the proposed intermediacy of aldehyde 66, The third product, 4,66-dideuterocholest-4-ene (gé), was found to have two deuterium atoms. The mass spectrum revealed a molecular ion at m/e 372 and an M-CH3 fragment at 357 [Figure 26]. Two carbon-deuterium stretching fre— quencies, at 2220 and 2100 cm-1, were apparent in the infra- red spectrum. The nmr spectrum indicated that one of the deuteriums is on carbon 4, since a vinyl hydrogen signal is not present. On the basis of mechanistic considerations the second deuterium is believed to be in the 63 position. As mentioned previously, LAH reduction of cholest-S- en—4-one (61) gave almost exclusively the 4a-alcohol. If attack of hydride (or deuteride) was predominantly, but not exclusively, from the B side of the molecule some 4B-alk- oxide §§ should also be formed in the reduction (equation 43). The 4B—alcohol can actually be isolated from the sodium borohydride reduction of enone 61377 The 4B-aluminumalkoxide (gg), since it is axial, is oriented favorably for an allylic rearrangement. The cyclic transition state necessary for this S i' type mechanism73, N illustrated in structure §§a, leads to a compound which has 49 \ x. x D” 67 o /I> §§’ v- \Al' I .. .1 \x \ A\. .// //’Al ~ I o \D 1 x ‘ ——~ Q. <44> D 86a ’W D D 85 the double bond in the observed 4-5 position. In the case of the LAD reduction this mechanism predicts the position of the second deuterium. A similar allylic rearrangement on the same system has been observed by Young and co-workers.79 The reaction of thionylchloride with alcohol §Z’proceeds exclusively by way of an S i' mechanism, the sole product being 6S-chloro- N cholest—4—ene (gg). OH m Cl 50 The proposed mechanism for the formation of cholest-4- ene (61) asserts that enone 61 is a common intermediate for the formation of both olefin 61'and cholest-5-en-4a—ol (62). Lithium aluminum hydride reduction of 61 did not afford any detectable olefin but gave only the 4a-alcohol (62). This may be because the exact conditions under which the tosylate reduction occurs cannot be duplicated in a simple LAH reduc- tion. The products of the lithium aluminum deuteride reduc— tion and the formation of a proposed intermediate (61), from the reaction of sodium hydride with the starting tosyl- ate (64), clearly substantiate the proposed mechanism. The over-all scheme for the formation of the three products of the lithium aluminum hydride reduction of 62 is shown in Scheme 2. 51 8H17 Scheme 2 C TSO l l 54 OH 1 \ + 65 0“M TSO \ EXPERIMENTAL General. Infrared spectra were recorded on a Perkin- Elmer 237B grating spectrophotometer, using sodium chloride cells. A Unicam SP. 800 spectrophotometer was used for ultraviolet spectra. Nuclear magnetic resonance spectra were taken on a JEOL Co. C-60H high resolution Spectrometer, in deuterochloroform using tetramethylsilane as an internal standard. Mass spectra* were determined on an LKB type 9000 mass spectrometer with an ionizing voltage of 70 ev and an ion source temperature of 230°. Glc inlet was via a 1% SE-30 column at 235°. Melting points were determined on a Koefler hot stage, and are uncorrected. Specific rotations were obtained on a Perkin-Elmer Model 141 Polarimeter. Brine refers to an aqueous saturated sodium chloride solution. * Mass spectra were obtained thru the courtesy of Dr. C. C. Sweeley, Biochemistry Department, Michigan State University. 52 53 36:27Toluenesulfonoxycholest:5-en-4B-ol (§§)' Tosylate 62'was prepared according to the method of Karrer56: mp 102-104°(d) [lit56 mp 111°(d)]7 ir (CHCla) 3580, 2950, 2860. 1175 cm-1; nmr (CDC13) 0 0.65 (s, 18-CH3), 0.30 (s, CH3), 0.90 (s, CB3), 1.25 (s, 19-C§3), 2.16 (s, 03), 2.43 (s, Ar-Cga), 4.24 (m, C-4 g), 4.45 (m, C-3 g), 5.65 (m, C-6 g), 7.6 (AB—q, Arig); one spot on tlc (silica gel G, CHCla). Lithium Aluminum Hydride Reduction of BprfToluene- sulfonoxycholest-5-en-4B-ol (64). To a flask equipped with reflux condenser and a magnetic stirrer, and containing 1.2 g (31 mM) of lithium aluminum hydride (Ventron) in 200 ml of anhydrous ether, 10.0 g (17.9 mM) of 62 in 100 ml of dry benzene was added in portions. The mixture was stirred for 1 hr at room temperature and then heated at reflux temperature for 20 hrs. The mixture was cooled and 6.1 ml of ethyl acetate was added followed by slow dropwise addi- tion of water until a precipitate formed. After stirring overnight the mixture was filtered and the filter cake washed with ether. The filtrate was washed twice with water, once with 5% sodium bicarbonate and once with brine, then dried over sodium sulfate. Evaporation of the solvent gave 6.6 g of crude product. Crystallization of the crude product from benzene af- forded a crystalline material of mp 123-124°, [a];5 --38.60 (g 0.5, CHC13); tlc (silica gel G, CHC13) one spot, Rf 0.36. Recrystallization from acetone-hexane did not change the melting point. 54 Ascending Dry Column (ADC) Chromatography of the LAH Reduction Product. ADC chromatography is a modification of methods described by Loev and Goodman,8° and Dev, g£_al.31 A 0.5 g portion of the crude product, which had been pre— absorbed on 2.5 g of alumina, was added to the top of 110 9 neutral alumina (Woelm, activity II, Brockman scale) which had been dry packed in a 2.5 x 45 cm glass column. An addi— tional 2.5 g of alumina was placed on top of the pre—adsorbed sample and the empty portion of the column was filled with cotton. The column was inverted into anhydrous ether and allowed to develop by the ascending solvent. The solvent front was kept just above the level of the solvent outside the column to insure a fairly slow development by capillary action. When the solvent had reached the top of the alumina packing, the column was righted and after removal of the cotton, was eluted with anhydrous ether in a normal fashion. Collection of 20 ml fractions gave the results shown in Table 2. Table 2. ADC chromatography of the LAH reduction product. Fract. Wt.(mp) Fract. Wt.(mp) Fract. Wt.(mp) mg mg mg 1. 48 5. 8 10. 4 2. 2 6. 13 11. 1 3. 90 (142°) 7. 53 (99°) 12. 1 4. 160 (142°) 8. 63 (99°) 13. 0 9. 55 (99°) Total 498 mg 55 Compound A. Recrystallization of fractions 3 and 4 (250 mg, 48% yield) from hexane gave cholest-5—en-4a—ol (6%): mp 141-142° [a];5 -53° (2 0.45, CHC13); [lit59,6° mp 144°, 143°, (a157'5 -55.0 (g 1.49, CHCl3): [61D -540 (g 1.0, CHClQ]: Rf 0.43 (neutral alumina act. II, ether); ir (CHC13) 3600, 2940, 2860, 1470, 1375, and 1070 cm"1 [Figure 10]; nmr (CDCl3) 6 0.68 (s, 18-cg3), 0.80 (s, cgs), 0.91 (s, cg3), 0.98 (s, 19—Cg3), 1.05-2.5 (methylene envelope) 2.05 (S, 05H), 4.25 (m, 1H, line width 21 cps, CHrOH), 5.8 (m, 1H, vinyl H) [Figure 13]; mass spectrum (70 eV, direct probe, ion source 230°) m/e (rel intensity) 388 (1.4), 387 (8.4), 386 (27), 371 (22), 368 (100), 353 (20), [Figure 21]. Compound B. Recrystallization of fractions 7-10 (175 mg, 33% yield) from methanol gave 3B—hydroxymethyleA—nor— cholest-5-ene (§§)= mp 99°; [a];5 -24.40 (g 0.5,.CHC13), [lit65 mp 99-1010; [a]D -26° (3 1.56, CHC13)]: Rf 0.32 (neutral alumina act. II, ether); ir (CHC13) 3600, 2950, 2860, 1465, 1380, and 1020 cm‘1 [Figure 11]; nmr (CDClé) 0 0.70 (S, 18—CH3), 0.82 (s, CH3), 0.90 (s, 2CH3), 1.0-2.4 (methylene envelope) 2.05 (s, OH), 2.7 (m, 1H, C-3 H), 3.62 (d, 2H, g.= 7.2 cps, carbinol-H), 5.5 (m, 1H, Vinyl H) [Figures 14 and 19]; mass Spectrum (70 eV, direct probe, ion source 230°) m1; (rel intensity) 389 (5), 387 (30), 388 (100), 372 (9), 371 (30), 355 (45), [Figure 22]. 56 Compound C. Preparative thin layer chromatography of ADC chromatography fraction 1 (48 mg) on one 20 x 20 cm glass plate with a 1.5 mm layer of silica gel PF254, using hexane as the developing solvent, showed a band of Rf 0.80. Elution of this band with ether and crystallization from acetone gave 15 mg (3% yield) of cholest-4-ene (61): mp 80-810; [01:7 +670 (g_0.6, ch13), [lit61 mp 81°, (Q1;3 +670 (CHCl3)]; ir (CHCla) 2940, 2860, 1470, and 1375 cm"1 [Fig— ure 12]; nmr (CDC13) 0 0.67 (s, 18-CH3), 0.80 (s, CH3), 0.90 (s, CH3), 1.00 (s, 19-CH3), 1.05-2.4 (methylene envel— ope), 5.25 (m, vinyl-H) [Figure 15]; mass spectrum (70 eV, glc inlet, ion source 230°) mge (rel intensity) 372 (4.6), 371 (30), 370 (100), 355 (25) [Figure 23). Lithium Aluminum Deuteride Reduction of 35:2;Toluene- sulfonoxycholest-5-en-4B-ol (§$)' The same procedure and work-up used in the lithium aluminum hydride reduction was employed for the lithium aluminum deuteride reduction of 62; Separation by ADC chromatography afforded the three products in yields similar to those obtained in the lithium aluminum hydride reduction of 66. A) 4B~Deuterocholest-S-en—4a—ol (66): mp 141—142° (from hexane); ir (CHCl3) 2110 cm-1 (C-D); nmr (CDCla) identical with nmr spectrum of 23,except for the absence of the 0 4.25 (CHfOH) signal [Figure 16]; mass spectrum (70 eV, direct probe, ion source 230°) gig (rel intensity) 398 (4), 388 (22), 387 (72), 372 (42), 369 (100), 354 (23), [Figure 24]. 57 B) BB-Hydroxydeuteromethyl-A-norcholest-5-ene (§é)‘ mp 99° (from methanol); ir (CHCla) 2140 cm.1 (C—D); nmr (CDC13) two doublets at 0 3.56 and 3.62 (combined area 1 H, g.= 7.2 and 7.2 cps) [Figures 17 and 20]; mass Spectrum (70 eV, direct probe, ion source 230°) mge (rel intensity) 389 (5), 388 (30), 387 (100), 373 (9), 372 (30), 355 (50), [Figure 25]. C) 4,65-Dideuterocholest-4-ene (66): mp 99° (from acetone); ir (CHcla) 2220, 2100 cm‘1 (C-D); nmr (CDc13) vinyl hydrogen signal present in the nmr Spectrum of 6Z'at 0 5.25 is absent in the spectrum of 66 [Figure 18]; mass spectrum (70 eV, glc inlet, ion source 230°) mZe (rel in- tensity) 374 (5), 373 (31), 372 (100), 357 (31). 26]. [Figure Jones Oxidation of Cholest—5-en-4a-ol (66). Jones reagent75 (8N, 0.85 ml) was added dropwise to a solution of 1.125 g (2.92 mM) of alcohol (62) in 100 m1 of acetone at 20°. The solution was allowed to warm to room temperature and then diluted with an equal volume of water. This solu- tion was extracted with ether and the ether extract was washed with water (2x), a saturated sodium bicarbonate solu- tion, and with brine. Evaporation of the solvent and crys- tallization from acetone gave 0.9 g (81%) of pure cholest-5- en-4-one (61): mp ill-112°; uv max (EtOH) 241 mu (6 6,000), [lit74 mp Ill—112°, uv max (EtOH) 241 mu (6 7,200)]7 ir (CHC13) 1675 and 1620 cm-1. 58 Reaction of Sodium Hydride with 3B-proluenesulfonoxy- cholest—5-en-4a-ol (62). To a stirred solution of 500 mg (0.9 mM) of tosylate 62’in 25 ml of tetrahydrofuran (THF) under a nitrogen atomosphere, 41.0 g (0.9 mM) of sodium hydride on mineral oil (52.8% NaH) was added in small por- tions. The mixture was heated at reflux for 3 hrs, cooled, and one equivalent of water was added dropwise. The result- ing solution was filtered and evaporated to dryness at re- duced pressure. The brown residue which resulted was dis- solved in ether and the ether solution was washed with water (2x), 5% sodium bicarbonate, and brine, then dried over sodium sulfate. Evaporation of the solvent gave 0.45 g of crude product. Column chromatography on 20 g of silica gel (E. Merk), starting the elution with benzene and progressing to methylene chloride, afforded 165 mg (48%) of cholest-S- en-—4-one (61): mp Ill-112° (from acetone); ir (CHCla) identical with infrared spectrum of 61 obtained from Jones oxidation of cholest-5—en——4a-ol (62). Lithium Aluminum Hydride Reduction of Cholest—5-en:f 3:0ne (61). To a Slurry of 380 mg (10 mM) of lithium alum- inum hydride in 5.8 ml of refluxing anhydrous ether, 200 mg (0.52 mM) of ketone 61'in 3.0 ml of dry benzene was added in three portions. The resulting slurry was heated at re— flux for 24 hrs. The usual work-up gave a crude product that tlc indicated contained cholest-5-en-—4a—ol (62) but no cholest-4-ene (61). Crystallization of the crude product 59 from hexane gave 190 mg (95%) of 62: mp 140—1410; ir (CHC13) identical with known 62. FIGURES 60 61 Figure 1 moss in ‘ to 7.0 ”00" 10.0 no mo Figure 2 ,- —- IB‘I - nu. 62 .1- 9“" a I.“ 63 Figure 6 7.0 ,. . c ta: -1...) 7 , . l . t . _ I..." _ . . _ _ . Iv, 1.161 '5‘, f. .. . m m m _. F . m _ _ 1 _ r i .2 _ , . . _ u _ __ )fllll..nl _ . _ . (I) Figure 7 rf'vT ,. 1-- . HD—-<—----- -- ----—-.—-Ih— - - . -_.——- »¢- .9..- 64 - .. -.._-- ,- “- , . . .. u... . >4 _-- A J. t I § T . , -o a u - \ O 0 D » -_ -.o .. 7 n o o. I u a- 3 . .1. p. .5... 7».r1.£.w.. I ‘Q V O I t L 3 Y Y I I... i —+—- l I I AJL H 0 s I‘ll)? 3.1.1.: . i 0 l I t . 2-..... h... .. I i ‘. “‘3’ ! ".1- ? -. .-——.> 4. .. "“1? ! J . i... r -. J. A_1; [IO 7| 7... o i L. I 3 7 k .... o ; - . ----*--. -‘ 1.-.-.. . i ‘ ' .' .ltutt Q. .900 I l n I —- >-.—.—.—- I -1 .m . l “up-om ‘00-“. Ind»- I I . n c...- 1 0 c I..- a l l ’was?- .0— - 4.... K . I n L LL A I Y .._-.r-— o 3 3 L.-. _. l I I .. i, I é“..- I 1 I ‘ 3 i 7 I I I IL4.‘ A I I 7 _i___ : I ! v t I Z 1 9 .4- i I 4; l L . l I . t -a- w~g~~uuqr-.u .. . f 1 _.. ---...——.—&—.-- >— u - i --..- -4..- L ! a i ._..J'.......4...; 4...;- _+_.. .i § ;_. #7 J; . LLJ JJ ..;.. 3..."... ._...._-;.+.__ ..... . , . , .. . . . . 00:0. 1.“... .nonur‘. Isl n70.71- . , . . i l 3 . . . . 9‘0..—0: I 0. FF) _...,...._—‘.' .— , . 1 L ..4(._.. .._£. 5 2 L A l -3 L. Figure 9 I) L 3 , -.,—'+'+.—-‘:‘«»:+ 3’1.) . .- I -— . ‘ '1- .- 5: l . ;» 7...;U ... .-.—. 65 Figure 10 I non I 3°00 I mo woo Figure 11 ........., mo mo mo no “DI "I.“ Figure 13 66 93: a" 3». :;- -. .1: m ‘ nub“ = a. z :00 2 no :50 a 08 EIIIFYTIIIII'IIITIYTIIIYI1I[VIYITITWWIIIIIIIYHWIIUII‘ITTIW][117117fl?1fi111111¥mm 1 i ' A A 1 - i l I I I O 0 7 O 0 no. 0 Figure 14 I saw: 00: 3:3 3:0 75: 7:0 '5: mo 53 "5 “Q ~ a a 3‘ I” “ ”I «no - m a no no I. ”I ____../\ . \ 11111111111111111111111111111111144111111114L11111J1UJJ ll11111]lllLllLlllllLJlllUlllll C t“ 0 Figure 15 I 52'. 40.) 35" 300 231‘ 2. '3.’ 100 .‘J ‘S «no - a an no 30 ID I” a a name no 7. M no «no no lfl P. rTIIWTI[IYVIII!ITlIII‘[117111111]IIIYIITIIITITWIIIIUIIWTWTIIl7] ‘ I . A.__-,_._AAA ‘IAAAA‘. A 1'“ ‘VVY vvw—u—v‘v“ 1411111lAililliiilninnlillllin11111111111111 ‘ 1111:111l11:ll11111111111111111111111Alli O ‘ 0 $1.] 0” F gure 16 m a as a an no no no. 50 t?” ‘0 U 7' fl = I. I. I8 O m ' 'JJI" LELILLELLEI - ~ a m m Y . . ‘1 1 1 1 l A 11 i “YT-111.171 I Tfj (I 1': Ll] . . nllfiniilinnill‘nfi1jfiTronlliLilTTLAI{11111117111111 3 I I I O 7 C Figure 17 - i m an an no 30 700 ”0 IN 5. 00! u - ‘ - a. ’ ” I. I, . a m . 7. ~ . a In an 7. m H IIrVIIrIrLItrIIIIIIIVItv[IVIILrIT‘WTm-WFFWWA .... , . . . , r , , . .. I ,-... ., 1. , , , I . ,. ...,. . ' ‘ ;;.;;.J l lI.‘I.‘Illllllill.Ill]II‘lllll‘lllllllllllllllllll LLI‘LI1L.lllL‘lLllllElllll'llLllllll l | 68 Figure 19 F“ 0(pr i F : P 2' w) i , f: ’1'". "" ; _:_‘ ; 3 Figure 20 CIIDOH r -_-- )— o. 0- on ‘1. FIGURE 2| 69 FIGURE ‘22 - Ioo- FIGJRL: 23 368 386 370 - x I , - m \. g Cb {Q “ GU 3 °” W Z 8 < 355 _j an E 353 I 306 3” 0.I ' ! I I I. ‘ | I] ' III I I ' I l | I I I 350. 390 350 390 350 ' 350 . . m , /e FIGURE 24 FIGURE 25 FIGURE 26 100 7 369 ~ I ~ , 307 '37: 1 " W.” " U u / ‘ lg. D '0“ 3.7 D o“ D D Z < D Z 3 355 m < _j 372 55" g 354 o- I I l I I! II I [I I l ' I I ' I I I I I l I I 350 390 350 390 350. ' 390 , m/ \ REFERENCES REFERENCES Reviews: a) J. Sicher, "Progress in Stereochemistry," Vol. 3, Butterworths Press, Washington, 1962, p. 202; b) A. C. Cope, M. Martin, and M. A. McKervey, Quart. 52!: (London), 26, 119 (1966); c) F. Sorm, Pure Appl. Chem., 2) 533 (1961). C. A. Grob and P. W. Schiess, Angew. Chem. (Int. Ed.), 6, 1 (1967). C. A. Grob in "Theoretical Organic Chemistry" (Papers presented to the Kekule Symposium, London, Sept. 1958) Butterworth and Co. Ltd., London, 1959, p. 114 ff. S. Searles and M. Gortatowski, J. Amer. Chem. Soc., 16. 3030 (1953). A. Slawjanow, J. Russ. Phys. Chem. Soc., 62) 140 (1907); Chem. Abstr., 1) 2077(1907)T R. D'Arcy, C. A. Grob, T. Kaffenberger, and V. Kras- nobajew, Helv. Chim. Acta., 22) 185 (1966). H. 0. House, "Modern Synthetic Reactions," W. A. Benjamin, Inc., New York, 1965. A. Eschenmoser and A. Frey, Helv. Chim. Acta., 66) 1660 (1952). E. J. Corey, R. Mitra, and H. Uda, J. Amer. Chem. Soc., 66) 485 (1964). M. Tanabe and D. Crowe, Tetrahedron Letts., 2955 (1964). H. Weston, Helv. Chim. Acta., 61) 575 (1964). P. S. Wharton, Y. Sumi, and R. Kretchmer, J. Org. Chem., 66) 234 (1965). E. L. Eliel, J. Chem. Ed., 61, 126 (1960). 70 (14) (15) (16) (21) (22) (23) (24) (25) 71 C. A. Grob, H. Kiefer, H. Lutz, and H. Wilkens, Tetrahedron Letts., 2901 (1964). P. S. Wharton and G. Hiegel, J. Org, Chem., 66) 3254 (1965). S. J. Cristol and W. P. Norris, J. Amer. Chem. Soc., 16) 632, 2645 (1953); E. Grovenstein and D. Lee, ibid., 7.5.: 2639 (1953). R. C. Cookson, J. Henstock, and Hudec, J. Amer. Chem. Soc., §§u 1060 (1966). D. Todd, "Organic Reactions," Vol. IV, John Wiley and Sons, Inc., New York, 1948, p. 378. H. H. Szmant and C. M. Harmuth, J. Amer. Chem. Soc., 66) 2909 (1964). D. H. Gustafson and W. Erman, J. Org. Chem., 26) 1665 (1965). N. Kishner, J. Russ. Phys. Chem. Soc., 26) 582 (1911); Chem. Abstr., 6) 347 (1912). N. J. Leonard and S. Gelfand, J. Amer. 3269, 3272 (1955). Chem. Soc., 11, P. S. Wharton, and L. Krebs, 22,. 958 (1964). S. Dunny, J. Org. Chem. R. C. Clayton, H. B. Henbest, and M. Smith, J. Chem. Soc., 1982 (1957). J. B. Hendrickson, Tetrahedron, 1) 82 (1959). E. J. Core and A. Hortmann, J. Amer. Chem. Soc., £1, 5736 (1965 . J. A. Marshall and G. Bundy, Chem. Commun., 854 (1967). S. Sharma, R. Srivastava, and D. Devaprabhakara, Can. J. Chem., 66) 84 (1968). D. J. Cram and N. L. Allinger, J. Amer. Chem. Soc., 16) 2518 (1956). J. A. Marshall and G. Bundy, J. Amer. chem. Soc., 66/ 4291 (1966). P. Heinback, Angew. Chem. Int. Ed., 6, 595 (1966). (32> <33) <34) (35) (36) (42) (43) (44) (45) (46) 72 W. S. Johnson, C. D. Gutsche, and D. Banerjee, J. Amer. Chem. Soc., 12) 5464 (1951). H. 0. Brown and R. F. McFarlin, J. Amer. 62, 5372 (1958). Chem. Soc., J. Bobbitt, "Thin-Layer Chromatography," Reinhold Publishing Corp., New York, 1963. A. R. H. Cole, in "Technique of Organic Chemistry," Vol. XI, A. Weissberger, Ed., Interscience Publ., 1963, p 133. L. M. Jackman, "Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry," Pergamon Press, Long Island City, N.Y., 1959, p. 35. W.-S. Johnson, D. Allen, R. Hindesinn, G. Sausen, and R. Pappo, J. Amer. Chem. Soc., §2x 2181 (1962). N. S. Bhacca and D. H. Williams, "Applications of NMR Spectroscopy in Organic Chemistry," Holden-Day, Inc., San Francisco, 1964, p. 47-52, 77—85. P. S. Wharton and D. H. Bohlen, J. Org. Chem., 26, 3615 (1961); p. s. Wharton, ibid., 22' 4781 (1961'). L. F. Fieser and M. Fieser, Synthesis," John Wiley and Sons, 1967, p. 1147. "Reagents for Organic Inc., New York, H. E. Simmons, E. Blanchard, and H. Hartzler, J. Org. Chem., 31, 295 1966 ; P. G. Gassman, and F. V. Zalar, ibid.,~§ifl 166 1966 R. Silverstein and G. C. Bassler, "Spectrometric Identification of Organic Compounds," 2nd ed, John Wiley and Sons, Inc., New York, 1967, p. 9. J. H. Beynon and A. E. Williams, "Mass and Abundance Tables for Use in Mass Spectroscopy," Elsevier, Amsterdam, 1963. C. Djerassi, T. Grossnickle, and L. High, J. Amer. Chem. Soc., 16) 3166 (1955). A. I. Scott, "Interpretation of the Ultraviolet Spectra of Natural Products," Macmillan Co., New York, 1964, p. 28, 34. W. G. Dauben and G. Berezin, J. Amer. Chem. Soc., 89,, 3449 (1967). (47) (43) (49) (52) (53) (54) (55) (56) 73 G. Lock, Monatsh. Chem., 66) 802 (1954). F. Sondheimer and D. Rosenthal, J. Amer. Chem. Soc., §Q, 3995 (1958). D. J. Cram, M. Sahyun, and G. Knox, J. Amer. Chem. Soc., 62, 1734 (1962). Reference 40, p. 873. R. K. Hill and T. R. Conley, J. Amer. Chem. Soc., 66, 1734 (1962). L. F. Fieser, M. Fieser, and R. Chakranarti. §;_§EE£° Chem. Soc., 11) 2226 (1949). A special prep. from Aldrich Chemical Company, Milwaukee, Wis. Inc., L. Durham, D. McLeod, and J. Cason, Coll. Vol. IV, John Wiley and Sons, Inc., 1963, p. 510. L. J. Bellamy, "The Infrared Spectra of Complex Molecules,‘I John Wiley and Sons, Inc., New York, 1958, p. 268. P.-Karrer, H. Asmis, K. Sareen, and R. Schwyzer, Helv. Chim. Acta., 221 1022 (1951). H. Schmidt and P. Karrer, Helv. Chim. Acta., 32, 1371 (1949). D. H. R. Barton and W. Rosenfelder, J. Chem. Soc., 2875 (1951). D. N. Jones, J. R. Lewis, C. W. Shoppee, and G. Summers, J. Chem. Soc., 2876 (1955). E. J. Becker and E. S. Wallis, J. Org. Chem., 26, 353 (1955). J. Broome, B. Brown, A. Roberts, and A. White, J. Chem. Soc., 1406 (1960). Reference 38, p. 13. J. Emsley, J. Feeney, and J. Sutcliffe, "High Resolu- tion Nuclear Magnetic Resonance Spectrosc0py,” Pergamon Press, London, 1965, p. 363. 39, W R. J. Abraham and H. J. Bernstein, Can. J. Chem., 216 (1961); T. Schaefer, ibid., 66/ 1678 (1962). "Organic Synthesis," (72) (73) (80) (81) 74 G. H. Whitham and J. Wickramasinghe, J. Chem. Soc., 1655 (1964). W. G. Dauben and J. Ross, J. Amer. Chem. Soc., 61/ 6521 (1959). Y. Mazur and M. Nussim, J. Amer. Chem. Soc., 22, 3911 (1961). F. Winternitz and A. Paulet, Bull. Soc. Chim. (France), 1460 (1960). D. H. R. Barton, J. Chem. Soc., 1027 (1953). L. F. Fieser and X. Dominguex, J. Amer. Chem. 16, 1704 (1953). Soc., K. Heyns, W. Walter, and H. Scharmann, 2057 (1960). Chem. Ber., 22, F. A. Hochstein, J. Amer. Chem. Soc., 11) 305 (1949). W. G. Dauben and J. Eastham, J. Amer. Chem. Soc., 16) 3260 (1951). A. Butenandt and G. 11) 397 (1944). Ruhenstroth-Bauer, Chem. Ber., K. Bowden, I. Heilbron, E. H. R. Jones, J. Chem. Soc., 39 (1946). B. Wedon, H. Nace and G. Crosby, J. Org. Chem., 66/ 834 (1968). D. Lavie, Y. Kashman, and E. Glotter, Igtrahedron, 22, 1103 (1966). E. Gould, "Mechanism and Structure in Organic Chemr istry," Holt, Rinehart and Winston, Inc., New York, 1959, p. 296. R. E. Ireland, T. I. Wrigley, and W. G. Young, J. Amer. Chem. Soc., 66) 4604 (1958). B. Loev and M. Goodman, Chem. and Ind. (London), 2026 (1967). V. K. Bhalla, U. R. Nayak, and S. Dev, 26, 54 (1967). J. Chromatog., M)W|))l)ll|fllll)) iiWIHWIWIIWIUIWEs 31293 03175 3449