UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthetic studies in salamander alkaloids Paisley, Joseph Kenneth 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1973_A1 P33.pdf [ 9.85MB ]
Metadata
JSON: 831-1.0060149.json
JSON-LD: 831-1.0060149-ld.json
RDF/XML (Pretty): 831-1.0060149-rdf.xml
RDF/JSON: 831-1.0060149-rdf.json
Turtle: 831-1.0060149-turtle.txt
N-Triples: 831-1.0060149-rdf-ntriples.txt
Original Record: 831-1.0060149-source.json
Full Text
831-1.0060149-fulltext.txt
Citation
831-1.0060149.ris

Full Text

SYNTHETIC STUDIES IN SALAMANDER ALKALOIDS by JOSEPH KENNETH PAISLEY B.Sc. (Hon.), Queen's Un iver s i ty , Be l fa s t , 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of CHEMISTRY We accept th i s thes is as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January, 1973 In presenting t h i s thes is i n p a r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f ree ly ava i lab le for reference and study. I further agree that permission for extensive copying of t h i s thes is for s cho lar ly purposes may be granted by the Head of my Department or by h i s representat ives . I t i s understood that copying or p u b l i c a t i o n of th i s thes is for f i n a n c i a l gain s h a l l not be allowed without my wri t ten permiss ion. Department of The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada ABSTRACT Part I, describes an e f f i c i e n t four step method to e f fect an unsymmetric r ing cleavage between carbons 2 and 3 in the A r ing of 17g-acetoxy-5g-androstan-3-one (81b). Bromination of 81b followed by treatment with sodium acetate in re f lux ing acet ic acid gave 2e,17B-diacetoxy-5g-androstan-3-one (159) in good y i e l d . Subjecting 159 to hydroxylamine hydrochloride-sodium acetate in re f lux ing methanol afforded ant i 17g-acetoxy-2g-hydroxy-58-androstan-3-one oxime (187) in 75-85% crude y i e l d . Beckmann fragmen-tat ion of 187 by employing thionyl ch lor ide furnished 17g-acetoxy-2-oxo-2,3-seco-5s-androstane-3-nitrile (lj>5) in over 80% pur i f i ed y i e l d . Mechanistic studies on the formation of 159 from 17g-acetoxy-4g-bromo-5e-androstan-3-one (158) indicated that neither 4a,173-diacetoxy-5s-androstan-3-one (175a) or 43,178-diacetoxy-58-androstan-3-one (175b) can be intermediates and that the intermediate i so lated by Satoh and Takahashi must be 2a-acetoxy-53-cholestan-3-one (174). Part I I, describes attempts to elaborate 195 to the 17B-hydroxy isomer of samandarine 47b. Treatment of 195 with re f lux ing isopropenyl acetate in the presence of concentrated sulphuric acid y ie lded a mixture of c i s and trans 2,17B-diacetoxy-2,3-seco-58-androst- l -ene-3-nitr i le (223a) and (223b) in 62-58% pur i f i ed y i e l d . Ozonization of t h i s mixture followed by reduction gave 178-acetoxy-l-oxo-2,3-seco-A-nor-58-androstane-3-nitri l - i i i -(262) in 86% y i e l d . A w i t t i g reaction on 262_with subsequent ace ty l a -t ion afforded 178-acetoxy-2,3-seco-5g-androst- l-ene-3-nitr i le (57a) in ca_. 65% pur i f i ed y i e l d . Attempts to construct 47b from 57a proved unrewarding. However, most recent ly , Shimizu has converted 57a to 47b in three steps. - iv -TABLE OF CONTENTS Page INTRODUCTION 1 1. General 1 2. Therapeutic Potential 10 3. Chemical History 12 DISCUSSION 30 1. General Plan 30 Part I 45 a. Insertion of a hetero atom between carbons 2 and 3 45 b. Incorporation of unsaturation between carbons 2 and 3 83 c. Funct ional izat ion of carbon-2 91 d. Unsymmetric r ing cleavage between carbons 2 and 3 111 Part II , 122 Attempted formation of the bicycloxazoladine skeleton 122 EXPERIMENTAL 190 BIBLIOGRAPHY 259 LIST OF TABLES Table Page I Salamandra Alkalo ids 5 I I Toxic substances with t h e i r LD 5 0 8 I I I C-17 Subst i tut ion Pattern in Salamandra A lka lo ids 23 I V The e f fec t of a and 6 branching on the rate of E2 e l iminat ion and on the amount of o l e f i n formed. The reactions were between the a l ky l bromide and OEt 9 133 - v i -LIST OF CHARTS Chart Page I Structure E luc idat ion of the A lka lo ids 14 II Structure E luc idat ion of the Alka lo ids 15 III Pa r t i a l Synthesis of Samandaridine 21 IV Biosynthesis of the Side Chain in Salamandra A lka lo ids 24 V Hara - Oka Synthesis 27 VI Shimizu Synthesis 29 VII Stereoselect ive Synthesis of the B icyc looxazol id ine 47a from Intermediate 46_ 32 VIII Mechanism for the Formation of the B icyc looxazol id ine 47a from Acetal 6J_ 34 IX Possible Synthetic Pathways to the Bicyclooxazol id ine 47b 40 X Structural Aspects of Anti-form Oximes 47 XI Total Synthesis of Samanine (8_) 50 XII Pyro lys i s of N-nitrosolactam 96_ 56 X I I la N-nitrosolactam Investigations 57 XII lb N-nitrosolactam Investigations 61 XIV Baeyer -V i l l i ge r Investigations 65 XV Possible Ef fects of a Base on Tosyloxy Ester 103b 75 XVI Relat ive Rates for the Reaction of Some Substituted Pyridines with Methyl Iodide 80 - v i i -Chart Page XVII Baeyer -V i l l i ge r Oxidation Studies 82 XVIII Proposed Synthesis of Compound 86b 89 XIX Mechanistic Aspects of the Bromination of Ketones . 95 XX Synthetic Scheme for the Formation of Cyanoolefin 57a_ 124 XXI Second General Scheme for the Formation of Cyanoolefin 57a 137 XXII Poss ible Iodo Azide Adducts of Enol Acetates 223a and 223b 151 XXIII Formation of Cyanoolefin 57a via_a Wit t ig Reaction . 158 XXIV Elaboration of Cyanoolefin 57b to Az i r i d i ne 273 ... 167 - vi i i -ACKNOWLEDGEMENTS I wish to express my sincere thanks to Dr. Larry Weiler fo r his encouragement and excel lent guidance throughout the course of th is research and during the preparation of th i s manuscript. My thanks are extended to John Balsevich, Stuart Huckin, John Kingston, Ba l a j i Rao, Frank Skinner, and Ron Warneboldt for t he i r many worthwhile discussions. I wish to thank Beatr ix Kriszan fo r her i l l u s t r a t i v e work and Phy l l i s Moore f o r typing the manuscript. The f i nanc ia l support from the National Research Council of Canada (1969-72) is g ra te fu l l y acknowledged. F i n a l l y , I wish to thank the many facu l ty and s t a f f members of th i s department who have made my stay at U.B.C. a rewarding and enjoyable experience. INTRODUCTION 1. General The domain of natural product synthesis has provided an exce l -lent tes t ing ground for the development of the fundamental p r inc ip les of organic chemistry. Each synthetic endeavour usual ly brings to l i g h t a new p r i nc ip l e and extends the boundaries of ex i s t i ng chemical knowledge. In general, synthetic studies in the f i e l d of natural product chemistry have played a key ro le in the advancement and shaping of organic chemistry, e spec i a l l y , in terms of reaction mechanisms, stereo-chemistry, and synthetic methods.^ For example, synthetic studies d i rected towards the synthesis of vitamin B 1 2 have led d i r e c t l y to the development of the p r i nc ip le of conservation of o rb i t a l symmetry. These facts coupled with the increasing demand for more e f fec t i ve therapeutic agents provided the d r i v ing force to tackle a synthetic problem. I t was f e l t that the salamander a l ka l o i d s , a unique class of 3 stero ida l a lka lo ids , would provide a te s t i ng ground for new synthetic reactions and they could possess therapeutic po ten t i a l . These s tero ida l a lka lo ids are divided into three groups accord-ing to t he i r ske leta l features. Group one possess a b icyc looxazol id ine skeleton. Six a l ka l o i d s ; samandarine (1_), samandarone (2), samandaridine (3_), - 2 -- 3 -- 4 -O-acetyl-samandarine (4), samandenone (5_), and samandinine (6) exh ib i t th i s novel cons t i tu t i on . Group two contain a carbinolamine skeleton. Two a l ka lo id s ; cycloneosamandione (7a) and cycloneosamandaridine (7b) d i sp lay th i s type of skeleton. Samanine (8) const i tutes group three. It i s worthy of mention that there is a d i f ference in a l ka lo id content between the two subspecies of Salamandra, namely, S^ . maculosa taeniata (S.m.t.) endemic to western Europe and S_. maculosa maculosa (S.m.m.) endemic to south-eastern Europe. An out l ine of the a lka lo ids and the i r occurrence in the two subspecies are presented in Table I (see Page 5 ); the table shows the three groups of a lka lo ids with in the salamander fami ly. The work described herein i s concerned with the a lka lo ids in group one. Despite t he i r apparent s imp l i c i t y the elaboration of the s tero ida l framework to the b icyc looxazol id ine skeleton of fers a chal leng-4 ing synthetic problem. Hara and Oka have described the to ta l synthesis of samandarone (2_) and th i s const ituted a formal synthesis of samandarine (1_) and samandaridine (3_). However, i t was recognized that the i r synthetic approach had l i m i t a t i o n s , for example, t h e i r synthesis y ie lded only mi l l igram quant i t ies of samandarone (2) and the i r overal l synthetic sequence was rather lengthy. The work presented herein describes several approaches directed towards gaining a general method of entry into the b icyc looxazol id ine skeleton of general type 1_. It was hoped that the l im i ta t i on s of the Hara - Oka synthesis would be overcome by developing a new synthetic approach which would be both e f f i c i e n t and short. Since these a lka lo ids occupy such a conspicuous pos it ion among the natura l ly occurring s tero ida l a lka lo ids i t i s in order to ind icate a - 5 -Table I. Salamandra Alkalo ids Group Name Formula m.p. Occurrence 1. Samandarine C 1 9 H 3 1 N 0 2 188° S.m.t. Samandarone C 1 9 H 2 9 N 0 2 190° S.m.t.; S.m.m. Samandaridine C 2 1 H 3 1 N O 3 290° S.m.t.; S.m.m. O-Acetyl-Samandarine C21H33NO3 159° S.m.m.; S.m.t. Samandenone C 2 2 H 3 i N 0 2 191° S.m.t.; S.m.m. Samandinine C 2 l + H 3 9 N0 3 170° S.m.m. 2. Cycloneosamandi one C 1 9 H 2 9 N 0 2 119° S.m.t.; S.m.m. Cycloneosamandaridi ne C 2 ( t H 3 1 N0 3 282° S.m.t.; S.m.m. 3. Samanine C i 9 H 3 3 N0 197° S.m.t. - 6 -general type of c l a s s i f i c a t i o n . Heterosteroids may be divided into nuclear 5 and extranuclear categories. The former contain heteroatoms l i k e nitrogen, oxygen, and sulphur in the basic s tero ida l nucleus, and in the l a t t e r , the heteroatom forms part of a r ing system, groups or a side chain attached to the nucleus. Steroids possessing nitrogen are designated azasteroids. The number of natura l l y occurring extranuclear azasteroids are considerable. Several categories of s tero ida l a lka lo ids belonging to th i s group are wel l 6 7 8 known. These include a lka lo ids present in Solanum , Veratrum , Holarrhena , 8 8 8 9-11 Buxus , Sarcococca , and Pachysandra . Martin-Smith et al_., have reviewed the l i t e r a t u r e concerning the s tero id a lka lo ids of Funtumia, Pa rava l l a r i s , 12 Chonemorpha, F r i t i l a n a , a n d Malouetia. Recently, Witkop and coworkers have investigated the venom of the Columbian arrow poison f rog , Phyllobates  aurotaenia. They i so lated batrachotoxinin A (9) which retained only 1/500 of the t o x i c i t y of the o r i g ina l venom. Nevertheless, i t i s s t i l l as tox i c as strychnine. It i s very unusual fo r a s tero ida l a l ka l o i d to be so extremely tox i c . From the pharmaceutical viewpoint, batrachotoxin and the related congeners in the venom are in terest ing because they or a - 7 -synthetic var ia t ion could have medicinal app l icat ions . For example, s t ruc tu ra l l y modified compounds of batrachotoxinin A (9) could have lower t o x i c i t y and increased b io log ica l a c t i v i t y . These p o s s i b i l i t i e s and facts provided impetus to work with heterosteroids. In add i t i on , the structure of batrachotoxinin A (9) i s s t r i k i n g l y novel. Wehrli and 13 coworkers have recently reported the pa r t i a l synthesis of batracho-tox in in A (9). With regards to natu ra l l y occurring nuclear azasteroids, the only example where nitrogen forms an integral part of the stero id nuclear skeleton i s of the compounds obtained from the parotid and skin glands 3 of salamanders. It has been known since ant iqu i ty that salamanders are venomous animals. Over a century ago Za l e s k y^ i so lated a poisonous substance from the skin glands of salamanders which he designated saman-darine though i t i s now known to be a mixture of a l ka lo id s . Schopf and 15 Braun examined the a l ka lo id mixture and in 1930 they i so lated the major a lka lo ids presented in Table I (see Page 5 ). The ske leta l features of these a lka lo ids coupled with t he i r animal o r i g in made them s t r i k i n g l y unique. These were the f i r s t of a very small number of a lka lo ids found in animals to date. In terest ing ly , several tox ins ; bu fo ta l i n , pumil iotoxin C, dehydrobufotenine, tet radotox in , batrachotoxin, and samandarine have attracted attention because of t he i r diverse chemical, tox i co log i ca l 1fi 17 18 22 and pharmacological properties. In p a r t i c u l a r , Faust and Gessner have examined the toxicology and pharmacology of samandarine (1_). Poison-ing with samandarine f i r s t causes convulsions followed by i r regu la r pa lp i ta t i on and f i n a l l y pa ra ly s i s . Death occurs very quickly because of - 8 -TABLE I I. Toxic substances with t he i r LD 5 0 Substance LD 5 0 (yg/kg.) Batrachotoxin 2 Tetradotoxin 8 Bufota l in 400 Curare 500 Strychnine 500 Samandarine <3400 ( L D 1 0 0 ) * Sodium Cyanide 10,000 * L D 1 0 0 i s greater than a LD 5 0 - 9 -primary resp i ratory paralys i s without damaging the heart. Samandarine i s tox i c to a l l higher animals, f i s h , b i rds , and mammals. Even the salamanders die i f t he i r own venom enters t he i r blood. The le tha l dose i s 19 mg/kg. for the f rog , 3.4 mg/kg. for the mouse and 1 mg/kg. for the rabb i t . The r e l a t i v e t o x i c i t y of samandarine i s out l ined in Table II (see Page 8 ). Ch'an su, the dried venom of a common Chinese toad, and extracts of the Mediterranean plant Sci11 a maritima have received varied app l i ca -t ion in pr imi t i ve medical pract ice. The l a t t e r has been used from ca_. 23 3500 B.C. p r i n c i pa l l y for i t s d i u r e t i c and heart e f fects . In 1785, Withering recognized the therapeutic e f fec t of the cardiac glycosides 24 obtained from D i g i t a l i s species. Wieland ejt al_., investigated the extracts from the European toad Bufo vulgar i s which led to the i s o l a t i on of bufota l in and bu fa l i n . The cardiac action of bufa l in has been found almost equal to that of d i g i tox i gen in . It i s i n teres t ing to mention that 25 extensive studies of Ch'an su, p a r t i c u l a r l y by Meyer and his colleagues, has led to the i d e n t i f i c a t i o n of a number of re lated bufadienolides in 17 18-22 th i s mater ia l . Faust and Gessner " have reported that the cardiac action of the Salamandra and the D i g i t a l i s toxins are s im i l a r . In view of the increase of cardiovascular diseases th i s resu l t could have con-siderable importance in therapeutic treatment of these diseases. I t i s noteworthy that there i s a d i f ference between the Salamandra and the D i g i t a l i s toxins insofar as the arrest of the heart in d iasto le i s not compensated for by atropine in the administration of samandarone but i s for d i g i t a l i s . However, the Salamandra toxins are not used in medical pract ice because of t he i r u nava i l a b i l i t y and the lack of data on t he i r b io log ica l a c t i v i t y . - 10 -2. Therapeutic Potential Having selected a synthetic target which appears to demon-strate therapeutic potent ia l i t i s now in order to consider the general pr inc ip les of such a program. In retrospect, there are two governing p r inc ip le s which determine the therapeutic potent ia l of a synthetic endeavour. F i r s t l y , the prerequis i te of c l i n i c a l evaluation i s a v a i l a b i l i t y of tes t ing mater ia l . To the synthetic chemist i t i s , therefore, important to develop high over -a l l e f f i c i ency in the synthetic sequence. Consequently, each synthetic operation must be assessed with regard to economy of e f f o r t . Samandarone (2) i s neither read i l y ava i lab le by synthesis (vide  supra) nor from the natural source. Secondly, i t i s through the technique of molecular modif icat ion in drug design that the f u l l potent ia l of any system can be r ea l i z ed . These two general p r i n c i p l e s , therefore, d irected the approach to the synthetic problem. In order to gain a method of entry into the b icyc looxazol id ine system i t was planned to u t i l i z e a z i r i d i ne aldehyde 1_0 as a "key" i n t e r -mediate (vide i n f r a ) . It was hoped that a common precursor could serve OH io - 11 -for the elaboration to the sulphur and oxygen analogues 1J_ and 1_2_ respect ive ly , to give the sulphur and oxygen analogues of samandarone (2_). From a l o g i s t i c viewpoint the ideal precursor would be an o l e f i n of general type 1_3, since there are numerous, e f f i c i e n t , methods for con-vert ing an o l e f i n to the appropriate three membered heterocycles. By OH th i s approach i t i s easy to conceive a wide range of p laus ib le nuclear analogues and s t r u c tu r a l l y modified compounds re lated to samandarone (2). In add i t ion, manipulating and transforming the inherent functional group in the D r ing gives further freedom for s t ructura l variance. - 12 -Once a v e r s a t i l e , p r a c t i c a l , method of synthesis has been rea l i zed i t would be in s t ruc t i ve to ascertain i f the a c t i v i t y i s dependent upon the maintenance of the stereochemical i n t eg r i t y of various asymmetric centres in these a lka lo id s . If the t o x i c i t y is lowered or the a c t i v i t y increased by a l te r i ng t he i r stereochemical features or functional groups then they could have medicinal value in the treatment of cardiovascular diseases. S im i l a r l y , the various nuclear analogues may manifest low t o x i c i t y or increased a c t i v i t y and thus have therapeutic potential as wel1. 3. Chemical History " The formulation of a synthetic plan const itutes a prominent role in the successful synthesis of a natural product. The f i r s t step focuses on s impl i fy ing the synthetic problem with respect to the s t ruc -tura l and stereochemical features in the target molecule. In general, there are three sources which can provide valuable knowledge for th i s so lut ion. F i r s t l y , de ta i l s regarding the i s o l a t i on and s t ructura l e luc idat ion . Secondly, studies on the mode of biosynthesis and f i n a l l y , i f a va i l ab le , a descr ipt ion of previous synthesis of the target molecule or related molecules. Structure Determination Schbpf and his col leagues, in the 1930's, reported the i s o l a t i o n , separation, and pu r i f i c a t i on of several a lka lo ids (Table I, Page 5 ) - 13 -from the tox ic skin gland secretions of the alpine salamander. The gross structure 1_ was determined for samandarine in 1963 through chemical, 3 spectroscopic, and X-ray crysta l lograph ic studies. Structural invest igat ions of a newly i so lated natural product begin with the determination of the elementary composition. The com-pos it ion of samandarine was C 1 gH 3 iNo 2 . Samandarine (1_) i s a co lour les s , c r y s t a l l i n e , saturated secondary amine containing a secondary hydroxy 15 26 group which upon chromic acid oxidation affords samandarone (2_). ' The second oxygen occurs in an ether l inkage and there are two C-methyl. groups. Hofmann degradation of N-methylsamandarine methiodide (14) provided the f i r s t ins ight into the s t ructura l framework of these a lka lo ids (CHART I, Page 14). ' This degradative reaction yielded 1_5 which possesses a l l the carbon atoms of the methiodide, thus proving that the nitrogen atom of samandarine (1_) is present in a r i ng . Ca ta l y t i c hydrogenation of 1_5 afforded 16_. Compound 15^  exhibited s t a b i l i t y towards a l k a l i but on warming with d i l u t e sulphuric acid i t adds one molecule of water forming 17_. Chromic acid oxidation of 17_ furnished the lactone samandesone (18.). Furthermore, warming 27_with acet i c anhydride y ie lded the quaternary acetate 20 of the s ta r t ing material and 1_5_. This re su l t can be ra t iona l i zed as an a l ky l a t i on of the t e r t i a r y amine by an i n t e r -mediate acetate 1_9. These resu l t s established that the nitrogen and the ether oxygen atom are joined to the same carbon atom in samandarine. A series of chemical reactions provided for the elaboration of pa r t i a l formula 1_4 to formula 24_ (CHART I I , Page 15 ). Treatment of samandarine (1_) with l i th ium aluminum hydride y ie lded samandiol 2J_. While samandarine (1) i s stable towards lead tet raacetate, samandiol 21_ - 14 -CHART I . Structure E luc idat ion of the Alka lo ids CH, ,H3C-W I M ' •N O 1£L / / (CH 3 ) 2 N o 15 (CH 3 ) 2 N OH O H C - ^ / H 30 (CH 3 ) 2 N P 17 II 16 CCH 3) 2N O HOCH^ / CH 17 A c 20 (CH 3 ) 2 N P ACOC<4 c/ CrO, rT-(CH 3) 2N O ^ C H 1 8 (CH, ) 9 N W O + '3'2 e OAc 19 2 0 - 15 -CHART II. Structure Elucidat ion of the Alka lo ids - 16 -reacts with 1 mole of the reagent with formation of 1 mole of formalde-hyde. Thus, samandiol 2J_ must possess a — NH—CH2—CHOH— grouping as samandiol 2J_ does not contain a primary hydroxy group. In th i s reaction the Sch i f f base 22^  which i s f i r s t formed eliminates 1 mole of formalde-hyde and undergoes r ing closure to 23_. Combining a l l these resu lts 91 f i n a l l y led to pa r t i a l formula 21_ for samandiol and 24_ for samandarine. The next major phase in the e luc idat ion of the carbon skeleton of samandarine enta i led dehydrogenation of samandiol with selenium at 28 320-340°. The main product was i so lated as a c r y s t a l l i n e compound, CisHie- The u l t r a - v i o l e t spectrum of th i s showed i t to be a 1,2-dimethyl-5,6-cyclopentenonaphthalene (25J which was synthesized by the method shown in Chart II (see Page 15 ) . From the pa r t i a l formulae 24_ and 25_ one could deduce for samandarine a steroid skeleton with r ing A containing an oxazol id ine system. On th i s basis i t i s possible to wr ite down three t r i a l structures (26,27_,28) in which the pos it ion of the secondary hydroxy group i s s t i l l uncertain (see Page 17 ). The infrared spectrum of samandarone, however, revealed that the keto group formed by oxidation of the hydroxy group i s in a f i ve membered r ing (band at 1740 cm" 1 ) - Hence r ing D contains the hydroxyl f unc t i ona l i t y . X-ray analysis of samandarine hydrobromide, by the heavy atom 3 method, gave the structure 1_ corresponding to pa r t i a l structure 27_. In add i t ion , i t was apparent that samandarine (1_) has the same stereochemical configuration as the cho l i c acids. The salamander a lka lo ids are the f i r s t and only representatives with such a skeleton in the stero id 29 a l ka lo id s . It i s thus i n s t ruc t i ve to consider the d i s t i n c t i v e features of the related compounds. - 18 -O Samandarone (2) Samandarone i s the main a l ka lo id in S^ . maculosa maculosa; in S^ . maculosa taeniata i t i s one of the minor a l ka lo id s . Chromic acid oxidation of samandarine (1_) affords samandarone; conversely, i t i s reduced s te reo spec i f i c a l l y by sodium and alcohol to samandarine (1_). X 0-acetyl-samandarine (4) The structure of 0-acetyl-samandarine was elucidated from the infrared spectrum and from chemical invest igat ions. On saponi f icat ion i t affords samandarine (1_) and acety lat ion of the l a t t e r forms 0 - ace t y l -samandarine."^ - 19 -O Samandenone (5_) The infrared indicated the presence of an oxazol idine system. 31 Elemental analysis and spectroscopic methods gave structure 5_. Samandinine (6_) The structure of th i s base resu l t s from infrared and mass 32 spectral data. - 20 -Samandaridine (3) Chemical invest igat ions and the infrared spectrum revealed 26 the presence of a f i v e membered lactone r i ng . The structure was elucidated by X-ray analys is of the hydrobromide. The stereochemistry of samandaridine was confirmed by pa r t i a l synthesis (CHART I I I , Page 33 21). N-acetylsamandarone 29_ was condensed with g l yoxy l i c acid to give a mixture of two carboxyl ic acids 30 and 31_. Reduction of 30 with sodium borohydride followed by c a t a l y t i c hydrogenation proceeded stereo-s p e c i f i c a l l y to y i e l d samandaric acid (32). D i lute mineral acid lactonises 32 to afford samandaridine (3_). Biosynthesis Insight into the biogenesis of natural products has led to the development of "b iogenetic-type" synthesis. Probably, the most outstand-ing case in the f i e l d of a lka lo ids is the famous Robinson tropinone synthesis which demonstrated to the synthetic chemist that complex molecules could be constructed by u t i l i z i n g simple synthetic methods - 21 -CHART I I I. Pa r t i a l Synthesis of Samandaridine - 22 -under very mild condit ions. Furthermore, biogenetic synthesis are usual ly very short and e f f i c i e n t . This provides impetus to pattern a synthetic program along biogenetic l i n e s . Consequently, the mode of biosynthesis of the salamander a lka lo ids could guide the d i rec t ion of the synthetic approach. The biosynthesis of the salamander a lka lo ids has attracted considerable attent ion since t he i r structure was f u l l y elucidated in 1963. It has been well documented that cholesterol serves as a common 34 precursor in s tero id metabolism. For instance, cholesterol i s a key intermediate in the formation of testosterone, a s tero id sex hormone. Structural analys is of the a lka lo ids revealed that they possess a s tero id nucleus. In add i t i on , i t i s apparent that s t ructura l var iat ions in these compounds are due to the C-17 subst i tut ion pattern. Accordingly, they can be c l a s s i f i e d into three categories as i l l u s t r a t e d in Table III (see Page 23 ) . Further, the skin gland secretion has been found to 35 contain considerable quant it ies of cho les te ro l . As a re su l t of these considerations i t was proposed that cholesterol 33_ could be a key b io -synthetic intermediate. Since samandenone (5_) and samandaridine (3) bear substituents at C-17 they could represent intermediate stages in the biosynthetic pathway from cholesterol to the a lka lo ids without a side chain f unc t i ona l i t y at C-17. Indeed, Habermehl e_t al_., used 1 4 C labe l led acetate and cholesterol to demonstrate that the salamander a l ka lo id s or ig inate from acetate v i a cho les te ro l . Hence, i t seems a t t r ac t i ve to conjecture that a cho l i c acid intermediate 34 (CHART IV, Page 24) i s degraded by oxida-t ion to the isopropyl group which i s cha rac te r i s t i c of samandenine (5) - 23 -Table I I I. C-17 Subst i tut ion Pattern in Salamandra Alkalo ids Carbon-17 having 2 H's Samandarine (1_) Samandarone (2) 0-acetyl-samandarine (4) / C H 3 Carbon-17 having — C H ^ C H 3 Samandinine (6_) Samandenone (5) Carbon-17 having — C H ? — C — \ / Samandaridine (3) - 24 -CHART IV. Biosynthesis of the Side Chain in Salamandra A lka lo ids 37 - 25 -and samandinine (6_). The formation of samandaridine (3_) could be envisaged by further oxidation of 34_ to a malonic acid 36_ followed by decarboxylation and r ing closure to generate the appropriate lactone 3_7. It should be noted that in a s im i l a r fashion the skin glands of Bufo (toad) could generate the bufotal ine d ienol ide r ing 38 from carboxyl ic acid 34 without 37 loss of carbon atoms. These concepts are portrayed in Chart IV (see Page 24 ) . The other in te res t ing biosynthetic feature, which i s at a speculative l e v e l , embodies the elaboration of the 5g-steroid A r ing to the cha rac te r i s t i c b icyc looxazol id ine system. This enta i l s the reduction of the A 5 double bond, and insert ion of nitrogen between carbons 2 and 3 with the appropriate introduct ion of an ether l inkage between carbons 1 and 4. In th i s regard one can envisage a r ing f i s s i o n between carbons 2 and 3 followed by oxidation at the appropriate react ive s i t e s which sets the stage for an enzyme catalysed r ing closure to the natural skeleton. On the basis of t h i s conjecture i t , therefore, becomes apparent that the f i r s t major synthetic hurdle would be to e f fec t an unsymmetric r ing cleavage between carbons 2 and 3 in the A-r ing of 5g-stero ids . Assuming the foregoing, the task of d i f f e r e n t i a t i n g the chemical r e a c t i v i t y at carbons 2 and 3 should be workable. Hence, there i s now the p o s s i b i l i t y of performing the required types of transformations at s pec i f i c centres with ordinary chemical reagents. In f a c t , th i s genera? behaviour was simulated by developing a method to e f fec t an unsymmetric r ing cleavage between carbons 2 and 3 in the A r ing of 17e-acetoxy-5e-38 androstan-3-one. In summary, the biosynthetic map, although speculat ive, directed the approach to the synthetic problem on hand. - 26 -Synthesis 4 In 1967, Hara and Oka reported the to ta l synthesis of samadarone (2). This represented the f i r s t synthesis in the area of salamander a l ka l o id s . In t he i r synthesis they adopted the strategy of constructing a "key" intermediate 46_ from compound 39 v ia l-formyl-5/3-A-norandrost-l-en-17/?-ol (42) (CHART V, Page 27). They found that treatment of 46_with 75% acet ic acid at 100° afforded compounds 47a A and 48_. Compound 47a was transformed to samandarone (2) by the sequence of reactions depicted i n Chart V (see Page 27). As noted e a r l i e r , the Hara-Oka synthesis had l im i t a t i o n s . For example, t he i r synthesis y ie lded only mi l l igram quant i t ies of samandarone (2). However, the impract ical nature of t he i r synthesis does not detract from t he i r achievement but emphasizes the need for further work in th i s area. 39 Most recent ly , Shimizu has reported a new synthetic approach to samadarine-type a lka lo ids as depicted in Chart VI (see Page 29). The hydroxymethylene der ivat ive of 17/3-hydroxy-5/?-androstan-3-one (53) was elaborated to compound 57a by u t i l i z i n g the procedure of Autrey and S c u l l a r d ^ . The "key" reaction involved the conversion of 58_ to the 17#-hydroxy isomer of samandarine 47b in ca 60% y i e l d by employing sodium borohydride in re f lux ing isopropanol. Since a n i t r i l e group i s not a normal target of sodium borohydride reduct ion, Shimizu speculated that th i s reaction proceeded v ia formation of the c y c l i c amidine 59_ or iminoester 60, which would then undergo sodium borohydride reduction with concomittant c y c l i z a t i on to compound 47b. 41 It i s worthy of note that Eggart, Pascual, and Wehrli have reported the synthesis of the 5a-isomer of 47b in another attempt to make the s im i la r r ing system. 3. NaOH - 28 -I • / « N 6 o DISCUSSION 1. General Plan As previously mentioned, th i s thes is describes several attempts to gain a general method of entry into the bicyclooxazol id ine skeleton of general type I. It was hoped that an e f f i c i e n t method of synthesis would be developed which would provide material f o r b io log ica l studies and that new synthetic reactions would be tested. Compound 1_ poses a s t ructura l and stereochemical problem which is concentrated in the A r ing of the stero ida l framework. The synthetic problem const i tutes the elaboration of the stero ida l A r ing to the b icyc looxazol id ine skeleton. The desired subst i tut ion pattern in the D r ing could be achieved by employing appropriate synthetic 4 procedures. - 31 -Interest ing ly , structure J_ possesses nine asymmetric centres. Of these, seven are contiguous and range over the A, B, C, and D r ings. In planning the synthesis testosterone was chosen as the s ta r t ing material since i t i s a read i l y ava i lab le s tero id with a ve r s a t i l e D r ing sub-s t i tuent . Thus, the stereochemical problem associated with a synthesis of the b icyc looxazol id ine system reduces to f i x i n g the r e l a t i v e stereo-chemistry at carbons 1, 4, and 5 (Structure J_) with respect to the ex i s t ing centres. The close proximity of carbons 1, 4, and 5 of fers the opportunity to create these new centres, with the desired stereo-chemistry, under the inf luence of ch i r a l factors present in the precursors. F i r s t l y , the c i s re la t ionsh ip between the carbon-10 methyl and carbon-5 hydrogen was establ ished at the outset of the synthesis by 42 employing Liston s procedure to hydrogenate testosterone . Secondly, i t was hoped that the stereochemistry of the ether bridge between carbons 1 and 4 in the f i n a l product would be f ixed in the f i n a l stages of the synthetic pathway. With regard to the stereochemical problem on hand i t i s pertinent to examine the work of Hara and Oka^ 3. In 1969, they reported the stereoselect ive conversion of 46_ to 47a as i l l u s t r a t e d in 43 Chart VII (see Page 32). In summary, they accomplished the construc-t ion of the b icyc looxazol id ine skeleton by mild hydrolysis of a one to one mixture of the epimers, 2-benzylamino-3,3-ethylenedioxy-2,3-seco-5$-^ androstane-1, 173-diol (46J with 75% acet ic ac id . This reaction gave 3-benzyl-3-aza- la,4 a-oxido-A-homo-58-androstan-17e-ol (47a) and another substance 48 which was subsequently found to be a mixture of compounds 48a and 48b. Treatment of compounds 48a and 48b with hydrogen ch lo r i de -methanol y ie lded the methoxy acetal 61_. The act ion of 20% hydrochloric - 32 -CHART VII. Stereoselect ive synthesis of the b icyc looxazol id ine 47a — from intermediate 46 H 4 7 a R = PhCH 2 4 7 b R=H 61 - 33 -acid in aqueous acetone on compound 6J_ afforded 47a. The reaction mechanism (CHART VI I I , Page 34 ) which they invoked for th i s l a t t e r conversion involved the el iminat ion of the methoxy group and the concerted attack of a water molecule at the e lec t ron -def i c ient carbon-1 atom, giving r i se to inversion at the carbon-1 pos i t ion of compound 6J_, followed by b i c y c l i z a t i on of the resu l tant seco-oxy-aldehyde 62a to produce the thermodynamically stable system 47a. It i s worthy of mention that compound 48b did not undergo b icyc l i z a t i on to give compound 63_. In order for th i s to occur i t would be necessary for the "A" r ing of 48 b to be in a boat conformation. On the other hand inversion at carbon-1 pos it ion with formation of compound 47a presumably occurs due to the fol lowing thermodynamic factors . In compound 48b there is a severe non-bonded interact ion between the equatorial side chain and the carbon-11 making i t thermodynamically less stable than compound 62b. In add i t ion, the side chain in compound 62b i s ax ia l and the opportunity ex i s t s for i n t r a -molecular hydrogen bonding between the primary amine and the hydroxy group of the hemiacetal. The energy gained in the formation of th i s type of hydrogen bond could be as strong as 5kcal/mole. With the formation of compound 62b a favourable s i tuat ion has been reached for c y c l i z a t i o n . Removal of the benzyl group from 47a was achieved by c a t a l y t i c hydro-genation which gave 47b in quant i tat ive y i e l d . In view of the foregoing resu l t s i t was p laus ib le to conceive the stereoselect ive synthesis of the samandarine nucleus v ia an intermediate c lo se ly analagous to 46. From the outset, the hypothetical intramolecular c yc l i z a t i on (eq. I, Page 35) const ituted the central feature of the projected synthesis. There i s analogy for th i s type of intramolecular c y c l i z a t i o n . In 1969, Wasserman - 35 -I O 4 7 b and Barber described the thermal rearrangement of 6,s-epoxy ketones to the [3.2.1] b i c y c l i c system. Thus, 6,7-epoxy-2 heptanone (64) may be thermally transformed into l -methyl-7,8-dioxabicyclo [3.2.1] octane (65) in 75% y i e l d . They extended and used th i s general type of thermal 6 4 , R ( = R 2 = H J S S . R ^ C2H5-, R 2= H 6 § , R, = R 2 = H 6 7 , ^ = C 2 H 5 i R 2 = H 6 8 , ^ = H i R 2= C 2 H 5 - 36 -rearrangement to synthesis brevicomin, the p r i nc ip le sex a t t r ac t of the western pine beetle. Thus, cis-6,7-epoxy-nonan-2-one (66_) could be thermally transformed into a mixture of exo-6-ethy l - l -methyl -7-d ioxa-bicyclo-[3.2.1] octane (67) (90%) and the corresponding endo isomer 68 (10%). The exo isomer was ident i ca l with brevicomin. Although the mechanistic de ta i l s of th i s carbonyl epoxide rearrangement remain to be explored, i t seems c lear from the above resu l t s that during thermolysis of the 6,e-epoxy ketones the epoxide r ing undergoes r ing opening pre-dominantly with inversion of conf igurat ion. In add i t ion, Demole and 45 46 47 Wuest , and others ' have reported the acid catalysed c yc l i z a t i on of 6,e-epoxy ketones. For example, epoxy ketone 69_ i s converted to a 1:9 mixture of 70. and 7]_ in the presence of jJ-toluenesulphonic acid while with s i l i c i c acid 70 and 71 are formed in the r a t i o of 3:1. 71 - 37 -The d i s t i n c t i v e features of the thermal epoxy ketone rearrange-ment are the e f f i c i ency and s te reo se lec t i v i t y of the conversion. Con-sequently, the a z i r i d i ne aldehyde 1_0 was proposed as a possible "key" intermediate for gaining a method of entry into the b icyc looxazol id ine skeleton. The scheme i s portrayed, in general terms, by the hypothetical I O 4 7 b cyc l i z a t i on of lfJ to 47b (eg. I ) . In p r i n c i p l e , t h i s transformation represents the nitrogen analogue of the 6,e-epoxy ketone rearrangement. 48 Doughty and his colleagues have prepared oxazolidines by an i n t e r -molecular reaction in f a i r y i e l d . For example, a z i r i d i ne 72^  reacted with acetaldehyde (73) to afford oxazol idine 74_. It is well authenticated V ~ 7 + C H , C H O *- > N 3 7 2 7 3 7 4 - 38 -that intramolecular cyc l i za t i on s are energet ica l ly more favourable than intermolecular c y c l i z a t i on s . Hence reaction I appeared a t t r ac t i ve and potent i a l l y e f f i c i e n t . However, four products, 47b, c_, d_, and e_> could 43 45 46 47 be formed. In view of the work of Hara and Oka, and others ' ' i t OH OH H H 47d 4 7 e was f e l t that i f s t ructura l isomerism or stereochemistry proved to be a problem, control could be introduced by varying the reaction conditions or, u l t imate ly , by modifying the "key" intermediate 1_0. The theoret ica l cleavage of several bonds of the target mole-cule with appropriate funct iona l i za t ion afforded four probable candidates 75, 76, 77, and 78_, for eventual c y c l i z a t i on to the bicyclooxazol id ine - 39 -system 47b (CHART IX, Page 40). Several other intermediates were d i s -carded because of s t ructura l and chemical complexit ies. It was hoped that a z i r i d i ne 75_, bearing the appropriate f unc t i ona l i t y X, could give r i s e to the intermediate ITJ. As a re su l t a z i r i d i ne 75_ was recognized as an important s t ructura l un i t through which our pathway could pass. It i s worthy to note that compounds 77_ and 78 are c lo se ly analogous to the "key" intermediates 46_ and 58_ employed in the Hara - Oka and Shimizu syntheses, respect ive ly . I f the a z i r i d i ne c yc l i z a t i on OH OAc approach f a i l e d i t could possibly be modified to embody an intermediate s im i l a r to e i ther compounds 76_, 77_ or 78_. In a c t u a l i t y , the o r i g ina l strategy did suf fer a reversal in th i s d i r ec t i on . One approach to the synthesis of a z i r i d i ne 75_ involved the formation of precursor 79_with the appropriate func t i ona l i t y X. I t was planned to construct o l e f i n 79_ from an intermediate of general type 80_ which could be derived from 17e-acetoxy-5£-androstan-3-one (81 b) by e f fec t ing an unsymmetric cleavage between carbons 2 and 3 in the A r i ng . It i s p a r t i c u l a r l y important that the f unc t i ona l i t i e s X and Y would set the stage for incorporating the o l e f i n i c bond of 79^ in an e f f i c i e n t manner. - 41 -81a, R = OH. 81 b, R = OAc This prospective synthetic plan, therefore, was concerned with two major object ives. The f i r s t object ive would be to develop a method to e f fec t an unsymmetric cleavage between carbons 2 and 3 in the A r ing of 56-steroids. The second major phase of the program would const i tute the elaboration of the r ing cleavage product to the b icyc looxazol id ine skeleton v ia one of the "key" intermediates which have been alluded to. Accordingly, th i s thes is i s divided into two parts. Part I, describes three general approaches directed towards e f fec t ing the desired r ing cleavage react ion. The th i rd approach which was investigated proved to be the method of choice. Part I I, describes attempts to elaborate - 42 -cyanoaldehyde 195 to the 176-hydroxy isomer of samandarine 47b. 195 47b I t i s now in order to turn from the foregoing general analys is to s pec i f i c synthetic de ta i l and development. A review of the l i t e r a t u r e revealed that two general methods could be employed to e f fec t cleavage of the 2,3 bond in the A r ing of 5g-steroids. F i r s t l y , Schmidt reaction of ketone 81b and Beckmann rearrangement of the oximes of ketone 81b effected r ing expansion with incorporation of nitrogen to furnish two 49 isomeric lactams 82c and 83c in a r a t i o of ca_. 1:1. Unfortunately, 8 2 a . R = OH 8 2 b. R = OTS 8 2 c , R = OAc 8 3 a . R = OH 8 3 b. R = OTS 8 3 c . R= OAc - 43 -50 the separation of lactams 82c and 83c i s a most d i f f i c u l t task. Even i f separation could be e f f e c t i v e l y achieved the required lactam can only be procured in less than 50% y i e l d . In short, th i s does not represent an e f f i c i e n t process. S im i l a r l y Baeyer -V i l l i ge r oxidation of 3-oxo-61 steroids gave a mixture of lactones. Secondly, ozonolysis of 2-hydroxy-51 methylene 53_ led to a symmetric cleavage of r ing A, equation II. OH C lea r l y , neither of these general procedures were compatible with the present needs. However, three general p r inc ip le s embedded with in these transformations were recognized. F i r s t , with regards to the Beckmann rearrangement and the Baeyer -V i l l i ge r react ion , the insert ion of a hetero atom between carbons 2 and 3 sets the stage for a f a c i l e unsym-metric r ing opening. Second, a f te r considering transformation II i t was decided to e f fec t an oxidative cleavage between carbons 2 and 3. Again, the stage would be set for unsymmetric r ing cleavage as depicted in equation I I I . F i n a l l y , funct iona l i za t ion of carbon-2 with subsequent introduction of a convenient "handle" at carbon-2 could lead to r ing 52 cleavage in the desired fashion. Accordingly, these three general concepts set the course in the f i r s t phase of the projected synthesis. The th i rd concept, namely, the funct iona l i za t ion of carbon-2 proved to be the method of choice. - 45 -Part I a. Insertion of a hetero atom between carbons 2 and 3 50a 53 Oka and Hara ' in 1968 reported the separation of syn and anti isomers of several steroidal 3-ketoximes and succeeded in d i f f e r -ent ia t ing between the geometrical isomers by means of n.m.r. spectro-scopy using chemical s h i f t s of the protons attached to the a-carbon 53a atoms. Analysis of the methylene protons adjacent to the oximino carbon atom in c y c l i c ketoximes had not been demonstrated except for a few cases of simple der ivat ives of cyclohexanone and cyclohexenone. These s igna l s , in the region of T8.2 to T 7 . 5 , are not c l e a r l y separated from the s ignals of other methylene or methine protons. In the case of a,6-unsaturated stero idal ketoximes., the assignment was due to the chemical s h i f t of the o l e f i n i c proton rather than the methylene protons. 53 a In 1968, Oka and Hara reported the analysis of the 100 MHz n.m.r. spectra of the methylene protons adjacent to the oximino carbon atom of 3-oxosteroid oxime der i vat i ves . The compounds used were pure isomeric forms of oximes and 0-methyloximes of testosterone, 17a-methyl-testo-sterone, 19-nortestosterone, 17a-ethyl-19-nortestosterone, A-nortesto-sterone, and 5 a - and 5B-androstanolones. F i r s t l y , the syn isomers of six-membered unsaturated 3-oxosteroid oximes had a four proton unresolved mul t ip le t at T7.9 to T7.6 which was assigned to C-2 and C-6 methylene hydrogen atoms. On the other hand, the corresponding ant i isomers showed one proton sh i f ted downfield from the three-proton unresolved mul t ip le t by 90 Hz. Its coupling pattern, - 46 -doublet (J = 18.5 Hz, centred at T7.0) of t r i p l e t (J = 4.8 Hz) indicated the signal to be due to the C-2a (equatorial) hydrogen atom. Consequently, the three proton mul t ip let at T 8.0 to T7.6 was assigned to C-2B (ax ia l ) and C-6 methylene hydrogen atoms. In contrast, the C-l methylene hydrogens of the f i ve membered unsaturated s tero ids , for example, A-nortestosterone were observed as a two proton s ing le t at ca. T 7.5 for e i ther syn or anti isomers, and only a 3 Hz downfield s h i f t was found for the anti isomer. The C-6 a l l y l i c methylene hydrogens (T7.74 and T7.52) are coupled to one another (J = 13 Hz) and show further s p l i t t i n g from the o l e f i n i c and the C-7 53a methylene hydrogens. Hara and Oka considered the d i f ference between s i x - and five-membered ring systems to be due to a d i f f e ren t anisotropic e f fec t of the oxygen atom, as i s read i l y understood when molecular models are constructed (CHART X, Page 47). A C-2 a (equatorial) hydrogen of a six-membered r ing may be influenced strongly by the anisotropic e f fect o because of i t s shorter distance (ca. 2.3A) from the oxygen atom than that o of the C-2g (ax ia l ) one (ca. 3.4A). For the five-membered r i n g , C - l a o and 3 hydrogens may be equidistant from the oxygen atom (cja. 3.1 A) and no d i f ference in chemical s h i f t between these two hydrogens i s observed. 53a Hara and Oka applied these observations to the assignment of the geometrical isomers of unsymmetrical saturated oxime der ivat i ves . They quant i tat ive ly separated 173-hydroxy-5g-androstan-3-one oximes into the syn and ant i isomers 87_ and 88, respect ive ly . The n.m.r. spectrum of pure syn oxime 87_ displayed a pa i r of doublets at about T 7 ( i n tens i t y one proton) assigned to carbon-4 equatorial hydrogen from the AMX coupling pattern. The n.m.r. spectrum of the ant i oxime 88 had a pa i r of t r i p l e t s at about T7 ( i n tens i t y one proton) assigned to carbon-2 equatorial hydrogen CHART X. - 47 -Structural Aspects of anti-form oximes i - f o r m - 48 -OH ant i 8 8 from the AMX2 coupling pattern. The n.m.r. spectrum of the i n i t i a l mixture indicated that the syn-anti composition was in accord with the y ie ld s obtained a f te r chromatography. In terest ing ly , the pure syn or ant i oxime, 87^  or 88_, was transformed into a mixture of oximes by heat-ing to t h e i r melting points or warming in polar solvents. Furthermore, O H R s y n 8 7 8 2 a , R = O H they effected the stereoselect ive synthesis of 17B-hydroxyl-3-aza-A-homo-58-androstan-4-one (82a) v ia s p e c i f i c Beckmann rearrangement of 53b 50a the pure syn oxime 87 in almost quantiat ive y i e l d . Oka and Hara u t i l i z e d these resu l t s to e f fec t the to ta l synthesis of samanine (8) as - 49 -represented in CHART XI (see Page 50). Af ter separation of the oximes 89 and 9fJ they converted the ant i oxime 90 into an equi l ibr ium mixture of syn and anti forms. The syn oxime 89 was obtained in almost quant i -t a t i ve y i e l d from the ant i isomer 90 by repet i t i on of th i s procedure followed by s i l i c a gel column chromatography. In contrast, the mixture of lactams 91_ and 92 prepared from the oximes 89 and 90 gave a s ing le product on t . l . c . The separation of lactams £1_ and 92 proved so d i f f i -c u l t that i t was not appl icable to the spec i f i c synthesis of samanine (8 j . However, Habermehl ejt aj_. in 1969 reported the separation of both isomers i n gram amounts by using alumina. Nevertheless, these resu l t s were encouraging insofar as they suggested that the separation of the syn and anti oximes 89 and 9fJ could be u t i l i z e d in developing an e f f i c i e n t method to e f fec t r ing cleavage in the required fashion. Since 53a Hara and Oka did not report quant i t ies of mater ia l , the r e l i a b i l i t y and e f f i c i ency of t he i r approach was tested on a large scale. To th i s end, hydrogenation of testosterone (93_) under ac id i c conditions gave ketones 81a and 94_ in ca. a 3:1 r a t i o as indicated by - 50 -CHART XI. Total synthesis of Samanine (8) - 51 -42 vapour phase chromatography of the acetylated mixture. Several H 2/Pd 8 1 a . R = OH 81 b, R = OAc 9 4 a . R= OH 49a rec ry s ta l ! i z a t i on s afforded ketone 81a in 22% y i e l d as a white c r y s t a l l i n e s o l i d , m.p. 138-140° ( l i t . 5 4 m.p. 139-140°). Treatment of ketone 81a with hydroxylamine hydrochloride^ and sodium acetate in re f lux ing methanol fo r three hours afforded a mixture of oximes 87_ and 88_, in 60% y i e l d , as a white c r y s t a l l i n e s o l i d , m.p. 210-214° ( l i t . 5 3 a m.p. 211-213°). Of note was the d i s -appearance in the inf rared spectrum of the saturated carbonyl absorp-t ion at 1708 cm" 1 and the appearance of a weak absorption at 1650 cm" 1 - 52 -due to the oxime f unc t i ona l i t y . The n.m.r. spectrum of the mixture of oximes 87_ and 88_ had s ignals at x9.03 and x 9 . 27 as two three-proton s ing lets due to the carbon-19 and carbon-18 t e r t i a r y methyl groups, respect ive ly . A one-proton t r i p l e t at x6.35 (J = 9 Hz) was assigned to the proton adjacent to the 1 78 - hydroxy group and a one-proton mul t ip le t appearing at ca_. x7 could be assigned to the carbon-4 and carbon-2 equatorial hydrogens of the syn and anti isomers 8_7 and 8 8 , respect ive ly. T.L.C. analys is of the reaction product with a var iety of solvent systems showed a s ingle spot. However, t . l . c . examination of the product on s i l i c a gel with benzene/ethyl acetate (4/1) as the solvent system revealed the presence of two components. Presumably, the syn and anti oximes 87_ and 88 had been separated as in accordance 53a with the work of Oka and Hara . The i s o l a t i on of the syn and anti oximes was achieved by employing preparative t . l . c . on s i l i c a ge l . The i so lated y i e l d s in t h i s case were very low. Attempts were made to separate the oximes 87 and 88_ by column chromatography on s i l i c a ge l . - 53 -Elution with benzene/ethyl acetate (4/1) gave a c r y s t a l l i n e s o l i d , m.p. 53a 211-213° ( l i t . m.p. 211-212°), and a slower moving compound, m.p. 210-213° ( l i t . 5 3 a m.p. 211-213°) in ca_. a 1:1 r a t i o . 5 0 a However, the n.m.r. spectrum of the f i r s t f rac t i on had a one proton unresolved mul t ip let at ca_. x7 which did not appear to be cha rac te r i s t i c of e i ther the syn or anti oximes. In add i t i on , the C-19 and C-18 t e r t i a r y methyl groups were c l ea r l y evident at x9.05 and x9.28, respect ive ly , as two sharp three-proton s ing lets and a t r i p l e t at x6.40 (J = 9 Hz) was assigned to the proton adjacent to the 173- hydroxy group. S im i l a r l y , the n.m.r. spectrum of the second f ract ion displayed a one proton mul t ip le t at ca. x7 which was not cha rac te r i s t i c of e i ther the syn or anti oximes. Further, the n.m.r. spectrum had two sharp three-proton s ing lets at x9.05 and x9.27 due to the C-19 and C-18 t e r t i a r y methyl groups, respect ive ly . Hence, n.m.r. spectroscopy indicated that both deuterochloroform solutions contained a mixture of the syn and ant i oximes 87_ and 88. I t appeared that isomerization had taken place in the n.m.r. tube. Even more annoying was the knowledge that they were insoluble in carbon te t rach lo r ide , benzene.and other non-polar solvents in which isomerization would not be expected to occur. At th i s stage, i t appeared that n.m.r. spectroscopy was not an e f fec t i ve means to d i s t ingu i sh between the syn and ant i oximes 87_ and 88. In summary, our resu lts indicated that the chromatographic separation of the syn and anti oximes 8_7 and 88^  was not amenable to large scale preparations. However, since about 200 mg. of the separated syn and anti oximes 87_ and 88 had been obtained by small scale column chromatography, i t was decided to attempt the Beckmann rearrangement of the syn oxime Q 7 50a,53b - 54 -Toward th i s end, treatment of syn oxime 87_ with three molar equivalents of p-toluenesulphonyl chlor ide in about 200 molar equivalents of dry pyridine at room temperature for two days afforded lactam 82a in 65% y i e l d . A small sample was r e c r y s t a l l i z ed to constant m.p. 241-242° O H R N O H s y n§7_ 8 2 a , R = O H ( l i t . m.p. 242-244°). The crude reaction product was homogeneous on t . l . c . ( s i l i c a gel and alumina). The infrared spectrum of the crude product indicated that the Beckmann rearrangement had indeed occurred. Thus, the infrared spectrum of 82a had a strong absorption at 1660 cm" 1 due to the lactam carbonyl and a broad absorption at 3440 cm" 1 due to the 178- hydroxy group and the NH stretching v ibrat ion of the lactam group. Anti oxime 88_ in acetone at room temperature was converted into an equi l ibr ium mixture of syn and anti forms as evidenced by t . l . c . on s i l i c a gel with the appropriate solvent system. Because of the d i f f i c u l t chromatographic separation of the syn and anti oximes 8_7 and 88_ no attempt was made to recover the syn oxime 87_ from the anti oxime 88_ by using s i l i c a gel chromatrography. Although the Beckmann rearrangement had been effected th i s pa r t i cu la r type of approach was rejected because of the poor overa l l y i e l d of lactam 82a based on ketone 81a, - 55 -coupled with the d i f f i c u l t chromatographic separation of the syn and anti oximes 87 and 88_, respect ive ly , 55 At th i s point, Uskokovic and his colleagues reported a new and highly promising fragmentation reaction (see CHART XII, Page 56 ). They found that pyrolys i s of N-nitrosolactam 96_ at 125° afforded the 56 diazolactone 97_, which fragmented with extrusion of nitrogen to give a mixture of compounds 98 and 99_ in 50 and 30% y i e l d , respect ive ly , 55 Uskokovic and his colleagues ra t i ona l i zed the formation of these pro-ducts by invoking two d i f f e ren t concerted fragmentation paths a and b. Path a, could probably be i n i t i a t e d by abstract ion of the C - l l hydrogen by e i ther one of the ester oxygens which could afford acid 98^  (R = H). Path b, could be envisaged as a nuc leoph i l i c attack of one of the ester oxygens on the C-10 carbon to give lactone 99_. They proceeded to t rans-form lactone 99 to ester 98_ (R = OMe) via the hydroxy acid 100 in f i v e simple steps. Consequently, i t was hoped that lactam 82a could be converted into o l e f i n i c ester 104a as depicted in Chart X I I la (see Page 57 ). Such a sequel represents the unsymmetric cleavage of the 2,3 bond with incorporation of an o l e f i n i c bond as in accordance with the general plan. The f i r s t problem associated with such a sequel involved the separation of lactams 82a and 83a. However, separation of the isomeric lactams on s i l i c a gel proved f u t i l e . Only pa r t i a l separation could 50b 57 be achieved by using alumina. ' Others have found d i f f i c u l -50a t i e s with the separation of 4-aza- and 3-aza-lactams. This - 57 -CHART XI I la. N-nitrosolactam Investigations N 2 0 4 O N — N 8 2 a , R= OH 8 2 b. R = OTS 8 2 c . R = OAc 101a. R = OH 1Q1 b. R = OTS 1Q1c.R= OAc 1Q4a. R = OH; X= Me 1Q4b, R=0TS-,X = IV1e l Q 4 c . R = OAciY = IV1e lQ4d, R = OTSvX = H 1Q3a. Y = OH 1Q3bT Y = OTS 1Q3c, Y = I JLQ2ji, R = OH 1 Q 2 b . R =OTS t Q 2 c . R = OAc - 58 -82a. R = OH 83a, R = OH chromatographic study was not pursued any further since the thermal fragmentation reaction was the "key" feature of th i s approach. I t was decided to launch synthetic invest igat ions with the mixture of lactams 82a and 83a and i t was f e l t that i f a su i table route was establ ished by u t i l i z i n g the mixture of lactams 82a and 83a i t would be appl icable to lactam 82a. Treatment of the mixture of oximes 87_ and 88 with three molar equivalents of p-toluenesulphony1 chlor ide in about 200 molar equivalents of dry pyridine at room temperature for f i ve days afforded lactams 82b and 83b, in 65% y i e l d , as a white c r y s t a l l i n e s o l i d , m.p. 200-202°. The product was homogeneous on t . l . c . ( s i l i c a ge l ) . 82 b. R = OTS 83b, R = OTS - 59 -In add i t ion, the spectral properties of the mixture of lactams 82b and 83b were in accord with the assigned structures. Thus, the infrared spectrum of compounds 82b and 83b showed strong absorptions at 1190 cm" 1 (doublet) due to the tosy late group and at 1660 cm" 1 due to the lactam carbonyl. A weak absorption was evident at 3400 cm" 1 due to the N-H stretching v ibrat ion of the secondary amide. The n.m.r. spectrum of the mixture of lactams had three three-proton s ing lets at T9.20, T9.0, and T7.53 due to the C-18, C-19 t e r t i a r y methyl groups, and the 176-tosyl methyl group, respect ive ly . Other assignable s ignals in the n.m.r. spectrum appeared at T5.70 as a one-proton t r i p l e t ( J = 10 Hz) due to the C-17 proton and at T3.50 one exchangeable proton. A AB pa i r of doublets (J = 8 Hz) appeared at x2.64 and t2.17 due to the four aromatic protons of the 178-tosylate group. F i n a l l y , the mass spectrum had a molecular ion peak at ^ 459. In view of the previous work i t was annoying and somewhat surpr i s ing to r ea l i ze that the 176- hydroxy f unc t i ona l i t y had been con-50a verted into a tosy late group. Repeating the Beckmann rearrangement several times under various conditions revealed that tdsy lat ion always occurred when working with greater than 200 mg quant it ies of oximes 87_ and 8JL With considerable quant i t ies of lactams 82b and 83b on hand, i t 82 b. R = OTS 83b, R = OTS - 60 -was decided to continue our studies with these substances hoping at a l a t e r stage to reintroduce the 176- hydroxy group. Thus, treatment of lactams 82b and 83b with dinitrogen tetrox ide afforded in quantiative y i e l d a mix-ture of N-nitrosolactams 101b and 105b as an unstable yel low powder (CHARTS C O XII l a and XH Ib , Pages 57 and 61). Of note was the disappearance in the infrared spectrum of the weak band at 3400 cm" 1 and the appearance of strong absorptions at 1720 cm" 1 (N-nitrosolactam) and 1415 cm" 1 (N itroso). T.L.C. analys is of the crude reaction product revealed the presence of one broad spot. The time was now at hand to tes t the f e a s i b i l i t y of the a n t i c i -pated fragmentation react ion. In the event, pyrolys i s of a small sample of the mixture of N-nitrosolactams 101b and 105b at 125° for two minutes afforded two major products as indicated by t . l . c . ana lys i s . When the pyro lys i s was car r ied out with larger amounts of material the resu l t s were quite reproducible. The inf rared spectrum of the crude reaction product displayed a strong absorption at 1720 cm" 1 . The n.m.r. spectrum of the crude product indicated the absence of o l e f i n i c protons. F i n a l l y , the mass spectrum showed a mole-cu lar ion peak at -^ 460. This spectroscopic data coupled with previous work in th i s area, tended to suggest that lactones 102b and 106b were being formed. 56,59,60 p Q r e x a m p - ] e j N a c e et_al_. converted lactam 109 to lactone 110 by 59a employing a procedure analagous to that used by White. The formation of the O, O H 1 0 9 H 1 1 0 - 61 -CHART X I l i b . N-nitrosolactam Investiqations 1Q8Q. R-= H-. X=Me 1Q7g. Y = OH 1Q6g . R = OH 1Q8 b. R = OTS; X = Me 1Q7b, Y = OTS 1 0 6 b , R = OTS 1 Q 8 c . R=OAc;X=Me 1p7c , Y = I 1 Q 6 c . R = OAc 1Q8d T R= OTS-,X = H - 62 -lactones 102b and 106b was confirmed by the fact that the products from the pyrolys i s reaction could be corre lated with the lactones derived 51 6 from the Baeyer -V i l l i ge r oxidation of the appropriate 3-oxo-5e-steroid. ' Since the o l e f i n i c acids 104d and 108d (CHARTS XHIa and XII Ib, Page 57 and 61 ) were not being generated i t was decided to study the fragmentation reaction under various conditions in order to suppress the formation of the lactones and induce the desired fragmentation. At lower temperatures the reaction was too slow to be of any u t i l i t y . At elevated temperatures several side products emerged which thwarted the i so l a t i on of the major components in a high state of pur i ty . Af ter t ry ing several chromatographic systems two side products were i so lated which did not d isp lay any o l e f i n i c character. In add i t ion , the f r ag -mentation reaction was effected in various solvents, for example, toluene, dimethylformamide, and hexamethylphosphoramide. The addit ion of acid to the reaction medium did not a l t e r the mode of fragmentation but tended to furnish more undesirable contaminants. In summary, by varying reaction conditions attempts to produce o l e f i n i c acids 104d and 108d were to no a v a i l . E s sent ia l l y the reaction products always consisted of lactones 102b and 106b. It became c l e a r l y apparent that the most e f f i c i e n t method to obtain these type of compounds would be v ia the Baeyer -V i l l i ge r oxidation of the appropriate 3-oxo 5B-steroid. Lactone 102c could then be converted to the o l e f i n i c ester 104a by employing procedures analogous 55 to the work of Uskokovic and his colleagues. Before proceeding to invest igate the Baeyer -V i l l i ge r approach, experiments were carr ied out to convert the 17B-tosylate group into the 178- hydroxy f unc t i ona l i t y in order to perform the necessary chemical co r re l a t i on . However, the removal of the tosy late group was found to be - 63 -a formidable task. Under mi ld l y basic condit ions, fo r example, the lactones suffered r ing opening. Unfortunately, attempts to remove the to sy l a te group by employing more drast ic methods only led to i l l - d e f i n e d products bearing a tosy late group. In order to circumvent these d i f f i -c u l t i e s i t was proposed to mask the 178- hydroxy group as an acetate f unc t i ona l i t y at the beginning of the sequence. As a r e su l t , N-nitroso-lactam acetates 101c and 105c were prepared (CHARTS XI I la and X I I lb , Pages 57 and 61 ). The infrared spectrum of 101c and 105c had strong absorption bands at 1720 cm" 1 due to the N-nitrosolactam carbonyl and the 173- acetate f unc t i ona l i t y . The nitroso bands were evident between 1385 and 1360 cm" 1 . As before, the thermal fragmentation of compounds 101c and 105c gave the.mixture of lactones 102c and 106c. The infrared spectrum of 102c and 106c had a band at 1720 cm" 1 due to the lactone carbonyl and the 178- acetoxy group. Of note was the disappearance of the nitroso bands at 1385 cm" 1 and 1360 cm" 1 . The n.m.r. of 102c and 106c had two three-proton s ing lets at T9.20 and x8.97 due to the carbon-18 and carbon-19 t e r t i a r y methyl groups, respectively, and a three proton mu l t ip le t (between x6.20 - x5.20) could be att r ibuted to the proton adjacent to the 178- acetoxy group and the methylene protons adjacent to the lactone oxygen atom in the "A" r i ng . The mass spectrum of the mixture of lactones 102c and 106c had a molecular ion peak at — 348. Having obtained lactones 102c and 106c the Baeyer-Vi11iger oxidation of ketone 81b was carr ied out. Burckhaldt and R e i c h s t e i n ^ 0 reported that Baeyer -V i l l i ger oxidation of 3-oxo 58-steroids gave a s ingle lactone, namely, a 3-oxo-4-oxa-A-homo compound. On the other 61b hand Hara and his colleagues re-examined the Baeyer -V i l l i ger oxidation - 64 -R H 81b, R = OAc of 3-oxo 58-steroids and showed that the re su l t ing product was a mixture of lactones although the infrared spectrum and other physical constants of the product seemed to suggest i t to be a s ingle compound. Evidence for i t being a mixture were provided by the hydrolysis of the lactone, e s t e r i f i c a t i o n with diazomethane, oxidation with chromium t r i o x i de and separation of the d iac id by pa r t i a l e s t e r i f i c a t i o n in the presence of a d i l u t e ac id . They studied the act ion of perbenzoic acid on 178-acetoxy-58-androstan-3-one (81b) and found that both lactones 102c and 106c were formed but they did not report y i e ld s (CHART XIV, Page 65 )• Re invest i -gating the Baeyer -V i l l i ge r oxidation of ketoacetate 81b with m-chloro-perbenzoic acid in chloroform for two days at room temperature gave a mixture of lactones 102c and 106c in 85% y i e l d as a white c r y s t a l l i n e s o l i d , m.p. 118-125°. A small sample was r e c r y s t a l l i z e d from methanol I Q 2 c t R = OAc 106c_, R = OAc - 66 -m.p. 207-208°, for ana lys i s . The infrared and n.m.r. spectral properties of the crude lactones were ident i ca l with the products derived from the thermal fragmentation reaction (CHART XII l a and XH Ib , Pages 57 and 61 ). In add i t ion , the i r th in layer and vapour phase chromatographic behaviour were i d e n t i c a l . The v .p .c . showed that the Baeyer -V i l l i ge r oxidation pro-duct was a mixture of lactones in ca. a 4:1 r a t i o , 6 ^ while the corres-ponding mixture from the fragmentation reaction exhibited ca_. a 2:3 53a r a t i o . Thus, these resu lts indicated that the proposed synthetic scheme as outl ined in CHART XII la had proved unproductive. The synthetic pathway suffers from two major drawbacks. F i r s t l y , the a v a i l a b i l i t y of pure lactam 82a appeared to o f fe r a d i f f i c u l t problem. The economy of R 8 2 a . R = OH th i s process was not a t t r a c t i v e . Secondly, the central feature of the sequel proved unpromising since the o l e f i n i c esters 104c and 108c were not formed by the thermal fragmentation reaction and instead only lactones 102c and 106c were obtained (CHARTS XHIa and XII Ib, Pages 57 and 61 ). I t was,therefore,decided to abandon t h i s type of approach. As previously noted, however, an a lternate method evolved from these studies, namely, the p o s s i b i l i t y of u t i l i z i n g lactone 102c. In view of the d i s t r i bu t i on - 67 -of the isomeric lactones in the crude Baeyer -V i l l i ge r oxidation product the most pressing problem was to ascertain the structure of the major lactone. Undoubtedly, th i s re su l t would determine the v a l i d i t y of the Baeyer -V i l l i ger approach. In f ac t , the s t ructura l problem was solved by taking advantage of part of the synthetic plan (CHART XII l a , 102c -> 104a, Page 57). This also provided the opportunity to test part of 55 the proposed synthetic pathway. Uskokovic and his colleagues proved structure 111 by the degradation into the o l e f i n i c ester 112. In the sequel (CHART X l l l a , Page 57) , lactone 102c would afford 104a. On the other hand lactone 106c would furnish 108a (CHART XH Ib , Page 61 ). Attempts to develop a p rac t i ca l chromatographic method for e f fect ing - 68 -separation by preparative t . l . c . on s i l i c a gel or alumina with various solvent systems gave at the very best only pa r t i a l separation. S im i l a r l y , with lactones 102a and 106a only pa r t i a l separation could be achieved. However, a f te r several f r ac t i ona l c r y s t a l l i z a t i o n s separation of 102c and 106c was achieved as evidenced by t . l . c . and v.p.c. examination, but th i s method was rejected because of poor y i e l d s . Inevitably, the proposed chemical operations would have to be performed on the mixture of lactones 102c and 106c. It was now hoped that the re su l t ing mixture of o l e f i n i c esters 104a and 108a would be more amenable to chromatographic separation. Ring opening of the lactones 102c and 106c was effected by t reat ing the mixture with 5% sodium hydroxide in re f lux ing methanol for two hours. The reaction mixture, a f te r work up, was e s t e r i f i e d with diazomethane to afford the hydroxy methyl esters 103a and 107a, in 75% y i e l d , as an o i l (CHARTS XHIa and XII Ib, Pages 57 and 61 ). The spectroscopic data of th i s material were in accord with the assigned structures. The pert inent spectral features in the n.m.r. were as fo l lows: s ing lets were evident at x9.02 and T8.93 due to the C-l9 - 69 -t e r t i a r y methyl groups of hydroxy methyl esters 103a and 107a, respect ive ly , in an approximate r a t i o of 1:3. In contrast, the C-18 t e r t i a r y methyl groups appeared as a s ing let at T9.27 for each compound. In add i t ion, a s ing let and mul t ip le t at T6.32 and x6.35 ( t o t a l l i n g s ix protons) could be att r ibuted to the methyl ester f unc t i ona l i t y and the protons adjacent to the hydroxy groups, respect ive ly , in compounds 103a and 107a. F i n a l l y , the mass spectrum indicated a molecular ion peak at 338. Tosylation of the crude hydroxy methyl esters 103a and 107a 64 with tosyl ch lor ide in pyr id ine, at room temperature, afforded the tosyloxy esters 103b and 107b (CHARTS XH Ia and XII l b , Pages 57 and 61). The spectral data of 103b and 107b were in agreement with the assigned structures. Thus, the infrared spectrum of 103b and 107b displayed an absorption band at 1190 cm" 1 (doublet) due to the tosy late func t i ona l i t y and a broad absorption band was evident at 3400 cm" 1 due to the 17g-hydroxy group. The n.m.r. spectrum of 103b and 107b exhibited a four-proton AB pair of doublets at ^2.68 and f2.20 (J = 9 Hz) and a three-proton s ing let at T7.55 which could be att r ibuted to the tosy late funct ion-a l i t y . In add i t i on , the methylene protons adjacent to the tosy late group was evident as a two-proton mul t ip le t at T5.90. F i n a l l y , the mass spectrum indicated a molecular ion peak at ^ 492. The next step in the projected synthesis involved the sub s t i -65 tut ion of the tosy late group with iodide . However, th i s reaction proved more d i f f i c u l t than had been ant ic ipated. Treatment of the tosyoxy esters 103b and 107b with 5% sodium iodide in acetone at room temperature for twenty-four hours under N 2 returned copious amounts of s ta r t i ng mater ia l . Af ter considerable experimentation i t was found that the - 70 -reaction time, temperature, and concentration of sodium iodide were most c r i t i c a l in th i s react ion. Thus, treatment of tosyloxy esters 103b and 107b with 10% sodium iodide in re f lux ing acetone for s i x t y hours under N 2 and protected from l i g h t furnished the iodo esters 103c and 107c as a yellow o i l i n f a i r y i e l d (CHARTS XII la and XI l i b , Pages 57 and 61). T.L.C. examination of the crude reaction product revealed one broad spot with baseline contaminants. The mass spectrum indicated a molecular ion peak at j^-464 and a prominent peak at ^-337. Hence, the introduction of iodide has been achieved. Uskokovic and his 55 66 colleagues, and others have found that the act ion of s i l v e r f l uo r ide in pyridine on primary iodides e f fect s the e l iminat ion of hydrogen iodide. For example, treatment of iodide 113 with s i l v e r f l uo r ide in pyridine 66h affords o l e f i n 114 in good y i e l d . AcO 1 O A n AcO 114 Experiments were car r ied out to generate the o l e f i n i c esters 104a and 108a from iodo esters 103c and 107c by employing s i l v e r f luor ide i n pyr id ine. This react ion , however, took an unexpected course which was 55 in marked contrast to the work of Uskokovic and his colleagues. The ole-f i n i c esters 104a and 108a were not i so lated but an isomeric mixture of - 71 -lactones which was chromatographically and spectroscopica l ly ident ica l with lactones 102a and 106a (CHARTS XH Ia and X I l i b , Pages 57 and 61 ). Presumably, a nuc leoph i l i c attack by one of the ester oxygens had occurred on carbon 2 with displacement of iodide in a S^2 type fashion to afford lactones 102a and 106a. Furthermore, s i l v e r OH f l uo r ide could f a c i l i t a t e in the removal of the iodide. It i s note-worthy that the nature of neighbouring group reactions was f i r s t e l u c i -dated through experiments involv ing pa r t i c i pa t i on of the carboxylate groups, pa r t i c u l a r l y through studies of the hydrolysis and a lcoholys i s of the anions derived from a -ha locarboxyl ic ac ids. As the distance between the halogen atom and the carboxylate group i s increased, the c y c l i c structure involved in d i rec t intramolecular displacement becomes much less strained and reaction by displacement becomes favoured. Thus the conversion of y-bromovalerate 115 in water to the five-membered 68a r ing lactone 116 proceeds by d i r ec t displacement, and the same i s probably true for the formation of 6-butyrolactone 118. 6 8 1 3 I n teres t ing ly , as i s the case with other types of c y c l i z a t i o n , formation of rings of seven or more members en ta i l s some d i f f i c u l t y , due to the low p robab i l i t y - 72 -CH3CHBr(CH2)2COO 115 H2Q 116 for c o l l i s i o n between the opposite ends of a long cha in l i ke molecule. When £-bromocaproic acid (119) is treated with Ag 20 in water, ordinary so lvo lys i s to give the e-hydroxy acid 120 competes with neighbouring-group p a r t i c i p a t i o n . 6 8 c Anion 121 gives only a hydroxy a c i d . ^ 8 d Whether CH.CHBrCH COO 117 H 20 O Me' 118 Br(CH 2)5COOH 119 Ag 20 H 2 0 O 0 +HO(CH2)sCOO 120 Br(CH 2) 6COO~ • HO(CH2)6COOe (no lactone) H 20 121 122. - 73 -or not an ester linkage functions as a neighbouring group depends la rge ly upon how i t i s s i tuated with respect to the reaction centre in the sub-s t ra te . The so lvo ly s i s of bromo ester 123 and the deamination of amino ester 124 re su l t in inversion of configuration about the a-carbon, suggest-ing that there i s neg l i g ib le pa r t i c i pa t i on by the ester group when the reaction centre i s alpha to the carbonyl of the ester. On the other hand, CH 3 CH BrCOOEt CH 3 CH (NH-J COOEt 123 124 ester pa r t i c ipa t i on has been observed in a number of subst i tut ions in which the reaction centre gives r i s e to an intermediate containing a f i v e -membered r ing . For example, the reaction of trans-2-acetoxycyclohexyl tosy late (125) with acetate in g l a c i a l acet ic acid gives trans-1 ,2-diacetoxycyclohexane (127). The product i s not the c i s -d iaceta te which - 74 -would be expected i f acetate ion attacked the r ing on one side while the tosy late ion departed from the other. The predominant product i s the trans-diacetate 127, ind icat ing that the acetoxy group is capable of preserving the configuration about a carbon atom 2 while th i s centre 69 i s subject to nuc leoph i l i c attack. The proposed intermediate i s the c y c l i c carbonium ion 126 which may suffer attack at carbon-1 or carbon-2. In e i ther case, the trans diacetate 127 should re su l t . On the basis of these resu lts i t appeared reasonable to presume that the seven membered lactones 102a and J_06a (CHARTS XHIa and XII Ib, Pages 57 and 61 ) had been formed by intramolecular displacement of iodide by one of the oxygens of the ester group. By u t i l i z i n g the tosyloxy esters 103b and 107b (CHARTS XHIa and X I l i b , Pages 57 and 61 ) in a base catalyzed el iminat ion i t was •t. hoped that the d i f f i c u l t i e s attendant with iodides 103c and 107c would be overcome. However, i t i s possible that a base catalyzed r ing closure as outl ined in CHART XV (see Page 75 ) or the displacement of the tosy late func t i ona l i t y by a nuc leoph i l i c reagent could occur. With regard to th i s l a t t e r reaction i t i s in order to consider the ef fects of changes in the base and medium on e l iminat ion versus s ub s t i t u t i on . ^ 0 Strong bases benefit e l iminat ion as against subs t i tu t ion . With a high concentra-t ion of strong base in a nonionizing solvent, bimolecular mechanisms are favoured, and E2 predominates over S^2. At low base concentrations, or in the absence of base altogether, in ion iz ing solvents, unimolecular mechanisms are favoured, and the S^l mechanism predominates over the E l . Some anions which do not promote el iminat ion in p rot ic solvents, where they are surrounded by a solvent s h e l l , do so in aprot ic solvents, - 75 -CHART XV. Possible ef fects of a base on tosyloxy ester 103b OH TS O 103 b ' TS O MeO - 76 -where t he i r a f f i n i t y for protons i s not s a t i s f i e d by the solvent. For example, l i th ium chlor ide in dimethyl formamide dehydrohalogenates many compounds. Ehmann and his co l leagues^ found that treatment of tosylate 128 with l i th ium chlor ide in dimethylformamide afforded o l e f i n 129 in treatment of tosy late 130 with l i th ium carbonate in dimethylacetamide gave o l e f i n 131 in low y i e l d . The reaction of tosyloxy esters 103b and 1 3 0 131 - 77 -and 107b with l i th ium chlor ide and l i th ium carbonate in dimethylformamide and dimethylacetamide, respect ive ly , led only to recovered s ta r t ing 73 material and diverse minor products. Wintersteiner and Moore effected dehydration of 3B-acetoxy-cholestane-7a-ol (132) by reaction with tosyl chlor ide in ref lux ing pyridine and obtained o l e f i n 133 in good y i e l d . 132 133 Treatment of tosyloxy esters 103b and 107b with ref lux ing pyridine for three hours under nitrogen afforded compounds 103a and 107a (CHARTS XII l a and X I l i b , Pages 57 and 61 ) in ca. 50% y i e l d and recovered s ta r t ing mater ia l . This re su l t could be interpreted i f one assumed that the pyridinium tosylates were formed which on work up gave the appropriate alcohols or there was a nuc leophi l ic attack on sulphur. Several other workers have observed the formation of alcohols in dehydrotosylation reactions. For example, the action of ref lux ing c o l l i d i n e on tosylate 74 134 affords o l e f i n 135 and small amounts of alcohol 136. Z ieg ler 75 and Bharucha reported that treatment of d i tosy la te 137 with potassium - 78 -Me-. TSO Me co 11 id ine/A Me... HO" acetate in aqueous dimethylformamide gave o l e f i n i c alcohol 138. As a TSO" JL3Z O Me KOAC,DMF,HPO 138 re su l t very dry pyridine was used in the above react ion. I t i s note-worthy that tosy late 139 in dimethylsulphoxide at 80° affords alcohol 141 in 38% y i e l d . ^ In order to prevent e i ther a nuc leoph i l i c attack on sulphur or the d i r ec t displacement of the tosy late func t i ona l i t y i t would be necessary to employ a hindered base. I t has been shown that - 79 -bimolecular subst i tut ions may be retarded when the s t e r i c requirements of the attacking reagent become excessive. For instance, in CHART XVI (see Page 80) are l i s t e d the r e l a t i ve rates for the reaction of some subst ituted pyridines with methyl i o d i d e . ^ Indeed, re f lux ing c o l l i d i n e 78 has been employed to e f fec t dehydrotosylation in good y i e l d . El imination i s favoured over subst i tut ion by increasing the temperature. Huffman 78a and his colleagues converted compound 142 into o l e f i n 143 by employing c o l l i d i n e at 189° for four hours. The action of re f lux ing c o l l i d i n e - 80 -CHART XVI. Relat ive rates for the reaction of some substituted  pyridines with methyl iodide .072 . 0 0 0 2 - 81 -on tosylate 144 fo r two hours afforded o l e f i n 145 in 65% y i e l d . 78b ISO*"' C0 2Me collidine/A H H 144 145 Thus, exposure of a mixture of tosyloxyesters 103b and 107b (CHART XVII, Page 82) to re f lux ing c o l l i d i n e for four hours under nitrogen afforded only o l e f i n 108a in ca. 30% i so lated y i e l d as a c lear o i l . This reaction was monitored by t . l . c . on s i l i c a ge l . Four hours represented the minimum, time for complete consumption of s ta r t ing mater ia l . Incomplete reaction gave lower y ie ld s of 108a and led to i s o l a t i on d i f f i c u l t i e s . O lef in 108a was i so lated by preparative t . l . c . on s i l i c a ge l . The spectroscopic properties of 108a were in agreement with the assigned structure. The infrared spectrum of 108a had absorption bands at 900 cm" and 1630 cm" 1 a t t r ibuted to the methylene exocycl ic double band. The sa l i en t feature in the n.m.r. of 108a was a broad two-proton doublet (J = 4 Hz) at T5.35 due to the v iny l protons on carbon 4. The mass spectrum of 108a indicated a molecular ion peak at ^ 320. Various other products and insoluble so l ids were i so lated from the reaction mixture. However, the infrared spectrum of these products did not display any o l e f i n i c features and they were not further character ized. From th i s series of reactions i t appeared that Baeyer-Vi l1iger oxidation of ketoacetate 81b gave predominantly lactone 106c. Thus, - 82 -Baeyer -V i l l i ge r Oxidation Studies lQ3b !Q4a - 83 -sequel 8TJb to 104a (CHART XVII, Page 82) did not demonstrate promise. This precluded making further explorations with the Baeyer -V i l l i ge r approach. From the pract i ca l aspect, the general plan involv ing the insert ion of a hetero atom proved un f ru i t f u l since at each synthetic step isomeric mixtures were encountered, which were e i ther inseparable or d i f f i c u l t to separate, and low y ie ld s were encountered in several instances. b. Incorporation of Unsaturation Between Carbons 2 and 3 It was next planned to invest igate the second approach, namely, the preparation of a A2 s tero id of type 85. Pr io r to th i s invest igat ion the only preparation of A 2 -53-steroid free from the A3 isomer involved 79 dehydration of the cyanohydrin 147. Normally, the formation of A 3 8 5 a . R = O H 8 5 b. R = O A c products from 3-oxo 58-steroids could be at t r ibuted to the preferent ia l 4? Rl Rf) enol i sat ion of these ketones towards carbon-4. > 3 l> o u p o r example, 42 Liston found that under enol acety lat ing conditions 17g-acetoxy-5g-androstan-3-one (81b) afforded a mixture of enol acetates 85b and - 84 -146 in ca_. a 1:3 r a t i o . When th i s mixture was subjected to equ i l i b ra t i ng OAc. 85a, R = OH 85 b. R = OAc' 4? Rl conditions enol acetate 146_ was formed exc lus ive ly . Djerassi has suggested that two main factors operate in determining the d i rec t ion of enol acety lat ion and probably eno l i sat ion in 3-oxo 5a - s te ro id s . The f i r s t i s s t e r i c and involves angular methyl group interact ions 42 whi le the second i s hyperconjugative. L iston u t i l i z e d H i l l ' s method of ca lcu la t ing H-H and CH3-H non-bonded interact ions to demonstrate that the d i rec t ion of enol i sat ion of the 3-oxo 5g-stero ids i s governed by s t e r i c forces in the absence of any hyperconjugative e f f ec t . In ca lcu la t ing the r e l a t i ve s t a b i l i t i e s of the two eno l ic forms of 17e-acetoxy-53-androstan-3-one (81b) a number of assumptions were made: - 85 -(1) the A r ing of the 3-oxo 53-steroid assumes the c l a s s i c a l ha l f - cha i r conformation in the enol ic form; (2) the B r ing of the stero id assumes a geometrically perfect chair form; (3) nonbonded interact ions are add i t i ve ; and (4) the acetate group has no influence on the r e l a t i v e s t a b i l i t y of the two enol acetates. None of the f i r s t three assumptions would be expected to hold in a l l cases since large nonbonded i n t e r -actions could cause ske leta l deformations; however, the interact ions involved in these cases are r e l a t i v e l y small and the d i s to r t ions are considered minimal. Using H i l l ' s procedure the calculated di f ference between the two enols i s 1.91 k cal/mole, which corresponds to an equi l ibr ium mixture of 96% A 3 -eno l and 4% A 2 -eno l at room temperature. Experimentally, 93.5% A 3 -and 6.5% A 2 -eno l were found under equ i l i b ra t i ng conditions at room temperature. These resu l t s suggest that s t e r i c forces are the dominant factors which govern enol acety lat ion and eno l i sa t ion . Under k i n e t i c a l l y contro l led conditions i t may be possible 83 to trap enol acetate 85b as the major product. Attempts to trap enol acetate 85b by quenching a mixture of the enolate anions with acetyl ch lor ide proved u n f r u i t f u l . The n.m.r. spectrum of the resu l t ing crude product indicated that enol acetate 146 had been formed. At th i s stage, i t was decided to study the synthesis of 79 n i t r i l e 148 as reported by Nathansohn. Exposure of cyanohydrin 147 to phosphorus oxychloride in ref lux ing pyridine effected dehydration to 79 afford o l e f i n 148 in good y i e l d . Thus, cyanohydrin T47 was prepared - 86 -OAc OAc N C OH H H 147 1 4 8 by treatment of ketoacetate 81b with acetonecyanohydrin at room temper-ature for t h i r t y hours. However, in our hands i t was found that the OAc action of phosphorus oxychloride on cyanohydrin 147 in re f lux ing pyridine afforded inso luble material and several minor compounds. As a r e s u l t , the stereochemical features of the addit ion and dehydration reactions were examined. In the addit ion reaction the s t e r i c f ac to r , which i s connected with the a c c e s s i b i l i t y of the carbonyl centre, would indicate that the cyanide nucleophile would approach from the e face to give cyanohydrin 147a. Furthermore, the reaction i s carr ied out under equ i l i b ra t i ng conditions (room temperature for t h i r t y hours) and, therefore, O 147 H 81b - 87 -NC OH H 147a the thermodynamically more stable product w i l l be formed. The conform-at ional free energy dif ferences for cyclohexanol and cyanocyclohexane 84 are 0.52 (pyridine) and 0.25 (tetrahydrofuran) k cal/mole, respect ive ly. These values and the above considerations tend to suggest that cyano-hydrin 147a would be the predominant product in the reaction of acetone-cyanohydrin with ketoacetate 81b. In the dehydration reaction the . s tereoelectronic factor requires t r an s -d i ax i a l e l im inat ion . For example, the action of phosphorus oxychloride on compound 150 in pyr idine gives 85a o l e f i n 151 in quant i tat ive y i e l d . On the other hand, the equatorial Me HO H H 1 5 0 151 alcohol 152 forms the phosphate ester 153 when subjected to phosphorus oxychloride in p y r i d i n e . * ^ Hence, cyanohydrin 147a would not be expected to read i l y undergo dehydration. I f the 3-hydroxy group of - 88 -compound 147 has the g conf igurat ion, i t appears that two trans-d iax ia l , el iminations are poss ib le; namely, dehydration of 147 to afford 148 and 149. OAc Next, attent ion was focused on a communication by Caspi et a l . , which described a convenient and prac t i ca l method for the introduction of unsaturation at carbon 1 in 3-oxo 5g-steroids (CHART XVIII, 81a —»»155, 87 Page 89). Birch reduction of ketone 155 and subsequent trapping of the enolate anion with acetyl ch lor ide might be expected to furn ish enol acetate 85b (CHART XVIII, Page 89 ) . 8 3 c ' 8 3 d Condensation of ketone 8 U with ethyl formate in dry benzene at room temperature for t h i r t y hours y ie lded the hydroxymethylene der i vat i ve 53 in 60% y i e l d as a white - 89 -CHART XVIII. Proposed synthesis of compound 86b 85Ji - 90 -c r y s t a l l i n e s o l i d , m.p. 153-159° ( l i t . m.p. 157-163°). Exposure of hydroxymethylene 5_3 to dichlorodicyanoquinone in re f lux ing benzene for t h i r t y hours y ie lded 2-formyl-17B-hydroxy-5 3-androst-l-en-3-one (154) in ca_. 30% pur i f i ed y i e l d as a c lear o i l . Several attempts to induce c r y s t a l l i z a t i o n f a i l e d . I t i s worthy of mention that the introduction of unsaturation between carbons 1 and 2 has been achieved in 2-hydroxy-methylenes of the A/B trans ser ies by the action of dichlorodicyanoquinone The spectroscopic data of 154 were in accord with the assigned structure. For example, the infrared spectrum of 154 had absorption bands at 1720 cm" and 2750 cm" 1 due to the aldehyde f unc t i ona l i t y . The a,B-unsaturated carbonyl was evident at 1670 cm" 1 . Although the required degree of unsaturation in the A r ing had been achieved and some of the d i f f i c u l t i e s associated with th i s transformation had been unravelled the pur i f i ed y i e l d s were s t i l l poor even when-operating on a small scale. This i s 86 in sharp contrast to the work of Caspi and his colleagues. For example, they recorded a 60% y i e l d of 154 in large scale preparative work. The next step demanded the removal of the formyl group. In a series of small scale t r i a l experiments i t was found that treatment of aldehyde 154 with chlorotris(tripheny1phosphine)rhodium in re f lux ing benzene fo r three hours gave a f te r work up a c lea r o i l which consisted of trace amounts of s ta r t ing material and two new compounds as evidenced by t . l . c . on s i l i c a ge l . The inf rared spectrum of the crude material tended to suggest the presence of an a,B unstaurated carbonyl com-pound. Thus, the reaction was car r ied out on a larger scale and ketone 155 was i so lated by chromatographic means. However, a f te r t r y -ing various chromatographic systems i t was found that the i so lated ketone - 91 -155 was always contaminated with triphenylphosphine oxide. In add i t ion , a l l attempts to induce c r y s t a l l i z a t i o n f a i l e d . From the spectroscopic data there i s l i t t l e doubt that th i s substance had structure 155. Thus, the infrared spectrum displayed absorptions at 1660 cm" 1 due to the a,8 unsaturated carbonyl and at 840 cm" 1 which could be att r ibuted to the carbon-1 carbon-2 double bond. The pertinent spectral features in the n.m.r. of 155 were an AB pa i r of doublets at x4.14 and T3.15 (J = 10 Hz) due to the carbon-1 and carbon-2 o l e f i n i c protons. Although th i s synthetic route (CHART XVIII, Page 89 ) had presented several obstacles i t appeared l i k e l y that intensive further invest igat ions could have rendered a p rac t i ca l synthetic sequence. However, th i s approach was abandoned because of the low overal l y i e l d and the cumbersome methods of pu r i f i c a t i on which had been r ea l i z ed . c. Funct ional izat ion of Carbon-2 It i s well documentated that reactions of 3-oxo 56-steroids afford predominantly carbon-4 funct iona l i zed products except for the 42 51 80 formation of the 2-hydroxymethylene der i va t i ve . ' ' The 2-hydroxy-methylene der ivat ive was not d i r e c t l y appl icable to the present design since ozonolysis leads to r ing cleavage between carbons 2 and 3 in a 51 symmetric manner, equation I I. C l ea r l y , a method to d i f f e r en t i a t e carbons 2 and 3 would have to be found. As noted e a r l i e r , the formation of 4-substituted compounds can be ra t i ona l i zed by the preferent ia l 42 51 80 eno l i zat ion of 3-oxo 58-steroids towards carbon-4. u i t appeared highly probable that th i s general p r i nc i p l e could serve i n d i r e c t l y to - 92 -A c O N a O A c H O A c / A 1 5 6 , R = C 8 H 1 7 j 5 8 , R = O A c 157, R = C Q H 1 7 1 5 9 , R = OAC funct iona l i ze carbon-2 in the required manner. 89 Takahashi ejt al_., found that treatment of 4g-bromo-5g-cholestan-3-one (156) with sodium acetate in re f lux ing acet ic acid afforded 28-acetoxy-53-cholestan-3-one (157) in good y i e l d . Thus, O A c 81b 158 th i s reaction was extended to bromoketone 158. 17g-Acetoxy-5e-androstan-3-one (81b) was brominated by employing a procedure s im i l a r to that used 90 by Fieser e_t al_., Bromination of compound 81b in g l a c i a l acet ic acid gave, a f te r work up, a white c r y s t a l l i n e so l id m.p. 135-151°. A small - 93 -sample was r e c r y s t a l l i z ed from ether to afford cubes m.p. 174-175° 90 ( l i t . , m.p. 174-175°). The v.p.c. of the crude react ion product indicated the presence of two compounds in ca_. a 1:4 r a t i o . The i n f r a -red spectrum of the crude material had a broad carbonyl band at 1720 cm" 1 . The sa l i en t feature in the n.m.r. spectrum of the crude product was a one-proton doublet (J = 12 Hz) at T5.00 due to the carbon-4 ax ia l proton of 158. F i n a l l y , the mass spectrum indicated a molecular ion peak at 2-410 and a P+2 peak almost equal in i n tens i t y to the parent peak because of the presence of molecular ions containing the 8 1 Br isotope. These resu l t s suggested that bromoketone 158 was the major product. The other compound present i n the crude reaction product i s presumably e i ther bromoketone 160a, 160b or 160c. However, the B r reoq i 6 Q b 1 6 o c i s o l a t i on of pure bromoketone 158 was not necessary for the next step i n sequence. I t i s worthy to comment on the mechanistic aspects of the above bromination react ion. The stereochemistry of bromination of enols appears to be contro l led by two factors that may e i ther oppose 91 or re inforce one another. The stereoelectronic f ac to r , which i s appl icable to cyclohexanone der ivat ives may be i l l u s t r a t e d by the - 94 -accompanying equations (CHART XIX, Page 95 ). The energet ica l l y most favourable t r an s i t i on states for removal of an alpha proton, for example, 161 and 163 to form the enol 165 and for addit ion of a bromine atom (162 and 164) to form the product are those in which continuous overlap of the p o rb i t a l s involved i s poss ib le. I f the chair forms 161 and 162 are considered i t i s c lear that there should be preference for the removal of an ax ia l proton and addit ion of bromine at an ax ia l bond. However, the importance of t h i s preference should diminish in proportion to the degree to which the t r an s i t i on states 161 and 162 resemble the planar enol 165 rather than the ketones 166 and 167. The second factor of concern i s the s t e r i c interference that ex i s t s in the t r an s i t i on states fo r proton removal and bromine add i t ion. I t i s apparent that i f serious s t e r i c interact ions e x i s t in the c h a i r l i k e t r an s i t i on states 161 and 162, the eno l i zat ion and bromination may proceed v ia the boat-l i k e t r an s i t i on states 163 and 164 and s t i l l allow continuous p-orb ita l overlap. In the bromination of ketoacetate 81b there would be a severe non-bonded interact ion between bromine and the ax ia l protons at carbons 7 and 9 in the c h a i r l i k e t r an s i t i on state. Presumably, th i s i s s u f f i c i e n t to cause the enol to react with bromine v ia the boat - l i ke t r an s i t i on s ta te , leading to the equatorial bromoketone 158. In view of these considerations the ax ia l bromoketone 160c would not be expected to form. 92 Djerassi and his colleagues have demonstrated that bromination of the enol acetates derived from stero ida l A r ing ketones resu l t s in v i r t u a l l y the same product mixtures as are obtained from the ketones 42 themselves. Since Liston reported that enol acety lat ion of keto-acetate 81b gave r i s e to enol acetates 146 and 85b in ca_. a 3:1 r a t i o , and bromination of enol acetate 85b gave bromoketone 160a, the bromination R - 96 -of ketoacetate 81b' would be expected to afford bromoketones 158 and 160a in ca_. a 3:1 r a t i o , respect ive ly . I t i s , therefore, highly probable OAc 16Qg that bromoketone 160a i s present in the crude reaction product. Treatment of the crude bromination product with sodium acetate in re f lux ing acet ic acid afforded 28, 176-diacetoxy-5s-androstan-3-one (159) in ca_. 70% pu r i f i ed y i e l d as a white c r y s t a l l i n e s o l i d , m.p. 170-173°. This material was homogeneous on s i l i c a gel t . l . c . A small sample was r e c r y s t a l l i z ed from methanol, m.p. 161-162°, for analys i s . The spectroscopic data of th i s material were in complete accord with - 97 -the assigned structure. The infrared spectrum of 159 had an intense carbonyl band at 1720 cm" 1 due to the 23- and 178- acetoxy groups and the 3-oxo group. An intense broad band at 1150-1200 cm" 1 was evident for the C-0 stretching v ib rat ions . The pert inent spectral features in the n.m.r. spectrum of 159 were a double doublet (J = 5 and 13 Hz) at ca. x7.8 a t t r ibutab le to the 48-hydrogen, a broad t r i p l e t (J = 9 Hz) at x7.19 due to the 4a-hydrogen, and a double doublet (J = 6 and 14 Hz) at x4.85 which could be assigned to the 2a-hydrogen. It i s very import-ant to r ea l i z e that the s p l i t t i n g s of the four intense l ines of the X portion of an ABX spectrum do not necessar i ly represent the coupling constants and Jg^, although frequently th i s may be a good approxi -93 mation i f the chemical s h i f t between A and B i s larger than J ^ g . This can pose a serious problem in conformational ana lys i s , e spec ia l l y in steroids because the resonances due to both A and B may l i e with in the bounds of the methylene envelope and hence cannot read i l y be located. It has been observed that as the chemical s h i f t between A and B i s increased, the s p l i t t i n g s of the X resonance correspond more c lo se l y to the coupling constants 0 ^ and J g ^ . Therefore, i f a spectrum i s determined at both 60 MC and 100 MC without observing any change in the pattern due to the X proton, i t i s probable that the s p l i t t i n g s are good approximations to the coupling constants. This fol lows since the chemical s h i f t between A and B has been increased by a factor of 1.67 in passing from the 60 MC to the 100 MC determination. The method, of course, f a i l s i f 6AB=0 and i s of dubious value i f the chemical s h i f t 93 between A and B i s only a few cyc les . This treatment was applied to compound 159. No change in the ABX pattern of the 2a-proton was observed. - 98 -The probab i l i t y that the chemical s h i f t between the l a - and 13- protons in th i s compound i s only a few cycles seems small since ax ia l and 93b equatorial protons are usual ly separated by at least 0.5 ppm. There-fo re , i t appears l i k e l y that the s p l i t t i n g s are good approximations to the coupling constants. F i n a l l y , the mass spectrum of 159 indicated a molecular ion peak at j^- 390. This product was not unexpected in l i g h t 89 90 of previous subst i tut ion reactions of s tero ida l bromoketones. ' For 90 example, Fieser et aJL , found that treatment of the bromoketone 169 with potassium acetate in re f lux ing g l ac i a l acet ic acid gave a 1:1 mixture of acetoxyketones 170 and 171. It i s worthy of mention that : H OAc 171 Zinc-acet ic acid reduction of diacetate 159 gave 81b in low y i e l d , which - 99 -94 indicated that no acyloxy-ketone exchange had occurred. Rosenfield R 4 C Q _ - . — Zn/acetic 81b, R = OAc has suggested that the configuration of the acetoxy group i s important in these Zn-acetic acid reductions, and el iminat ion occurs in good y i e l d only i f the group i s a x i a l . Furthermore, bromination of 159 with bromine in acet ic acid gave a compound whose n.m.r. spectrum was con-s i s tent only with structure 173. The sa l i en t features in the n.m.r. OAc 159 173 - 100 -spectrum of 173 were a doublet at T5 .02 due to the 4 a - hydrogen and a double doublet (J = 5 and 13 Hz) at T 4 . 7 9 due to the 2 a - hydrogen. In add i t i on , the spectrum was determined at both 60 Mc and 100 Mc. No 93 change in the ABX pattern afforded by the 2 a - proton was observed. Therefore, i t appears l i k e l y that the s p l i t t i n g s are good approximations to the coupling constants. Thus, a p rac t i ca l method had been developed for the introduction of an acetate f unc t i ona l i t y at carbon 2 in 17e-acetoxy-5e-androstan-3-one (81b). In view of the present aims and the ready synthetic a v a i l -a b i l i t y of diacetate 159 i t was decided to examine the chemical behaviour of 159. Before embarking on these explorations i t i s in order to consider the mechanistic aspects of transformation IV. R R I n teres t ing ly , at the beginning of th i s thes is i t was mentioned that synthetic studies in the f i e l d of natural product chemistry have played a prominent ro le in the development of reaction mechanisims. There are several examples of cine subst i tut ions in s tero ida l chemi s t r y^ but the above reaction i s notable for the good y i e l d and pur i ty of pro-duct. This general type of transformation, equation IV, has gained - 101 -the attent ion of several researchers. F i r s t l y , Fieser and Cox have demonstrated that a-bromoketones can undergo subst i tut ion at a- or ex-pos i t i on , equation V. Thus, th i s led us to consider the intermediacy O of the 4a- or 40- acetoxy compound in transformation IV. Secondly, 94 Warnhoff e_t al_., have found that a-acetoxycyclohexanone transfers the acetate group to the a'- carbon above 220°, equation VI, and s im i l a r 94 98 99 rearrangements have been found to occur at lower temperatures, ' HO equation VII. This further enhances the p o s s i b i l i t y that the 4-acetoxy isomer may be an intermediate in reaction IV. F i n a l l y , Satoh and 97 / x Takahashi treated bromoketone 157 with (a) potassium acetate-acet ic a c i d , (b) potassium acetate-dioxan, (c) t r iethy lamine-acet ic acid at - 102 -90-95°, (d) potassium pivalate-dioxan at 70°, and (e) ACO N Me^-dioxan at room temperature. Samples were taken from each reaction mixture at in terva l s and the progress of the reaction was followed by t . l . c . and by using a Varian HR-220 n.m.r. spectrometer to observe the change in the s ignals due to the methyl protons of the acetoxy groups. It was suggested that 2a-acetoxy-5B-cholestan-3-one (174) was formed f i r s t in each of the above cases. Since the isomenzation of the 2a-acetoxy to the 26-acetoxy der ivat ive i s r e l a t i v e l y fast for the cases (a) and (c ) , i t was impossible to i so la te the i n i t i a l product, 2 a -acetoxy-53-cholestan-3-one (174). In cases (b), (d), and (e) , however, th i s pro-duct was e a s i l y obtained. I t was reported that the 2 a-acetoxy der ivat ive 174 was produced almost s t e reo spec i f i c a l l y in 2.5 hours for method (d) and in f i ve days for method (e), and was gradually isomerized to the 26-acetoxy der i vat i ve 157 when the reaction was continued for longer periods of time. In order to .rational ize the i n i t i a l formation of the 2a-97 acetoxy der i vat i ve 174, Satoh and Takahashi invoked that the reaction proceeds in a trans-S^2' manner in which the leaving group i s trans to the entering group, unl ike the ordinary S^2' react ion, where a c i s re la t ionsh ip obtains. Since the a- side of the r ing A of a 56-steroid - 103 -is less favoured than the 8- side for nuc leoph i l i c attack, they considered that the conformation of the intermediate must be responsible for the unexpected attack at the a- face. F i e s e r ^ and C l a r k e ^ reported that eno l izat ion took place during the aceto lys i s of 2a-bromo and 68-bromo der ivat i ves . I f 156 undergoes eno l i z a t i on , i t i s possible that a con-formation w i l l re su l t in which r ing A i s r e l a t i v e l y f l a t with respect to r ing B, and hence nucleophiles may attack at the a- face. In order 97 to test th i s p o s s i b i l i t y Satoh and Takahashi prepared 48-bromo-58-cholest-2-en-3-ol acetate (176) by enol acety lat ion of the 48-bromo-3 ketone 156_ since the enol der ivat ive was not stable enough to i s o l a te . 176 Three possible conformers 176a, 176b and 176c were considered for compound 176 (see Page 104 ). After examining the n.m.r. spectrum of 97 176 Satoh and Takahashi assumed that enol acetate 176 had the boat-B conformation. This conformation provides a favourable environment for the nucleophile to attack at carbon-2. Furthermore, a- attack occurs more read i l y than 8- attack at th i s pos it ion because of the s t e r i c 97 e f f ec t of the 10-methyl group. In summary, Satoh and Takahashi - 104 -176c - 105 -concluded that the 2a-acetoxy der ivat ive 174 was formed as the product of a trans-S.,2' reaction and then isomerized to the more stable 28-isomer. It was, therefore, planned to determine i f the 4B- or 4a-acetoxy compounds 175b and 175a, were intermediates in transformation IV and to determine i f e i ther 175b or 175a was re lated to the intermediate 97 i so lated by Satoh and Takahashi. Since these workers started with OAc OAc 175a, 4 a —OAc 175 b, 4/9—OAc 5g-cholestan-3-one, and 173-acetoxy-5B-androstan-3-one was chosen as s ta r t ing material d i rec t comparisons could not be made. - 106 -Enol acety lat ion of compound 81b, by employing conditions analogous to those used by Liston gave enol acetate 146 in excel lent 42 y i e l d . Epoxidation of the enol acetate 146 with m-chloro 81 b 1 4 6 acid - sodium bicarbonate gave the e-epoxide 177. Assignment of OAc OAc 146 177 the e-configuration to the epoxidation product was made on the fol lowing bases. The g-face appears to be the s t e r i c a l l y more access ible d i rect ion 102 for peracid attack. The n.m.r. spectrum of 177 had a s ing let at x6.93 due to the proton on carbon-4. An examination of the Dreiding model of 177 indicated that the dihedral angle between the hydrogens on carbon-4 and carbon-5 i s ca. 100°. The dihedral angle between the hydrogens on carbon-4 and carbon-5 in the isomeric a-epoxide was estimated to be 50°. In an extensive study of stero idal epoxides and episulphides i t was found - 107 -that the coupling constant could approach zero only f o r dihedral angles of 70-100° while a dihedral angle of 50° was expected to y i e l d a coupling 103 constant of at least 2 Hz. A l so, the chemical s h i f t of the carhon-19 methyl protons in 177, T9 .13 , agreed c lo se ly with that of the carbon-19 104 methyl protons of 38> 48-oxidG-58-cholestane, x9.14. Pyrolys is of 177 at 160° for 5 minutes gave 4 a , 178-diacetoxy-53-androstan-3-one (175a) in C£. 80% y i e l d . The s a l i en t feature of the 177 1 7 5 a , 4 a — O A c n.m.r. spectrum of 175a, which suggested i t s s t ructure, was a one-proton doublet (J = 8 Hz) at x4.59. This was assigned to the 4 g - hydrogen of 175a which probably i s in a boat conformation due to the severe i n t e r -action of the 4 a - acetoxy group with carbon-7 and carbon-9 in the cha i r conformation. On re f lux ing i n acet ic acid - sodium acetate, which were conditions for reaction IV, 175a was converted c leanly to the 4 $ - acetoxy isomer 175b. The 4 3 - isomer was also obtained by treatment of 177 with HCl in ether. The 4 8 - acetoxy compound 175b had in i t s n.m.r. spectrum a one-proton doublet (J = 12 Hz) at x 4 . 4 8 which was assigned to the 4 a -hydrogen. These epoxide rearrangements pa ra l l e l those of the 2 a , 3 a -105 oxido-38-acetoxycholestane. When 175b was subjected to the conditions for reaction IV i t was recovered unchanged. - 108 -These experiments would ind icate that neither 175a nor 175b can be an intermediate in reaction IV. In f a c t , 175a and 175b did not rearrange to e i ther 2 - acetoxy isomer on thermolysis at 160°. Hence, 97 the intermediate i so lated by Satoh and Takahashi must be the 2 a -acetoxy compound and i t does not ar i se v ia the 4 a - or 4 B - acetoxy isomer. B o r d w e l l ^ 9 has catalogued some of the possible pathways by which a cine subst i tut ion such as reaction IV can occur. The f i r s t p o s s i b i l i t y , a S^2 subst i tut ion at carbon-4 followed by a S^ i 1 rearrange-ment v ia the enol was discarded by the above re su l t s . An S^l pathway, equation VI I I , i s a p o s s i b i l i t y . A S^ i ' rearrangement of the bromo enol 1 5 8 HO 159 (VIII) 178a to 178b followed by a S..2 reaction i s also a p o s s i b i l i t y , equation IX. - 109 -However, Liston has found that the bromoketone 158 and 17B-acetoxy-2g-bromo-5B-androstan-3-one (160a) were not interchanged or equ i l ibrated even in HBr-HOAC. This would suggest that the rearrangement of 178a to 178b does not occur under our condit ions. A second type of S^ i 1 reaction i s shown in equation X; but, th i s does not require the i n t e r -mediacy of the 2a- isomer. A f i n a l p o s s i b i l i t y i s a S w 2 ' reaction of 178a 159 (X ) 178a. After synthesizing 17$-acetoxy-2g-bromo-5g-androstan-3-one (160a) i t was c leanly converted to 159 in re f lux ing acet ic acid - sodium acetate at a rate comparable to reaction IV. This suggested that the mechanism depicted in equation VIII was operat ing; namely, both bromoketones gave the same intermediate 179. However, more data i s required to substantiate t h i s . Mechanistic aspects of the reaction of a,a'-dibromoketones and - no -180c o o H 3 C CH. H3C CHXOONa 3 H3C H3C OCOCH 3 CH-H 3 C 180d 180e - m -iron carbonyl have recent ly been reported by Noyori et_ al_. They suggested that i n i t i a l reduction of the dibromide 180a with Fe 2(C0)g produces the i ron enolate 180b which el iminates bromide ion to form the key oxya l l y l - Fe(II) intermediate 180c (see Page 110). Evidence for the intermediacy of 180c during the reduction of dibromoketones was obtained by trapping with nucleophiles. The reduction of 180a in the presence of sodium acetate gave the acetoxy ketone 180d (60%) along with the unsaturated ketone 180e (20%). The above mechanistic considerations of a,a'-dibromoketones tend to suggest that bromoenols may undergo an S^l reaction as proposed in equation VII I. d. Unsymmetric Ring Cleavage Between Carbons 2 and 3 Having rea l i zed the funct iona l i za t ion of carbon-2, experiments were carr ied out to e f fect an unsymmetric r ing cleavage between carbons 2 and 3. K i n g J ^ Clutterbuck et al_. and o t h e r s ^ " have demonstrated that a-hydroxyketones are oxidized smoothly even in the cold by per iodic a c i d , equation XI. In add i t ion, treatment of d io l 181 with periodic R — C O C H O H — R / — — — • RCOOH + R 'CHO (XI) HI acid gives aldehyde 1 8 2 . ^ In l i g h t of these resu l t s i t appeared highly - 112 -HO o ! HIOz H H0 2 C > 181 182 probable that exposure of hydroxyketone 183 to per iodic acid would afford compound 184. However, the action of mild base on diacetate OAc OAc H IQ , H 1 8 3 159 produced two isomeric hydroxy ketones 183 and 185 in c§_. a 1:1 r a t i o . AcO OAc mild base - 113 -This was revealed by the t . l . c . and the n.m.r. spectrum of the crude reaction product. The n.m.r. spectrum displayed two sharp s ing lets ( t o t a l l i n g three protons) at T9.25 and T9.20 due to the carbon-18 t e r t i a r y methyl groups while the appearance of absorptions at T8.97 and T8.90 as two s ing lets ( t o t a l l i n g three protons) could be at t r ibuted to the carbon-19 t e r t i a r y methyl groups. A l l attempts to prevent th i s unfavourable react ion--the Lobry de Bruyn-Alberda Ekenstein transformation by employing very mild hydrolysis conditions proved u n f r u i t f u l . For example, treatment of diacetate 159 with d i l u t e aqueous methanolic sodium bicarbonate gave hydroxyketones 183 and 185. On the other hand, the hydrolysis of acetoxyoxime 186 would furnish hydroxyoxime 187. Thus, exposure of diacetate 159 to hydroxylamine hydrochloride in methanol sodium acetate under re f lux ing conditions fo r three hours afforded two major compounds as indicated by t . l . c . on s i l i c a ge l . The n.m.r. and infrared spectra of th i s material suggested that pa r t i a l hydrolysis of the 2-acetoxy group had occurred to give a mixture of 186 and 187. As a r e su l t , the reaction was continued un t i l the 2-acetoxy group had been completely hydrolysed. The course of the hydrolysis was followed by t . l . c . and n.m.r. spectroscopy. The crude hydroxyoxime was obtained as - 114 -a white c r y s t a l l i n e s o l i d , m.p. 176-182°, in 75-85% y i e l d . This crude material was su i table for preparative work in the subsequent steps. A small sample was r e c r y s t a l l i z ed from methanol for ana lys i s , m.p. 213-215°. The pert inent spectral features in the n.m.r. of 187 were as fo l lows. A double doublet at x7.21 (J = 4 and 14 Hz) was assigned to the 4$- hydrogen which suggested that the oxime had the stereochemistry shown in 187. As noted previously, the n.m.r. spectrum of syn oxime 87 had a pair of doublets (J = 5 and 15 Hz) at ca_. x7 due to the 48-hydrogen. A second double doublet at x5.80 (J = 5 and 13 Hz) could be att r ibuted to the 2a- hydrogen. The n.m.r. spectrum of 187 was deter-mined at both 60 Mc and 100 Mc. No change in the patterns afforded by the 2a- and 48- hydrogens were observed. It, therefore, appears l i k e l y that the s p l i t t i n g s are good approximations to the coupling constants. Hence, the r ing A acetate had been hydrolysed and the hydroxy group was s t i l l at carbon-2. Structure 187 was further corroborated by the mass spectrum which indicated a molecular ion peak at ^ 363. Concurrently, the hydrolysis of diacetoxyoxime 186 was attempted by two other methods. F i r s t l y , treatment of diacetoxyoxime 186 with various bases afforded several products. For example, hydrolysis of 112 the oxime group and the 178-acetoxy group had also occurred. Kataoka demonstrated that treatment of acetoxyoxime 188 with concentrated sulphuric acid followed by neut ra l i za t ion gave hydroxyoxime 189. Hence, - 115 -experiments were carr ied out to e f fec t the hydrolysis of diacetoxyoxime 186 under s im i l a r condit ions. However, the re su l t i ng crude product consisted of numerous compounds as evidenced by t . l . c . Since a su itable method had been developed for obtaining hydroxyoxime 187 th i s did not necessitate further invest igat ing the above hydrolys is . I t i s possible that the isomeric hydroxyoxime 190 could have been formed. Autrey et al_. j ^ 0 * 5 found that oximation of the methoxy ketone OAc 192 afforded only the anti oxime 193. The convincing c r i t e r i o n for 1 9 3 - 116 -stereochemistry was the observation of the formation of a chelate, presumed to be 194, on the addit ion of d i l u te ethanol ic cupric n i t r a t e to an ethanol so lut ion of the oxime.^ 0 l 3 In l i g h t of th i s work and our 194 n.m.r. data i t appeared l i k e l y that only hydroxyoxime 187 had been formed. With hydroxyoxime 187 in hand, i t was f e l t that the opportunity existed for a Beckmann fragmentation react ion , equation x i . ^ k > 1 1 3 (XII) This would give r i s e to a desired unsymmetric cleavage of r ing A in which the termini of the cleaved bond would be l e f t in d i f f e ren t ox ida-t ion states and, therefore, could be separately modified. The l i t e r a t u r e suggested several methods to accomplish the 113 Beckmann fragmentation. Of pa r t i cu l a r relevance was the work of - 117 -Autrey et al_., which described the Beckmann fragmentation of anti oxime 193 to the o le f in s 196 and 197 in ca_. a 1:1 r a t i o . The fragmenta-1 9 3 TSCL/Py MeO 1 9 6 + MeO MeS 1 9 7 t ion of hydroxyoxime 187 to cyanoaldehyde 195 was studied under a var iety of condit ions. F i r s t l y , the t r ad i t i ona l use of tosy l ch lor ide in pyr idine was examined. Thus, treatment of crude hydroxyoxime 187 with tosyl ch lor ide in re f lux ing pyridine for f i ve hours afforded cyanoaldehyde 195 as a c lear o i l in 20% pu r i f i ed y i e l d . Although the y i e ld s were low the reaction had evident ly taken the desired reaction course. Of addit ional s i gn i f i cance was the fact that hydroxy lactam 191 was not detected. OAc not pbserved - 118 -However, according to Grob et_ al., the Beckmann fragmentation i s not s tereospec i f ic and thus th i s observation of fragmentation could not be used to assign structure to the hydroxyoxime 187. Several attempts were made to improve the y i e ld s of the above Beckmann fragmentation reaction by monitoring the react ion, by t . l . c , at regular time i n te r va l s . A f ter considerable experimental e f f o r t , however, the pu r i f i ed y i e ld s of cyano-aldehyde 195 were s t i l l ca_. 20%. In add i t ion , the i s o l a t i on of f a i r l y pure cyanoaldehyde 195 from the dark crude product posed a d i f f i c u l t problem. Preparative t . l . c . on s i l i c a gel with CHCl3/EtoAC (5/1) as the solvent system was found to be the most p rac t i ca l method to i s o l a te f a i r l y pure cyanoaldehyde. In general, the above procedure gave a very sluggish Beckmann fragmentation react ion , equation XII (see Page l l 6 ) . S im i l a r l y , Autrey ejtal_. , ^ o b reported low y ie ld s and i s o l a t i on d i f f i -c u l t i e s when they employed tosyl ch lor ide and pyr id ine. The low y i e l d was ra t iona l i zed on the basis that the re su l t ing cyanoaldehyde 195 was very susceptible to ox idat ion. Even an ana l y t i ca l sample of cyanoaldehyde was found to undergo rapid autox idat ion, even in the presence of nitrogen. In summary, the i n e f f i c i e n t Beckmann fragmentation react ion, equation XI I, and the unava i l ab i l i t y of f a i r l y pure cyanoaldehyde thwarted progress 113c f o r a considerable time. Shoppee ejt al_., demonstrated that t r e a t -ment of 5-hydroxy-5a-cholestan-6-on oxime (198) with thionyl ch lor ide gave exce l lent y i e l d s of cyanoketone 199. Thus, a f te r several t r i a l experiments, b r i e f exposure of hydroxyoxime 187 to thionyl chlor ide at -20° followed by treatment with 3N potassium hydroxide with subsequent - 119 -/ C 8 H 1 7 N H O I9-8- 199 ether extract ion afforded cyanoaldehyde 195 in over 80% i so lated y i e l d a f te r chromatography, as a white c r y s t a l l i n e s o l i d , m.p. n0-112°C. O A c 1 9 5 The t . l . c . of the crude reaction product showed the presence of pre-dominantly one product with polar baseline contaminants. Furthermore, the n.m.r. spectrum of the crude material suggested high product pur i ty . However, the pu r i f i c a t i on of cyanoaldehyde 195 was found to be a very temperamental process. The i so lated y ie ld s by chromatographic methods varied from 25-60%. Fortunately, subsequent synthetic studies obviated th i s d i f f i c u l t y . A f ter careful chromatography on s i l i c a gel hydroxyn i t r i le 200 and o l e f i n i c n i t r i l e 201 were i so lated as minor components of the crude reaction product. Under the separation condit ions, presumably, - 120 -the opportunity existed for anion formation to give 195 which could undergo an internal a ldo l - type c yc l i z a t i on to af ford hydroxyni t r i le 200 and subsequent dehydration would furnish the o l e f i n i c n i t r i l e 201. As OAc 201 a r e su l t , the chromatographic separation of crude cyanoaldehyde was tested on a very iner t support, namely, f l o r i s i l . In th i s manner, cyanoaldehyde 195 was i so lated in greater than 80% y i e l d and neither hydroxyn i t r i le 200 or o l e f i n i c n i t r i l e 201 were detected. The i so lated cyanoaldehyde 195 c r y s t a l l i z ed almost immediately to afford beautiful cubes, m.p. 110-112°C. The spectroscopic data of th i s material were in complete accord with the assigned structure. Thus, the infrared spectrum of 195 exhibited weak absorptions at 2250 crrf 1 due to the n i t r i l e f unc t i ona l i t y and at 2740 cm" 1 due to the aldehydic C-H stretching v i b r a -t i o n . In the n.m.r. spectrum of 195 a double doublet (J = 1 and 3 Hz) - 121 -appeared at x.20 due to the aldehyde proton which i s coupled to the two 115 adjacent d iastereotopic protons. The n.m.r. spectrum of 195 was determined at both 60 Mc and 100 Mc. No change in the pattern afforded by the aldehydic proton was observed at d i f f e ren t f i e l d strengths. I t , therefore, seems l i k e l y that the s p l i t t i n g s are good approximations to the coupling constants. F i n a l l y , the mass spectrum of 195 had a molecular ion peak at j^-345. In summary, a convenient and p rac t i ca l method had been developed to e f fec t an unsymmetric r ing cleavage between carbons 2 and 3 in 173-acetoxy-5B-androstan-3-one. Of addit ional value was the fact that i t seemed highly probable that the aldehyde or n i t r i l e group could be separately modified as in 116 accordance with the general plan. The f i r s t synthetic object ive had therefore been rea l i zed . 122 Part II Attempted formation o f the bicycloxazoladine skeleton Having achieved the desired r ing cleavage react ion , the stage was set for entry into the second phase of the program which i s the elaboration of cyanoaldehyde 195 to a b icyc looxazo l id ine. From the OAc OH N O outset, i t was planned to construct the c ruc ia l intermediate 1_0 from a common o l e f i n i c precursor c f . 79 which could ar i se from the unsymmetric r ing cleavage product. Hence, subsequent synthetic studies with cyano-OAc 7 9 - 123 -rS aldehyde 195 would be pr imar i l y concerned with the formation of cyanoolefin 57a (CHART XX, Page 124). I n i t i a l invest igat ions with cyanoaldehyde 195 were concerned with the preparation of cyanoiodide 203b v ia hydroxyn i t r i le 202 since the v i n y l i c side chain could be introduced by e f fec t ing the e l iminat ion of hydrogen iodide from cyanoiodide 203b as depicted in CHART XX (see Page 124). I t could be reca l led from previous experiments that t r e a t -ment of iodo esters 103c and 107c with s i l v e r f luor ide in pyr idine afforded lactones 102a and 106a (CHARTS XHIa and XH Ib , Pages 57 and 61 ). However, i t was hoped that the e l iminat ion of hydrogen iodide from cyanoiodide 203b by employing s i l v e r f l uo r ide would not be obstructed 55 66 by n i t r i l e pa r t i c i pa t i on . ' To t h i s end, treatment of cyanoaldehyde 195 with sodium boro-hydride in ethanol at room temperature for three hours afforded hydroxy-n i t r i l e 202 (CHART XX, Page 124) in 85% y i e l d as a s o l i d . T.L.C. analys is of t h i s material indicated that i t was su i tab le for use in subsequent steps. The infrared spectrum of th i s material had a broad band at 3450 cm" 1 due to the carbon-2 hydroxy group and of note was the absence of the weak band at 2740 cm" 1 due to the aldehydic C-H stretching v i b ra t i on . The sa l i en t feature in the n.m.r. spectrum of 202 was a t r i p l e t (J = 10 Hz) at T6.30 due to the protons adjacent to the carbon-2 hydroxy group. F i n a l l y , the mass spectrum of 202 possessed a molecular ion peak at j | 347. In p r i n c i p l e , the next step forward involved the dehydration of hydroxyni t r i le 202_ (CHART XX, Page 124). Treatment of hydroxyn i t r i le 202 with tosyl chlor ide in dry pyridine at room temperature for twenty-64 four hours afforded cyanotosylate 203a in 81% y i e l d . The t . l . c . of - 124 -CHART XX. Synthetic scheme for the formation of cyanoolefin 57a OAc - 125 -th i s material indicated predominantly one new compound. Of note was the appearance in the infrared spectrum of a strong doublet at 1190 cm" 1 due to the tosy late f unc t i ona l i t y and the disappearance of the broad hydroxy band at 3450 cm" 1 . Treatment of cyanotosylate 203a (CHART XX, Page 124) with sodium iodide in dimethylformamide y ie lded 65 cyanoiodide 203b in ca_. 70% y i e l d . The t . l . c . of th i s material indicated the formation of a new compound with baseline contaminants. The inf rared spectrum of the crude reaction product did not have absorp-t ion bands due to the tosylate f unc t i ona l i t y . In add i t i on , the mass spectrum of 203b had a molecular ion peak at ^ 457 with a prominent peak at 2- 330 due to cleavage of the R-I bond. I t appeared that the overa l l conversion of cyanoaldehyde 195 to cyanoiodide 203b had proceeded in a sa t i s fac to ry manner. Unfortunately, treatment of cyanoiodide 203b (CHART XX, Page 124) with s i l v e r f l uo r ide in pyr idine at room temperature afforded pre-dominantly recovered s ta r t ing mate r i a l , with no trace of an o l e f i n i c compound. Every precaution was now observed; for instance, act ivat ion cc of s i l v e r f l u o r i de , use of dry pyr id ine, exclusion of oxygen, vigorous s t i r r i n g and performing the reaction in the dark. In short, a l l attempts to el iminate hydrogen iodide from cyanoiodide 203b by employing s i l v e r f luor ide f a i l e d even a f te r long reaction times. It should be mentioned that cyanoiodide 203b was very sens i t ive to l i g h t and a i r . These resu l t s with s i l v e r f l uo r ide were very annoying since the major factors hinder-ing the conversion of cyanoiodide 203b into cyanoolefin 57a were not c l e a r l y apparent. At th i s stage, a pertinent communication by Kato and H i r a t a ^ appeared in the l i t e r a t u r e . In pa r t i c u l a r , they reported the dehydro-- 126 -halogenation of di iodide 204 with 1,5-diazabicyclo[5.4.0]undec-5-ene to afford acetoxy-diene 205 in 39% y i e l d , as a c lear o i l . 1,5-Diaza-DBu H OAc 2 Q 4 2Q5 118 bicyclo[5.4.0]undec-5-ene (DBU), and 1,5-diazabicyclo[4.3.0]non-5-119 ene (DBN) have been shown to be very ve r s a t i l e dehydrohalogenating agents. As both are much more react ive than the amines generally used, much milder conditions can be employed. Accordingly, the reaction of cyanoiodide 203b (CHART XX, Page 124) with DBU and DBN was examined over a wide temperature range. At lower temperatures (below 70° ) impure s ta r t ing material was recovered. When the reaction was carr ied out at temperatures above 100° there was a marked tendency for hydroxyn i t r i le 120 202 to be formed. Taurins et al_., have reported that treatment of a-bromoketone 206 with pyridine under re f lux ing conditions y ie ld s the pyridinium s a l t 207. Furthermore, the action of pyridine on bromide P y / A 2 0 6 2Q7 - 127 -s a l t may displace the pyridine r ing from the quaternizing group. For example, substituted a l ky l a t i on of primary alcohols has been car r ied out 123 with alkoxymethyl pyridinium sa l t s . In view of these considerations i t appeared reasonable to assume that 1,5-diazabicyclo[5.4.0]undec-5-ene (DBU) and 1,5-diazabicyclo[4.3.0]non-5-ene (DBN) had displaced the iodide in a S^2 manner to give a quaternary s a l t which on work up afforded hydroxyn i t r i le 202. As a r e s u l t , experiments were car r ied out with OAc 2 Q 2 dimethylaminotrimethylstannane, a powerful dehydrochlorinating reagent, which i s notable for e f fec t ing dehydrochlorination under mild condit ions. For example, the reaction of n-butyl ch lor ide (210) with dimethylamino-- 128 124 trimethylstannane at 40° gives trans but-2-ene (211) in over 80% y i e l d . CH 3 (CH2)3 Cl 2 1 0 (CH 3) 3SnN(CH 3) 2 CH, V CH=CH \ 211 CH, Treatment of cyanochloride 203c with dimethyl aminotrimethylstannane at 40° for four hours afforded several products as evidenced by t . l . c . The infrared spectrum of the crude reaction product, however, did not indicate any o l e f i n i c products. Because of th i s re su l t and since considerable quant it ies of cyanochloride 203c (CHART XX, Page 124) had been obtained, experiments 125 were carr ied out to e f fec t dehydrochlorination by employing c o l l i d i n e . For example, the action of re f lux ing c o l l i d i n e on bromoketone 212 gives 125a 213 in excel lent y i e l d . In contrast, when cyanochloride 203c was - 129 -subjected to re f lux ing c o l l i d i n e under nitrogen, impure s ta r t i ng material was recovered. It i s worthy of mention that ethyldicyclohexylamine has been recommended as a base for the dehydrochlorination of primary 126 ch lor ides . This base was not employed because of s t e r i c factors and the results from previous experiments with c o l l i d i n e . At th i s stage, other well documented synthetic procedures 127 ~ for introducing the v i n y l i c side chain were considered. The presence of the cyano f unc t i ona l i t y , however, severely l im i ted the number of promising methods. Since the dehydrotosylation of compound 107b had been effected by employing c o l l i d i n e , experiments were carr ied out to prepare cyanoolefin 57a by a s im i l a r procedure. A l l attempts to produce Me0 2 C 1Q7b, Y = OTS OAc any of cyanoolefin 57a by th i s method f a i l e d . 1 Synder and Soto have studied the action of potassium _t-butoxide in dimethylsulphoxide on sulphonate esters of primary a l i pha t i c a lcohols. They reported that the sulfonate esters of primary a l i p ha t i c 129 alcohols afforded predominantly _t-butyl esters . Nace has demonstrated that hexamethylphosphoramide i s an excel lent solvent for converting secondary sulphonates to the corresponding o l e f i n s . These reports - 130 -prompted an examination of the action of potassium t-butoxide on cyano-tosy late 203a in dimethylsulphoxide and hexamethylphosphoramide over a wide temperature range. As noted e a r l i e r , strong base and high temperatures favours e l iminat ion over subs t i tu t ion . No useful resu l t s viere obtained with dimethyl sul phoxide as the reaction medium. On the other hand, treatment of cyanotosylate 203a with potassium t_-butoxide in hexamethylphosphoramide at 160° under N 2 gave two major products as indicated by t . l . c . Of note was the appearance in the infrared of the OAc 215 crude product of absorption bands at 930 cm" 1 and 1000 cm" 1 due to out-of-plane C-H bending v ibrat ions of the v i n y l i c side chain. A weak absorption band was evident at 1630 cm" 1 which could be at t r ibuted to the C=C stretching v ibrat ion of the v i n y l i c side chain. Absorptions at - 131 -2250 cm 1 and 1720 cm 1 indicated that the n i t r i l e and 176-acetate groups had remained i n t ac t . The pert inent spectral features in the n.m.r. of the crude reaction product were as fo l lows. A broad s ing le t t o t a l l i n g three protons appeared at T9.20 due to the carbon-18 t e r t i a r y methyl protons of compounds 57a and 215 while two s inglets t o t a l l i n g three protons at T8.83 and x8.75 could be att r ibuted to the carbon-19 t e r t i a r y methyl protons of compounds 57a_ and 215, respect ive ly. A broad mul t ip le t between x5.10 and x4.20 t o t a l l i n g ca_. two and hal f protons was evident for the o l e f i n i c protons of 57a and the carbon-17 protons of 57a and 215. In f a c t , the n.m.r. s ignals which have been assigned to the above o l e f i n i c compound 57a were in accord with the n.m.r. s ignals of pure cyanoolefin 57a prepared subsequently. The assumption that compound 215 had been formed also was made on mechanistic grounds. Heusler, 130 Wieland, and Wettstein have reported the conversion of ketotosylate 216 to compound 217 by employing potassium i -butox ide. In view of th i s re su l t i t appeared reasonable to assume that the a n i t r i l e anion had displaced the tosy late group in a Sw2 type react ion. Even a f te r con-- 132 -siderable experimentation the maximum y i e l d of cyanoolefin 57a, prepared from cyanotosylate 203, was always less than 40%. In general, the incorporation of the double bond by e f fec t ing the el iminat ion of HX from a compound of general type 218 had proved rather unpromising. One of the major d i f f i c u l t i e s was that subst i tut ion OAc 218 5 7 g occurred at carbon-2. I t i s now appropriate to discuss the e f fect s of 127 changes in the substrate on e l iminat ion versus subs t i tu t ion . The carbon containing the leaving group i s referred to as the a-carbon and the carbon which loses the proton as the 3-carbon. Under second-order conditions a branching increases e l iminat ion to the point where t e r t i a r y substrates undergo few S^2 react ions. For example, Table IV (see Page 133) shows resu l t s on some simple a lky l bromides. Two reasons may be presented for t h i s trend. One i s s t a t i s t i c a l : as a branching increases, there are usual ly more hydrogens for the base to attack. The other i s that a branching presents s t e r i c hindrance to attack of the base at the carbon. 6 Branching also increases the amount of E2 e l iminat ion with respect to S^2 subs t i tu t i on , as shown in Table IV (see Page 133), not because el iminat ion is faster but because the S^2 mechanism i s so great ly slowed due to s t e r i c factor s . These considerations tend to - 133 -Table IV The e f fec t of a and 3 branching on the rate of E2 e l iminat ion and on the amount of o l e f i n formed. The reactions were between the a lky l bromide and OEt 6 . 1 2 7 Substrate Temperature O lef in % CH 3CH 2Br 55° 0.9 (CH 3) 2CHBr 25° 80.3 (CH 3) 3CBr 25° 97 CH 3CH 2CH 2Br 55° 8.9 CH3CHCH2Br 1 CH3 55° 59.5 - 134 -indicate that the introduction of unsaturation into a compound of general type 218 by a base catalysed el iminat ion of HX i s rather unfavourable. OAc -HX However, most recently Uskokovic and his colleagues reported the synthesis of 9-epi-quinine and 9-epi-quinidine in which they converted chlor ide 219a to o l e f i n 220b. The most e f f i c i e n t method to introduce 2 1 9 a , R = C 2 H 5 22Q a . R = H 219 b , R = H 2 2 Q b,R = C 2 H 5 the double bond consisted of a reaction sequence involv ing saponif icat ion of 219a, dehydrochlorination of acid 219b with potassium t-butoxide in - 135 -dimethylsulphoxide/benzene, and e s t e r i f i c a t i o n of the re su l t i ng acid 220a to give the desired N-benzoylhomomeroquinene ethyl ester 220b. Hence, OH 2 2 2 converting the ester f unc t i ona l i t y to the acid group presumably prevented an intramolecular c y c l i z a t i on reaction s im i l a r to reaction XIII from occurr ing. The p o s s i b i l i t y would also ex i s t fo r the formation of a lactone. However, c y c l i z a t i on to an eight-membered r ing lactone has low p robab i l i t y . On the other hand, the action of potassium t-butoxide on chloroacid 221 may form the seven-membered r ing lactone 102a rather than e f fec t dehydrochlorination to afford compound 222. Dehydrochlorination of acid 219b involves the e l iminat ion of hydrogen chlor ide adjacent to a t e r t i a r y centre while in compound 221 dehydrochlorination would occur adjacent to a quaternary centre. Several methods had been employed to synthesize compound 57a which formally represents an ox id i zat ion of carbon-1. I t was now con-- 136 -C H 3 ( C H 2 ) 4 C H 2 C H O (CH 3C0) 20 C H 3 C 0 2 K / A *- C H 3 ( C H 2 ) 4 C H = C H — O C O C H 3 2 2 6 B r 2 / C C l 4 C H 3 ( C H 2 ) 4 C H — C H — O C O C H 3 I I B r B r C H 3 O H C H 3 ( C H 2 ) 4 C H — C H ( O C H 3 ) 2 B r H 3 Q /A C H 3 ( C H 2 ) 4 C H C H O B r 2 2 7 tempiated that the introduction of an appropriate substituent at carbon-1 (oxidation process) could f a c i l i t a t e the introduct ion of the double bond. For example, halohydrins can be converted into the corresponding 132 o l e f i n i c product by zinc dust/acetic ac id . Hence, the generation of the double bond might be accomplished by reduction of hydroxybromide 225 (CHART XXI, see Page 137). Examination of the l i t e r a t u r e revealed that the d i r e c t bromination of aldehydes i s often complicated by a 133 134 competing reaction with the aldehyde C-H bond. Bedoukian has reported the formation of the bromoaldehyde 227 by bromination of the enol acetate 226 followed by the action of methanol and subsequent acid treatment. As a r e su l t , experiments were carr ied out to invest igate - 137 -CHART XXI. Second general scheme for the formation of cyanoolefin 57a OAc cis 223a trans 223 b 5 7 a - 138 -OAc H H 1 9 5 CIS 2 2 3 a t r a n s 2 2 3 b the enol acety lat ion of cyanoaldehyde 195. I t was rather surpr i s ing to 2 2 8 2 2 9 f ind that the conversion of an aldehyde into the corresponding enol acetate has been the subject of only 1imited experimentation. Grafen 135 and his colleagues treated aldehyde 228 with potassium acetate in acet ic anhydride at 135° for s ix hours to afford enol acetate 229. Oka - 139 -OAc and Hara employed isopropenyl acetate containing sulphuric acid to convert aldehyde 5_2 to the enol acetate 52a. The action of isopropenyl 231 acetate containing p-toluenesulphonic acid on aldehyde 230 gave enol acetate 231 in 35% y i e l d as reported by Moffett et_ aj_. . Cameron et - 140 -137 al_., demonstrated that treatment of aldehyde 232 with isopropenyl acetate in the presence of pyridine gave enol acetate 233. At t h i s juncture i t AcO C H O O A c OAc/Py 2 3 2 2 3 3 was considered by us that the chemical behaviour of aldehydes and ketones 135 134 were comparable. Since Grafen and his colleagues, Bedoukian, and others have employed potassium acetate - re f lux ing acet ic anhydride for enol acetylat ions with varying degrees of success attempts were made to form the enol acetate of the cyanoaldehyde 195 under s im i la r condit ions. The dark reaction product, however, consisted of numerous products accompanied by insoluble mater ia l . A number of t r i a l experiments were, therefore, i n i t i a t e d to ascertain the most favourable conditions for th i s react ion. For example, iospropenyl acetate - concentrated sulphuric a c i d , ^ ' ^ acet ic anhydride - perch lor ic a c i d , ^ isopropenyl acetate -136 141 142 p-toluenesulphonic a c i d , ' Edwards-Rao reagent, and isopropenyl 137 acetate - pyridine were employed. The reactions were ca re fu l l y monitored by t . l . c . on s i l i c a ge l . In summary, the reaction of cyanoaldehyde 195 with re f lux ing isopropenyl acetate - concentrated sulphuric acid was of s u f f i c i e n t interest to warrant further investigatons. It was found that the action of acet ic anhydride - perch lor ic acid in ethyl acetate on cyano-- 141 -aldehyde 195 gave compound 234 with several minor components as indicated by t . l . c . analys i s . The infrared spectrum of th i s crude product indicated OAc OAc the absence of the aldehyde C-H stretch ing v ibrat ion at 2750 cm" 1 . Of note was the appearance of a strong carbonyl band at 1750 cm" 1 which could be at t r ibuted to the acet ic acy la l carbonyls. The n.m.r. spectrum of the crude product had three three-proton s ing lets at ca .^ T 8 due to the acet ic acy la l group and the 17g-acetoxy f unc t i ona l i t y . A one-proton t r i p l e t (J = 7 Hz) appeared at T3.14 due to the proton adjacent to the 143 acet i c acy la l group. Most recent ly , Andersen et al_., have found that the action of acet i c anhydride - perch lor ic acid in ethyl acetate on an a l i p ha t i c aldehyde gave predominantly the corresponding acet ic a c y l a l . With the other reaction conditions an overwhelming mixture of compounds were formed and i t appeared that extensive oxidation had occurred. Employ-ing isopropenyl acetate - concentrated sulphuric acid conditions attempts were made to improve the y i e l d by varying the react ion parameters. Of overr id ing importance, however, was the awareness that cyanoaldehyde 195 was very susceptible to ox idat ion. Under the general reaction conditions i t would read i l y undergo ox idat ion. Undoubtedly th i s would lower y ie ld s and present pu r i f i c a t i on d i f f i c u l t i e s . Attempts to obviate these problems - 142 -by adding an i n h i b i t o r and continuously bubbling nitrogen through the reaction so lut ion proved promising. The addit ion of hydroquinone, 2 ,6 -d i - te r t -buty l -pheno l , or 2,5-di-tert-butyl-p-benzoquinone to the reaction prevented appreciable ox idat ion. However, i t was most c r i t i c a l to employ more than 100 mg. of i n h i b i t o r to 500 mg. of cyanoaldehyde otherwise appreciable amounts of dark in t ractab le material was formed. The u t i l i t y of the i nh i b i t o r was governed by i t s chromatographic behaviour. 2,5-Di-tert-butyl-p-benzoquinone was found to be the most e f fec t i ve and p rac t i ca l i n h i b i t o r in terms of preventing oxidation and chromatographic separation from the reaction product. It i s noteworthy that the enol acety lat ion was very sens i t ive to reactant concentration, reaction time, and temperature. The i s o l a t i on of enol acetates 223a and 223b (CHART XXI, Page 137) as a mixture was achieved by column chromatography on s i l i c a gel with CHC13 as the eluent. In summary, the enol acetates could be obtained in 62-68% i so lated y i e l d when the reaction was carr ied out by employing isopropenyl acetate - concentrated sulphuric acid under care-f u l l y contro l led conditions in the presence of 2 , 5 -d i - t e r t - bu t y l - p -benzoquinone. T.L.C. analysis of th i s material indicated the presence of two compounds in ca_. a 1:1 r a t i o . The spectroscopic properties of the mixture of enol acetates 223a and 223b were in accord with the assigned structures. Of note was the appearance in the infrared spectrum of an intense band at 1745 cm" 1 due to the v i n y l i c acetate carbonyl. A weak band at 1660 cm" 1 was evident for the carbon-1 carbon-2 double bond. The sa l i en t features in the n.m.r. spectrum of the mixture of enol acetates were a broad s ing let at T9.20 due to the carbon-18 t e r t i a r y methyl groups of 223a and 223b while two sharp s ing lets at T8.80 and x8.70 ( t o t a l l i n g - 143 -three protons) could be att r ibuted to the carbon-19 t e r t i a r y methyl groups of 223a and 223b. Two AB pa i r of doublets ( t o t a l l i n g two protons) appeared at T5.42, T4.74 and x3.04, T2.96 ( J = 8 and 13 Hz) due to the carbon-1 and carbon-2 o l e f i n i c protons of the c i s and trans enol acetates. The n.m.r. spectrum of the i so lated mixture of enol acetates (small scale) suggested that the c i s and trans enol acetates 223a and 223b were present in ca_. a 1:1 r a t i o . However, t h e i r separation was unnecessary since in subsequent steps the mixture of enol acetates would afford compound 57a (CHART XXI, Page 137). F i n a l l y , the mass spectrum of the mixture of enol acetates had a molecular ion peak at ^ 387. Having achieved the enol acety lat ion of cyanoaldehyde 195 i t was decided to brominate the enol acetates in a manner analogous to that employed by Bedouk ianJ 3 4 ( see Page'135) Preliminary studies d i rected along these l i n e s , however, proved un f ru i t f u l and led only to complex mixtures. I t i s worthy of note that bromine may be introduced alpha to a n i t r i l e group by d i r e c t bromination, which could be a complicating 144 factor in th i s react ion. For example, Robb and Schultz reported that the action of bromine on n i t r i l e 235 gave bromonitr i le 236. One B r 2 3 5 2 3 6 - 144 -important aspect of th i s study was the construction of the mixture of enol acetates 223a and 223b in reasonable quant i t ies . Hydroboration of the enol acetates followed by treatment with 145 acet ic anhydride could give cyanoolefin 57a. However, the d i r ec t i ve e f fect of the acetoxy group i s small and both a - and s- boron i n t e r -mediates would be formed. The s i tuat ion i s also complicated by the greatly enhanced tendency for the intermediate to undergo e l iminat ion, 146 147 with subsequent rehydroboration. Lewis and Pearce have reported that when N-cyclohex- l -enylpiper id ine (237) was treated with diborane in tetrahydrofuran and the re su l t ing organoborane 238 was subjected to re f lux ing diglyme in the presence of a carboxyl ic a c i d , cyclohexene (239) was formed. This procedure has been applied to a number of enamines of 2 3 9 c y c l i c and acyc l i c ketones and the alkenes have been obtained in y i e ld s of greater than 80%. It i s worthy of note that they did not report the formation of alkenes from the corresponding aldehydes v ia hydroboration - 145 -of the enamines Nevertheless, experiments were carr ied out to prepare cyanoolefin 57a from enamine 240 v ia hydroboration. Enamine 240 was prepared by employing conditions analogous to that used by Herr and 148 Heyl. Treatment of cyanoaldehyde 195 with piper id ine in benzene under re f lux ing condit ions using a Dean Stark trap gave enamine 240 in good y i e l d . T.L.C. analys is of the crude product indicated one major com-ponent. The infrared spectrum of the crude product had a weak band at 1640 cm" 1 due to the carbon-1 carbon-2 double bond and a weak band at 960 cm" 1 due to the out-of-plane o l e f i n i c C-H bending v ibrat ions. The sa l i en t features in the n.m.r. of the crude product were a three-proton s ing let at T8.87 due to the carbon-19 t e r t i a r y methyl group of 240 and a two proton AB pair of doublets (J = 15 Hz) at 5.84 and 4.24 due to the carbon-1 and carbon-2 o l e f i n i c protons of 240, respect ive ly . Treat-- 146 -ment of enamine 240 with diborane in tetrahydrofuran with subsequent re f lux ing in diglyme in the presence of acet ic acid did not give cyano-o l e f i n 57a. T.L.C. analys is of the crude product indicated the presence of one major compound with baseline contaminants. However, the i n f r a -red spectrum of the major component did not indicate any v i n y l i c side chain cha rac te r i s t i c s . Of note was the appearance in the inf rared spectrum of a very weak n i t r i l e band at 2250 cm" 1 . I t i s documented that diborane can e f fec t the reduction of n i t r i l e s to primary amines. In summary, the hydroboration approach had proved unpromising and, therefore, i t was planned to further invest igate the u t i l i t y of the mixture of enol acetates 223a and 223b. OAc 2 2 3 a . c i s 2 2 3 b t t r an s The formation of the enol acetates formally represented the oxidation of carbon-1 with the introduction of the required degree of unsaturation. On th i s basis two possible pathways became apparent for advancing the general plan. The f i r s t approach f u l l y explo ited the 149 unsaturated nature of the enol acetates. Hassner ejt al_., have studied the addition of iodine azide to o le f in s followed by l i th ium aluminum hydride reduction to afford a z i r i d i ne s . As a re su l t of th i s work i t was - 147 -hoped that iodine azide would add to enol acetates 223a and 223b. The p o s s i b i l i t y could then ex i s t for the reduction of the iodo azide adduct 241 to give an a z i r i d i ne of general type 242. The mode of addit ion OAc 2 4 2 would presumably govern the course of the reduction react ion. The stereochemical aspects of th i s addit ion reaction are worthy of consider-149c a t ion . Hassner has suggested that iodine azide addit ion involves - 148 -e lect roph i1 ic attack on the o l e f i n with formation of a three-membered-r ing iodonium ion , equation XIV. Once a three-membered-ring i s formed, H R H R \ / v T j—N—solvent • I +J(solvent) + N 3 (solvent) / \ / \ H R H R (XIV) H R N— C I C 1 / \ H R opening w i l l occur from the back side re su l t i ng in trans addit ion of iodine azide. A good probe fo r the mechanism i s often the regiochemistry of a react ion. The addit ion of iodine azide to terminal o l e f i n s gives a three-membered-ring ion itermediate 243 which undergoes opening v ia the lower energy t rans i t i on state 244 rather than 246 when R can s t a b i l i z e an inc ip ient pos i t ive charge. I f these considerations are v a l i d , then an - 149 -N3 9) Nf R C H — C H 2 - RCH — C H 2 - RCH CH? \/ V/ \/ S+I I £ 4 4 . 2 4 3 2 4 6 2 4 5 2 4 7 C H OCCH = CH 2 4 8 2 5 2 2 electron-withdrawing R group in 243 should de s tab i l i ze t rans i t i on state 244 r e l a t i ve to 246. In f a c t , Hassner demonstrated that ester 248_ leads to a 20:30 mixture of 247:245 (R = C0 2 C 2 H 5 ) . In addit ion to the e lect ron ic 149c factors one must consider s t e r i c factor s . Hassner found that addit ion iodine azide to t-butylethylene (249) gave exc lus ive ly 250. Thus the large t -buty l group exerts a strong s t e r i c e f fec t in the opening of the three-membered-ring iodonium ion. On the basis of these considerations i t was f e l t that the addit ion of iodine azide to enol acetates 223a and - 150 -X CH=CH, 2 4 9 IN, CH — C H 2 N 3 2 5 0 223b would give predominantly, i f not exc lu s i ve l y , an adduct of general type 241. I t i s i n te res t ing to note that since two new asymmetric centres are being created in both the c i s and trans enol acetates 223a and 223b there ex i s t s the p o s s i b i l i t y for the formation of eight new compounds (CHART XXII, Page 151). In view of the above discussion i t appeared l i k e l y that compounds 241-a,b,c, and d would be formed. However, the reduction reaction would give r i se to only two az i r i d i ne s 242a and 242b. Therefore, the stereochemical problem of the addit ion reaction did not cause concern. - 151 -- 152 -In order to f ind general reaction conditions which would give the smooth addit ion of iodine azide to enol acetates 223a and 223b a series of model experiments were i n i t i a t e d . To th i s end cyclohexanone (251) was converted into enol acetate 252. Treatment of enol acetate 251 a OAc 2 5 2 2 5 4 - 153 -252 with iodine azide under conditions s im i l a r to that employed by Hassner 149 et al_., gave ca. 50% addit ion as indicated by t . l . c . and infrared spectroscopy. On the other hand, when the concentration of iodine azide ( in s i tu ) was doubled complete addit ion occurred. The t . l . c . of the crude product on s i l i c a gel suggested the presence of one major component with polar baseline contaminants. The crude product was chromatographed on s i l i c a gel in benzene to afford a pure iodo azide adduct. The spectro-scopic properties of th i s adduct were in accord with structure 253. The infrared spectrum of 253 had a strong band at 2120 cm" 1 due to the azide group and a carbonyl band at 1740 cm*"1. The n.m.r. spectrum of 253 had a one-proton t r i p l e t (J = 6 Hz) at T5.37 which could be at t r ibuted to the proton adjacent to the iodide group. Although th i s assignment of structure was ambiguous, in view of e lec t ron ic considerations i t was f e l t that 253 would be formed rather than 254, since opening of iodonium ion 255 would proceed v ia t r an s i t i on state 256 where the acetoxy group e 2 5 5 2 5 6 can s t a b i l i z e the inc ip ient pos i t i ve charge. Treatment of aldehyde 257 with potassium acetate in re f lux ing acet ic anhydride gave enol acetate 258 in f a i r y i e l d . Subjecting enol acetate 258 to 2.3 molar equivalents of - 154 -iodine azide ( in s i tu ) for two hours gave iodo azide 259. The t . l . c . of the crude product indicated predominantly one compound. The crude product was chromatographed on s i l i c a gel i n CHC13 to give an iodo azide in 63% y i e l d . The spectroscopic properties of iodo azide 259 were in agreement with the assigned structure. Thus, the inf rared spectrum of 259 had a strong absorption band at 2120 cm" 1 due to the azide band and a carbonyl band at 1740 cm" 1 . The n.m.r. spectrum of 259 displayed a one~proton mul t ip le t at x5.90 due to the proton adjacent to the iodide group and a pa i r of doublets (J = 4.5 Hz) appeared at x4.13 and x4.0 which could be att r ibuted to the proton adjacent to the azide and acetate f u n c t i o n a l i t i e s . Since the v.p.c. of the crude enol acety lat ion product indicated a mix-ture of c i s and trans isomers the addit ion of iodine azide to the mixture - 155 -of enol acetates 258 could give two iodo azide diastereoisomers 259a and 259b which may account for the appearance of a pair of doublets in the n.m.r. spectrum of the iodo azide adduct. I t was now hoped that H H 259a 259b the reduction of iodo azides 253 and 259 would give r i s e to az i r id ines 260 and 261, respect ive ly . At th i s po int, i t was important to bear in mind that with iodo azide 241 i t would be necessary to s e l ec t i ve l y reduce the azide group in the presence of the n i t r i l e f unc t i ona l i t y . I t i s well documented 150 that l i th ium aluminum hydride can reduce n i t r i l e s to primary amines. 149 However, in l i g h t of the work of Hassner ejt al_., i t was decided to study the reduction of iodo azides 253 and 259 with l i th ium aluminum hydride in order to ascertain the v a l i d i t y of th i s approach. Treatment of iodo azide 253 with l i th ium aluminum hydride in d iethy l ether for twelve hours at room temperature followed by the action of 20% aqueous sodium hydroxide gave predominantly one product as indicated by v.p.c. The infrared spec-trum of a v.p.c. sample showed the absence of both azide and carbonyl bands and the product had a strong band at 3400 cm" 1 . Concurrently, the l i th ium aluminum hydride reduction of iodo azide 259 was carr ied out. This reduction reaction gave a product which had s im i l a r spectral features to the above product. However, the y ie ld s of crude product in these reduction reactions were very low. This could possibly be at t r ibuted - 156 -to the v o l a t i l e nature of the re su l t ing compounds and to t h e i r high s o l u b i l i t y in water. In general, i t was f e l t that the reduction of the iodo azides should be further investigated with more se lect i ve 151 reducing agents. Smith et al_., have reported that sodium borohydride in re f lux ing isopropanol can reduce azide groups to the corresponding amine. Since n i t r i l e f unc t i ona l i t i e s are not usual ly reduced with 152 sodium borohydride, i t was decided to study the reduction of iodo azides 253 and 259 under these condit ions. For example, treatment of iodo azide 25_3with an excess of sodium borohydride in re f lux ing i s o -propanol for three hours gave a product whose infrared spectrum showed the absence of the azide band at 2120 cm" 1 and of the carbonyl band at 1740 cm" 1 . However, the y i e ld s of crude product again were very low. S im i l a r l y , the reduction of iodo azide 259 with sodium borohydride gave low y ie lds of crude product. Attempts to i s o l a te products from other reduction reactions were not f r u i t f u l . Very l im i ted studies on the addit ion of iodine azide to other enol acetates and the subsequent reduc-t ion of the adducts d id not y i e l d pos i t i ve re su l t s . The addit ion of iodine azide to the mixture of enol acetates 223a and 223b by employing conditions analogous to that used in the model studies gave predominantly s ta r t ing mater ia l . On the other hand, addit ion under more d ra s t i c conditions led to a complex mixture of products. As a r e s u l t , th i s type of approach was abandoned and an a l ternate pathway was considered. This second pathway involved the ozonolysis of enol acetates 4 135 153 223a and 223b with subsequent reductive work up ' ' to afford cyano-154 aldehyde 262. Wi t t ig reaction on cyanoaldehyde 262 could then give r i s e to cyanoolefin 57a(CHART XXIII Page 158). Although the removal and - 157 -reintroduct ion of carbon-2 did not seem a t t r ac t i ve from a l o g i s t i c view-point, th i s approach, nevertheless, represented a possible pathway which would circumvent the problems associated with the dehydration of hydroxy-n i t r i l e 202, the bromination of the enol acetates 223a and 223b, and the generation of the double band v ia hydroboration of enamine 240. This OAc AcO' OAc H 2 2 3 a . c i s 2 2 3 b , t r an s scheme (CHART XXIII, Page 158) was therefore subjected to experimental scrut iny. Ozonolysis of enol acetates 223a and 223b under conditions 135 analagous to that employed by Grafen and his colleagues followed by z i nc -ace t i c acid work up gave cyanoaldehyde 262 accompanied by several contaminants as evidenced by t . l . c . and n.m.r. spectroscopy of the crude product. Several attempts were made to pur i f y cyanoaldehyde 262_ by chromatographic methods. Because of the sens i t ive nature of cyanoaldehyde 262 the i so lated y i e ld s were very low. The crude cyanoaldehyde 262 appeared to be very susceptible to aer ia l ox idat ion. In add i t i on , th i s material was found to give poor resu l t s in the next step of the projected synthesis. The ozonolysis was then carr ied out in various solvents for d i f f e ren t reaction times and several reductive work up procedures were 155 156 examined; for example, dimethyl sulphide, palladium - hydrogen, and sodium sulphite 157 In contrast to the e a r l i e r experiments, the - 158 -CHART XXIII. Formation of cyanoolefin 57a v i a a W i t t i g reaction H 5 7 a - 159 -ozonolysis of enol acetates 223a and 223b in ethyl acetate with subsequent standing at dry ice acetone temperatures for t h i r t y - f i v e minutes followed by the action of 5% aqueous sodium s u l f i t e in aqueous methanol fo r two and one-half hours gave cyanoaldehyde 262 in ca. 86% y i e l d and in a good state of pur i ty . I t i s worthy of mention that these conditions must be adhered to r igorous ly otherwise there i s a marked tendency for impurit ies to ar i se and lower y i e l d s . The spectroscopic properties of th i s material were in complete accord with the assigned structure. Of note was the disappearance in the inf rared spectrum of the enol acetate carbonyl band at 1740 cm" 1 and the appearance of a weak band at 2725 cm" 1 due to the aldehydic C-H stretching v i b ra t i on . In the n.m.r. spectrum of 262 a three-proton s ing le t appeared at x8.97 due to the carbon-19 t e r t i a r y methyl group and a sharp s ing let at T.46 could be att r ibuted to the aldehydic proton. The mass spectrum of 262 had a molecular ion peak at ^ 3 3 1 . Cleavage of the C-H bond next to the oxygen atom gave a strong peak at ^ 330. F i n a l l y , a small sample was converted into i t s 2.4 D.N.P. der i va t i ve , m.p. 186-188°, fo r analys i s . Having obtained considerable quant it ies of aldehyde 262 i t was next planned to invest igate the formation of cyanoolefin 57a v ia a Wi t t i g reaction (CHART XXIII, Page 158). Corey et_al_. , 1 5 8 have reported a Wi t t i g condensation reaction with a t e r t i a r y aldehyde using methylsu l f iny l 159 carbanion - dimethylsulphoxide. However, the condensation of cyano-aldehyde 262 with methylenetriphenylphosphorane under conditions analogous 158 to that employed by Corey et al_., led to a mixture of compounds. The major component was i so lated by preparative t . l . c . in low y i e l d . The infrared spectrum of th i s material showed the absence of the n i t r i l e band - 160 -at 2250 cm" 1 . I t i s i n teres t ing to note that Nagata st_al_. had effected a Wi t t ig condensation with cyanoketone 263 although in t h i s case the opportunity may not ex i s t for the y l i d e to react with the n i t r i l e group. In add i t i on , i t has been reported that the resu l t ing 2 6 3 y l i d e in the Wi t t i g reaction can react with a neighbouring group. For •J go instance, Taub and his colleagues have found that in the Wit t ig con-densation with aldehyde 264 the y l i d e 265 reacts with the ester group. OMe - 161 -The y ie ld s of desired o l e f i n were very low. S im i l a r l y , the resu l t ing y l i d e 266 could react with the n i t r i l e group. This type of reaction 2 6 6 could explain the formation of diverse products not having a n i t r i l e f unc t i ona l i t y . Furthermore, sodium borohydride reduction of cyanoaldehyde 262 in ref lux ing isopropanol gave one major component whose infrared spectrum indicated the absence of the n i t r i l e band at 2250 cm" 1 . Mass 2 6 2 2 6 7 spectral evidence suggested that compound 267 had been formed. The Wit t ig condensation reaction was studied in various solvents and the phosphorane was generated by employing other well documented procedures. It was hoped that the above i n te r fe r i ng reaction could possibly be prevented by carry ing out the condensation in a non polar solvent. Generation of the - 162 -phosphorane by the action of n-butyl l i th ium on (methyl)-triphenylphosphonium bromide in benzene followed by the addit ion of cyanoaldehyde 262 with subsequent s t i r r i n g at room temperature for two days y ie lded cyanoolefin 57a in ca_. 25% i so lated y i e l d . Although the i so lated y i e l d was low the reaction had taken the desired reaction course. In order to optimize the y i e l d s , the reaction times and reactant concentrations were var ied. Fortunately, the progress of the reaction could be followed by t . l . c . on s i l i c a ge l . Af ter considerable experimental e f f o r t in th i s d i rec t ion the average y i e l d of pu r i f i ed cyanoolefin was s t i l l low. Of pa r t i cu l a r importance, however, was the fact that several compounds had been i so lated which did not bear an acetate group. A survey of the Wi t t i g reaction with various ketosteroids indicated that functional groups, for example, 1 CO hydroxy and acetoxy,decreased the y i e l d . At t h i s stage, the low y i e l d of cyanoolefin 57a was ascribed to the presence of an addit ional react ive s i t e , namely, the 178-acetoxy group and the hindered nature of the carbonyl f u n c t i o n a l i t y . ^ 3 Ireland e_t al_. J 6 4 have reported the conden-sation of t e r t i a r y aldehyde 268 with methylenetriphenylphosphorane to afford compound 269 in ca_. 75% y i e l d by employing a large excess of phosphorane reagent. When, in the event, a large excess of methylenetr i -2 6 8 2 6 9 - 163 -phenylphosphorane was employed for short reaction times compound 57b was i so la ted. Subsequent acety lat ion of compound 57b gave cyanoolefin 57a OAc in ca_. 50% pu r i f i ed y i e l d as a white c r y s t a l l i n e s o l i d , m.p. 132-135°. A small sample was sublimed at 125° (.1mm. pressure) to y i e l d beautiful needles, m.p. 134-135°, for ana lys i s . The spectroscopic properties of 57a were in complete accord with the assigned structure. Thus, the infrared spectrum of 57a had bands at 920 cm" 1 and 980 cm" 1 due to out-of-plane o l e f i n i c C-H bending v ib ra t ions . A weak band appeared at 1630 cm" 1 due to the carbon-1 carbon-2 double bond stretch ing v ibrat ion and a band at 2250 cm" 1 indicated that the n i t r i l e group had remained in tac t . The n.m.r. spectrum of 57a possessed a three-proton mu l t ip le t - 164 -between T5.10 and T4.20 due to the carbon-1 and carbon-2 o l e f i n i c protons. F i n a l l y , the mass spectrum of 57a had a molecular ion peak at ^ 329 with a prominent peak at ^-301. 165 More recent ly , Corey e_t al_., have described conditions for e f fect ing a Wi t t ig reaction with a hindered aldehyde in good y i e l d . Employing Corey's reaction conditions with a large excess of methylene-t r i phenyl phosphorane afforded cyanoolefin 57b in ca_. 65% pu r i f i ed y i e l d . The spectroscopic data of 57b were also in accord with the assigned structure. Thus, the infrared spectrum of 57b had bands at 920 cm" 1 , 990 cm" 1 , and 1630 cm" 1 due to the o l e f i n i c side chain and a broad band at 3400 cm" 1 was ascribed to the 178-hydroxy group. In the n.m.r. spectrum of 57b a three-proton mu l t ip le t at T5.0 - 4.0 could be at t r ibuted to the carbon-1 and carbon-2 o l e f i n i c protons and a one-proton t r i p l e t (J = 9 Hz) at T6.34 was ascribed to the proton adjacent to the 17B-hydroxy group. La s t l y , the mass spectrum of 57b_ possessed a molecular ion peak at 2. 287 with a prominent peak at ^ 269. Acety lat ion of compound 57b with acet ic anhydride in pyridine afforded 57a in quant i tat ive y i e l d . In summary, cyanoolefin 57a had been obtained in ca_. 34% over-a l l y i e l d based on cyanoaldehyde 195. At the beginning of th i s thes is OAc 1 9 5 57Q. - 165 -i t was mentioned that an o l e f i n i c compound could serve as a common pre-cursor. Cyanoolefin 57a represents such a precursor where -CH2X i s a n i t r i l e group (See 79_, Page 41). The preceeding synthetic work had therefore proceeded along l ines in accordance with the general plan. It should be reca l led that the next phase of the projected synthesis involved the elaboration of the cyanoolefin 57b to an a z i r i d i n e of general type 75 (CHART IX, Page 40). As noted previous ly, Hassner 149 ejt aj_., reported that the addit ion of iodine azide to t -buty l ethylene (249) gave only compound 250. Subsequently, l i th ium aluminum hydride 149 reduction of 250 afforded the appropriate a z i r i d i ne in exce l lent y i e l d . H In l i g h t of th i s work i t was proposed that the addit ion of iodine azide to 57a followed by reduction would afford an a z i r i d i ne of general type 242. It i s important to note that compounds possessing a primary iodo group 149 are susceptible to hydrogenolysis viith l i th ium aluminum hydride. As a resu l t of these considerations two possible pathways became apparent. F i r s t l y , d i rec t addit ion of iodine azide to cyanoolefin 57a - 166 -followed by reduction could give a z i r i d i n e of general type 242. This approach would necessitate the se lect i ve reduction of the azide group in the presence of the n i t r i l e f unc t i ona l i t y . Undoubtedly, employing l i th ium aluminum hydride as the reducing agent the reduction of the n i t r i l e group would be a complicating factor . On the other hand, i t was hoped that sodium b o r o h y d r i d e , ^ ' 1 5 2 sodium bis(2-methoxyethoxy)-aluminum h y d r i d e , 1 6 6 or aluminum/amalgam 1 6 7 would s e l ec t i ve l y reduce the azide f unc t i ona l i t y . A l t e r na t i v e l y , the n i t r i l e group could be converted into an acetal group which would be iner t to l i th ium aluminum hydride. Hence, the addit ion of iodine azide to compound 272 followed by l i th ium aluminum hydride reduction would furn ish a z i r i d i ne of general type 273_ (CHART XXIV, Page 167). - 167 -CHART XXIV. Elaboration of cyanoolefin 57_b to a z i r i d i ne 273 - 168 -Since the f i r s t pathway appeared a t t rac t i ve and po ten t i a l l y e f f i c i e n t the addit ion of iodine azide to cyanoolefin 57a was examined. In contrast to the behaviour of enol acetates 223a and 223b, treatment of cyanoolefin 57a with 3 molar equivalents of iodine azide in a c e t o n i t r i l e fo r f i ve hours at room temperature gave an adduct 270 (see Page 166) in ca_. 80% crude y i e l d . The progress of the reaction was followed by infrared spectroscopy. T.L.C. analysis of the crude pro-duct suggested e s sent i a l l y one compound. Of note was the appearance in the infrared spectrum of a broad band at 2200 cm" 1 due to the azide. The o l e f i n i c bands at 920, 980, and 1630 cm" 1 were not present. The n.m.r. spectrum of the crude product had a broad s ing le t ( t o t a l l i n g three protons) at x9.20 due to the carbon-18 t e r t i a r y methyl groups of compounds 270a and 270b while the carbon-19 t e r t i a r y methyl groups appeared as two sharp s ing lets of about equal i n tens i t y at x8.74 and x8.62. The stereochemical factors in th i s addit ion reaction are worthy I I H H 2 7 Q g 27Qb of comment. Since one new asymmetric centre has been created the p o s s i b i l i t y ex i s t s fo r the formation of two compounds, namely 270a and 270b. Examination of a model of cyanoolefin 57a suggested that both - 169 -modes of addit ion were favourable as indicated by the n.m.r. spectrum of the addit ion product. I f compounds 270a and 270b were formed in ca. a 1:1 r a t i o , the reduction reaction would presumably give r i s e to az i r i d ines 242a and 242b. Since the prime object ive at th i s point 2 4 2 a 2 4 2 b was to tes t the f e a s i b i l i t y of th i s type of approach the above stereo-chemical problem did not cause concern. Before attempting to s e l e c t i v e l y reduce the azide func t i ona l i t y in compound 270 a model compound was studied; namely, iodo azide 275. CN IN, 2 7 4 The s e l e c t i v i t y of various reducing agents were examined; for example, sodium bis(2-methoxyethoxy)-aluminum hydr ide, 1 6 6 aluminum amalgam,167 Pt/ 151 152 and sodium borohydride. ' In summary, the se lect i ve reduction of - 170 -the azide group in the presence of the n i t r i l e f unc t i ona l i t y could not be achieved by employing any of these reagents. Nevertheless, iodo azide 2_70_ was treated with excess sodium borohydride in re f lux ing i s o -propanol f o r three hours. T.L.C. Analysis of the crude product indicated minor quant i t ies of s t a r t i ng material and a major compound. The i n f r a -red spectrum of the crude product indicated that the azide and n i t r i l e groups had been p a r t i a l l y reduced. Several attempts were now made to s e l e c t i ve l y reduce the azide group with sodium bis(2-methoxyethoxy)-aluminum hydride. Employing 3 molar equivalents of sodium bis(2-methoxy-ethoxy)aluminum hydride also caused pa r t i a l reduction of both azide and n i t r i l e groups as suggested by in f rared spectroscopy. The reaction between phosphorus nucleophiles and azides (Staudinger reaction) gives phosphine imines, and the subsequent reaction of these compounds with a lky l halides gives dialkylamino-169 phosphonium s a l t s . A lky l or arylphosphines react with a l ky l halides read i ly to produce phosphonium s a l t s . 1 7 ^ With very few e x c e p t i o n s , 1 7 1 no attent ion has been given to incorporating both the phosphine imine and the hal ide function in to one molecule with the p o s s i b i l i t y of 172 creat ing a c y c l i c compound. Recently, Hassner et aj_., found that 2- iodoalkyl azides react with triphenylphosphine or tr imethyl phosphite P(CH ) f 6 53 R-2 7 6 0«-lj>—(OCH 3) 2 N R 2 7 7 - 171 -to afford in good y i e l d an az i r i d i ne der ivat ive of type 276 and 277, respect ive ly . They showed that 2-iodoalkyl azides react with t r i v a l en t phosphorus nucleophiles almost exc lus ive ly at the azide funct ion. These reactions are thought to involve i n i t i a l nuc leoph i l i c attack on the terminal azide nitrogen to give an intermediate of type 278. This can then undergo loss of nitrogen to give y l i d e 279 which cyc l i zes to 280a with displacement of iodide anion. A l t e rna t i ve l y , c yc l i z a t i on to 281 or 282 may precede loss of nitrogen, although the rate of nitrogen loss from azide phosphine adducts of type 278 i s believed to be faster than the i r rate of formation. Lithium aluminum hydride reduction of 280a leads to the formation of the free az i r i d ine 280b. This i s in agreement with - 172 -the general p r i nc ip le that reduction of phosphonium sa l t s l iberates the most 173 electronegative group from the phosphorus. Bai ley e_t al_., obtained s im i l a r results in the l i th ium aluminum hydride reduction of te t raa l ky l phosphonium s a l t s . As a re su l t , i t was contemplated that iodo azide of type 270 could react with triphenylphosphine or trimethylphosphite to give a z i r i d i ne der ivat ive of general type 283 and subsequent reduction or hydrolysis could lead to az i r i d i ne of general type 242. I t was hoped that sodium borohydride, e © 242 mild hydrolysis condit ions, or dry HCl could l i be ra te the cyanoazir idine 175 242. Before attempting to prepare a z i r i d i ne of type 242 the reaction of iodo azides 275 and 284 with phosphines "wasinvestigated. Iodo azide 275 - 173 -2 8 8 reacted with triphenylphosphine in benzene at room temperature to afford 285 as a white insoluble s o l i d in quant i tat ive y i e l d . The spectro-scopic properties of 285 were in accord with the assigned structure. The infrared spectrum of 285 had a band at 2250 cm" 1 due to the n i t r i l e f unc t i ona l i t y and of note was the disappearance of the azide band at 2120 cm" 1 . The n.m.r. spectrum of 285 had a two-proton mul t ip let at T7.0 which could be at t r ibuted to the t e r t i a r y a z i r i d i n y l CH absorptions. Of s ign i f icance was the absence of the mu l t ip le t at x6.0 due to the - 174 -protons adjacent to the iodide and azide groups. The 15 aromatic protons of 285 appeared as a mul t ip le t at T2.20. I t i s noteworthy that the n.m.r. spectra of a z i r i d i ne protons (C-H) are shielded compared to C-Hs in open chain amines due to the magnetic f i e l d generated in three-membered r i n g s . 1 7 6 Tert ia ry a z i r i d i n y l CH absorptions usual ly occur at T7.6 -8.10 but these chemical s h i f t s vary from T7.20 to T8.50 depending on substituents and t h e i r stereochemistry. Hassner et_ al_., demonstrated that P(Ph 3) as a N substituent ef fects ca. 0.8 - 1.2 ppm downfield 172 s h i f t . In add i t ion , iodo azide 284 was treated with tr iphenylphos-phine to afford a z i r i d i ne der ivat ive 286 as a white s o l i d . The n.m.r. and inf rared spectra of 286_ were c lo se ly s im i l a r to that of 285. Having achieved the formation of a z i r i d i ne der ivat ives 285 and 286 experiments were carr ied out to l i be ra te az i r i d i ne s 287 and 288, respect ive ly , by employing mild hydrolysis c o n d i t i o n s . 1 7 ^ However, the action of methanol, methanol/water, or methanol polyphosphoric acid on compound 285 y ielded recovered s ta r t ing mater ia l . Exposure of 285 to dry HCl in ether or chloroform gave s ta r t ing mater ia l . On the other hand, reduction of compound 285 with sodium borohydride in 95% ethanol at room temperature appeared to af ford a z i r i d i ne 287 as evidenced by t . l . c . and infrared spectroscopy. However, the i so la ted y ie ld s of 287 were low. The infrared spectrum of 287 indicated the absence of aromatic bands and the appearance of a weak band at 3350 cm" 1 could be att r ibuted to the N-H stretch of the a z i r i d i ne group. Attempts to l i be ra te a z i r i d i ne 288 by hydrolysis methods proved unrewarding. However, the act ion of sodium borohydride on a z i r i d i ne der ivat ive 286 gave a z i r i d i ne 288, in low y i e l d . The infrared spectrum of 288 was very s im i l a r to the i n f r a -red spectrum of 287. - 175 -As a resu l t of these studies, the reaction of iodo azide 270 with triphenylphosphine in benzene at room temperature was invest igated. OAc 2 7 Q In contrast to iodo azides 275 and 284 no insoluble phosphonium s a l t was observed to form during the course of th i s react ion. The progress of the reaction was followed by infrared spectroscopy. The inf rared of the crude material indicated the absence of the azide band at 2120 cm" 1 . However, the n i t r i l e band at 2250 cm" 1 also appeared to have decreased considerably. The t . l . c . of the crude o i l showed the presence of several compounds with baseline contaminants and the disappearance of s ta r t i ng mater ia l . In order to r a t i ona l i z e these resu l t s i t was assumed that triphenylphosphine reacted with iodo azide 270 to af ford 283b, because of s t e r i c interact ions between the phosphorus groups and the carbon-19 t e r t i a r y group. Presumably, the opportunity would ex i s t for the y l i d e H 2 8 3 b - 176 -283b to react with the n i t r i l e f unc t i ona l i t y . Carrying out the reaction at higher temperatures (ca_. 45° ) gave an overwhelming mixture of pro-ducts. It i s worthy of comment that azides and n i t r i l e s can react under various conditions to afford t e t r a z o l e s w h i c h could be a complicating factor in the case of iodo azide 270. Because of these serious setbacks i t was proposed to examine an a l te rnat i ve sequence, to which reference has already been made (CHART XXIV, Page 167). The f i r s t step involved the conversion of cyanoolefin 57b into o l e f i n i c acetal 272. Toward th i s end, treatment of cyanoolefin 57b with diisobutylaluminum hydride (2.5 equivalents) in benzene at room temperature for one hour followed by d i l u t e acet ic acid y ielded 1 yo o l e f i n i c aldehyde 271 as a c lear o i l in ca. 75% y i e l d . The t . l . c . 5 7 b 271 of the crude product indicated e s sen t i a l l y one compound. The spectro-scopic properties of 271 were in accord with the assigned structure. The infrared spectrum of 271 had bands at 910 cm" 1 , 990 cm" 1 and 1630 cm" 1 due to the o l e f i n i c side chain. Of note was the appearance of a strong carbonyl band at 1720 cm" 1 and of a weak band at 2750 cm" 1 due to the CH stretching v ibrat ion of the aldehydic group. The sa l i en t - 177 -features in the n.m.r. spectrum of 271 were a three-proton mul t ip le t at T5.4 - 4.0 due to the carbon-1 and carbon-2 o l e f i n i c protons and a one-proton t r i p l e t (J = 2 Hz) at T.47 was att r ibuted to the aldehydic proton. O l e f i n i c aldehyde 271 was found to be very sens i t ive to d i l u t e mineral ac id . For instance, d i l u t e hydrochloric acid effected a Prins H 2 8 5 reaction to g ive, presumably, t r i o l 284 and o l e f i n i c d i o l 285. LeBel 180 et_ aj_., reported that o l e f i n i c aldehyde 286 appeared to be unstable while o l e f i n i c aldehyde 287 seemed to be stable. In add i t ion, i t i s well documented that certa in unsaturated aldehydes l i k e c i t r o n e l l a l • A * T J ! • 4.- 181 , . . T 181b,182 read i ly undergo acid-catalyzed eyenzat ions . Johnson et a l . , - 178 -2 8 6 2 8 7 have found that po l yo l e f i n i c acetals can also undergo ac id - cyc l i za t i on s to form mono and b i c y c l i c products in very good y i e l d . As a r e s u l t , i t was ant ic ipated that o l e f i n i c acetal 272. might undergo an acid-catalyzed 183 c y c l i z a t i on as well under too strong condit ions. Nagata et a l . , have reported the conversion of an aldehyde into the corresponding acetal under very mild conditions at room temperature. For example, the action of a c a t a l y t i c amount of £-toluenesulphonic ac i d , ethylene g l y c o l , and dry methylene ch lor ide on an aldehyde of type 288 at room temperature for twenty four hours gave acetal 289 in high y i e l d . Thus, treatment of o l e f i n i c aldehyde 271_ (CHART XXIV, Page 167) with ethylene g l y c o l -2 8 8 2 8 9 - 179 -methylene chlor ide in the presence of a trace of n-toluenesulphonic a c i d , with vigorous shaking, at room temperature for twenty-four hours y ie lded o l e f i n i c acetal 272 in 70% pu r i f i ed y i e l d . The spectroscopic properties of th i s material were in accord with the assigned st ructure. Of note was the disappearance in the inf rared spectrum of the weak band at 2725 cm" 1 and of the carbonyl band at 1720 cm" 1 . The o l e f i n i c bands were evident at 910 cm" 1 , 990 cm" 1 , and 1630 cm" 1 . In the n.m.r. spectrum of 272 a f ive-proton mu l t ip le t appeared at ca_. x6.3 due to the carbon-17 a -proton and the methylene protons of the acetal group. A one-proton t r i p l e t (J = 5 Hz) was evident at T5.32 due to the methine proton of the acetal f unc t i ona l i t y and the o l e f i n i c protons appeared as a three-proton mul t ip le t at ca_. x5.3 - 4.0. F i n a l l y , the mass spectrum of 272^ had a molecular ion peak at j 334. Having constructed o l e f i n i c acetal 272 the formation of a z i r i d i ne of type 273_ was now invest igated (CHART XXIV, Page 167). The addit ion of iodine azide to o l e f i n i c acetal 272 was found to be a very sluggish react ion. Employing Hassner's standard conditions returned copious amounts of s ta r t i ng mater ia l . On the other hand, employing more d ra s t i c conditions afforded only trace amounts of s ta r t ing material and several new compounds as indicated by t . l . c . The infrared spectrum of the crude product indicated a carbonyl band at 1740 cm" 1 and a broad azide band at 2100 cm" 1 . It appeared that the 17$-hydroxy group had been oxidized to a carbonyl f unc t i ona l i t y . This d i f f i c u l t y could, however, have been obviated i f the 17g-hydroxy group had been masked as an acetate funct ion-1 OA a l i t y . Eaton has reported that the act ion of pyridinium bromide perbromide on ketal 290 afforded dibromide 290a. Furthermore, acetals - 180 -185 are sometimes brominated d i r e c t l y . In view of these considerations the iodinat ion of the acetal group could be a complicating factor in the addit ion of iodine azide to o l e f i n i c acetal 272. Also, the acetal group may be pa r t i c i pa t i ng in t h i s react ion. At th i s stage, the capacity of the carbonyl f unc t i ona l i t y of compound 262 to undergo addit ion reactions was invest igated. 187 Thus, OAc OAc 2 9 1 the act ion of nitromethane on cyanoaldehyde 262 was studied. 188 It was hoped that a hydroxy n i t r o compound of general type 291_ would be obtained. 39 As previously mentioned, Shimizu converted hydroxy azide 58 - 181 -to the b icyc looxazol id ine skeleton 47b by employing sodium borohydride in re f lux ing isopropanol. Furthermore, the reductive c y c l i z a t i on of 189 cyano n i t r o compounds have been reported by several workers. 190 Hester has reported that hydrogenation of compound 292 over 10% palladium-on-carbon in ethyl acetate gave compound 293. Buckely and - 182 -Nickel in methanol afforded lactam 295. Hence, i t appeared l i k e l y H 2 9 4 2 9 5 that the reductive c y c l i z a t i on of hydroxy n i t r o compound 291 to com-pound 47b could be effected by employing sodium borohydride or by c a t a l y t i c hydrogenation. 2 6 2 291 Fieser e_t al_., have effected the condensation of n i t r o -benzaldehyde with nitromethane in the presence of a trace of t r i e t h y l -amine. However, treatment of cyanoaldehyde 262 with nitromethane in the presence of a trace of t r i e t h y l amine at room temperature for twenty four hours returned s ta r t ing mater ia l . On the other hand, performing the reaction in the presence of excess t r i e t h y l amine for f i ve days at - 183 -room temperature, in the dark under an atmosphere of n i t rogen, y ie lded hydroxy n i t r o compound of general type 291 in ca. 70% pu r i f i ed y i e l d , m.p. 186-188°. The spectroscopic properties of 291 were in accord with OAc 2 6 2 291 the assigned structure. The infrared spectrum of 291 had a broad band at 3400 cm" 1 due to the carbon-1 hydroxy group, and intense bands appeared at 1560 cm" 1 and 1380 cm" 1 due to the symmetrical and asymmetri-cal stretching v ibrat ions of the n i t ro f unc t i ona l i t y , respect ive ly . The n.m.r. of 291 had two three-proton s ing lets at x9.21 and T8.95 due to the carbon-18 and carbon-19 t e r t i a r y methyl groups, and a one-proton doublet (J = 3Hz) appeared at x6.36 due to the proton adjacent to the carbon-1 hydroxy group. A three-proton mul t ip le t was evident at x4.60 due to the protons adjacent to the 17p-acetoxy and n i t r o groups. F i n a l l y , the mass spectrum of 291 had a weak molecular ion peak at ~^ 392. The stereochemical aspects of the above reaction are worthy of considerat ion. Since one new asymmetric centre has been created there i s the p o s s i b i l i t y - 184 -for the formation of two compounds; namely, 291a and 291b. It was a n t i -cipated that the thermodynamically more stable compound would be formed since the reaction was performed under equ i l i b ra t i ng condit ions. Examin-ation of molecular models of 291a and 291b suggested that the formation of 291a and 291b was equal ly favourable. However, the t . l . c . and m.p. of the i so lated product tended to ind icate the presence of one compound. In add i t i on , the appearance of a sharp three-proton s ing let at T8.95 due to the carbon-19 t e r t i a r y methyl group tended to suggest the presence 43 of one compound. In l i g h t of the work of Hara and Oka (see CHART V, Page 27 ), and Sh imizu 3 9 (see CHART VI, Page 29 ), i t was f e l t that the reductive c y c l i z a t i on of compound 291a would afford 47b and compound 291b would give an uncyclized product which could subsequently be con-verted into 47b. Treatment of hydroxy n i t r o compound 291 with sodium boro-hydride in re f lux ing isopropanol fo r twenty s i x hours afforded predom-- 185 -inant ly compound 267. T.L.C. analys is of the crude product on s i l i c a gel with ethyl acetate as eluent indicated one major compound, R^  0.75. V.P.C. analysis (column D, 250°, 45 ml/min) indicated one compound (Retention time, 2.8 min). Preparative t . l . c . of the crude product gave 26_7 in ca_. 50% y i e l d as a c r y s t a l l i n e s o l i d , m.p. 136-140°. Of note was the disappearance of the n i t r o , n i t r i l e and acetate bands in the inf rared spectrum of 267. The mass spectrum of 267 had a molecular ion peak at ^ 294 with a prominent peak at ^- 276 due to loss of water. Ev ident ly, sodium borohydride had effected a ret ro -a ldo l reaction to give cyanoaldehyde 262 which subsequently was reduced to afford compound 267 a f te r work up. I t was found that sodium borohydride reduction of cyanoaldehyde 262 in re f lux ing isopropanol y ie lded compound 267 ( t . l . c , v . p . c , i . r . and mass spectral s tudies) . Attempts were made to prevent the ret ro -a ldo l reaction from occurring by performing the reduction in the presence of acet ic ac id . However, t h i s led to a complex mixture of - 186 -products as indicated by t . l . c . ana lys i s . Cata l y t i c hydrogenation of compound 291 over 10% palladium-on-charcoal in methanol for f i v e hours y ie lded cyanoaldehyde 262 and hydroxy n i t r i l e 296. T.L.C. analys is of the crude product on s i l i c a gel with a mixture of chloroform and ethyl acetate (5:1, v/v) as the eluent indicated two major compounds and several minor components. Acid extrac-t ion of the crude product, followed by neutra l i zat ion and subsequent d iethy l ether extract ion was performed. T.L.C. examination of the d i -ethyl ether so lut ion suggested the absence of basic compounds in the crude product. V.P.C. analys is (column D, 250°, 45 ml/min) of the crude product indicated two compounds (Retention times 9 and 18 min). The infrared spectrum of the crude product had a n i t r i l e band at 2250 cm" 1 and of note was the absence of the n i t ro bands at 1560 cm" 1 and 1380 cm" 1 . Sodium borohydride reduction of 262 at room temperature for twenty four hours gave compound 296. The t . l . c . and v.p.c. of 296 was i dent i ca l with the t . l . c . and v.p.c. of one of the products derived from the c a t a l y t i c hydrogenation of hydroxy n i t r o 291. The infrared spectrum of 296 had bands at 3450 cm" 1 , 2250 cm ' 1 and 1720 cm" 1 due to the hydroxy, OAc 2 6 2 NaBH 4 r.t - 187 -n i t r i l e and acetate f u n c t i o n a l i t i e s , respect ive ly . The sa l i ent feature in the n.m.r. spectrum of 296 was a two-proton double doublet (J = 11 Hz) at T6.53, 6.49 due to the protons adjacent to the carbon-1 hydroxy group. F i n a l l y , the mass spectrum of 296 had a molecular ion peak at ^ 333. Performing the c a t a l y t i c hydrogenation of 291 in the presence 193 194 of oxa l i c or acet ic ac id with various cata lysts appeared to e f fec t a ret ro -a ldo l type reaction as indicated by t . l . c , v.p.c. and infrared studies. It i s documented that hydrogenation of hydroxy n i t r o compounds OAc OAc NC 2 9 1 can be complicated by a reversal r e a c t i o n . 1 8 ^ Corey et_ al_. have employed aluminum amalgam to reduce a n i t r o group to an amine. Accord-ing ly , treatment of hydroxy n i t ro compound 291 in d iethyl ether-methanol at 0° with aluminum amalgam did not y i e l d compound 297 but afforded - 188 -cyanoaldehyde 262 as indicated by t . l . c , v . p . c , and infrared studies. OAc H 2 N H 2 9 7 A further extension of th i s work would be to protect the carbon-1 196 hydroxy group of compound 291 as an acetate f unc t i ona l i t y , followed by reduction and subsequent c yc l i z a t i on to afford the b icyc looxazol id ine 47b. OH 4 7 b - 189 -In summary, the o r i g ina l goal to develop an e f f i c i e n t method of synthesizing compound 47b had not been f u l l y r ea l i zed . However, our synthetic invest igat ions led to the preparation of 17e-acetoxy-2,3-39 seco-5g-androst - l -ene-3-n i t r i le (57a) which Shimizu has recent ly converted to the 178-hydroxy isomer of samandarine 47b. OAc 47 b EXPERIMENTAL Melting points, which were determined on a Kof ler hot stage, and bo i l ings are uncorrected. Optical rotat ions were recorded at the sodium D l i n e using a Perkin-Elmer Model 141 Automatic Polarimeter. U l t r a v i o l e t spectra were measured in methanol so lut ion on a model SP800 spectrophotometer. The infrared spectra were recorded on a Perkin-Elmer model 700 and were ca l ib rated using the 1601 cm" 1 band of polystyrene. The 'H n.m.r. spectra were recorded in deuterochloroform solut ion on e i ther a Varian T-60 or Varian HA-100. Line posit ions are given in the Tiers x sca le , with tetramethyls i lane as internal standard; the m u l t i -p l i c i t y , integrated peak areas and proton assignments are indicated in parentheses. The mass spectra were obtained using an Atlas CH-4 mass spectrometer and high resolut ion determinations were performed on an AEI MS-9 mass spectrometer. Microanalyses were performed by Mr. Peter Borda, Un ivers i ty of B r i t i s h Columbia. The vapour phase chromatography (v.p.c.) analyses were performed with a Varian Aerograph 90-P-3 using column A, 8 f t . x h i n . column of 5% Fluoro S i l i cone (QF-1) on 60-80 mesh Diaport "S", column B, 5 f t . x h i n . column of 3% SE 30 on Chromosorb w, and column C, 5 f t . x h i n . column of 10% carbo-wax on 60-80 mesh Chromosorb W, or with a Perkin-Elmer Model 900 using - 191 -column D, 6 f t . x — i n . column of 8% SE 30 on 80-100 mesh Chromosorb W. The spec i f i c column used, along with the column temperature and c a r r i e r gas (helium) f low-rate ( in ml/min) are indicated in parentheses. S i l i c a Gel GF-254 and a Woelm neutral alumina were used for th in layer chromato-graphy ( t . l . c ) . S i l i c a Gel PF-254 and Aluminum Oxide F-254 were used for preparative layer chromatography. S i l i c a Ge l , Woelm alumina, and F l u o r i s i l (100-200 mesh) were used for column chromatography. Preparation of 176-hydroxy-58-androstan-3-one (81a) A solut ion of testosterone (93_, 5.0 g, 0.017 moles) in 95% ethanol (100 ml) containing 3N hydrochloric ac id (8 ml) and 10% palladium-on-charcoal ca ta l y s t (250 mg) was hydrogenated at ambient pressure and temperature by s t i r r i n g in an atmosphere of hydrogen. After the uptake of hydrogen had ceased (400 ml, 30 min) the cata lys t was removed by f i l t r a t i o n and washed with acetone. The combined washings and ethanol ic so lut ion were concentrated to 20 ml under reduced pressure. The suspen-sion was extracted with ether (2 x 75 ml) , the organic layer was washed with d i l u t e aqueous sodium bicarbonate (2 x 20 ml) and saturated sodium chlor ide (2 x 20 ml) so lut ions , dr ied over sodium sulphate, and f i l t e r e d . The solvent was removed under reduced pressure to afford 4.98 g of a c r y s t a l l i n e residue, m.p. 123-128°. Rec ry s ta l l i za t i on of the residue from methanol afforded 1.43 g (21%) of 8_[a as a c r y s t a l l i n e s o l i d , m.p. 138-140°, [a]n°+34° (c=l, EtOH) ( l i t . 5 4 m.p. 139-140°, [ a ] n +32 .7° ) . Infrared (CHC1 3), 3450, 1705'cm" 1; - 192 -n.m.r. (CDC1 3), x9.22 ( s i ng le t , 3H, C-18 CH 3 ) , 8.95 ( s i n g l e t , 3H, C-19 CH 3 ), 6.35 ( t r i p l e t , IH, C-17 H^, J = 9 Hz); mass spectrum | ( r e l a t i ve i n t en s i t y ) , 290(100),275(10),273(10),248(32), 323(33),221(35). Preparation of 176-acetoxy-53-androstan-3-one (81b) The hydrogenation of testosterone (93_, 5.0 g, 0.017 moles) 42 was carr ied out by employing the procedure of Liston to y i e l d 4.98 g of a c r y s t a l l i n e residue, m.p. 123-127°. To a solut ion of th i s material (4.98 g) in dry pyr idine (20 ml) was added acet ic anhydride (5 ml). The solut ion was s t i r r e d at room temperature for 16 hr, then d i l u ted with water (200 ml) , and shaken for 15 min. The p rec ip i ta te was co l l e c ted , washed with IN hydrochloric acid (100 ml) and with water (100 ml) , and a i r dr ied. Rec ry s ta l l i za t i on from ether afforded 2.19 g (38%) of 176-acetoxy-56-androstan-3-one (81b), m.p. 142-144°, [ a ] 2 5 + 4 3 ° (c=l, MeOH) ( l i t . 4 2 m.p. 140-142°, [ a ] 2 5 + 4 5 . 2 ° ) . V.P.C. analys is (column A, 225°, 60 ml/min) of the crude product indicated 75% of 178-acetoxy-53-androstan-3-one (81b) and 25% of 178-acetoxy-58-androstan-3-one (Retention times 7.5 and 8.6 min, re spect i ve ly ) . Infrared (CHC1 3), 1720 cm" 1 ; n.m.r. (CDC1 3), T9.22 ( s i ng l e t , 3H, C-18 CH 3 ), 8.97 ( s i ng le t , 3H, C-19 CH3), 7.97 ( s i n g l e t , 3H, acetate), 5.37 ( t r i p l e t , IH, C-17 H , J = 9 Hz); - 193 -mass spectrum | ( re l a t i ve i n t en s i t y ) , 332(49),272(85) ,257(32),230(16), 214(12),160(20),148(35),42(100). Anal. Calcd fo r : C21H32O3 : C, 75.86; H, 9.70. Found: C, 75.86; H, 9.52. Preparation of a mixture of syn and anti 17e-hydroxy-5g-androstan-3-one oximes (87_) and (88_), respect ive ly A solut ion of 178-hydroxy-5B-androstan-3-one (81b, 801 mg, 2.76 mmoles), hydroxylamine hydrochloride (3.095 g, 0.044 moles) and sodium acetate t r ihydrate (4.0 g, 0.048 moles) in 90% methanol (60 ml) was refluxed for 2.5 hr and then allowed to coo l . The p rec ip i ta te was co l l e c ted , washed with water (75 ml) , d r i ed , and r e c r y s t a l l i z e d from methanol to afford 0.50 g(60%) of syn and ant i oximes 87_ and 88,respect ively, m.p. 210-214° ( l i t . 5 3 a m.p. 211-213°). Infrared (CHC1 3), 1650, 3350 cm" 1 ; n.m.r. (CDC1 3), x9.27 (broad s i ng l e t , 3H, C-18 CH3), 9.03 (broad s i n g l e t , 3H, C-19 C H 3 ) , ca. 7 (unresolved mu l t i p l e t , IH, C-4 ^ 87_ and C-2 88), 6.35 ( t r i p l e t , IH, C-17, H , J = 9 Hz) —a m e 257(70),216(75). mass spectrum ™-(relative i n t e n s i t y ) , 305(13) ,2 98(10) ,2 7 8(22) ,2 7 5(100), - 194 -Attempted separation of the mixture of syn and' ant i oximes 87_ and 88_ A mixture of syn and anti oximes 87_ and 88_ (100 mg, 0.32 mmoles) was chromatographed on a 20 x 20 cm s i l i c a gel coated p la te , adsorbant thickness 0.9 mm, using a mixture of benzene and ethyl acetate (4:1, v/v) as eluent. A f ter e l u t i o n , the band l y ing in the region Rf 0.60 - 0.68 was removed and extracted with ethyl acetate (40 ml). The solvent was removed by evaporation under reduced pressure to give 23.2 mg (23%) of anti oxime 88, m.p. 211-213° ( l i t . 5 0 a ' 5 3 a m . p . 211-212°). Infrared (CHC1 3), 1650, 3350 cm" 1 ; n.m.r. (CDC1 3), x9.28 ( s i n g l e t , 3H, C-18 CH3), 9.03 ( s i ng le t , 3H, C-19 CHj), ca_. 7 (unresolved m u l t i p l e t , IH), 6.35 ( t r i p l e t , IH, C-17 1^, J = 9 Hz). The band ly ing in the region R^  0.52 - 0.59 was removed and extracted with ethyl acetate (40 ml). The solvent was removed by evaporation under reduced pressure to afford 21.2 mg (21%) of syn oxime 87, m.p. 211-213° ( l i t . 5 0 a » 5 3 a m . p . 211-213°); Infrared (CHC1 3), 1650, 3350 cm" 1 ; n.m.r. (CDC1 3), T9.27 ( s i ng l e t , 3H, C-18 CH3), 9.05 ( s i ng l e t , 3H, C-19 CH3), ca_. 7 (unresolved m u l t i p l e t , IH), 6.36 ( t r i p l e t , IH, C-17 H^, J = 9 Hz). - 195 -Attempted separation of the mixture of syn and anti oximes 87_ and 88_ A mixture of syn and anti oximes 87_ and 88_ (100 mg, 0.32 mmoles) was dissolved in ethyl acetate (3 ml) and chromatographed on a column of s i l i c a gel (100 g). Elut ion with a mixture of benzene and ethyl acetate (4:1, v/v) afforded 40.0 mg (40%) of the anti oxime 88, m.p. 211-213° ( l i t . 5 0 a ' 5 3 a m . p . 211-212°), as the f i r s t f ract ion and then 41.0 mg (41%) of the syji oxime 87, m.p. 210-213° ( l i t . 5 0 a ' 5 3 a m.p. 211-213°). F i r s t Fract ion: Infrared (CHC1 3), 1650 cm" 1 , 3350 cm" 1 ; n.m.r. (CDC1 3), x9.28 ( s i ng le t , 3H, C-18 CHj), 9.05 ( s i ng le t , 3H, C-19 CH 3 ) , ca. : J = 9 Hz). _. 7 (unresolved mu l t i p l e t , IH), 6.40 ( t r i p l e t , IH, C-17 H^ Second Fract ion: Infrared (CHC1 3), 1650, 3350 cm" 1 ; n.m.r. (CDC1 3), T9.27 ( s i ng le t , 3H, C-18 CH 3 ), 9.05 ( s i ng le t , 3H, C-19 CH 3 ), ca. 7 (unresolved mu l t i p l e t , IH), 6.40 ( t r i p l e t , IH, C-17 H^, J = 9 Hz). Beckmann Rearrangement of syn 17e-hydroxy-58-androstan-3-one oxime (87) To a solut ion of syn oxime 87 (15.1 mg, 0.049 mmoles) in p y r i -dine (.8 ml, 9.8 mmoles) was added p-toluenesulphonyl chlor ide (27.6 mg, 0.148 mmoles). The solut ion was s t i r r ed at room temperature for 2 days; then ethyl acetate (3 ml) was added, followed by 5% aqueous sodium - 196 -bicarbonate (3 ml). The mixture was s t i r r ed for 5 min, then d i lu ted with 5% sodium bicarbonate solut ion (5 ml) and ethyl acetate (5 ml). The aqueous layer was separated and extracted with ether (2 x 5 ml) and the combined organic layers were dried over sodium sulphate. The solvent was removed under reduced pressure to afford 18.2 mg of a colour less foamy residue. C r y s t a l l i z a t i o n of the residue from methanol afforded 9.8 mg (65%) of 17B-hydroxy-3-aza-A-homo-5B-androstan-4-one (82a), m.p. 241-242° ( l i t . 5 3 6 m.p. 242-244°). T.L.C. analys is of t h i s material ( s i l i c a gel or alumina) in various solvent systems indicated the presence of one compound. Infrared (CHC1 3), 3440, 1660 cm" 1 . Attempted preparation of pure 17B-hydroxy-3-aza-A-homo-5B-androstan-4-one (87a) To a so lut ion of the syn and anti oximes 87_ and 88_ (200 mg, 0.65 mmoles) in pyridine (13.5 ml) was added p-toluenesulphonyl chlor ide (500 mg, 2.63 mmoles). The reaction mixture was s t i r r ed at room tempera ture for ca_. 2 days. The pyridine was removed under reduced pressure (below 70°) and the crude dark red o i l was dissolved in chloroform. The chloroform solut ion was washed with IN hydrochloric acid (5 x 2 ml) , IN sodium hydroxide (2 x 2 ml) so lut ion and saturated sodium chlor ide (5 x 2 ml) so lut ion. The organic layer was dried over sodium sulphate, f i l t e r e d , and the solvent removed by evaporation under reduced pressure. The resu l t ing crude o i l was dr ied under vacuum to y i e l d 210 mg of a mixture of crude lactams 82a and 83a as an o f f - red s o l i d . T.L.C. analys - 197 -on alumina, with benzene and n-pentanol (9:1, v/v) as the solvent system, indicated pa r t i a l separation of the lactams 82a and 83a. The crude material (200 mg) was chromatographed on a 20 x 20 cm alumina coated p late, adsorbant thickness 0.9 mm, using a mixture of benzene and n_-pentanol (9:1, v/v) as eluent. Af ter e l u t i o n , the bands l y ing in the region R f 0.81-0.65 and R f 0.55-0.62 were removed and extracted with chloroform. The solvent was removed by evaporation under reduced pressure to afford 40.0 mg (20%, R f 0.81-0.65) of an o i l and 100.3 mg (50%, R f 0.55-0.62) of a semi-so l id. T.L.C. examination of these products indicated that the f i r s t f r ac t i on (R^ 0.81-0.65) consisted of one major compound with a minor component while the second f rac t ion (R^ 0.55-0.62) consisted of a mixture of compounds. Attempted preparation of methyl 178-hydroxy-2,3-seco-58-androst-l-en-3-oate (104a) from a mixture of syn and ant i 17B-hydroxy-5e-androstan-3-one oximes 87 and 88 Attempted preparation of a mixture of 17e-hydroxy-3-aza-A-homo-5e-androstan-4-one (82a) and 178-hydroxy-4-aza-A-homo-53-androstan-3-one (83a) To a so lut ion of a mixture of syn and anti oximes 87_ and 88_ (800 mg, 2.62 mmoles) in pyridine (16 ml) was added p-toluene-sulphonyl ch lor ide (1.5 g, 0.007 moles). The reaction mixture - 198 -was s t i r r ed at room temperature for 3 days. The pyridine was removed by evaporation under reduced pressure (below 70°) and the crude o i l was dissolved in diethyl, ether (100 ml). The d iethy l ether solut ion was washed with IN hydrochloric acid (3 x 20 ml), IN sodium hydroxide (2 x 20 ml) and saturated sodium chlor ide so lut ion (3 x 20 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 1.10 g of a c r y s t a l l i n e s o l i d . Rec ry s ta l l i za t i on of th i s material y ie lded 789 mg (65%) of a mixture of 17g-tosyloxy-3-aza-A-homo-5e-androstan-4-one (82b) and 17e-tosyloxy-4-aza-A-homo-5e-androstan-3-one (83b), m.p. 200-202°, A M A X 225 m u. Infrared (CHC1 3), 3400, 1660, 1190 cm" 1 ; n.m.r. (CDC1 3), T9.20 ( s i ng le t , 3H, C-18 CH 3), 9.0 ( s i ng le t , 3H, C-19 CH 3 ), 7.53 ( s i ng le t , 3H, tosylate group), 5.70 ( t r i p l e t , IH, C-17 H^, J = 10 Hz), 3.50 (broad s i ng l e t , IH, exchangeable proton), 2.64, 2.17 (double doublet, *4H, tosy late group, J = 8 Hz); mass spectrum ^ ( re l a t i ve i n t e n s i t y ) , 459(26),287(72),272(15),91(100); Anal. Calcd. for : C 2 6H 3 7 N0^S: C, 67.94; H, 8.11; N, 3.04. .Found C, 67.92; H, 8.21. - 199 -Preparation of a mixture of N-nitroso-178-tosyloxy-3-aza-A-homo-53-androstan-4-one (101b) and N-nitroso-17B-tosyloxy-4-aza-A-homo-58-androstan-3-one (105b) Sodium acetate (500 mg) was added to carbon te t rach lo r ide (20 ml) saturated with nitrogen d iox ide. This mixture was cooled to -60° and then a mixture of lactams 82b and 83b (50 mg, 0.108 mmoles) was added. Af ter 5 min the reaction mixture was allowed to warm to room temperature. The solvent was removed under reduced pressure to afford 54.1 mg (101%) of a mixture of com-pounds 101b and 105b as an unstable yel low powder. The t . l . c . of th i s material on s i l i c a gel with ethyl acetate as the eluent showed one broad spot (R^ 0.82). Infrared (CHC1 3), 1720, 1415, 1385, 1360 cm" 1 ; n.m.r. (CDC1 3), T9.2 ( s i ng le t , 3H, C-18 C H 3 ) , 9.0 ( s i ng le t , 3H, C-19 C H 3 ) , 7.54 ( s i ng l e t , 3H, tosy late group), 5.7 ( t r i p l e t , IH, C-17 H ), 2.67, 2.24 (double doublet, 4H, tosy late group). Pyro lys i s of a mixture of N-nitroso-178-tosyloxy-3-aza-A-homo-5B-androstan-4-one (101b) and N-nitroso-178-tosyloxy-4-aza-A-homo-5B-androstan-3-one (105b) - 200 -A mixture of compounds 101b and 105b (50 mg, 0.102 mmoles) was heated at 125° under an atmosphere of nitrogen for 2 min to afford 48 mg (crude, 101%) of a mixture of 173-tosyloxy-3-oxa-A-homo-5e-androstan-4-one (102b) and 17e-tosyloxy-4-oxa-A-homo-5B-androstan-3-one (106b). T.L.C. analysis of th i s product on s i l i c a gel with ethyl acetate as the eluent indicated the presence of two compounds, R f 0.40 and R f 0.35. Infrared (CHC1 3), 1720, 1600, 1190 cm" 1 ; n.m.r. (CDC1 3), T9.2 (broad s i ng l e t , 3H, C-18 CH 3), 9.04 (broad s i n g l e t , 3H, C-19 CH 3 ), 7.27 ( s i ng le t , 3H, tosy late group), 5.80 ( t r i p l e t , IH, C-17 H j , 2.70, 2.24 (double.doublet, 4H, tosy late group); mass spect rum^ ( r e l a t i ve i n t e n s i t y ) , 460(1) ,289(1) ,288(3), 287(1),172(1),107(6),105(3),91(15),44(100); Moi. Wt. Calcd. for C 26H 360 5S: 460.2283. Found: (high resolut ion mass spectrometry): 460.2261. Pyro lys i s of a mixture of 17B-acetoxy-N-nitroso-3-aza-A-homo-56-androstan-4-one (101c) and 178-acetoxy-N-nitroso-4-aza-A-homo-5B-androstan-3-one (105c) A mixture of compounds 101c and 105c (50 mg, 0.13 mmoles) was added to bo i l i ng toluene (10 ml) under an atmosphere of nitrogen. A f te r 5 min the reaction mixture was cooled and the solvent removed under reduced pressure to afford 47 mg (100%) of a mixture of 17p-- 201 -acetoxy-3-oxa-A-homo-5B-androstan-4-one (102c) and 17g-acetoxy-4-oxa-A-homo-5B-androstan-3-one (106c). This crude product was ident ica l ( t . l . c , v . p . c , i . r . , and n.m.r.) with the products derived from the Baeyer-Vi l1iger oxidation of 17e-acetoxy-5g-androstan-3-one (81b). T.L.C. analys is of the crude product on s i l i c a gel with ethyl acetate as the eluent indicated the presence of two compounds, 0.55 and R^  0.60. V.P.C. analys is (column B, 300°, 45 ml/min) of the crude product indicated the presence of two compounds (Retention times, 18 and 31 min) in ca. a 3:2 r a t i o . Infr.ared (CHC1 3), 1720 cm" 1 ; n.m.r. (CDC1 3), T9.20 ( s i n g l e t , 3H, C-18 CH 3 ), 8.97 ( s i ng l e t , 3H, C-19 CH_3), 8.0 ( s i n g l e t , 3H, acetate), 6.20 - 5.20 (mu l t i p le t , 3H, C-17 ^ and C-2 H^, 102c, C-4a H^, 106c); mass spectrum ^ ( r e l a t i ve i n t en s i t y ) , 348(6),288(16),260(8),187(16), 147(16),133(23),94(100). Preparation of a mixture of 17e-acetoxy-3-oxa-A-homo-56-androstan-4-one (102c) and 178-acetoxy-4-oxa-A-homo-5g-androstan-3-one (106c) To a solut ion of 17g-acetoxy-5g-androstan-3-one (81b, 200 mg, 0.60 mmoles) in chloroform (15 ml) was added meta-chloroperbenzoic acid (155 mg, 0.90 mmoles). The reaction mixture was s t i r r e d in the dark at room temperature fo r 2 days. The solut ion was then poured into saturated sodium bicarbonate so lut ion (30 ml) and then extracted with chloroform (50 ml). The chloroform so lut ion was washed with - 202 -water ( 2 x 10 ml) and saturated sodium chlor ide so lut ion (2 x 10 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to af ford 180 mg (85%) of a mixture of compounds 102c and 106c as a c r y s t a l l i n e s o l i d , m.p. 118-125°. T.L.C. analys is of th i s product on s i l i c a gel with ethyl acetate as the eluent indicated the presence of two compounds, 0.58 and R f 0.61. V.P.C. analys is (column B, 300°, 60 ml/min) of the crude product showed two peaks (Retention times 18 and 31 min) in ca_. a 1:4 r a t i o . An ana l y t i ca l specimen was obtained by three rec ry s ta l1 i za t ions from methanol, m.p. 207-208°. Infrared (CHC1 3), 1720 cm" 1 ; n.m.r. (CDC1 3), x9.20 ( s i ng l e t , 3H, C-18 CH3), 8.97 ( s i ng le t , 3H, C-19 CH 3 ), 8.0 ( s i ng le t , 3H, acetate), 6.20 - 5.20 (mu l t ip le t , 3H, C-17 and C-2 Hj> 102c, C-4a Hj> 106c); mass spectrum I ( r e l a t i ve i n t e n s i t y ) , 348(15),288(25),260(9),94(100); Anal. Calcd. for C a i ^ O i , : C, 72.30; H, 9.25. Found: C, 72.02; H, 9.48. Attempted preparation of methyl 17B-hydroxy-2,3-seco-5B-androst-l-en-3-oate (104a) from a mixture of 178-acetoxy-3-oxa-A-homo-58-androstan-4-one (102c) and 17B-acetoxy-4-oxa-A-homo-58-androstan-3-one (106c) - 203 -Preparation of a mixture of methyl 2,17e-dihydroxy-2,3-seco-58-androstan-3-oate (103a) and methyl 4,17g-dihydroxy-3,4-seco-5g-androstan-3-oate (107a) A solut ion of compounds 102c and 106c (100 mg, 0.258 mmoles) in 5% methanolic sodium hydroxide (30 ml) was heated to re f l ux . A f te r 2 hr of r e f l ux i ng , the reaction mixture was cooled and the solvent removed under reduced pressure. The residue was d i s -solved in water (30 ml) and the so lut ion was ca re fu l l y a c i d i f i e d with d i l u t e acet ic a c i d , extracted with chloroform (2 x 30 ml). The combined chloroform extracts were washed with saturated sodium chlor ide (2 x 10 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to y i e l d 100 mg of a c r y s t a l l i n e s o l i d , m.p. 185-189°. To a solut ion of th i s material (100 mg, 0.30 mmoles) in methanol (30 ml) was added a so lut ion of diazomethane in ether (0.35 mmoles). A f ter s t i r r i n g the reaction mixture at room temperature for 5 min the solvent was removed under reduced pressure to y i e l d 66 mg (75%) of a mix-ture of hydroxy methyl esters 103a and 107a as an o i l . Attempts to induce c r y s t a l l i z a t i o n f a i l e d . The t . l . c . of th i s material on s i l i c a gel with ethyl acetate as the solvent system showed one broad spot, R f 0.80. Infrared (CHC1 3), 3600, 3450, 1720 cm" 1 ; n.m.r. (CDC1 3), T9.27 ( s i ng l e t , 3H, C-18 CH 3 ) , 9.02, 8.93 ( s i ng le t s , 3H, C-19 CH3), ca. 6.35 (mu l t i p l e t , 3H, C-2 Hj> and C-17 H j , 6.32 ( s i n g l e t , 3H, methyl ester group); - 204 -mass spectrum | ( re l a t i ve intensity),, 338(8), 320(11) ,306(76), 274(100),263(65),233(84). Mol. Wt. Calcd. for C H 0 : 338.2456 Found 20 31) i | (high resolut ion mass spectrometry): 338.2442 Preparation of a mixture of methyl 17g-hydroxy-2-tosyloxy-2,3-seco-5e-androstan-3-oate (103b) and methyl 17£-hydroxy-2-tosyloxy-2,3-seco-5g-androstan-3-oate (107b) To a so lut ion of compounds 103a and 107a (30 mg, 0.088 mmoles) was added p-toluenesulphonyl ch lor ide (17.1 mg, 0.089 mmoles). The reaction mixture was allowed to stand at 20° for 2 days. The pyridine was removed under reduced pressure and the re su l t ing residue dissolved in d iethy l ether (40 ml). The d iethy l ether solut ion was washed with IN hydrochloric ac id (4 x 10 ml) , IN sodium hydroxide (4 x 10 ml) , and saturated sodium ch lor ide solut ion (2 x 10 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to y i e l d 39.3 mg (88%) of a mixture of the tosyloxy methyl esters 103b and 107b as a c lea r o i l . Attempts to induce c r y s t a l l i z a t i o n f a i l e d . The t . l . c . of th i s material on s i l i c a gel with various solvent systems showed one broad spot. Infrared (CHC1 3), 3600, 3400, 1720, 1600, 1190 cm" 1 ; n.m.r. (CDC1 3), T9.29, 9.27 ( s i ng l e t s , 3H, C-18 CH 3 ), 9.07, 9.03 ( s i ng le t s , 3H, C-19 CH^), 7.55 ( s i n g l e t , 3H, tosy late group), - 205 -6.67 ( s i ng le t , 3H, methyl ester group), ca_. 6.4 (mu l t i p le t , IH, C-17 HJ, 5.90 (mu l t ip le t , 2H, C-2 H2 103b and C-4 H^ 107b), 2.68, 2.20 (double doublet, 4H, tosy late group, J = 9 Hz); mass spectrum ^ ( r e l a t i ve intensity),492(1),474(1),460(1),305(3), 304(4) ,288(3),287(4),234(12),233(30),215(30),187(20),91(100). Attempted preparation of methyl 17e-hydroxy-2,3-seco-5g-androst-l-en-3-oate (104a) A mixture of compounds 103b and 107b (160 mg, 0.32 mmoles) was added to c o l l i d i n e (10 ml). The reaction mixture was ref luxed for 4 hr under an atmosphere of nitrogen. The solvent was removed by evaporation under reduced pressure to afford a brown o i l which was dissolved in d iethy l ether (35 ml). The d iethy l ether solut ion was washed with saturated sodium bicarbonate (2 x 10 ml) and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to y i e l d 753 mg of a brown o i l which was chromatographed on a 20 x 20 cm s i l i c a gel coated p la te , adsorbant thickness 0.9 mm, using a mixture of benzene and ethyl acetate (1:1, v/v) as the eluent. Af ter e l u t i o n , the band l y ing in the region R f 0.55 - 0.60 was removed and extracted with ethyl acetate (50 ml). The solvent was removed under reduced pressure to afford 35.5 mg (extrapolated y i e l d , 32%) of methyl 17e-hydroxy-3,4-seco-5B-androst-4-en-3-oate (108a) as a c lear o i l . Attempts to induce c r y s t a l l i z a t i o n f a i l e d . - 206 -Infrared (CHC1 3), 3600, 3450, 1720, 1630, 900 cm" 1 ; n.m.r. (CDC1 3), T9.27 ( s i ng l e t , 3H, C-18 CH_3), 8.97 ( s i ng le t , 3H, C-19 CH 3 ), ca_. 6.5 (mu l t ip le t , IH, C-17 H j , 6.37 ( s i ng l e t , 3H, ester group), 5.35 (doublet, 2H, C-4 H^, J = 4 Hz); mass spectrum | ( r e l a t i ve in tens i ty ) ,320(5),234(6),233(25),215(12), 205(6),187(9),160(10),62(100); Moi. Wt. Calcd. for C 2 0 H 3 2 0 3 : 320.2351. Found (high resolut ion mass spectrometry): 320.2327. Attempted preparation of 3,178-diacetoxy-5g-androst-2-ene (85b) To a solut ion of tr iphenyl methane (133 mg, 0.56 mmoles) in dimethoxyethane under nitrogen was added potassium (22 mg, 0.55 mmoles). The resu l t ing red solut ion was s t i r r ed for 10 min at room temperature and then 17g-acetoxy-53-androstan-3-one (81b, 166 mg, 0.50 mmoles) was added. The reaction mixture was s t i r -red for 1 hr at room temperature, then acetyl ch lor ide (61.2 mg, 0.60 mmoles) was added, and the so lut ion was d i l u ted with d iethy l ether (30 ml). The d iethyl ether solut ion was washed with saturated sodium bicarbonate (3 x 20 ml) and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to afford 142 mg (75%, crude) of a c lea r o i l . This material was i dent i ca l with 3,178-diacetoxy-58-androst-3-ene (146) by infrared and n.m.r. spectroscopy. - 207 -Infrared (CHC1 3), 1750, 1725, 1685, 1660 cm" 1 ; n.m.r. (CDC1 3), x9.22 ( s i ng l e t , 3H, C-18 CH 3), 9.02 ( s i ng l e t , 3H, C-19 CH3), 7.97 ( s i ng l e t , 3H, acetate) , 7.89 ( s i ng le t , 3H, v iny l acetate), 5.36 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 4.95 (broad s i n g l e t , IH, C-4 H ^ . Attempted preparation of 178-hydroxy-5B-androst-l-en-3-one (155) from — 17g-hydroxy-58-androstan-3-one (81a) Preparation of 2-Hydroxymethylene-58-androstan-17e-ol-3-one (153) To a suspension of sodium hydride (31.2 mg, 1.3 mmoles) in dry benzene (5 ml) , under n i t rogen, was added absolute methanol (40 ml). The mixture was s t i r r ed and heated b r i e f l y to bo i l i n g . A f ter the mixture was cooled to room temperature, 17g-hydroxy-5$-androstan-3-one ( 8 l a , 200 mg, 0.68 mmoles) and ethyl formate (50.3 mg, 0.68 mmoles) were added. The mixture was s t i r r ed at room temperature under nitrogen for 30 hr. A f te r the careful addit ion of water (10 ml) to destroy the excess sodium hydride, the mixture was d i l u ted with water (5 ml) and d iethy l ether (25 ml). The ether-benzene layer was separated and re-extracted with water (10 ml). The combined aqueous layers were washed once with d iethy l ether (10 ml) and then neutral ized with carbon dioxide to pH 7. The mixture was f i l teredand the co l lected s o l i d washed thoroughly - 208 -with water (3 x 20 ml). Rec ry s ta l l i za t i on of th i s material from a c e t o n i t r i l e gave 131 mg (60%) of the hydroxymethylene der ivat ive 53 as a c r y s t a l l i n e s o l i d , m.p. 153-159° (evacuated sealed tube), [a]25+30.2° (c=l, MeOH), \ m x 290 my (e = 8000), ( l i t . 8 0 e m.p. 157-163°, [a]n+26.9°, X m a v 284 my (e = 7900)). Preparation of 2-formyl-17e-hydroxy-58-androst-l-en-3-one (154) To a solut ion of hydroxymethylene der i vat i ve 153 (100 mg, 0.31 mmoles) in dry benzene (22 ml) was added dichlorodicyanoquinone (70.4 mg, 0.31 mmoles). The mixture was refluxed under an atmos-phere of nitrogen for 30 hr. The solvent was removed by evapora-t ion under reduced pressure to afford 170 mg of a yellow o i l which was chromatographed on s i l i c a gel (5 g). E lut ion with ethyl acetate gave 30.6 mg (31%) of 2-formyl-17e-hydroxy-5e-androst-l-en-3-one (154) as a c lear o i l , x " ° H 245 my (e = 5760),, x " 5 H _ N a 0 H 306 my ( e = 10850), ( l i t . 8 6 x " ° H 249 my, e = 5700; x " ° H " N a 0 H 306 my, max max e = 10800): Infrared (CHC13), 3450, 2750, 1720, 1670, 1600 cm" 1 . Preparation of 17B-hydroxy-53-androst-l-en-3-one (155) To a solut ion of compound 1_54_ (15.4 mg, 0.047 mmoles) i n dry benzene (20 ml) was added chlorotris(triphenylphosphine)rhodium - 209 -(46.2 mg, 0.141 mmoles). The mixture was refluxed for 3 hr under nitrogen. The so l i d yellow ch lorocarbony lb i s - ( t r ipheny l -phosphine)-rhodium (m.p. 199-207°) was co l l ec ted . The solvent was removed by evaporation under reduced pressure to afford 34.1 mg of a yellow o i l which was chromatographed in a 5 x 20 cm s i l i c a gel coated p la te , adsorbant thickness 0.9 mm, using a mixture of chloroform and ethyl acetate (4:1, v/v) as eluent. Af ter e l u t i o n , the band l y ing in the region 0.70 - 0.75 was removed and extracted with ethyl acetate (50 ml). The ethyl acetate was removed by evaporation under reduced pressure to af ford 10.2 mg of 173-hydroxy-5e-androst-l-en-3-one (155) contaminated with triphenylphosphine oxide. Infrared (CHC1 3), 3350, 1660, 840 cm" 1 ; n.m.r. (CDC1 3), x9.24 ( s i n g l e t , 3H, C-18 CH3), 8.80 ( s i ng l e t , 3H, C-19 CH3), 6.34 ( t r i p l e t , IH, C-17 ^ , 0 = 9 Hz), 4.14, 3.15 (double doublet, 2H, v iny l protons, J = 10 Hz). Preparation of 17B-acetoxy-43-bromo-58-androstan-3-one (158) To a solut ion of 176-acetoxy-5p-androstan-3-one (81b, 9.197 g, 0.027 moles) in g l a c i a l acet ic acid (50 ml) was added a so lut ion of bromine (1.53 ml, 28.5 mmoles) in acet ic ac id (40 ml) over a period of 20 min with vigorous s t i r r i n g at 10°. Absorption of bromine was rap id , and 30 min l a t e r water (300 ml) was added and the mixture was allowed to stand for 1 hr at 10°. The prec ip i tated product was - 210 -co l l e c ted , washed with water (3 x 100 ml) and dissolved in d iethy l ether (300 ml). The organic layer was washed with saturated sodium chlor ide (2 x 20 ml) so lu t ion , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to give 10.44 g (92%) of a c r y s t a l l i n e s o l i d , m.p. 135-151°. V.P.C. analys is (column D, 250°, 45 ml/min) of th i s product indicated the presence of two com-pounds (Retention times 9.0 and 8.9 min) in ca. a 4:1 r a t i o . Two r e c r y s t a l l i z a t i on s from diethy ether gave compound 158 as a c r y s t a l -l i n e s o l i d , m.p. 174-175°, [a]D+43.0° (c=l, MeOH) ( l i t . 9 0 m.p. 174-175°, [a]D+44.7° ± 2° CHC13). Infrared (CHC13), 1730 cm" 1 ; n.m.r. (CDC13), x9.18 ( s i ng l e t , 3H, C-18 CH3), 8.90 ( s i n g l e t , 3H, C-19 C H 3 ) , 7.97 ( s i ng le t , 3H, acetate) , 5.37 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 5.0 (doublet, IH, C-4 H , J = 12 Hz); —ax mass spectrum ^ ( r e l a t i v e in tens i ty ) ,412(30) ,410(30) ,353(70) ,351 (70), 333(100),332(70),273(100),258(80),245(70),223(100). Preparation of 2e,17e-diacetoxy-58-androstan-3-one (159) A solut ion of crude 173-acetoxy-4g-bromo-5e-androstan-3-one (158, 10.3 g, 0.025 moles) and anhydrous sodium acetate (50.2 g, 0.612 moles) in g l ac i a l acet ic acid (730 ml) was refluxed for 2.5 hr. Af ter cool ing the solut ion was poured into water (500 ml) and extracted with d iethy l ether (2 x 200 ml). The combined - 211 -ethereal extracts were washed with saturated sodium bicarbonate (4 x 50 ml) and saturated sodium chlor ide (2 x 50 ml) so lut ions , dr ied over sodium sulphate, and f i l t e r e d . The solvent was removed under reduced pressure to afford a c r y s t a l l i n e residue. Rec ry s ta l l i za t i on of the residue from methanol gave 6.80 g (70%) of d iacetate 159, m.p. 170-173°. V.P.C. analysis (column D, 250°, 45 ml/min) indicated the presence of one compound (Retention time 16.8 min). The ana l y t i ca l specimen was obtained by two r e c r y s t a l l i z a t i o n s from methanol, m.p. 161-162°. Infrared (CHC1 3), 1720 cm" 1 (see Figure I, Page 250); n.m.r. (CDC1 3), x9.20 ( s i n g l e t , 3H, C-18 CH3), 8.95 ( s i n g l e t , 3H, C-19 CH3), 8.00 ( s i n g l e t , 3H, acetate) , 7.90 ( s i ng le t , 3H, acetate), ca. 7.8 (double doublet, IH, C-4 H^, J = 5 and 13 Hz), 7.19 (broad t r i p l e t , IH, C-4 H^, J = 9 Hz), 5.43 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 4.85 (double doublet, IH, C-2 H^, J = 6 and 14 Hz) (see Figure X,' Page 253 ); mass spectrum ^ ( r e l a t i ve i n t e n s i t y ) , 390(4) ,348(15),346(9),331(7), 330(30) ,304(10) ,288(14),270(12),55(100); (see Figure XIX . Page 256 ); Anal. Calcd. for C 2 3 H 3 4 0 5 : C, 70.74; H, 8.77. Found: C, 70.70; H, 8.82. - 212 -Z inc -acet ic acid reduction of 28,173-diacetoxy-5B-androstan-3-one (159) To a solut ion of diacetate 159 (100 mg, 0.25 mmoles) in g l a c i a l acet ic ac id (30 ml) was added dry zinc dust (500 mg). The reaction mixture was refluxed for 30 hr, then cooled and poured into water (50 ml). The aqueous solut ion was extracted with d iethy l ether (3 x 30 ml). The combined ethereal extracts were washed with water (3 x 10 ml ) , saturated sodium bicarbonate (3 x 10 ml) and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium s u l -phate and f i l t e r e d . The solvent was removed under reduced pressure to af ford 72.5 mg of a yellow o i l which was chromatographed on a 5 x 20 cm s i l i c a gel coated p l a te , adsorbant thickness 0.9 mm, using a mixture of chloroform and ethyl acetate (5:1, v/v) as the eluent. A f ter e l u t i o n , the band l y i ng in the region R^  0.52 - 0.59 was removed and extracted with ethyl acetate (50 ml). The solvent was removed by evaporation under reduced pressure to y i e l d 25.7 mg (extrapolated y i e l d , 31%) of a c lea r o i l . This material was ident ica l ( t . l . c , i n f r a red , n.m.r., and mass spectrum) with 178-acetoxy-58-androstan-3-one (81b). Preparation of 3,178-diacetoxy-53-androst-3-ene (146) To a solut ion of 178-acetoxy-58-androstan-3-one (81b, 250 mg, 0.75 mmoles) in isopropenyl acetate (5 ml) was added hydroquinone (30 mg) and concentrated sulphuric acid (10 y l ) under an atmosphere - 213 -of nitrogen. The reaction mixture was heated to a gentle re f lux and acetone formed during the reaction was co l l ec ted . Isopropenyl acetate was pe r i od i ca l l y added to maintain a solvent volume of ca. 5 ml. After re f lux ing under nitrogen for 2 hr, the reaction mixture was cooled to 0° and d i lu ted with d iethy l ether (30 ml). This organic mixture was washed with saturated sodium bicarbonate (3 x 25 ml) and sodium chlor ide (15 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was d i s t i l l e d to af ford 361 mg of a l i g h t yellow o i l . V.P.C. analys is (column A, 225°, 60 ml/min) of the crude product indicated the presence of 59% 3,173-diacetoxy-5e-androst-3-ene (146) and 23% 3,173-diacetoxy-5B-androst-2-ene (85b). (Retention times 12 and 14 min, re spect i ve ly ) . Infrared (CHC1 3), 1750, 1725, 1685 cm" 1 ; n.m.r. (CDC1 3), x9.22 ( s i n g l e t , 3H, C-18 CH3), 9.01 ( s i ng l e t , 3H, C-19 CH 3 ) , 7.98, 7.90 ( s i ng le t s , 6H, v iny l acetates) , 5.36 ( t r i p l e t , 2H, C-17 H^, J = 9 Hz), 4.95 ( s i n g l e t , IH, C-4 H ), 4.80 (mu l t i p le t , IH, C-2 H ). Equ i l i b ra t ion of enol acetates 146 and 85b__ The crude enol acetates were not pu r i f i ed but were equ i l ib rated immediately. The enol acetate mixture (150 mg) was dissolved in benzene (12 ml), carbon te t rach lo r ide (6 ml) and acet ic anhydride (2.4 ml). To th i s so lut ion was added 70% perch lor ic acid (30 y l ) - 214 -under an atmosphere of nitrogen. The reaction mixture was s t i r r ed under nitrogen at room temperature for 22 hr. Then the reaction mixture was poured into d iethyl ether (50 ml) and washed with water (25 ml) , saturated sodium bicarbonate (4 x 20 ml) and sodium chlor ide (20 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to y i e l d 121 mg of a l i g h t yel low o i l . V.P.C. analys is (column A, 225°, 60 ml/min) of th i s product indicated 88% of 2,178-diacetoxy-5B-androst-3-ene (146, Retention time 12 min) and 2% of 2,17B-diacetoxy-58-androst-2-ene (85b, Retention time 14 min). Infrared (CHC1 3), 1750, 1725, 1685, 1660 cm" 1 ; n.m.r. (CDC1 3), x9.22 ( s i ng l e t , 3H, C-18 % ) , 9.02 ( s i ng le t , 3H, C-19 CHj), 7.97 ( s i ng l e t , 3H, acetate) , 7 .89 ' ( s i ng le t , 3H, v iny l acetate), 5.36 ( t r i p l e t , IH, C-17 H^, J= 9 Hz), 4.95 ( s i ng le t , IH, C - 4 H ) . Preparation of 3a,178-diacetoxy-38,48-oxido-58-androstane (177)19? The equ i l ib rated enol acetate mixture containing predominantly compound 146 (53 mg, 0.14 mmoles) was dissolved in chloroform (5 ml). To th i s so lut ion was added m-chloroperoxybenzoic ac id (70 mg, 0.40 mmoles) and sodium bicarbonate (50 mg). The reaction mixture was s t i r r ed at 0° for 4 hr and then the reaction was allowed to stand at -5° for 40 hr. The reaction mixture was - 215 -poured into cold saturated sodium bicarbonate solut ion (10 ml) and extracted with d iethy l ether (3 x 10 ml). The organic layer was separated and washed with 20% sodium carbonate (3 x 15 ml) and saturated sodium chlor ide (10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . Removal of the solvent under reduced pres-sure afforded 46 mg (84%) of 8-epoxide 177 as a l i g h t yel low o i l , with no trace of the s ta r t i ng enol acetate 146 by v.p.c. Infrared (CHC1 3), 1720-1745, 860 cm" 1 ; n.m.r. (CDC1 3), T9.20 ( s i n g l e t , 3H, C-18 CH 3 ), 9.13 ( s i ng le t , 3H, C-19 CH 3 ), 7.97 ( s i ng l e t , 3H, acetate), 7.93 ( s i n g l e t , 3H, acetate), 6.93 ( s i ng l e t , IH, C-4 H ), 5.36 ( t r i p l e t , IH, C-17 H ). Preparation of 4a,17B-diacetoxy-5B-androstan-3-one ( 175a ) 1 9 7 3a,17B-diacetoxy-3B,4B-oxido-58-androstan-3-one (177, 25 mg, 0.064 mmoles) was pyrolyzed at 160° for 5 min under an atmosphere of nitrogen to afford 20.0 mg (80%) of compound 175a. V.P.C. analys is (column A, 225°, 60 ml/min) of th i s product indicated one compound (Retention time 24 min). Infrared (CHC1 3), 1740, 1725 cm" 1 ; n.m.r. (CDC1 3), x9.21 ( s i n g l e t , 3H, C-18 CHj), 8.90 ( s i ng l e t , 3H, C-19 CH 3 ), 7.97 ( s i ng le t , 3H, acetate), 7.85 ( s i ng l e t , 3H, acetate), 5.38 ( t r i p l e t , IH, C-17 H , J = 9 Hz), 4.59 (doublet, IH, C-4 H,,, —ot — p J = 8 Hz). - 216 -197 Preparation of 4g,17g-diacetoxy-53-androstan-3-one (175b) Hydrogen chlor ide was bubbled through a s t i r r ed so lut ion of 3a,178-diacetoxy-38,48-oxide-5g-androstane (177, 25 mg, 0.064 mmoles) in d iethyl ether at 12° fo r 5 min. The so lut ion was s t i r r ed for 30 min, then allowed to stand at -5° for 20 hr. The reaction mix-ture was d i lu ted with d iethy l ether (30 ml) , washed with saturated sodium bicarbonate (3 x 20 ml) and sodium ch lor ide (20 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to afford 20.0 mg (80%) of crude compound 175b. V.P.C. analys i s (column A, 225°, 60 ml/min) of the crude product indicated the presence of one compound (Retention time 24 min). Infrared (CHC1 3), 1740, 1725 c m " 1 ; ' n.m.r. (CDC1 3), x9.21 ( s i ng le t , 3H, C-18 CH 3 ), 8.93 ( s i n g l e t , 3H, C-19 CHj), 7.98 ( s i n g l e t , 3H, acetate) , 7.85 ( s i n g l e t , 3H, C-4 acetate) , 5.39 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 4.48 (doublet, IH, C-4 H , J = 12 Hz). Preparation of 48,173-diacetoxy-5g-androstan-3-one (J_75b_)l97 To a solut ion of 4a,17g-diacetoxy-5g-androstan-3-one (175a, 19 mg, 0.048 mmoles) in hexamethylphosphoramide (1.5 ml) was added a c rys ta l of p-toluenesulphonic acid and the reaction mixture was heated under nitrogen at 160° for 15 min. The react ion mixture was - 217 -then cooled to room temperature and d i luted with d iethy l ether (25 ml). This organic layer was washed with water (3 x 20 ml) , saturated sodium bicarbonate (2 x 15 ml) and sodium chlor ide (15 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 12 mg of compound 175b with no trace of compound 175a detectable by n.m.r. spectroscopy. 1^7 Formation of 4g,17g-diacetoxy-5g-androstan-3-one (175b) To a so lut ion of 4a,173-diacetoxy-5g-androstan-3-one (175a, 30 mg, 0.076 mmoles) in g l a c i a l acet ic acid (4.5 ml) was added sodium acetate (300 mg, 3.65 mmoles). The reaction mixture was refluxed under an atmosphere of nitrogen for 2 hr. The reaction mixture was then cooled to 0 ° , d i l u ted with cold water (25 ml) and extracted with d iethy l ether (3 x 25 ml). The d iethy l ether was washed with saturated sodium bicarbonate (3 x 20 ml) and sodium chlor ide (20 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 19 mg (63%) of 4e,17B-diacetoxy-5B-androstan-3-one (175b). n.m.r. (CDC1 3), T9.22 ( s i n g l e t , 3H, C-18 CH 3 ), 9.02 ( s i ng le t , 3H, C-19 CH 3 ), 9.79 ( s i ng le t , 3H, C-17 acetate) , 7.89 ( s i ng le t , 3H, C-19 H 3 ) , 5.36 ( t r i p l e t , IH, C-17 ^ , 0 = 9 Hz), 4.95 ( s i n g l e t , IH, C-4 H ). - 218 -Attempted preparation of 17g-acetoxy-2e-hydroxy-5e-androstan-3-one (183) To a so lut ion of 28,17B-diacetoxy-5g-androstan-3-one (159, 100 mg, 0.255 mmoles) in methanol (4 ml) and water (1 ml) was added sodium bicarbonate (100 mg). The reaction mixture was heated at 50° for 30 min and then s t i r r e d for ca_. 1 hr at room temperature. The solvent was removed under reduced pressure and the resu l t ing residue was d i s -solved in water (10 ml) and extracted with d iethy l ether (2 x 10 ml). The combined ethereal extracts were washed with saturated sodium chlor ide ( 2 x 5 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 75 mg (84%, crude) of a mixture of 173-acetoxy-2e-hydroxy-53-androstan-3-one (183) and 173-acetoxy-3a-hydroxy-5g-androstan-2-one (185). T.L.C. analys is of the crude product on s i l i c a gel with a mixture of chloroform and ethyl acetate (5:1, v/v) as the eluent indicated the presence of two compounds, R f 0.75 and R f 0.78. Infrared (CHC1 3), 3350, 1720 cm" 1 ; n.m.r. (CDC1 3), T9.25, 9.20 ( s i ng l e t s , 3H, C-18 CH 3 ) , 8.97, 8.90 ( s i ng le t s , 3H, C-19 CH,), 6.60-5.80 (mu l t i p le t , IH, C-2 H , 183 and C-3 H_, 185), 5.40 ( t r i p l e t , IH, C-17 H^, J = 9 Hz). Preparation of ant i 173-acetoxy-28-hydroxy-53-androstan-3-one oxime (187) A solut ion of 2B,17B-diacetoxy-5g-androstan-3-one (159, 5.10 g, 0.013 moles) hydroxylamine hydrochloride (21.0 g, 0.302 moles), and sodium - 219 -acetate t r ihydrate (27.5 g, 0.202 moles) in 90% methanol (400 ml) was refluxed for 48 hr and then the methanolic solut ion was concentrated to 20 ml under reduced pressure. Water (100 ml) was added to the so lut ion and the re su l t ing suspension was extracted with d iethy l ether (2 x 200 ml). The combined ethereal extracts were washed with aqueous sodium bicarbonate (2 x 30 ml ) , and saturated sodium chlor ide (2 x 20 ml) so lut ions , dr ied over Na2S0it and f i l t e r e d . The soivent was removed under reduced pressure to af ford 4.01 g (85%) of hydroxyoxime 187_, m.p. 176-182°. T.L.C. analys is of t h i s material on s i l i c a gel using a mixture of ethyl acetate and benzene (5:1, v/v) as eluent indicated one major compound, R f 0.78. The ana ly t i ca l specimen was obtained by three r e c r y s t a l l i z a t i o n s from methanol, m.p. 214-215°, [a ]p 5 +21.42°(c=.7, MeOH). Infrared (CHC1 3), 3550, 3350, 1660 cm" 1 (see Figure II , Page250 ) ; n.m.r. (CDC1 3), T9.25 ( s i ng l e t , 3H, C-18 CHj), 9.03 ( s i n g l e t , 3H, C-19 CHj), 7.21 (double doublet, IH, C-4 ^ , 0 = 4 and 14 Hz), 5.80 (doublet doublet, IH, C-2 H^, J = 5 and 13 Hz), 5.40 ( t r i p l e t , IH, C-17 H j , (see Figure XI , Page 253); mass spectrum | ( r e l a t i v e intensity),363(1) ,362(6),347(3),345(4),344(5), 334(17) ,333(78),316(6),303(3),271(3),43(100); (see Figure XX , Page 256); Anal. Calcd. for CziHaaNOi, : C, 69.39; H, 9.15; N, 3.85. Found: C, 69.62; H, 9.37; N, 4.02. - 220 -Preparation of 178-acetoxy-2-oxo-2,3-seco-5e-androstane-3-nitrile (195) To a so lut ion of anti 178-acetoxy-2e-hydroxy-5B-androstan-3-one oxime (187, 1.0 g, 2.75 mmoles) in pyr idine (45 ml) was added p-toluenesulphonyl ch lor ide (500 mg, 2.62 mmoles). The reaction mix-ture was ref luxed under an atmosphere of nitrogen for 5 hr. The so lut ion was cooled, poured into d i l u t e sodium bicarbonate so lut ion (50 ml) and extracted with d iethy l ether. The organic layer was sepa-rated and washed with water (4 x 20 ml) , saturated sodium bicarbonate (2 x 15 ml) , and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to y i e l d 985 mg of a brown o i l . This material was chromatographed on a 20 x 20 cm s i l i c a gel coated p l a te , adsorbant thickness 0.9 mm, using a mixture of chloroform and ethyl acetate (5:1, v/v). A f ter e l u t i o n , the band l y ing in the region Rp 0.62 - 0.68 was removed and extracted with ethyl acetate (100 ml). The solvent was removed by evaporation under reduced pressure to afford 190 mg (20%) of cyanoaldehyde 195 as a c lear o i l . Infrared (CHC1 3), 2750, 2250, 1720 cm" 1 (see Figure H I , Page250 ). n.m.r. (CDC1 3), T9.21 ( s i ng le t , 3H, C-18 C H 3 ) , 8.68 ( s i ng l e t , 3H, C-19 CH 3 ), 5.40 ( t r i p l e t , IH, C-17 ttj, 0.20 (double doublet, IH, CH0, J = 1 and 3 Hz) (see Figure XII, Page 253 ). mass spectrum ^ ( r e l a t i v e intensity),345(2),318(5),302(6),301(11), 290(4),258(7),242(8),241(16),43(100) (see Figure XXI , Page 256 ); Anal. Calcd. for C 2 i H 3 1 N 0 3 : C, 73.01; H, 9.04; N, 4.06. Found: C, 72.82; H, 9.16; N, 3.88. - 221 -Preparation of 17B-acetoxy-2-oxo-2,3-seco-5B-androstane-3-nitrile (195) Anti 173-acetoxy-2B-hydroxy-5g-androstan-3-one oxime (187, 1.0 g, 2.75 mmoles) was treated with d i s t i l l e d thionyl ch lor ide (10 ml) at -20° (methanol-ice). Af ter 1.5 min the re su l t ing colourless solut ion was at once slowly poured into a mixture of 3N potassium hydroxide (300 ml) and d iethy l ether (100 ml) at 0° . The organic phase was separated and the aqueous solut ion was extracted with d iethy l ether (2 x 50 ml). The combined ethereal extracts were washed with saturated sodium chlor ide (2 x 50 ml) s o lu t i on , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to af ford 1.10 g of an o i l y residue which was chromatographed on f l u o r i s i l (50 g). E lut ion with a mixture of chloroform and benzene (1:2, v/v) afforded 805 mg (85%) of cyanoaldehyde 195 as a c lear o i l which c r y s t a l l i z e d 90 on standing, m.p. 110-112°, [ a ] ^ +23.3° (c=l, MeOH). Cyanoaldehyde 195 was converted into i t s 2,4-DNP de r i va t i ve , m.p. 238-240°, for ana lys i s . Infrared (CHC1 3), 2750, 2250, 1720 cm" 1 (see Figure III , Page 250 ). n.m.r. (CDC1 3), x9.21 ( s i n g l e t , 3H, C-18 CH^), 8.68 ( s i n g l e t , 3H, C-19 CH 3), 5.40 ( t r i p l e t , IH, C-17 H^), 0.20 (double doublet, IH, CH0, J = 1 and 3 Hz) (see Figure XII , Page 253 ); mass spectrum j ( r e l a t i ve intensity),345(2) ,318(5) ,302(6),301(11), 290(4),258(7),242(8),241(16),43(100) (see Figure XXI , Page 256 ); Anal. Calcd. for C 2 7 H 3 5 N 5 0 6 : C, 61.70; H, 6.71; N, 13.32. Found: C, 61.90; H, 6.84; N, 13.13. - 222 -Attempted preparation of 178-acetoxy-2,3-seco-58-androst-l-ene-3 - n i t r i l e (57a) from 17B-acetoxy-2-oxo-2,3-seco-5B-androstane-3-n i t r i l e (195) Preparation of 17B-acetoxy-2-hydroxy-2,3-seco-58-androstane-3-n i t r i l e (202) To a solut ion of 17g-acetoxy-2-oxo-2,3-seco-5g-androstane-3-n i t r i l e (195, 80 mg, 0.23 mmoles) in ethanol (10 ml) was added sodium borohydride (4.4 mg, 0.46 mmoles). The reaction mixture was s t i r r ed for 3 hr at room temperature under an atmosphere of nitrogen. The solvent was removed by evaporation under reduced pressure, water (20 ml) and d iethy l ether (40 ml) were added to the residue. The organic layer was separated and the aqueous layer was extracted with d iethy l ether (2 x 20 ml). The combined ethereal extracts were washed with IN hydrochloric acid (2 x 20 ml) , saturated sodium bicarbonate (3 x 10 ml) and sodium ch lor ide (10 ml) so lut ions , dr ied over sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to y i e l d 80.2 mg of a c r y s t a l l i n e s o l i d . Rec ry s ta l l i za t i on of th i s material from methanol y ie lded 68.1 mg (85%) of compound 202, m.p. 139-144°. An ana l y t i ca l specimen was obtained by two recry s ta l1 i za t ions from methanol, m.p. 138-140°, [CC]Q0+34° (C=5, MeOH). Infrared (CHC1 3), 3450, 2250, 1720 c m ' i ; n.m.r. (CDC1 3), x9.25 ( s i ng l e t , 3H, C-18 C H 3 ) , 8.97 ( s i ng le t , 3H, C-19 CHj), 8.0 ( s i ng l e t , 3H, acetate),5.47 ( t r i p l e t , IH, C-17 H , - 223 -J = 9 Hz), 6.30 ( t r i p l e t , IH, C-2 H^, J = 10 Hz); mass spectrum ^ ( r e l a t i ve i n t e n s i t y ) , 348(6),347(23),329(6), 332(5),303(10),302(39),288(9),287(24),270(31),260(14), 43(100); Anal. Calcd. for C 2 1 H 3 3 N03 : C, 72.58; H, 9.57; N, 4.03. Found: C, 72.43; H, 9.56; N, 3.88. Preparation of 173-acetoxy-2-tosyloxy-2,3-seco-53-androstane-3-n i t r i l e (203a) To a solut ion of 173-acetoxy-2-hydroxy-2,3-seco-58-androstane-3 - n i t r i l e (202, 200 mg, 0.57 mmoles) in pyr idine (15 ml) was added p-toluenesulphonyl chlor ide (180.0 mg, 0.94 mmoles). The reaction mixture was s t i r r ed at room temperature f o r 2 4 h r and then cold saturated sodium bicarbonate (20 ml) solut ion and diethyl ether (40 ml) were added .The organic layer was separated and washed with .IN hydrochloric acid (3 x 10 ml) , saturated sodium bicarbonate (2 x 15 ml) and sodium chlor ide (10 ml) so lut ions , dried over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to afford 310.3 mg of a c r y s t a l l i n e s o l i d . This material was r e c r y s t a l l i z ed from diethyl ether to give 232 mg (81%) of compound 203a, m.p. 128-129°. An ana ly t i ca l specimen.was obtained 20 by three r e c r y s t a l l i z a t i on s from diethyl ether, m.p. 129-130 [CX]Q +37° (c=l, MeOH). Infrared (CHC1 3), 2250, 1720, 1600, 1190 cm" 1 ; - 224 -n.m.r. (CDC1 3), T9.24 ( s i ng le t , 3H, C-18 CH_3), 8.99 ( s i ng l e t , 3H, C-19 CH 3 ), 8.0 ( s i ng le t , 3H, acetate), 7.52 ( s i ng le t , 3H, tosy late group), 5.92 ( t r i p l e t , 2H, C-2 H^, J = 9 Hz), 5.40 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 2.60, 2.15 (double doublet, 4H, tosy late group, J = 10 Hz); mass spectrum ^ ( r e l a t i v e i n t en s i t y ) , 501 (10),440(16) ,346(15), 301(15),269(35),268(45) ,241(29),43(100); Anal. Calcd. for C 2 8 H 3 9 N0 5 S : C, 67.03; H, 7.83; N, 2.79; S, 6.39. Found: C, 67.02; H, 7.96; N, 2.89; S, 6.18. Attempted preparation of 178-acetoxy-2,3-seco-5e-androst-l-ene-3 - n i t r i l e (57a) To a solut ion of 17g-acetoxy-2-tosyloxy-2,3-seco-5g-androstane-3 - n i t r i l e (203a, 70 mg, 0.139 mmoles) in hexamethylphosphoramide (20 ml) was added dry potassium t>butoxide (15.0 mg, 0.13 mmoles) under an atmosphere of nitrogen. The reaction mixture was heated at 160° for 2 hr under an atmosphere of nitrogen and then cooled. The so lut ion was poured into cold water (100 ml) and extracted with d iethy l ether (2 x 50 ml). The d iethy l ether so lu -t ion was washed with water (6 x 20 ml) and saturated sodium chlor ide solut ion (2 x 10 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to y i e l d 78 mg of a brown o i l . T.L.C. analys is of th i s material on s i l i c a gel with chloroform and ethyl acetate (5:1, v/v) as the eluent - 225 -indicated the presence of two major compounds, 0.82 and R f 0.68, and polar baseline contaminants. Infrared (CHC1 3), 2250, 1720, 1630, 1000, 930 cm" 1 ; n.m.r. (CDC1 3), T9.20-(broad s i n g l e t , 3H, C-19 CH3), 8.83, 8.75 ( s i ng le t s , 3H, C-19 CH_3), 5.10-4.20 (mu l t ip le t , 2%H, v i n y l i c protons and C-17 H ). Attempted preparation of c i s and trans 2,17g-diacetoxy-2,3-seco-5g-and ro s t - l - ene -3 -n i t r i l e (223a) and (223b) 14? Preparation of Reagent B (10 2MHC10iJ. To absolute ethyl acetate (40 ml) was added 72% perch lor ic acid (0.05 ml, 0.58 mmoles) and acet i c anhydride (4.8 ml, 5.1 mmoles), and the solut ion was made up to 50 ml with ethyl acetate. 17e-Acetoxy-2-oxo-2,3-seco-5g-androstane-3-nitrile (195, 60 mg, 0.17 mmoles) was dissolved in 6 ml of reagent B and the reaction mixture allowed to stand for 5 min at room temperature. The so lut ion was then washed with saturated sodium bicarbonate (2 x 10 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 65 mg (85%) of a c lear o i l . T.L.C. analys is of th i s material on s i l i c a gel with chloroform and ethyl acetate (5:1, v/v) as the eluent indicated the presence of one major compound, R. 0.75. Infrared (CHC1 3), 2250, 1750, 1720 cm" 1 ; - 226 -n.m.r. (CDC1 3), x9.25 ( s i n g l e t , 3H, C-18 CH 3 ), 8.86 ( s i ng l e t , 3H, C-19 CHj), ca_. x8 (three s i ng l e t s , 9H, acetates) , 5.42 ( t r i p l e t , IH, C-17 H^, J = 9 Hz), 3.14 ( t r i p l e t , IH, C-2 H, J =7 Hz). Preparation of c i s and trans 2,173-diacetoxy-2,3-seco-5g-androst-1-ene - 3 - n i t r i l e (223a) and (223b) To a so lut ion of 17g-acetoxy-2-oxo-2,3-seco-5g-androstane-3-n i t r i l e (195, 612 mg, 1.774 mmoles) in isopropenyl acetate (15 ml) was added 2,5-di-tert-butyl-p-benzoquinone (60 mg) and 15 y l of concentrated sulphuric a c i d , and the mixture was heated to r e f l ux . During addit ion and re f l ux a stream of nitrogen was passed through the so lut ion. A f ter 36 hr of re f lux ing the mixture was cooled to room temperature, poured into d i l u t e aqueous sodium bicarbonate (15 ml) and extracted with d iethy l ether (2 x 100 ml). The combined ethereal extracts were washed with saturated sodium ch lor ide (2 x 20 ml) so lu t ion , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 930 mg of a brown o i l which was chromatographed on s i l i c a gel (45 g). E lut ion with chloroform gave 472 mg (68%) of a mixture of c i s and trans enol acetates 223a and 223b as a colourless o i l . T.L.C. analys is of t h i s material on a 20 x 5 cm s i l i c a coates p l a te , adsorbant thickness 0.1 mm, using chloroform as eluent indicated the presence of two compounds, R^  0.81 and R^  0.72, in a ca. a 1:1 r a t i o . Infrared (CHC1 3), 1745, 1725, 1660 cm" 1 (see Figure IV , Page 251 ); - 227 -n.m.r. (CDC1 3), T9.20 (broad s i n g l e t , 3H, C-18 CH 3 ), 8.80, 8.70 ( s i ng le t s , 3H, C-19 CH 3 ), 8.02 ( s i ng l e t , 3H, acetate), 7.93, 7.86 ( s ing le t s , 3H, v iny l acetate), 5.42, 4.74 (two doublets, IH, CiH, J = 8 and 13 Hz, re spec t i ve l y ) , 3.04, 2.96 (two doublets, IH, C 2H, J = 8 and 13 Hz, respect ively) (see Figure XI I I, Page 254 ); mass spectrum | ( r e l a t i ve i n t en s i t y ) , 387(25),360(9),359(35),345(52), 344(45),327(34),318(22),311(46),303(36),302(54),301(23),285(45),49(100) (see Figure XXII, Page 257 ); Mol. Wt. Calcd. for C 2 3 H 3 3 N0 1 + : 387.2409. Found (high resolut ion mass spectrometry): 387.2389. Preparation of cyclohexenyl acetate (252)198 To cyclohexanone (9.3 g, 0.094 moles) was added acet ic anhydride (20.4 g, 0.20 mole) and p-toluenesulphonic acid (0.1 g). The reaction mixture was refluxed for 4 hr. During the heating per iod, acet ic acid along with some acet ic anhydride was allowed to d i s t i l l o f f , care being taken to keep the d i s t i l l a t i o n temperature below 125° in order to avoid excessive losses of acet ic anhydride. The crude black o i l was d i s t i l l e d through a 10 cm Vigreux column to y i e l d 4.2 g (32%) of enol acetate 252 b.p. 76-78720 m m ( l i t . 1 9 ? b.D. 74-76°/17 mm). The V.P.C. analys is (column C, 100°, 60 ml/min) of th i s mater ia l , indicated the presence of one compound (Retention time 4.8 min).• - 228 -Preparation of l -azido-2- iodocyclohexyl acetate (253) Sodium azide (300 mg, 4.6 mmoles) in a c e t o n i t r i l e (4 ml) was cooled i n ice-methanol and treated with a solut ion of iodine monochloride (382 mg, 2.3 mmoles) in a c e t o n i t r i l e (2 ml). The mixture was s t i r r ed for 5 - 10 min then treated with a solut ion of the enol acetate 252 (140 mg, 1.0 mmoles) in a c e t o n i t r i l e (2 ml). Af ter s t i r r i n g the re -action mixture at room temperature fo r 2 hr a l l the enol acetate appeared to be used (by v . p . c ) . The reaction mixture was then treated with water (5 ml) and extracted with d iethy l ether (4 x 5 ml). The combined ethereal extracts were washed with d i l u t e sodium thiosulphate u n t i l co lour les s , then water (2 x 5 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 212 mg of a l i g h t yel low o i l which was chromatographed on s i l i c a gel (5 g). E lut ion with benzene afforded 200 mg (60%) of l -azido-2- iodocyclohexyl acetate (253). T.L.C. analys is of th i s material on s i l i c a gel with benzene as the eluent indicated one compound, 0.85. 0 Infrared (CHC1 3), 2120, 1740 cm" 1 ; n.m.r. (CDC1 3), T8.89 ( s i n g l e t , 3H, acetate) , 5.37 ( t r i p l e t , IH, J = 6 Hz); mass spectrum | ( re la t i ve i n t e n s i t y ) , 309(19.6),267(28.7) ,225(48.1), 224(10.3),222(7.9),180(11.5),140(45.7),128(46.9),127(50.3),112(48.8), 98(55.6),85(17.8),84(45.9),43(100). Anal. Calcd. for C 8 H 1 2 I N 3 0 2 : C, 31.09; H, 3.91; N, 13.59; I, 41.05. Found: C, 30.95; H, 3.95; N, 13.73; I, 41.20. - 229 -Preparation of pentenyl acetate (258) To pentanal (6.0 g, 0.069 moles) was added potassium acetate (1.0 g, 0.01 moles) and acet ic anhydride (15.0 g, 0.147 moles). The reaction mixture was ref luxed for 10 hr and then poured into d iethy l ether (100 ml). The d iethy l ether solut ion was washed with water (4 x 20 ml) and saturated sodium bicarbonate so lu t i on , dr ied over sodium sulphate and f i l t e r e d . The f i l t r a t e was removed under reduced pressure to afford a dark brown o i l . D i s t i l l a t i o n of th i s material gave 3.37 g (38%) of the enol acetate 258 , b.p. 148-150° ( l i t . 2 °°b.p. 148-149°). Infrared (CHC1 3), 1740, 1600 c m ' 1 . Preparation of l -azido-2- iodopentyl acetate (259) | y a Sodium azide (1.20 g, 18 moles) in a c e t o n i t r i l e (16 ml) was cooled i n ice-methanol and treated with a solut ion of iodine monochloride (1.53 g, 8.4 mmoles) in a c e t o n i t r i l e (9 ml). The mixture was s t i r r ed for 5 - 1 0 min, then treated with a solut ion of the enol acetate 258 (470 mg, 3.6 mmoles) in a c e t o n i t r i l e (8 ml). The reaction mixture was s t i r r ed at room temperature for 2 hr, then treated with water (20 ml) and extracted with d iethy l ether (4 x 20 ml). The combined ethereal extracts were washed with d i l u t e sodium thiosulphate un t i l co lour les s , then water (2 x 20 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 700 mg of a yel low l i q u i d which was chromatographed on s i l i c a gel (10 g). - 230 -Elut ion with chloroform y ie lded 730 mg (63%) of compound 259 as a c lear o i l . Infrared (CHC1 3), 2120, 1740 cm" 1 ; n.m.r. (CDC1 3), x7.80 ( s i ng l e t , 3H, acetate), 5.90 (mu l t i p le t , IH, C-2 H), 4.13, 4.0 (double doublet, IH, C-l H, J = 4.5 Hz); mass spectrum | ( r e l a t i v e i n t e n s i t y ) , 297(.4) ,255(1.9) ,209(1.6) ,183(1.0), 170(1.5),154(2.3),128(22.1),127(16.3),43(100), Anal. Calcd. for C 7 H 1 2 I N 3 0 2 : C, 28.30; H, 4.07; N, 14.14; I, 42.71. Found: C, 28.25; H, 4.15; N, 14.20; I, 42.52. Attempted preparation of 2,17e-diacetoxy-2-azido-l- iodo-2,3-seco-5g-androstane-3-n i t r i le (241) Sodium azide (29.4 mg, 0.45 mmoles) in a c e t o n i t r i l e (2 ml) was cooled in ice-methanol and then treated with a solut ion of iodine mono-chlor ide (33.7 mg, 0.208 mmoles) in a c e t o n i t r i l e (3 ml). The mixture was s t i r r ed for 5 - 10 min then treated with a so lut ion of the enol acetates 223a and 223b (35 mg, 0.09 mmoles) in a c e t o n i t r i l e (2 ml). The reaction mixture was s t i r r ed at room temperature for 2 hr, then treated with water (6 ml) and extracted with d iethy l ether (2 x 5 ml). The combined ethereal extracts were washed with d i l u t e sodium thiosulphate u n t i l co lour les s , then water (2 x 5 ml ) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure - 231 -to afford 32 mg of a brown o i l . The infrared spectrum and t . l . c . analysis of th i s product indicated predominantly s ta r t ing mater i a l . Preparation of 17e-acetoxy-l-oxo-2,3-seco-A-nor-5e-androstane-3-n i t r i l e (262) A solut ion of c i s and trans 2,17B-acetoxy-2,3-seco-5g-androst-l-e n e - 3 - n i t r i l e (223a and 223b, 510 mg, 1.318 mmoles) in ethyl acetate (20 ml) at dry ice-acetone temperature was treated with ozone (50 ml/min) f o r 20 min. The resu l t ing dark blue solut ion was allowed to stand at dry ice-acetone temperature for 35 min and then the excess ozone was removed with a stream of nitrogen. The solvent was removed under reduced pressure to give an o i l y residue. Methanol (30 ml) and aqueous sodium su lphite (5%, 70 ml) was added to the residue. The mixture was allowed to stand at room temperature for 2.5 hr and then concentrated (70 ml). The aqueous solut ion was then extracted with d iethy l ether (3 x 50 ml). The combined ethereal extracts were washed with saturated sodium bicarbonate (2 x 50 ml) and saturated sodium chlor ide (2 x 50 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 378 mg (86%) of compound 262 as a c lear o i l . T.L.C. analys is of th i s material on s i l i c a gel with a mixture of chloroform and ethyl acetate as the eluent indicated one compound, Rp 0.73. V.P.C. analys is (column D, 250°, 45 ml/ min) indicated one compound (Retention t ime, 9 min). Compound 262 was converted into i t s 2,4-DNP de r i va t i ve , m.p. 186-188°, for ana lys i s . - 232 -Infrared (CHC1 3), 2725, 1720 cm" 1 (see Figure V , Page 251); n.m.r. (CDC1 3), x9.24 ( s i ng l e t , 3H, C-18 CH 3 ), 8.97 ( s i ng le t , 3H, C-19 CH 3 ), 8.01 ( s i ng l e t , 3H, acetate), 5.41 ( t r i p l e t , IH, C-17 H^), x.46 ( s i ng le t , IH, CHO) (see Figure XIV, p a g e 254 ) ; mass spectrum ^ ( re la t i ve i n t en s i t y ) , 331(15),330(8) ,303(15),302(20),301(15), 270(12),243(36),242(70),241(20),202(15),201(24),200(15),43(100) (see Figure XXIII, Page 257 ); Anal. Calcd. for C 26H33N 5 0 6 : C, 61.04; H, 6.50; N, 13.68. Found: C, 60.90; H, 6.70; N, 13.48. Preparation of methylenetriphenylphosphorane Sodium hydride (25 mmoles as a 50% dispersion in mineral o i l ) in a 50 ml three-necked f la sk was washed with several portions of n-pentane to remove the mineral o i l . The f la sk then was equipped with rubber stopples, a re f lux condenser f i t t e d with a three-way stopcock, and a magnetic s t i r r e r . The system was a l t e rna t i ve l y evacuated and f i l l e d with nitrogen; dimethyl sulphoxide (28 ml) was introduced v ia syr inge, and the mixture was heated at 75-80° for ca_. 45 min. The resu l t ing solut ion of methylsu l f iny l carbonion was cooled in an ice-water bath, and (methyl)-triphenylphosphonium bromide (8.9 g, 0.025 moles) in warm dimethyl sulphoxide (25 ml) was added. The re su l t i ng solut ion of the y l i d e was s t i r r ed at room temperature for 10 min before use. - 233 -Attempted preparation of 17B-acetoxy-2,3-seco-5g-androst-l-ene-3- n i t r i l e (57a) To a so lut ion of 178-acetoxy-l-oxo-2,3-seco-A-nor-5g-androstane-3-n i t r i l e (262, 30 mg, 0.090 mmoles) in dimethyl sulphoxide (5 ml) was added methylenetriphenylphosphorane (100 y l , 0.10 mmoles). The reaction mixture was heated at 50° for 6 hr. The solut ion was cooled and then poured into water (20 ml). The aqueous phase was extracted with pentane (4 x 25 ml). The combined f ract ions were washed with water (3 x 25 ml) and saturated sodium chlor ide solut ion (2 x 25 ml) , dried over sodium sulphate and f i l t e r e d . The solvent was removed by evaporation under reduced pressure to afford 15 mg of a brown o i l . Preparative layer chromatography of t h i s product on a 20 x 5 cm s i l i c a gel coated p l a te , adsorbant thickness 0.9 mm, using chloroform as e luent, did not y i e l d any o l e f i n i c compounds by infrared studies. Preparation of 17g-acetoxy-2,3-seco-5g-androst- l -ene-3-nitr i le (57a) n_- Butyl l i th ium (2.1M, 1.3 ml, 2.7 mmoles) was added to (methyl ) - t r i -phenylphosphonium bromide (1.00 g, 2.8 mmoles) in dry benzene (50 ml) under an atmosphere of nitrogen. A f te r 1.5 hr, 176-acetoxy-l-oxo-2,3-seco-A-nor-5g-androstane-3-nitr i le (262, 160 mg, 0.48 mmoles) was added and the solut ion was s t i r r ed for 6 hr at room temperature. Water (50 ml) was then added, and the organic layer separated. The aqueous phase was extracted with benzene (3 x 30 ml). The combined - 234 -benzene extracts were washed with saturated sodium chlor ide (2 x 20 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford a brown o i l (630 mg) which was chromato-graphed on s i l i c a gel (36 g). E lut ion with a mixture of chloroform ethyl acetate (3:1, v/v) afforded a c lear o i l which was dissolved in acet ic anhydride (3 ml) and pyridine (0.4 ml). The solut ion was s t i r r ed at room temperature for 15 hr, and then the solvent was removed under reduced pressure to give an o i l y residue which was taken up in d iethy l ether (40 ml). The d iethy l ether so lut ion was washed with IN hydro-ch l o r i c acid (2 x 10 ml) , saturated sodium bicarbonate (2 x 20 ml) and sodium chlor ide (2 x 10 ml) so lut ions, dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to give 81.2 mg (51%) of compound 57a as a c r y s t a l l i n e s o l i d , m.p. 132 - 135°. Compound 57a was sublimed at 125°, 0.1 mm pressure, to afford needles, m.p. 134-135°, [a]p 5+31.2° (c=.8, MeOH). Infrared (CHC1 3), 2250, 1720, 1630, 980, 920 cm~i (see Figure VI Page 251 ); n.m.r. (CDC1 3), T9.20 ( s i ng l e t , 3H, C-18 C H 3 ) , 8.83 ( s i ng le t , 3H, C-19 C H 3 ) , 5.41 ( t r i p l e t , I H , C-17 H j , 5.10-4.20 (mu l t i p le t , 3H, v i n y l i c protons) (see Figure XV , Page 254 ); mass spectrum ^ ( r e l a t i ve i n t e n s i t y ) , 329(15),303(10) ,302(7),301(20), 288(6),287(11),286(6),254(10),243(20) ,242(16) ,241(23),43(100) (see Figure XXIV , Page 257 ); Anal. Calcd. for C 2 i H 3 1 N 0 2 : C, 76.55; H, 9.48; N, 4.25. Found: C, 76.49; H, 9.54; N, 4.19. - 235 -Preparation of 17e-hydroxy-2,3-seco-58-androst- l-ene-3-nitr i le (57b) n_-Butyllithium (2.5M, 2 ml, 5.0 mmoles) was added to (methy l ) - t r i -phenylphosphonium bromide (1.80 g, 5.04 mmoles) in dry tetrahydrofuran (20 ml) at 0° under an atmosphere of nitrogen. Af ter 1.5 hr the mixture was cooled to - 78° , 17g-acetoxy-1 -oxo-2,3-seco-A-nor-5g-androstane-3-n i t r i le (262, 360 mg, 1.08 mmoles) in dry tetrahydrofuran (10 ml) was added, and the solut ion was s t i r r ed for 6.5 hr at room temperature. Water (4 ml) was then added, the tetrahydrofuran was evaporated, and the residue was dissolved in d iethy l ether (100 ml). The d iethy l ether solut ion was washed with saturated sodium chlor ide (5 x 60 ml) , dr ied over sodium sulphate and f i l t e r e d . Removal of the solvent under reduced pressure afforded 1.04 g of a brown o i l which was chromatographed on s i l i c a gel (36 g). E lut ion with a mixture of chloroform ethyl acetate (3:1, v/v) gave 205 mg (65%) of compound 57b as a s o l i d . Compound 57b was sublimed at 125°, 0.1 mm pressure, to y i e l d a c r y s t a l l i n e s o l i d , m.p. 129-131°. Infrared (CHC1 3), 3400, 2250, 1630, 920, 990 cm"*; n.m.r. (CDC1 3), T9.23 ( s i ng l e t , 3H, C-18 C H 3 ) , 8.80 ( s i ng l e t , 3H, C-19 CH*), 6.34 ( t r i p l e t , IH, C-17 H , J = 9 Hz), 5.0 - 4.0 (mu l t i p le t , 3H, v i n y l i c protons); mass spectrum I ( r e l a t i v e i n t e n s i t y ) , 288(32) ,287(100) ,272(20),269(15), 260(20),245(25),228(34); M. W. Calcd. for C 1 9 H 2 9 N0 : 287.2248. Found (high resolut ion mass spectrometry): 287.2254. - 236 -Preparation of 3-azido-4-iodocyclohexyl c a rbon i t r i l e (275) Sodium azide (2.38 g, 0.035 moles) in a c e t o n i t r i l e (20 ml) was cooled in ice-methanol and then treated with a solut ion of iodine mono-chlor ides (2.67 g, 0.016 moles) in a ce t on i t r i l e (10 ml). The mixture was s t i r r ed for 5 - 1 0 min then treated with a solut ion of cyclohex-3-enyl c a rbon i t r i l e (274, 1.57 g, 0.014 moles) in a c e t o n i t r i l e (10 ml). The reaction mixture was s t i r r ed at room temperature for 24 hr, then poured into water (40 ml) , and extracted with d iethy l ether (3 x 50 ml). The combined ethereal extracts were washed with d i l u t e sodium t h i o -sulphate un t i l co lour le s s , then with water (2 x 20 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 3.3 g (81%) of compound 275. Infrared (CHC1 3), 2250, 2100 cm" 1 ; n.m.r. (CDC1 3), tca_. 6 (mu l t i p le t , 2H, C-3 H_ and C-4 H); mass spectrum | ( r e l a t i ve i n t en s i t y ) , 276(70) ,149(100),127(22) ,107(29), 106(40),80(46),79(46),68(50),67(70),42(57). Mol. Wt. Calcd. fo r C 7 H 9 N 4 I : 275.9873. Found (high resolut ion mass spectrometry): 275.9858. Preparation of ( 2 -az ido - l - i odocyc lohexy l ) -aceton i t r i l e (284) Sodium azide (1.17 g, 0.018 moles) in a ce t on i t r i l e (20 ml) was cooled in ice-methanol and then treated with a so lut ion of iodine mono-chlor ide (1.46 g, 0.009 moles) in a c e t o n i t r i l e (10 ml). The mixture - 237 -was s t i r r ed for 5 - 1 0 min then treated with a so lut ion of cyclohex-1-eny laceton i t r i le (960 mg, 7.93 mmoles). The reaction mixture was s t i r r ed at room temperature for 24 hr then treated with water (20 ml) and extracted with d iethy l ether (2 x 20 ml). The combined ethereal extracts were combined and washed with d i l u t e sodium thiosulphate un t i l co lour les s , then with water (2 x 20 ml ) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to af ford 1.96 g (85%) of compound 284 as a yellow o i l . T.L.C. analysis of th i s material on s i l i c a gel with chloroform as the eluent indicated one compound, 0.85. Infrared (CHC1 3), 2250, 2100 c m ' 1 ; n.m.r. (CDC13), T6.20 - 5.60 (mu l t i p le t , IH ); mass spectrum | ( r e l a t i ve i n t e n s i t y ) , 290(6),163(64) ,134(75) ,133(75), 121(64),120(100),119(51),107(80),106(85),105(70),94(75),93(98) ,42(64). Preparation of 178-acetoxy-2-azido-l-iodo-2,3-seco-5e-androstane-3-n i t r i l e (270) Sodium azide (59.5 mg, 0.9 mmoles) in a c e t o n i t r i l e (4 ml) was cooled in ice-methanol and treated with a solut ion of iodine mono-chlor ide (92.4 mg, 0.57 mmoles) in a c e t o n i t r i l e (2 ml). The mixture was s t i r r ed for 5 - 10 min then treated with a solut ion of 173-acetoxy-2,3-seco-53-androst- l-ene-3-nitr i le (57a, 65 mg, 0.19 mmoles). - 238 -The course of the reaction was monitored by t . l . c . and infrared spectro-scopy. The reaction mixture was s t i r r ed at room temperature for 5 hr, then treated with water (10 ml) and extracted with d iethy l ether (2 x 20 ml). The combined ethereal extracts were washed with d i l u t e sodium thiosulphate u n t i l co lour les s , then with water (2 x 10 ml) , dried over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 80 mg (80%, crude) of compound 270 as a l i g h t yellow o i l . T.L.C. analys is of th i s material on s i l i c a gel with a mixture of chloroform and ethyl acetate (1:1, v/v) as the eluent indicated one major compound, 0.71, and baseline contaminants. Attempts to pur i fy the crude product by preparative layer chromatography f a i l e d . Infrared (CHC1 3), 2205, 220, 1720 cm" 1 ; n.m.r. (CDC1 3), T9.20 ( s i ng le t , 3H, C-18 CHj), 8.74, 8.62 ( s i ng le t s , 3H, C-19 CH3), 6.6 - 5.8 (mu l t i p l e t , 3H, C- l H and C-2 Hj,), 5.37 ( t r i p l e t , 1 H, C-17 H , J = 9 Hz). Attempted preparation of 173-acetoxy-l,2-imino-2,3-seco-5e-androstape-3-n i t r i l e (242) To a solut ion of 17B*acetoxy-2-azido-l-iodo-2,3-seco-58-androstane-3 - n i t r i l e (270, 10 mg, 0.02 mmoles) in isopropanol (5 ml) was added sodium borohydride (5 mg, 0.105 mmoles). The reaction mixture was refluxed for 3 hr and then the solvent was removed by evaporation - 239 -under reduced pressure. The residue v/as dissolved in water (10 ml) and extracted with d iethyl ether (3 x 5 ml). The combined ethereal extracts were washed with .IN hydrochloric acid (2 x 5 ml), saturated sodium bicarbonate ( 2 x 2 ml) and sodium chlor ide (5 ml) so lut ions. The solvent was removed by evaporation under reduced pressure to afford 4.3 mg of a c lear o i l . T.L.C. analys is of th i s material on s i l i c a gel with chloroform as the eluent indicated predominantly one compound, 0.65, and minor amounts of s ta r t ing mater ia l , R f 0.42. Infrared (CHC1 3), 2250, 2125, 1600 cm* 1 . Attempted preparation of N-(triphenylphosphonium iodide)-17e-acetoxy-1,2-imino-2,3-seco-5e-androstane-3-nitrile (283a) To a solut ion of 17B-acetoxy-2-azido-l-iodo-2,3-seco-5g-androstane-3 - n i t r i l e (270, 65 mg, 0.11 mmoles) in dry benzene (10 ml) was added triphenylphosphine (31 mg, 0.11 mmoles). The reaction mixture was s t i r r ed at room temperature for 24 hr. No phosphonium s a l t was observed to p rec ip i t a te . The solvent was removed under reduced pressure to afford a brown o i l . T.L.C. analys is of th i s material indicated a complex mixture of compounds. Infrared (CHC1 3), 2250, 1720, 1600 cm" 1 . - 240 -Preparation of 17g-hydroxy-2,3-seco-5g-androst-1-en-3-al (271) Diisobutylaluminum hydride (87 mg, 0.36 mmoles) in benzene (0.31 mis) was added to a solut ion of 17e-hydroxy-2,3-seco-53-androst-l - e n e - 3 - n i t r i l e (57b, 40 mg, 0.14 mmoles) in dry benzene (10 ml) under an atmosphere of nitrogen. The mixture was s t i r r ed for 1 hr at room temperature and then the complex was decomposed by careful addit ion of d i l u t e acet ic acid (7 ml, 10%). The resu l t ing mixture was s t i r r ed for 30 min at room temperature. The organic layer was extracted with benzene (2 x 25 ml). The combined benzene extracts were washed with saturated sodium bicarbonate (2 x 10 ml) and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 30.3 mg (74%) of o l e f i n i c aldehyde 27J_ as a c lear o i l . T.L.C. analys is of t h i s material on s i l i c a gel with a mixture of chloroform and ethyl acetate (2:1, v/v) as the solvent system indicated e s sen t i a l l y one compound, R,. 0.72. Infrared (CHC13), 3450, 2750, 1720, 1630, 990, 910 cm" 1 ; n.m.r. (CDC1 3), T9.30 ( s i ng l e t , 3H, C-18 CHg), 8.80 ( s i n g l e t , 3H, C-19 CH,), 6.47 ( t r i p l e t , IH, C-17 H ), 5.4 - 4.0 (mu l t i p le t , 3H, v iny l - a protons), 4.7 ( t r i p l e t , IH, CH0, J = 2 Hz); mass spectrum | ( r e l a t i v e i n t e n s i t y ) , 291(21),290(95),247(42) ,246(60), 233(15),231(22),229(16),221(100),220(50),219(100),218(64),202(24), 201(48) ,200(17). Mol. Wt. Calcd. for C 1 9 H 3 0 ° 2 : 290.2245. Found (high resolut ion mass spectrometry): 290.2271. - 241 -Preparation of 3,3-ethylenedioxy-2,3-seco-5B-androst-l-en-17e-ol (272) A mixture of 17B-hydroxy-2,3~seco-5B-androst-l-en-3-al (271 , 33 mg, 0.11 mmoles), p-toluenesulphonic acid monohydrate (1 small c r y s t a l ) , ethylene g lycol (0.6 ml), and dry methylene chlor ide (5 ml) was shaken in a f lask at room temperature for 24 hr, poured into d i l u t e sodium bicarbonate so lut ion (20 ml) and extracted with methylene chlor ide (2 x 12 ml). The combined methylene chlor ide extracts were washed with saturated sodium bicarbonate (2 x 10 ml) and sodium chlor ide (2 x 10 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to give 47.2 mg of a brown o i l which was chromatographed on f l o r i s i l (6 g). E lut ion with a mixture of chloroform and benzene (1:1, v/v/) afforded 246 mg (71%) of o l e f i n i c acetal 272 as a c lear o i l . T.L.C. analysis of th i s material on f l u o r o s i l indicated one major compound, R f 0.45. Infrared (CHC1 3), 3500, 1630, 990, 910 cm" 1 ; (see Figure VII , Page252 ); n.m.r. (CDC1 3), x9.37 ( s i ng le t , 3H, C-18 CH 3 ), 8.93 ( s i ng le t , 3H, C-19 CH 3 ), ca. 6.3 (mu l t ip le t , 5H, C-17 H and acetal protons), 5.32 ( t r i p l e t , IH, C2H_, J = 5 Hz), 5.3 - 4.0 (mu l t i p le t , 3H, v i n y l i c protons); (see Figure X Page 255 ); mass spectrum | ( r e l a t i ve intensity),335(9) ,334(40),248(38),1 75(100), 147(50),125(85); (see Figure XXV , Page 258 ); Moi. Wt. Calcd. for : C 2 1 H 3 t t 0 5 : 334.2507. Found (high resolut ion mass spectrometry): 334.2506. - 242 -Attempted preparation of 2-azido-3,3-ethylenedioxy- l- iodo-2,3-seco-58-androstan-173-ol Sodium azide (10 mg, 0.15 mmoles) in a c e t o n i t r i l e (4 ml) was cooled in ice-methanol and treated with a solut ion of iodine monochloride (11.6 mg, 0.071 mmoles) in a c e t o n i t r i l e (4 ml). The mixture was s t i r -red for 10 min, then treated with 3,3-ethylenedioxy-2,3-seco-53-androst-l-en-178-ol (272, 20 mg, 0.059 mmoles) in a c e t o n i t r i l e (3 ml). The reaction mixture was s t i r r ed at room temperature for 48 hr, then poured into water (5 ml) , and extracted with d iethy l ether (2 x 10 ml). The combined ethereal extracts were washed with d i l u t e sodium t h i o -sulphate u n t i l co lour les s , then with water (2 x 20 ml) , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to y i e l d 10 mg of a brown o i l . T.L.C. analys is of th i s material on s i l i c a gel with a mixture of chloroform and ethyl acetate (2:1, v/v) as the eluent indicated two major compounds, R f 0.85 and 0.75, several minor compounds, and baseline contaminants. Pu r i f i c a t i on by preparative layer chromatography on s i l i c a gel f a i l e d . Infrared (CHC1 3), 3400, 2100, 1740, 1620 cm" 1 . Preparation of 178-acetoxy-l-hydroxy-2-nitro-2,3-seco-5$-androstane-3 - n i t r i l e (291) A so lut ion of cyanoaldehyde (65.0 mg, 0.19 mmoles) in nitromethane (25.0 ml) at room temperature was treated with t r i e t h y l amine (6 ml) - 243 -under an atmosphere of nitrogen. The solut ion was s t i r r ed at room temperature for f i ve days in the dark. The solvent was removed under reduced pressure to afford 89.3 mg of a yellow o i l . This material (89.3 mg) was chromatographed on a 20 x 20 cm s i l i c a gel coated p la te , adsorbant thickness 0.9 mm, using a mixture of chloroform and ethyl acetate (5:1, v/v) as eluent. A f te r e l u t i o n , the band l y ing in the region Rf 0.75-0.85 was removed and extracted with ethyl acetate (50 ml). The solvent was removed by evaporation under reduced pressure to afford a c r y s t a l l i n e s o l i d . Rec ry s ta l l i za t i on from ethanol gave 55.2 mg (74%) of compound 29J_ as a c r y s t a l l i n e s o l i d m.p. 186-188°. An ana ly t i ca l specimen was obtained by two rec ry s ta l1 i za t ions from ethanol to afford prisms, m.p. 195-197°. Infrared (CHC1 3), 3450, 2260, 1725, 1560, 1380 cm" 1 (see Figure VIII Page 252 ); n.m.r. (CDC1 3), x9.21 ( s i ng le t , 3H, C-18 CH^), 8.95 ( s i ng l e t , 3H, C-19 CH 3 ), 6.36 (doublet, IH, C-l H, J = 3 Hz), 4.60 (mu l t ip le t , 3H, C-17 H^ and C-2 H^) (see Figure XVII , Page 255 ) ; mass spectrum ^ ( r e l a t i ve i n t e n s i t y ) , 392(1),346(1),303(5),302(15),260(26), 243(20),242(100),107(15),105(10),95(11),93(18),43(100) (see Figure XXVI Page 258 ). Anal. Calcd. for C 2 1 H 3 2 N 2 O 5 : C, 64.26; H, 8.22; N, 7.13. Found: C, 64.58; H, 8.35; N, 6.86. - 244 -Sodium borohydride reduction of 17e-acetoxy-l-hydroxy-2-nitro-2,3-seco-58-androstane-3-nitr i le (291) To a so lut ion of compound 291 (12 mg, 0.031 mmoles) in isopropanol (7 ml) was added sodium borohydride (12 mg, 0.31 mmoles). The reaction mixture was refluxed for 26 hr and then cooled to room temperature. The solvent was removed by evaporation under reduced pressure to afford a residue which was dissolved in .IN hydrochloric acid (10 ml) and diethyl ether (20 ml). The organic layer was separated and then the aqueous phase was neutra l ized with saturated sodium bicarbonate, and extracted with d iethy l ether ( 3 x 5 ml). The combined ether extracts were washed with saturated sodium bicarbonate (2 x 5 ml) and sodium chlor ide (2 x 5 ml) so lut ions , dr ied over sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 9.1 mg of a c lear o i l . T.L.C. analys is of th i s material on s i l i c a gel with ethyl acetate as eluent indicated predominantly one compound, R f 0.75. Acid extract ion of a so lut ion of the crude product in d iethy l ether, followed by n e u t r a l i -zat ion and subsequent d iethyl ether extract ion was performed. T.L.C. examination of the resu l t ing d iethyl ether solut ion indicated the absence of basic compounds. V.P.C. analysis (column D, 250°, 45 ml/min) indicated one compound (Retention time, 2.8 min). The crude product (8 mg) was chromatographed on a 5 x 20 cm s i l i c a gel coated p la te , adsorbant t h i c k -ness 0.6 mm, with ethyl acetate as eluent. After e lut ion the band l y ing in the region R f 0.7 was removed and extracted with ethyl acetate. The solvent was removed under reduced pressure to afford 4.5 mg (ca_. 50%) of compound 267 as a c r y s t a l l i n e s o l i d , m.p. 136-140°. - 245 -Infrared (CHC1 3), 3400 cm" 1 ; mass spectrum £ ( r e l a t i ve i n t en s i t y ) , 294(0.8) ,277(0.7),276(2.7) ,259(1.2) , 220(1.5 ) ,219(1.8 ) ,202(3.8 ) ,149(13.9) ,43(100) ; Moi. Wt. Calcd. for C 1 8 H 3 0 0 3 : 294.2194. Found (high resolut ion mass spectrometry): 294.2202. Sodium borohydride reduction of 178-acetoxy-l-oxo-2,3-seco-A-nor-5e-androstane-3-n i t r i le (262) To a solut ion of compound 262 (25.3 mg, 0.076 mmoles) in isopropanol (10 ml) was added sodium borohydride (25 mg, 0.65 mmoles). The reaction mixture was ref luxed for 26 hr and then cooled to room temperature. The solvent was removed by evaporation under reduced pressure to y i e l d 21.6 mg (crude, 9 6 J ^ ) of a c lear o i l . The t . l . c , v . p . c , infrared and mass spectrum of t h i s material were ident ica l with that of 267. Cata l y t i c hydrogenation of 178-acetoxy-l-hydroxy-2-nitro-2,3-seco-5g-androstane-3-n i t r i le (291) A solut ion of compound 291 (30.3 mg, 0.077 mmoles) in methanol (25 ml) was hydrogenated over 10% palladium-on-charcoal for 5 hr. The solut ion was f i l t e r e d and then the solvent was removed under reduced pressure to afford an o i l which was dissolved in d iethyl ether (20 ml). The d iethy l ether solut ion was washed with saturated sodium ch lo r ide , dr ied over - 246 -sodium sulphate and f i l t e r e d . The solvent was removed under reduced pressure to afford 21.6 mg of a c lear o i l . T.L.C. analys is of the crude product on s i l i c a gel with a mixture of chloroform and ethyl acetate (5:1, v/v) indicated predominantly two compounds, R f 0.73 and 0.61, and several minor components. Acid extract ion of the crude product, followed by neut ra l i za t ion and subsequent d iethyl ether extract ion was performed. T.L.C. examination of the d iethy l ether so lut ion indicated the absence of any basic compounds. V.P.C. analysis (column D, 250°, 45 ml/min) of the crude product indicated two compounds (Retention times 9 and 18 min). Infrared (CHC13), 3450, 2250, 1720 cm" 1 . Preparation of 173-acetoxy-l-hydroxy-2,3-seco-A-nor-5e-androstane-3-n i t r i l e (296) To a solut ion of 17g-acetoxy-l-oxo-2,3-seco-A-nor-5g-androstane-3 - n i t r i l e (262, 30.1 mg, 0.0909 mmoles) in ethanol (10 ml) was added sodium borohydride (4 mg, 0.10 mmoles). The reaction mixture was s t i r r ed at room temperature for 24 hr and then the solvent was removed under reduced pressure. The resu l t ing residue was dissolved in .IN hydrochloric acid (10 ml) and d iethyl ether (20 ml). The organic layer was separated and the aqueous phase was neutral ized with saturated sodium bicarbonate, and extracted with d iethy l ether (2 x 10 ml). The combined d iethy l ether extracts were washed with saturated sodium chlor ide solut ion (2 x 5 ml ) , dried over sodium sulphate and f i l t e r e d . The solvent was removed - 247 -under reduced pressure to y i e l d 28 mg of an o i l which was chromatographed on a 5 x 20 cm s i l i c a gel coated p l a te , adsorbant thickness 0.6 mm, with a mixture of chloroform and ethyl acetate (5:1, v/v) as the eluent. After e l u t i o n , the band ly ing in the region R^  0.6 was removed and extracted with ethyl acetate. The solvent was removed under reduced pressure to afford 18.2 mg (60%) of compound 296 as a c r y s t a l l i n e s o l i d , m.p. 139-141°. Infrared (CHC1 3), 3450, 2250, 1720 cm' 1 ( see Figure IX, Page 252); n.m.r. (CDC1 3), T9.22 ( s i ng le t , 3H, C-18 CH 3), 8.97 ( s i ng le t , 3H, C-19 CH3 ) , 7.98 ( s i n g l e t , 3H, acetate), 6.53, 6.49 (double doublet, 2H C-l H2. 0 = 11 Hz), 5.40 ( t r i p l e t , IH, C-17 hT, J = 10 Hz) (see Figure XVIII, Page 255); mass spectrum ^ ( re l a t i ve i n t e n s i t y ) , 333(1.5),318(1) ,304(9),303(42), 301(7),286(5),263(39),262(58),260(35),243(20),242(100),203(24),202(47), 201(47),188(6),187(10),186(6),177(8),161(10),107(35),105(23) (see Figure XXVII, Page 258). Moi. Wt. Calcd. for C 2 0 H 3 i 0 3 N ; 333.2303. Found (high reso lut ion mass spectrometry): 333.2297. Preparation of aluminum amalgam Aluminum (1 g) f o i l was thoroughly washed with petroleum ether and - 248 -cut into pieces of about 2 x 2 cm. Su f f i c i en t 2% aqueous sodium hydro-xide to cover the metal (10 ml) was added. After vigorous hydrogen evolution set i n , the solut ion was decanted and the metal washed repeatedly, and qu ick ly , with water; 0.5% aqueous mercuric ch lor ide (20 ml) was added and, a f te r 2 min, the solut ion was decanted and the metal washed repeatedly, and qu ick ly , with water and then 95% ethanol. Attempted reduction of 178-acetoxy-l-hydroxy-2-nitro-2,3-seco-5g-androstane-3-n i t r i le (291) by employing aluminum amalgam To a solut ion of compound 291 (5 mg, 10.2 mmoles) in d iethy l ether (5 ml) and methanol (3 ml) was added 100 mg of aluminum amalgam. The reaction mixture was allowed to stand at 0° for twenty four hr and then the reaction solut ion was decanted. The solvent was removed under reduced pressure to af ford a yel low o i l . T . L . C , - 249 -V.P.C. and inf rared studies indicated 173-acetoxy-l-oxo-2,3-seco-A-nor-5B-androstane-3-n i t r i le (262) as the predominant product. - 250 -FREQUENCY (cm- 1) FREQUENCY (cm- 1) FREQUENCY (cm-i) - 255 -Figure XIX J L _ i 100 so 200 2 5 0 Figure XX iSo" 1 300 3 5 0 m/e 4 O 0 J L 100 2 0 0 2 5 0 Figure XXI 360 3 5 0 m/e 4 0 0 —1—1—I m/e 4 0 0 too 150 -1 T i r 1 1 | 1 2 0 0 2 5 0 3 0 0 3 5 0 lOO 80-60-Figure XXII 3 5 0 m/e ACO TOO 150 200 250 300 Figure XXIII 1O0 ISO 200 250 360 T — 1 — r T 1 f Figure XXIV 100 T—•)—r—*-i r—r 150 — T — 1 — 1 1 r | 1 26o 250 350 m/e 4CO 300 n 1—*i—r — j 1 1 r-350 m/e 4O0 ioo7 8 0 -6 0 -4 0 -20 -100 8 0 6 0 2 0 1 0 0 150 1 — 1 1 1 j — 1 1 1 — 1 1——1—T—r ISO Figure XXV >00 5o" 3 0 0 3o~ m/e 4 0 0 Figure XXVI S o 2 5 0 2 0 0 1 3 0 0 Figure XXVII 350 m/e 4OO 2 5 0 3 0 0 3 5 0 m/e 4 0 0 BIBLIOGRAPHY 1. (a) L. F. Fieser and M. F ieser , " S te ro id s " , Reinhold Publishing Corp., New York, N. Y., 1959; (b) P. A. Hart in "Stero id Reactions", C. D jeras s i , Ed., Holden-Day, Inc., San Francisco, C a l i f . , 1963, chapter 3; (c) C. Fenselau in "Stero id Reactions", C. D jeras s i , Ed., Holden-Day, Inc., San Francisco, C a l i f . , 1963, chapter 13; (d) N. L. Wendler in "Molecular Rearrangements", P. de Mayo, Ed., Interscience Publ ishers, New York, N. Y., 1964, Volume I I , chapter 16; (e) J . Fried and J . A. Edwards, "Organic Reactions in Steroid Chemistry", Van Nostrand-Reinhold Co., 1972, Volume I and II. 2. R. B. Woodward, "A romat ic i ty " , Special Publ icat ion No. 21, The Chemical Society, London, 1967, p. 217. 3. Review of the salamander a l ka l o i d s : (a) C. Schopf, Experientia LZ, 285 (1961); (b) G. Habermehl, Naturwiosenchaften, §3 , 123 (1966); (c) G. Habermehl in "The A l k a l o i d s " , R. H. F. Manske, ed, Academic Press, New York, 1967, Volume IX, p. 427; (d) G. Habermehl in "Progress in Organic Chemistry", Butterworth, London, 1968, Volume VII, p. 35. 4. S. Hara and K. Oka, J . Am. Chem. S o c , 8 J , 1041 (1967). 5. H. Singh, V. V. Parashar, S. Padmanabhan, and R. B. Mathur, Indian  Journal of Pharmaceutical Education, 4, 2 (1970). - 260 -6. K. Schreiber in "The A l k a l o i d s " , R. H. F. Manske, Ed., Academic Press, New York, 1968, Volume X, p. 1. 7. S. M. Kupchan and A. W. By in "The A l ka l o i d s " , Academic Press, New York, 1968, Volume X, p. 193. 8. V. Cerny and F. Sorm in "The A l k a l o i d s " , R. H. F. Manske, Ed., Academic Press, New York, 1967, Volume IX, p. 305. 9. M. Martin-Smith and M. F. Sugrue, J . Pharm. Pharmacol., ] J , 569 (1964). 10. M. Alauddin and M. Martin-Smith, J . Pharm. Pharmacol., 14, 325 (1962). 11. M. Alauddin and M. Martin-Smith, J . Pharm. Pharmacol, 14, 469 (1962). 12. T. Tokuyama, J . Da l ly , B. Witkop, I. L. Kar le, and J . Kar le , J . Am.  Chem. S o c , 2D. 1917 (1968). 13. R. Imhof, Miss E. Gbssinger, W. Graf, H. Berner, Mrs. L. Berner-Fenz, and H. Wehrl i , Hel v. Chim. Acta, 5J, 1151 (1972). 14. S. Zalesky, Med. Chem. Unterss., Hoppe-Seyler, 1, 85 (1866). 15. C. Schopf and W. Braun, Ann. Chem., 514, 69 (1934). 16. J . Daley and B. Witkop, A ldr ich imica Acta, 2, (No. 4 ) , 3 (1970). 17. E. S. Faust, Arch. Expt l . Pathol. Pharmakol., 41, 229 (1898). 18. 0. Gessner in "Handbuch der experimentellen Pathologie" , (Erg. W.), Volume 6, p. 56. 19. 0. Gessner, Ber. Ges. Bef. Naturwiss. Marburg, 6J , 138 (1926). 20. 0. Gessner and F. Mollenhoff, Arch. Expt l . Pathol. Pharmakol., J6Z, 638 (1932). 21. 0. Gessner and W. Esser, Arch. Expt l . Pathol. Pharmakol., 122, 639 (1935) 22. 0. Gessner and G. Urban, Arch. Expt l . Pathol. Pharmakol., 187, 378 (1937) 23. F. M. Dean "Natura l ly Occurring Oxygen Ring Compounds", Butterworth and Co. L t d . , 1963. - 261 -24. H. Wieland and F. J . Wei l , Chem. Ber., 4J , 3315 (1913). 25. K. Meyer, Heh. Chim. Acta, 5J , 1097 (1969). 26. C. Schopf and K. Koch, Ann. Chem., 552, 37 (1942). 27. C. Schopf, H. K. Blbdom, D. K l e i n , and 6. Se i t z , Chem. Ber., 8 j , 372 (1950). 28. C. Schopf and D. K l e i n , Chem. Ber., §2 , 1638 (1954). 29. E. Wbl fe l , C. Schopf, G. Weitz, and G. Habermehl, Chem. Ber., 9J , 2361 (1961). 30. G. Habermehl, Ann. Chem., §7J, 164 (1964). 31. G. Habermehl, Chem. Ber., 9 J , 1439 (1966). 32. G. Habermehl, G. Haaf, and G. Volkwein, unpublished. 33. G. Habermehl, Chem. Ber., 9 J , 840 (1962). 34. E. Staple in "Biogenesis of Natural Compounds", P. Bernf ie ld (Ed.), Pergamon Press, New York, 1967, chapter 4. 35. G. Habermehl, Ann. Chem., 680, 104 (1964). 36. G. Habermehl, G. Volkwein, and A. Haaf, Chem. Ber., 100, (1967). 37. G. Habermehl, Chem. Ber., 9 J , 143 (1963). 38. J . K. Pa is ley and L. Wei ler, Tetrahedron Let te r s , 261 (1972). 39. Y. Shimizu, Tetrahedron Let te r s , 2919 (1972). 40. (a) R. B. Woodward, A. A. Patchett, D. H. R. Barton, D. A. J . Ives, and R. B. Ke l l y , J . Chem. S o c , 1131 (1957); (b) R. L. Autrey and P. W. Schul lard, J . Amer. Chem. S o c , 9J , 4917 (1968). 41. G. Eggart, C. Pascual, and H. Wehrl i , Hel v. Chim. Acta, 5jQ, 985 (1967). 42. A. J . L i s ton , J . Org. Chem., 3J, 2105 (1966). 43. S. Hara and K. Oka, Tetrahedron Let te r s , 1987 (1969). 44. (a) H. Wasserman and E. H. Barber, J . Am. Chem. S o c , 9J., 3674 (1969); (b) P i s . Abst., 3£k, 193 (1971). - 262 -45. E. Demole and H. Wuest, Hel v. Chim. Acta, §Q, 1314 (1967). 46. U. Scheidegger, K. Schaffner, and 0. Jeger, i b i d . , 4 J , 400 (1962). 47. Y. Gaoni, J . Chem. Soc., £, 2925 (1968). 48. J . B. Doughty, C. L. L a z z e l l , and A. R. C o l l e t t , J . Am. Chem. Soc., Z2, 2866 (1950). 49. (a) C. W. Shoppee and G. Kruger, J . Chem. S o c , 3641 (1961); (b) G. Habermehl and A. Haaf, Justus Liebigs Ann. Chem., 722 > 155 (1969); (c) For a review of the Beckmann rearrangement and Schmidt reaction of s tero id ketones, see H. Singh, V. V. Parashar, and S. Padmanabhan, J . S c i . & Ind. Res., 2J , 200 (1966). 50. (a) K. Oka and S. Hara, Tetrahedron Let te r s , 1189 and 1193 (1969); (b) G. Habermehl and A. Haaf, Tetrahedron Let te r s , 3815 (1969). 51. E. Caspi, Y. Shimizu, and S. N. Balasubrahmanyam, Tetrahedron, 2 J , 1271 (1964). 52. D. Lednicer in "Advances in Organic Chemistry", E. C. Taylor, Ed., Wiley-Interscience Publ i shers, Inc., New York, N. Y., 1972, Volume 8, pp. 209 - 225. 53. (a) K. Oka and S. Hara, Chem. Ind. (London), 911 (1968); (b) K. Oka and S. Hara, Chem. Ind. (London), 168 (1969). 54. M. G. Combe, H. B. Henbest, and W. R. Jackson, J . Chem. S o c , 2467 (1967). 55. M. Uskokovic, J . Gutzw i l l e r , and T. Henderson, J . Am. Chem. Soc., 9J , 203 (1970). 56. (a) R. Huisgen and J . Reinertshofer, Ann., 575, 174 (1952); (b) R. Huisgen and J . Reinertshofer, Ann., 575, 197 (1952). 57. (a) G. Habermehl and A. Haaf, Liebigs Ann. Chem., 722, 155 (1969); - 263 -(b) G. Habermehl and A. Haaf, Liebigs Ann. Chem., 723, 181 (1969). 58. (a) J . L. Riebsomer, Chem. Revs., 3J , 160 (1945); (b) R. Huisgen, Ann., 574, 185 (1951); (c) E. H. White, J . Am. Chem. Soc., ZZ, 6008 (1955). 59. (a) E. H. White, J . Am. Chem. Soc. , ZZ, 6011 (1955); (b) E. H. White, J . Am. Chem. S o c , ZZ, 6014 (1955). 60. H. R. Nace and A. C. Watterson, J . Org. Chem., 3J, 2109 (1966). 61. (a) L. Ruzicka, V. Prelog, and P. Meister, Helv. Chim. Acta . , 2 J , 1251 (1945); (b) S. Hara, N. Matsumoto, and M. Takeuchi, Chem. Ind. (London), 2086 (1962); (c) V. Burckhaldt and T. Reichstein, Helv. Chim. Acta . , 2 J , 821 (1942); (d) V. Burckhaldt and T. Reichstein, Helv. Chim. Acta . , 2 J , 1434 (1942); (e) L. F. Fieser and M. F ieser, " S te ro ids " , Reinhold Publ ishing Corp., New York, N. Y., 1959, p. 249. 62. (a) C. H. Hassa l l , Org. Reactions, 2. 73 (1957); (b) P. A. S. Smith in "Molecular Rearrangements", P. de Mayo, Ed., Wi ley- Inter-science, New York, 1963, Volume I, pp. 568 - 591; (c) H. 0. House, "Modern Synthetic Reactions", W. A. Benjamin, Inc., New York, 1965, pp. 123 - 128. 63. (a) A. D. Yoffee and P. Gray, J . Chem. S o c , 1412 (1951); (b) E. H. White and W. R. Feldman, J . Am. Chem. S o c , 7J , 5832 (1957). 64. W. S. Johnson, J . C. C o l l i n s , J r . , R. Pappo, M. B. Rubin, P. J . Kropp, W. F. Johns, J . E. P ike, and W. Bartmann, J . Am. Chem. Soc., 8 J , 1409 (1963). 65. (a) R. S. Tipson, M. A. Clapp, and L. H. Cretcher, J . Org. Chem., 12, 133 (1947); For reviews see (b) R. S. Tipson, Adv. Carbohydrate - 264 -Chem., §, 107 (1953); (c) D. J . Ba l l and F. w. Par r i sh , i b i d , 2 J , 233 (1968). 66. For a review see (a) R. J . Fe r r i e r , Adv. Carbohydrate Chem., ZQ, 67 (1965); (b) B. He l fer ich and E. Himmer, Ber., 6J , 1825 (1928); (c) B. He l fer ich and R. Mittag, Ber., J J , 1585 (1938); (d) L. Hough and B. A. Otter, Chem. Commun., 173 (1966). 67. E. S. Gould, "Mechanism and Structure in Organic Chemistry", Holt Rinehart and Wiston, Inc., New York, 1959, Chapter 14. 68. (a) P. A. Levene and T. Mori, J . B i o l . Chem., Z& 1 (1928). (b) H. Johansson and S. M. Hagman, Ber., 5J , 647 (1922); (c) C. S. Marvel, D. W. MacCorquodale, F. E. Kendal l , and W. A. Laz ie r , J . Am. Chem. Soc., 4J , 2838 (1924); (d) J . Braun and E. Beschke, Ber., 23, 4364 (1906). 69. (a) S. Winstein and R. E. Buckles, J . Am. Chem. S o c , 6 J , 2780, 2787, 2796 (1942); (b) S. Winstein, E. Grunwald, and L. L. Ingraham, J . Am. Chem. Soc., 7J, 821 (1948). 70. J . March, "Advanced Organic Chemistry: Reactions, Mechanisms, and St ructure" , McGraw-Hill Inc., New York, 1968, Chapter 17, pp. 743 - 747. 71. L. Ehmann, K. Heusler, Ch. Meystre, P. Wieland, G. Anner and A. Wettstein, Helv. Chim. Acta. , 4J , 2548 (1959). 72. B. Berkoz, A. D. Cross, M. E. Adame, H. Carpio, and A. Bowers, J . Org. Chem., 28, 1976 (1963). 73. 0. Winterstein and M. Moore, J . Am. Chem. Soc , 6J , 1503, 1507 (1943). 74. A. D. Cross, J . A. Edwards, J . C. Orr, B. Berkoz, L. Cervantes, M. C. Calzada, and A. Bowers, J . Med. Chem., §, 162 (1963). - 265 -75. P. Z ieg le r , K. R. Bharucha, Chem. and Ind. (London), 1351 (1955). 76. F. Jarreau, M. B. Tchoubar and R. Goutarel , B u l l . Soc. Chem.  France, 887 (1962). 77. H. C. Brown and D. G i n t i s , J . Am. Chem. Soc., 7J, 5376 (1956). 78. (a) M. H. Huffman, M. H. Lo t t , and A. T i l l o t s o n , J . B i o l . Chem., 217, 103 (1955); (b) R. T. B l i ckens ta f f and E. L. Foster, vh_ Org. Chem., 2J , 2883 (1961). 79. G. Nathansohn, U. S. Patent, 3,025,309 (1969). 80. (a) W. G. Dauben, R. A. M i c h i l i , and J . F. Eastham, J . Am. Chem.  Soc., 7J, 3852 (1952); (b) L. F. Fieser and X. A. Dominguez, i b i d . , 7J , 1704 (1953); (c) M. Rubin and B. H. Ambrecht, i b i d . , Z J , 3513 (1953); (d) B. Camerino, B. P a t e l l i , and R. S l i aky , Tetrahedron Let te r s , 554 (1961); (e) R. 0. C l i n ton , R. L. Clarke, F. W. Stonner, A. J . Manson, K. F. Jennings, and D. F. P h i l l i p s , J . Org. Chem., 22, 2800 (1962); ( f ) A. K. Rose, G. Mina, M. S. Manhas, and E. Rzucidlo, Tetrahedron Let te r s , 1467 (1963); (g) M. Garodetsky, E. Levy, R. D. Youssefyeh, and Y. Mazur, Tetrahedron, Z2, 2039 (1966); (h) L. F. Fieser and M. F ieser, " S te ro id s " , Reinhold Publishing Corp., New York, N. Y., 1959, Chapter 8. 81. B. Berkos, E. P. Chavez, and C. D jeras s i , J . Chem. S o c , 1312 (1962). 82. T. L. H i l l , J . Chem. Phys., ] J , 399 (1948). 83. (a) H. 0. House, "Modern Synthetic Reactions", W. A. Benjamin, Inc., New York, N. Y., 1965, Chapter 3, pp. 61 - 7 1 ; (b) H. E. Zimmerman in "Molecular Rearrangements", P. de Mayo (Ed.), Inter-science Publ ishers, New York, N. Y., Volume I, Chapter 6; (c) H. 0. House and V. Kramar, J . Org. Chem., 2J , 3362 (1963); - 266 -(d) H. 0. House and B. M. Trost, J . Org. Chem., 2Q, 1341 (1965). 84. See J.A. Hirsch in "Topics in Stereochemistry" N.L.Al l inger and E.L. E l i e l , Eds., Interscience Publ ishers, John Wiley & Sons, Inc., N.Y., 1967, p. 199. 85. (a) D. H. R. Barton, A. DaS. Campos-Neves, and R. C. Cookson, J . Chem. S o c , 3500 (1956); (b) D. H. R. Barton and W. J . Rosen-fe l de r , J . Chem. S o c , 1047 (1951). 86. Y. Shimizu, H. Mitsuhashi, and E. Caspi, Tetrahedron Let te r s , 4113 (1966). 87. 6. Stark, P. Rosen, and N. L. Goldman, J . Am. Chem. S o c , 8 J , 2965 (1961). 88. J . A. Edwards, J . C. Orr, and A. Bowers, J . Org. Chem., 2Z, 3378 (1962). 89. (a) T. Takahashi, Y. Satoh, and A. Hagitani, Nippon Nagaku Zasshi, § 3 . 974 (1968); (b) Chem. Abst r . , 7 J , 78247a (1969); (c) Y. Satoh, M. Mukotu, Y. Ogaki, T. Takahashi, T. Kimura, H. Aoki , and A. Hagitani, B u l l . Chem. Soc., Japan, 3 J , 855 (1966). 90. L. F. Fieser and M. Romero, J . Am. Chem. Soc., 7 J , 4716 (1953). 91. (a) E. J . Corey, J . Am. Chem. S o c , 7 J , 2301 , 3297, 4832 (1953); (b) E. J . Corey, J . Am. Chem. S o c , 7 J , 175 (1954); (c) E. J . Corey, T. H. Topie and W. A. Wozniak, J . Am. Chem. Soc., ZZ, 5415 (1956); (d) E. J . Corey and J . H. Burke, J . Am. Chem. S o c , ZZ, 5418 (1955); (e) E. J . Corey and R. A. Sneen, J . Am. Chem. Soc., 73, 6269 (1956). 92. R. V i l l o t t i , H. J . Ringold, and C. Djerass i , J . Am. Chem. Soc. , 82. 5693 (1960). 93. (a) N. S. Bhacca and D. H. Wil l iams, "Appl icat ions of N.M.R. - 267 -Spectroscopy in Organic Chemistry", Holden-Day, Inc., London, 1964, Chapter 3 and 6; (b) L. M. Jackman and S. S te rnhe l l , "Appl icat ions of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry", Pergamon Press, Oxford, Chapter 5, p. 238. 94. I. S. Y. Wang and E. W. Warnhoff, Chem. Commun., 1158 (1969). 95. R. Rosenfeld, J . Am. Chem. Soc., 7 J , 5540 (1957). 96. J . S. G. Cox, J . Chem. Soc., 4508 (1960). 97. J . Y. Satoh and T. T. Takahashi, Chem. Commun., 1714 (1970). 98. D. H. R. Barton, P. D. Magnus, and M. J . Pearson, Chem. Commun., 550 (1969). 99. J . C. Sheehan and R. M. Wilson, J . Am. Chem. S o c , 89_. 3457 (1967). 100. R. L. Clarke, K. Dobriner, A. Mooradian, and M. Ma r t i n i , J . Amer. Chem. S o c , U, 661 (1955). 101. K. L. Williamson and W. S. Johnson, J . Org. Chem., 2J , 4563 (1961). 102. K. Takeda, T. Okanishi, H. Osaka, A. Shimaoka, and N. Aezano, Chem. Pharm. B u l l . (Tokyo), 9, 388 (1961). 103. K. T o r i , T. Komeno, and T. Nakagawa, J . Org. Chem., 2J , 1136 (1964). 104. R. F. Zurcher, Hel v. Chim. Acta . , 4 J , 2054 (1964). 105. K. L. Will iamson, J . I. Coburn, and M. F. Herr, J . Org. Chem., 3 J , 3934 (1967). 106. (a) F. G. Bordwell, Acc. Chem. Res., 3, 281 (1970); (b) R. Noyori, Y. Hayakaua, M. Funakura, H. Takaya, S. Murai, R. Kobayashi, S. Tsutsumi, J . Am. Chem. S o c , 2 4 , 7202 (1972). 107. J . King, J . Chem. S o c , 1788 (1936). 108. P. W. Clutterbuck and F. Reuter, J . Chem. S o c , 1467 (1935). 109. (a) L. Malaprade, Compt. Rend., 195, 1395 (1932); (b) T. Posternak - 268 -and H. Pal laczek, Helv. Chim. Acta . , 2J , 1190 (1941); For reviews see (c) B. Sk larz, Quart Review, 21, 3 (1967); (d) G. Dryhurst, "Periodate oxidation of d io l and other functional groups", Pergamon Press L td . , Oxford, 1970. 110. R.B. Woodward et a i . , Tetrahedron, 2, 1(1958). 111. (a) R. U. Lemieux in "Molecular Rearrangements", P. de Mayo (Ed.), Interscience Publ ishers, New York, N. Y., 1964, Volume I I , Chapter 12; (b) C. A. Lobry de Bruyn and W. Alberda van Ekenstein, Rec.  t rav. chim., 12, 1 (1900) and e a r l i e r publ icat ions. 112. H. Kataoka, Jap. Pat. , 7002,575 (1970); Chem. Abstr . , U, 121083z (1970) . 113. (a) R. H. H i l l , J . Org. Chem., 2 J , 29 (1962); (b) C. W. Shoppee and S. K. Ray, J . Chem. S o c , 3774 (1963); (c) C. A. Grob, H. P. F ischer, W. Randenbusch and J . Zergenyi, Helv. Chim. Acta . , 4 J , 1003 (1964); For a review see (d) C. G. McCarty in "The Chemistry of the carbon-nitrogen double bond", S. Patai (Ed.), Interscience Publ ishers, Inc., New York, N. Y., 1970, Chapter 9; (e) D. Lendnicer in "Advances in Organic Chemistry:, E. C. Taylor (Ed.), Wiley-Interscience Publ ishers, Inc., New York, N. Y., 1972, Volume VI I I , pp. 233 - 235. 114. H. P. F ischer, C. A. Grob, and E. Renk, Helv. Chim. Acta . , 4 J , 2930 (1962). 115. (a) S. I to , H. Takeshita and M. Hiramai, Tetrahedron Let te r s , 1181 (1971) ; (b) C. W. M. Grant 's Ph.D. thes i s , Univers i ty of B r i t i s h Columbia, March 1972. 116. (a) "The Chemistry of the Carbonyl Groups", S. Patai (Ed.), Inter-- 269 -science Publ i shers, New York, N. Y., 1966; (b) "The Chemistry of the Cyano group", Z. Rappoport (Ed.), Interscience Publ i shers, New York, N. Y., 1970. 117. K. Kato and Y. H i ra ta , Chem. Commun., 1324 (1970). 118. H. Oediger and F. Mo l ler , Angew. Chem., 7 J , 53 (1967). 119. (a) E. Truscheit and K. E i t e r , Ann., 682, 65 (1962); (b) E. Vogel, R. Schubart, and W. A. B a l l , Angew. Chem., Internat. Ed., 2, 510 (1964); (c) H. Oediger, H. J . Kabbe, F. Mo l ler , and K. E i t e r , Ber., 22, 2012 (1966); (d) E. Vogel and F. G. Klamer, Angew, Chem., Internat. Ed., Z, 374 (1968). 120. G. Frangatos and A. Taurins, Can. J . Chem., 22, 410 (1961). 121. L. A. Paquette and N. A. Nelson, J . Org. Chem., 2J , 1085 (1962). 122. K. Schof ie ld , "Hetero - Aromatic Nitrogen Compounds", Butterworth and Co. L t d . , London, Chapter 6, p. 389. 123. V. N. Setkina, N. K. Baranetskaya, and D. N. Kursanov, B u l l . Acad.  S c i . (Div. Chem. S c i . ) , 667 (1955). 124. (a) K. Jones and M. F. Lappert, J . Chem. S o c , 1944 (1965); (b) D. J . Cardin and M. F. Lappert, Chem. Commun., 1034 (1967). 125. (a) C. W. Shoppee, S. K. Roy and B. S. Goodrich, J . Chem. S o c , 1583 (1961); (b) H. B. Henbest, E. R. H. Jones, A. A. Wagland, and T. I. Wrigley, J . Chem. S o c , 2477 (1955). 126. S. Hiunig and M. K i e se l , Ber., 21, 380 (1958). 127. J . March, "Advanced Organic Chemistry: Reactions, Mechanisms, and St ructure" , McGraw-Hill, Inc., New York, N. Y., 1968, Chapter 17. 128. C. H. Snyder and A. R. Soto, J . Org. Chem., 22, 742 (1964). 129. H. R. Nace, J . Am. Chem. S o c , 8 J , 5428 (1959). - 270 130. K. Heusler, P. Wieland, and A. Wettstein, Helv. Chim. Acta . , 42, 1586 (1959). 131. G. Grethe, J . Gutzwi l le r , H. L. Lee and M. R. Uskokovic, Helv.  Chim. Ac ta . , § § , 1044 (1972). 132. (a) S. Mon, J . Chem. Soc. Japan, ZQ, 303 (1949); (b) L. F. Fieser and R. E t to r re , J . Am. Chem. S o c , 7J , 1700 (1953); (c) L. F. Fieser and W. Y. Huang, i b i d . , 7 J , 4837 (1953). 133. H. 0. House, "Modern Synthetic Reactions", W. A. Benjamin, Inc., New York, N. Y., 1965, Chapter 6. 134. (a) P. Z. Bedoukian, Org. Syn., C o l l . Vol. 3, 127 (1955).; (b) P. Z. Bedoukian, J . Am. Chem. Soc., 6J , 1325 (1944). 135. P. Grafen, H. J . Kabbe, 0. Roos, G. D. Diana, Tsung-tee L i , and R. B. Turner, J . Am. Chem. Soc., 9J , 6131 (1968). 136. R. B. Moffatt and D. I. Weisblat, J . Am. Chem. Soc., 7 J , 2183 (1952). 137. A. F. B. Cameron, J . S. Hunt, J . F. Oughton, P. A. Wilk inson, and B. H. Wilson, J . Chem. S o c , 3864 (1953). 138. (a) R. B. Turner, J . Am. Chem. S o c , 22, 579 (1950); (b) M. Gut, Helv. Chim. Acta . , 3 J , 906 (1953). 139. (a) J . Fajkos and F. Sorm, C o l l . Czech., 2 J , 766 (1959); (b) R. Deghenghi and C. R. Engel, J . Am. Chem. S o c , §2 , 3201 (1960); (c) C. R. Engel, R. M. Haegerle, and R. Deghenghi, Can. J . Chem., 3 J , 1199 (1960). 140. D. H. R. Barton, R. M. Evans, J . C. Hamlet, P. G. Jones, and T. Walker, J . Chem. S o c , 747 (1954). 141. R. V i l l o t t i , H. J . Ringold, and C. D jeras s i , J . Am. Chem. S o c , 82, 5693 (1960). - 271 -142. B. E. Edwards and P. N. Rao, J . Org. Chem., 3J , 324 (1966). 143. N. H. Andersen, H. Uh, S. E. Smith, and P. G. M. Wuts, Chem.  Commun., 956 (1972). 144. C. M. Robb and E. M. Schultz, Org. Syn., C o l l . Vo l . 3, 347 (1955). 145. H. C. Braun and 0. J . Cope, J . Am. Chem. Soc., 8 J , 1801 (1964). 146. H. C. Braun and L. Sharp, J . Am. Chem. Soc., 9J , 2915 (1968). 147. J . W. Lewis and A. A. Pearce, Tetrahedron Let te r s , 2039 (1964). 148. M. E. Herr and F. W. Heyl, J . Chem. S o c , 3627 (1952). 149. (a) F. W. Fowler, A. Hassner, and L. A. Levy, J . Am. Chem. S o c , §2, 2077 (1967); (b) A. Hassner, G. J . Matthews, and F. W. Fowler, J . Am. Chem. Soc , 9J , 5046 (1969); For a review see (c) A. Hassner, Accounts of Chemical Research, 4, 9 (1971). 150. (a) M. Rabinovitz in "The Chemistry of the Cyano Group", Z. Rappoport (Ed.), Interscience Publ ishers, John Wiley & Sons, 1970, Chapter 7; (b) P. A. S. Smith, "The Chemistry of Open-Chain Nitrogen Compounds", W. A. Benjamin, Inc., New York, N. Y., 1965, Volume 1, Chapter 5. 151. P. A. S. Smith, J . H. H a l l , and R. 0. Kan, J . Am. Chem. S o c , § J , 485 (1962). 152. (a) P. A. S. Smith, "The Chemistry of Open-Chain Nitrogen Compounds", Benjamin, Inc., New York, 1965, Volume 1, p. 71; (b) J . March, "Advanced Organic Chemistry; Reactions, Mechanisms, and St ructure" , McGraw-Hill, Inc., New York, 1968, Chapter 6, pp. 683 - 684. 153. For a review of ozonolysis see J . S. Belew in "Ox idat ion " , Marcel Dekker, Inc., New York, Volume I, Chapter 6, p. 259. 154. For reviews see (a) A. Maercker, Org. Reactions, H , 270 (1965); (b) S. T r i ppet t , Quart. Rev. (London), U , 406 (1963); (c) H. 0. - 272 -House, "Modern Synthetic Reactions", W. A. Benjamin, Inc., New York, N. Y., 1965, Chapter 8, pp. 245 - 256. 155. J . J . Pappas, W. P. Keaveney, E. Gancher, and M. Berger, Tetra- hedron Let te r s , 4273 (1966). 156. R. Aneja, S. K. Mukerjee, and T. R. Seshadri, Chem. Ber., 9J , 297 (1960). 157. R. H. Call ighan and M. G. W i l t , J . Org. Chem., 2J , 4912 (1961). 158. E. J . Corey and E. Hamanaka, J . Am. Chem. Soc., § J , 2758 (1967). 159. R. Greenwald, M. Chaykovsky, and E. J . Corey, J . Org. Chem., 2J , 1128 (1963). 160. W. Nagata, T. Sugasaua, M. Narisada, T. Wakabayashi, and Y. Hayase, J . Am. Chem. S o c , 8J , 1483 (1967). 161. D. Taub, N. N. G i r o t r a , R. D. Hoffsommer, C. H. Kuo, H. L. S lates, S. Weber and N. L. Wendler, Tetrahedron, 2J , 2443 (1968). 162. (a) F. Sondheimer and R. Mechoulam, J . Am. Chem. Soc., 7 J , 5029 (1957); (b) G. R. P e t t i t , B. Green, G. L. Dunn, and P. Sunder-Plassmann, J . Org. Chem.,. 3 J , 1385 (1970). 163. (a) J . E. McMurry, J . Am. Chem. S o c , £Q, 6821 (1968); (b) W. F. Erman and L. C. Stone, J . Am. Chem. Soc., 9J , 2821 (1971); (c) F. Kida, H. Uda, and A. Yoshekoshi, Chem. Commun., 1335 (1969). 164. R. E. Ireland and L. N. Mander, J . Org. Chem., 3 J , 689 (1967). 165. E. J . Corey, D. E. Cane, and L. L i b i t , J . Am. Chem. Soc , 9J, 7016 (1971). 166. V. Bazant, M. Capka, M. Cerny, V. Chralovsky, K. Koch loef l , M. Kraus, and J . Malek, Tetrahedron Let te r s , 3303 (1968) 167. R. B. Woodward, K. Heusler. J . Go s te l i , P. Naegel i, W.Oppolzer, R. Ramage, S. Ranganathan, and H. Vabruggen, J . Am. Chem. Soc., 8 J , 852 (1966). - 273 -168. (a) For a general review on the reactions of azides with nucleo-ph i l e s , see G. L'abbe, Ind. Chim. Beige, 3J , 519 (1968); (b) H. Staudinger and M. Meyer, Helv. Chim. Acta, Z, 635 (1919); (c) J . E. L e f f l e r and R. D. Temple, J . Am. Chem. Soc., p j , 5235 (1967), and references c i ted there in. 169. H. Zimmer and G. Singh, J . Org. Chem., 2J, 483 (1963). 170. G. Wi t t i g and V. Schollkopf, Chem. Ber., 8 J , 1318 (1954). 171. K. Ponsold and H. Groh, Chem. Ber., 9J , 1009 (1965). 172. A. Hassner and J . E. Ga l le , J . Am. Chem. Soc., 7 J , 3733 (1969). 173. W. J . Bai ley and S. A. Buckler, J . Am. Chem. Soc., 7 J , 3567 (1957). 174. For a review of Phosphorus-Nitrogen chemistry, see E. Fluck in "Topics in Phosphorus Chemistry", Intersicence Publ ishers, New York, N. Y., 1967, Volume 4, P- 291. 175. (a) A. B. Burg and P. J . S lo ta , J . Am. Chem. Soc., §£, 2145 (1960); (b) Z. Shrowaczewska and P. Mastalerz, Chem. Abstr . , §fj, 7075 (1956). 176. S. J . Bro i s , J . Org. Chem., ZI, 3532 (1962). 177. T. Sheradsky in "The Chemistry of the Azido Group", S. Patai (Ed.), Interscience Publ ishers, John Wiley & Sons, New York, 1971, Chapter 6. 178. (a) L. I. Zakharkin and I. M. Khor l ina, Chem. Abstr., 5J, 8040f (1958); (b) A. E. G. M i l l e r , J . W. B i s s , and L. H. Schuartzman, J . Org. Chem., 2J, 627 (1959); (c) S. Trofimenko, J . Org. Chem., 2J, 3046 (1964). 179. (a) For reviews, see (a) E. Arundale and L. A. Mikeska, Chem. Rev., 5J 505 (1952); (b)J.Roberts in " F r i eda l -C ra f t s and Related Reactions", 01 ah (Ed.), Interscience Publ ishers, Inc., New York, 1963, Volume 2, pp. 1175 - 1210; (c) J . March, "Advanced Organic Chemistry: - 274 -Reactions, Mechanisms, and St ructure " , McGraw-Hill, Inc., New York, 1968, Chapter.16, pp. 711 - 713. 180. N. A. LeBel and E. G. Banucci, J . Org. Chem., 3 J , 2440 (1971). 181. (a) Y. Naves and P. Ochsner, Helv. Chim. Acta. , 4 J , 51 (1964); (b) W. S. Johnson, Accounts of Chemical Research, 1, 1 (1968). 182. W. S. Johnson and R. B. K inne l , J . Am. Chem. Soc., 8 J , 3861 (1966). 183. W. Nagata, T. Wakabayashi, M. Narisada, Y. Hayase, and S. Kamata, J . Am. Chem. Soc., 9J , 5740 (1971 ). 184. P. E. Eaton, J . Am. Chem. Soc., 8 J , 2344 (1962). 185. J . March, "Advanced Organic Chemistry: Reactions, Mechanisms, and St ructure" , McGraw-Hill, Inc., New York, 1968, Chapter 12, pp. 458 - 460. 186. (a) L. A. Paquette and M. K. Scott, J . Am. Chem. Soc., 9J, 6751 (1972); (b) L. A. Paquette and M. K. Scott, J . Am. Chem. Soc., " 9J , 6760 (1972). 187. R. L. Reeves in "The Chemistry of the Carbonyl Group", S. Patai (Ed.), Interscience Publ ishers, John-Wiley & Sons, New York, M. Y., 1966, Chapter 12. 188. (a) For a review, see G. Jones, Org. Reactions, ] i , 204-599 (1967); (b) J . March, "Advanced Organic Chemistry: Reactions, Mechanisms, and St ructure " , McGraw-Hill, Inc., 1968, pp. 697 - 698. 189. (a) P. N. Rylander in "Ca ta l y t i c Hydrogenation over Platinum Metals " , Academic Press, New York, 1967, Chapters 11 and 12; (b) M. F re i fe lder in " P r ac t i ca l Ca ta l y t i c Hydrogenation", Wiley-Interscience, Inc., New York, 1971, Chapters 10 and 12. 190. J . B. Hester, J r . , J . Org. Chem., 2J , 1158 (1969). - 275 -191. 6. D. Buckley and T. J . E l l i o t t , J . Chem. Soc., 1508 (1947). 192. L. F. Fieser and W. H. Daudt, J . Am. Chem. S o c , 6 J , 2248 (1946). 193. E. Schmidt and R. Wilkendorf, Ber., 5J , 389 (1919). 194. B. L. Zen i tz , E. B. Macks, and M. L. Moore, J . Am. Chem. S o c , ZD, 955 (1948). 195. J . Corey, N.H. Andersen, R.M. Carlson, J . Paust, E. Vedejs, J . Vlattas and R.E.K. Winter, J . Am. Chem. Soc., gQ, 3245 (1968). 196. G.B. Bachman and R.J. Maleski, J . Org. Chem., 37, 2810 (1972). 197. This experiment was performed by Mr. R.B. Warneboldt. 198. This experiment was performed by Dr. L. Weiler. 199. P.Z. Bedoukian, J . Am. Chem. S o c , 67_, 1430 (1945). 200. L.E. Schniepp and H.H. Ge l l e r , J . Am. Chem. Soc., 67, 54 (1945). 201. M. Smith in "Reduction", R. L. Augustine, Ed., Marcel Dekker, Inc., New York, N. Y., 1968, Chapter 2, p. 95. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0060149/manifest

Comment

Related Items