I u 111 11 1 11 11 1 1 1 1|. 1 11 1 1 1151 3351""! 1‘ ‘1111 1' U “1‘ III.'1'I 1,1'11I1I1ll'1I 1151111' .1II'.'11'1'1"1"” 1' '1 1"--. .I ' ' \m I.......' I"1‘ . IIIIII..IIIL' I'...IIIMI'I1I.' '.n.'...I...II'.I'.".‘fl. u."1 II. 'IIm...I..II . '1" '171111HI T17511'11'1'1' 11II11111. “1119111111 1.11 1.111 11111 111111111 11111 IIII1.111111 ...11 1 5 111111 ”I" "511739 IA),(163 1'1 I111 51.13 1111 'III 11 '11711' '11 ,II11117I1II51' 111111121 : 1'11'1"I'I.I' 1“"11II‘11 :. I111 501. 11.111 I. 1151' .II 'W'I'. II1' &11'110' 1" '1hI'1'1'1II 1 '1' WI. I ’ _ ' ' 1 _ 1 I'm111I1 I 'J.11 111111 111 11151 1I' 11111 1111111111 1 1.11 1 111 1 1 _ I .. I'~".'-' : 1"“ 51:} ~‘ ~..1 If? 3?I. IIIIIIIII 7111'1u'jldi111'5'11' 1' 7"7" 17111; 1 I11] I .. '5' 1.1111ng ., A ”'1' 1‘4"“ ‘1 L 1' 1 )1-111 1'IIII'7111'1'11111 1 1 1111 11301111 1; '1?” 11 19:1,.1-{i'13hf'54 E'fi? /II 'II '11I11'11' 'I ‘7’1’I'I5I1 5.“! 1m I- 13:31“: .157? 'I"'1"I1,.I'I 5::1L'11'11A'Z'1'1F1rl'5'151 '1’1'150 ' II'I'1I'.'="'-I I". '5'5i2I""-"‘""7'13'g'1r'1'1'1 "'1. "1' Ii" 12"". " '1'“ ' .' «I .1III 111 1117 1.111 ”1v '1711'111‘ II111 I'“'5 111 1-1 11,1111 1: 11111111111111.11111111111111111 {111/11111111111111.1111 4 1 ,1 1 II '5 1517'1‘ -I11I',.1"'1""1 '11" ' 'I. .7111": 1.111%.“ ' I II 11715111111‘1'111'! 11111j—fl1'11111111g1111hd 1111 11111111 1111111111 111 Mfiflyfi ':'1I1I III.’ '1117" 1'11"'1'I.11i1' I1I1"‘ .. I111 'I‘I‘I1I11'Ik 1I1‘1 5'“ 11h 1'1117'Vm1 11111017411. '11" ”11 IIIII'11111j-1111111,1a1 111;)” I1IIJI11'1'111 I 111111111311“: II :1, 11%. .11'1, 1'1111111I1 111 1 "I11" 11111111'11'1'1' 1111211111151 1&1111Iw11111'111111 11 171' 11111.11? 11%114 '1' ('LJ v " _ ,1 1 1, '1'1 1"?i'lll1'11111l1 '. I ”“1'151'111111'111111'1'1'1 1111111111,“:‘11 11 1 111' 111.1104“ .111 13,111)“ {1:};1'11‘ Lg!’1r1.¥..'1.1.1_.: .2 .14 '1. . "i. ‘11 .1: 1 fl \ A 1 I " '11.11"'1"1" '1" . '1' ' 11 1 .' 31431“ I1 115511., 51.71;! -__’..._.111 1 1 11 .51 5‘55}? H1» .‘ - . I.I.".i"II'I"I1 I7'I' 1'5"“1'. '1“ H’ '1' ‘fI-‘I‘I 454% 'II.";'II':'.-L~i-§.III5"9*~‘III.- , . . 1 I . .. I551 "51115.51" '11' I111" 1111111 . . .II'1,,11\7;-11.L_f["11'§1. pi." , l ' " 'I1 11I|1I111 I11123' . 11111111I1‘I I15 1111111111151 51111111 :11‘1‘10 151.1»1'11411811111 11111111111111.31111ji 1011' "5 11111111|1 '1' 1.01 [11- 1 101 1" 1'.'|| 1 1 111:1 I;';17'7111111I111I111I1(. 1111II111 111": 1 1151111'1'1 1'111 11 . “11:1; 11 11111 1 1. 11111.11 I l .2, .I 111' II I 1211;: I 11711. 1I 1111551 ' I I ' l1 "'1""'111 ”111111111 "" I I" :IIJ:'I 1111111 111111111 1 I 1 1'1111 I 1|" 11 I '7'1-1‘01'. "611.11 _ I'jw I"1":?'m,%:g'1-{LI1; . 1 1 I I 1 I I1 I»? 1 1|II' 1 1111 11.1111 1.1 ‘.l‘ 1 1111111 11111111 11111111 11 I. ' I 111 3.111 '.I‘1‘-11'1I1.11. 111111'1111 ,111 I1111111111 IIIIh-I “‘21:" .r.-,1 "'1'1'11'1' 1 ' I I’ll(:1'[1|"’ " 11.111111" . " “"1 5'111‘M|\111I511""11'111'.’1§1'1'11"l'1111‘“1.111. 1 1 .‘1 l1‘ ,t, ' ' 111541141 1" 1 W1 1'11fi11.1:1:'jl111l11‘11 I1111 ‘ 1 ' ‘1 111‘111”1l1 1 111.1111 ?$${."gd (5 fl .1. 1 181' 11‘ 1 5" II1I II 'I 1. ' ' 1;. I7. 'I ' ' III"1IIIW'1'1'1" 'I III1- .-.‘ "1'1"" 1" 11111"qu 1 ' , . .. I .. II...I.I:IIIII .II . '111|IIII ,1 11.1 1 ' 1 1 1111 I "I" {.‘1. .1 '1‘1I'1 "1111111 1'1 111,151" 11:11111111fl| 1I177I '1'111'1'1 11111H11111I ,1-1I 1 1111 141111 1.1 1 I 11. 111111 11. 1. I 1 I 111111111 1'101l1C'11111 11111111111111 1'. ‘ . 1. I ' . . .‘ ' "'. I" II 1 1 I 1'1 .11 1'1 1' 'I 1151» 11 ;.|.5.'1 1 ' ., I7| 111151'11‘17'111'J‘I’f‘ 1M I " 1"I111I1-".1"". I "'111 L1" ' ' ""1111 "'15 "1' 1' 111 'II '1 II. II .1'111'51111' ”"3 " 1 1 . 1 1 . 1 I 1 1 II11 I . .11 11 )1: \lgyuuumfluwxw“Emmi W LIBRARY . . .Michigan State Univusity ovmwa FINES ARE 25¢ PER DAY PER ITEM Return to buck drop to remove this checkout from your record. A HIGHLY STEREOSELECTIVE SYNTHESIS OF COMPOUNDS RELATED TO (+)-DISPARLURE, THE SEX ATTRACTANT OF THE GYPSY MOTH By Barbara Ann Duhl-Emswiler A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1979 ABSTRACT A HIGHLY STEREOSELECTIVE SYNTHESIS OF COMPOUNDS RELATED TO (+)-DISPARLURE, THE SEx ATTRACTANT OF THE GYPSY MOTH BY Barbara Ann Duhl-Emswiler A highly stereoselective synthesis of Eisg7R,8S-epoxy-2- methyloctadecane, known as (+)—disparlure, and its enantiomer can be carried out in 15-20% overall yield from p-tolyl-L- methylsulfinate. This ester is converted to a t-butylalkyl- sulfoxide via a Grignard reaction followed by an arylalkyl exchange. The a-sulfoxo anion generated from this sulfoxide with n-butyl lithium can be alkylated in a highly stereoselec- tive fashion, employing the chirality of the sulfoxide to induce asymmetry at the alpha carbon. The mixture of dia- stereomeric B—hydroxysulfoxides thus formed can be separated chromatographically, and the desired compound reduced to the corresponding sulfide. Epoxidation can then be effected by treatment of the B-hydroxysulfide with Meerwein's reagent and subsequent displacement of the sulfonium ion by base. Variation of the alkyl groups employed in the Grignard reaction and the alkylation produces the two enantiomers of disparlure, and could be used for the production of homologues or analogues. The episulfide analogues of both (+) and (-)-dispar1ure can be produced directly from the epoxides, with inversion of configuration at each center, by treatment with 3- methylbenzothiazole-Z-thione and trifluoroacetic acid. The optical purity of the B-hydroxysulfide precursors to (+) and (-) disparlure were determined by the use of chiral shift reagent in the 1H NMR to be in the range of 95- 96% depending on the batch studied. The entire sequence is amenable to scale—up and if several of the undesired side products along the route are recycled, the synthesis is an economically viable approach to kilogram quantities of disparlure or its analogues. to John With All My Love 11 ACKNOWLEDGMENTS I would like to thank all of those whose contributions to the gypsy moth project were invaluable, especially Dr. D. G. "Uncle Don" Farnum for his friendship and guid- ance, Dr. Ring Cardé and the U.S.D.A. for financial support, and my colleagues whose work on the chemistry of the project led to its successful conclusion, including Houston and Elene Brown, AnJa Cardé, Ross and Chrysa Muenchausen, Jill Hochlowski, and John Emswiler. I am also grateful for the moral support provided by such friends as the current Farnum group, Jim Larson, Kim Chamberlin, and Larry Klein, In addition, I owe a great debt to the "old" Farnum group for getting me into all of this, particularly Al Hagedorn, Tom Reitz, Chuck Geraci and the late Bill Chambers. Additional financial support was generously provided to me during my graduate career by Dr. John Boezi of the Department of Biochemistry and Dr. Norman Good of the Department of Plant Pathology. Finally, I would like to say thanks also to the Fish Monger people (and all of the nice fishes I have known), and to Pepper Cory and the wonderful women of the Calico Cutters, all of whom made my life here a Joy. iii TABLE OF CONTENTS Chapter LIST OF TABLES. LIST OF FIGURES INTRODUCTION. RESULTS AND DISCUSSION. Synthetic Strategy. . . Synthesis of z-menthyl—p- toluenesulfinate (ég) . . . . . Preparation of p—tolyl-undecyl- sulfoxide ( Z) and Retolyl-6— methylhepty sulfoxide (ég). . Preparation of tfbutylundecyl- sulfoxide ( ) and tabutyl-6- methylhepty sulfoxide (éé). . . The Preparation of the Diastereomers of 2-methyl-8-hydroxyoctadecan-7- yl-t-butylsulfoxide (3Ha and ) and 2-methyl-7-hydroxyocta ecane- -yl-t- butylsulfoxide (32a and R). . . .‘— The Preparation of 7S,88-2-methyl- 8-hydroxyoctadecane-7-yl-t7butyl- sulfide (35) and 7S,88-2-methyl-7- hydroxy-B-yl-tfbutylsulfide (Ag). Preparation of 7R,8S (+)-dis- parlure and 7S,8R (-)-disparlure. Preparation of 7S,8R, cis-7,8- epithio-Z-methyloctadecane and 7R,88,cis-7,8-epithio—2-methyl- octadecane. . . . . . . . . . iv Page . Vii .Viii . 12 . l2 . 18 . 2A 31 3A 35 Chapter Page Determination of Enantiomeric Purity of Disparlure Precursors. . . . . . . . . . . . . Al EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . UN General . . . . . . . . . . , . . . . . . . . . . uu Preparation of p—toluenesulfinic acid (g1) . . . . . . . . . . . . . . . . . . . . AA Preparation of pgtoluenesulfinyl chloride (g8) . . . . . . . . . . . . . . . . . . A5 Preparation of ethyl p- -toluene- sulfinate (g3). . . . . . . . . . . . . . . . . . “6 Preparation of fi-menthyl-pg toluenesulfinate (3Q) from (£3) . . . . . . . . . A7 Preparation of 6-methylheptyl-p- tolylsulfoxide (§g) . . . . . . . . . . . . . . . U8 Preparation of tebutyl-6—methyl- heptylsulfoxide ($3). . . . . . . . . . . . . . . “9 Preparation of 7S,88-2-methyl-8- hydroxyoctadecane-7-yl-t-butyl- sulfoxide ( A , cis Precursor) and its 78, 8R D astereomer (3A p, trans Precursor). . . . . . . . . . . . . . . . . 50 Preparation of 78, 88- -2-methyl— -8- hydroxyoctadecane—7- yl- -t- -butyl- sulfide (SQ). . . . . . . . . . . . . . . . . 51 Preparation of (+)-Disparlure (l§). . . . . . . . . . . . . . . . . . . . . . . 52 Preparation of l-bromoundecane (3g).......................53 Preparation of Undecyl-pftolyl- sulfoxide (31).. . . . . . . . . . . . . . . . 5A Preparation of t-butyl-undecyl- sulfoxide (é8). . . . . . . . . . . . . . . . . 55 Chapter Page Preparation of 6-methylheptyl Tosylate. . . . . . . . . . . . . . . . . . . . . 56 Preparation of l-iodo-6—methyl- heptane . . . . . . . . . . . . . . . . . . . . . 57 Preparation of 6-methylheptan—l- al (up. .. ...............57 Preparation of 7S,8S-2—methyl-7- hydroxyoctadecane-8-yl—tgbutyl- sulfoxide (33a, (-)-cis Precursor . . . . . . . . 58 Preparation of 7S, 88— 2-methyl- -7- hydroxyoctadecane—B-yl- t-buty1- sulfide (ug). . . . . . . . . . . . . . . . . 60 Preparation of 7S,8R (-)-disparlure (gk). . O O O O O O O O O O O O O I O O O O O O O 6]- Conversion of 7S,8R-2-methyl—8- hydroxyoctadecane-7-yl-t-butyl- sulfoxide ( Ab, trans pFecursor) into 78, 88- -methyl-8-hydroxy- octadecane— 7- -yl- t-butylsulfoxide (3&4, gi§_precursor). . . . . . . . . . . . . . . 62 Preparation of racemic cis-7,8—epithio- 2-methyloctadecane. . . . . . . . . . . . . . . . 6“ REFERENCES. . . . . . . . . . . . . . . . . . . . . . 65 vi LIST OF TABLES Table Page 1 Some Optically Active Insect Pheromone Components. . . . . . . . . . . 3 vii Figure LIST OF FIGURES The initial synthesis of racemic disparlure. . . . . . . . . . . . . . . The first stereoselective synthesis of (+)-disparlure . . . . . . . . . A highly stereoselective synthesis of (+)-disparlure . . . . . . . . . . . An asymmetric synthesis of an op- tically active epoxide from an optically active sulfoxide. . . . . . . The synthesis of 7R,88 (+)-disparlure as developed by the Farnum group. . . . The proposed synthesis of 7S,8R (-)-disparlure. . . . . . . . . . . . The Phillips synthesis of (l)- menthyl-p:toluenesulfinate. . . . . . The stereochemistry of the forma- tion of the pftolylalkylsulfoxide precursors to (+) and (-)- disparlure. . . . . . . . . . The stereochemical course of the conversion of pgtolylalkylsulfoxides to Egbutylalkylsulfoxides . . . . . . . viii Page 10 1A 15 l7 19 22 22 Figure 10 11 l2 13 it: Possible conformations of the two configurations of the a-sulfoxo- anion of a tabutylalkylsulfoxide. The inversion of configuration of a hydroxyl—bearing carbon via nucleophilic displacement by super- oxide ion . . . . . . . . Synthesis of (+) disparlure by Pirkle, 22 al.. . . . . . . . Mechanism for KSCN reaction with epoxides. . . . . . . Proposed mechanism for epoxide to episulfide interconversion. . . . ix Page 26 37 HO INTRODUCTION Insect behavior has been the subject of considerable research in the last fifteen to twenty years. Many aspects of insect behavior, including aggregation, mating, and trail marking, have been found to be influenced or regulated by chemicals secreted by the insect. One of the most intensely investigated of these behavioral as- pects, because of its implications for pest control, is the control of reproduction by sex attractants, called pheromones. These are substances secreted by the adult insect, often the female, for the purpose of attracting a mate of the correct species from a distance. At close range, other compounds, known as excitants, are used by some species to stimulate mating behavior. Because the pheromone system of a given insect is normally species specific, such systems have very real potential for use as population control agents which might augment or re- place conventional pesticides in certain cases. Conse— quently, much attention has been turned toward the isola- tion, identification, and synthesis of the sex attractants for various insect species, particularly those which are considered economic or environmental pests. The isolation and identification of the sex attractant for any particular species of insect poses a challenge due to the minute amounts of material secreted by a single insect. In addition, for some species the active attrac— tant is not a single compound, but is rather a mixture of several components in a precise ratio to one another. A change in that ratio might in fact produce the attractant for another species. With the advent of modern instrumen- tal methods for performing separations and analyses on small quantities of material, the task has become some— what easier, and the pheromone systems for a number of insects have been identified.l’2 Many of the compounds which were first identified as insect pheromones were straight chain olefins or dienes with an alcohol or acetate functionality.3 These were relatively easy to synthesize for positive verification of their activity by field testing. However, an increas- ingly large number of attractants are being identified which are more complex molecules, often optically active, and which show much less attractancy in their racemic forms.2 Insect species which make use of chiral pheromone components include a number of economically significant pests, such as the bark beetles of the family Scolytidae,“ the dermestid beetles of the genus Trogoderm (which infest 6 stored grain),5 the boll weevil, and the gypsy moth.7 The compounds serving as pheromones for some of these Table 1. Some Optically Active Insect Pheromone Components. Structure Insect Ref. H CH3 m Coleoptera: Scolytidae A ips paraconfusus a“ cg 3*Oh ips paraconfusus A Mac /CH2 C H3 «(I Hi \ /' ips paraconfusus A Coleoptera: Dermestidae fifi’fifi;\V”\f=\/\v/\/o\/** Trogoderma inclusum 5 do, Him 56“ m": Trogaderma inclusum 5 W/** Coleoptera: Curculionidae (a; Anthonomus grandis um\ (Boll weevil) 6 a» H q; Lepidoptera Lymantria dispar 16a (Gypsy moth) species are shown in Table I. The gypsy moth (Porthetria dispar), in its larval form, is a serious defoliator of shade and forest trees in the 8 northeastern part of the United States. It became es- tablished there after being imported from abroad for the purpose of silk production. A steady migration of the gypsy moth is occurring toward the south and west, and it is now considered a serious economic pest. A single caterpillar can consume one square foot of leaf surface per day. In most cases, a hardwood tree cannot survive more than two successive defoliations.8 The damage is done entirely by the larvae; the adults do not feed after emergence from pupation. An adult female gypsy moth is incapable of flight, being egg laden, and instead relies upon the release of a pheromone to attract a male for mating, after which she lives only a short time. A male may mate several times, but also enjoys only a brief lifespan as an adult.8 Numerous control measures have been used against the gypsy moth, especially since 1958 when the use of DDT was phased out. Several species of predators have been imported to prey upon the insect in its various develop- mental stages; viral and bacterial diseases have been investigated, and the insecticide Carbaryl (Sevin) can be employed.8 The use of insecticides against the moth has met with opposition from environmentalists, and at best has only served to minimize damage and slow the spread of the pest. This spread has been accelerated recently by the increasingly large numbers of campers whose vehicles serve as carriers for the egg masses of the gypsy moth. It was hoped that the identification and synthesis of the sex attractant might prove useful in controlling the populations of this pest. Prior to the identification of the active attractant, crude benzene extracts of the abdominal sections of female moths were used as bait in traps designed to monitor popu- lations. In 1961, Beroza and coworkers reported that they had identified the pheromone as lO-acetoxy-9-hexadecen-l- 01, which they called gyptol,9 1. A homologue of gyptol, 0A HOW/1W 9tho| l l2-acetoxy-9-octadecen-l-ol, called gyplure, 2, was also )1 (3 ”c,I’A\xI’fl\\v’/F‘xz’flP\Jfl==‘V/’L\\/”\\v/’~‘\z/’ gypluro 2 ~ reported to show activity as an attractant.10 However, reinvestigation showed that the attractancy stemmed not from either of these compounds, but from contamination by trace amounts of a substance with "extraordinarily high biological activity."11 Field use of the benzene extracts was resumed until in 1969 the actual pheromone was isolated from 78,000 virgin female moths and identified as gi§r7,8- epoxy-2-methyl-octadecane.7 This compound, christened disparlure, 9, was synthesized in its racemic form by disparlure 3 N epoxidation of the corresponding olefin, (Z)-2-methyl—7- octadecene, 6, as shown in Figure l, and was found to be ..; / 3': ,. peroxide 1) Pha P 2) g— Buli 3) undcconol di‘wflu'. ASE, WNW 3 __ N 85$ cis 6 ~ Figure l. The initial synthesis of racemic disparlure. attractive to male gypsy moths.7 Epoxide derived from the trans olefin proved to be inactive in field testing, neither attracting males itself, nor interfering with the attractancy of the gig olefin derived epoxide.7 Since then, several other groups have devised syntheses 12’1“ The synthetic route of Bestmann of racemic disparlure. and Vostrowsky12 differed from the original route only in that the 6-methylheptyl bromide, 5, required for the Wittig reaction was prepared by hydrogen bromide treatment of the alcohol formed by the reaction of isoamyl magnesium bromide with oxetane. Other more novel routes have utilized olefin metathesis reactions,3 double Kolbe electrolysis,13 and 1” for the synthesis of the a-silyl alkyl lithium reagents olefin precursor to disparlure. In 197A, Marumo and coworkers published the first synthesis of optically active disparlure.15 This synthesis, which makes use of S-(+)-glutamic acid, Z, for the control of the stereochemistry at C-7 of disparlure, is shown in Figure 2. This route was not stereospecific, and produced 7R,8S(+)-disparlure, 1%, which was contaminated with 5.8% of its enantiomer. However, upon field testing, this material showed more activity than racemic disparlure, whereas 7S,8R-(-)—disparlure (prepared by essentially the same route) showed much less attractancy than racemic material. These results led to the postulation of the existence of a chiral receptor site in the insect, and .OLSHLQOHoIA+V mo mfimocuczw o>HuOOHomoopoum ummfim on& .m opswfim .Hmfipmuws mcfippmom Song 0 OHLpoEEzmm wepocoo* m. 95.59:. 3 3 .3 t 0/ Sr :0 2 N05%.: a mega :2: #2 C 2:0 1(208 w. $3 A. 35020331 + O vxnuz IO :onxu :9. 2mm. 2 N —. n‘w // *2, 01 a FF 9: AN ’20: .o .3 A. I Omxh h :«8 uuo: AI. 4| /\J\ a Sex :2 suggested that the natural pheromone is 7R,8S (+)-disparlure. 16 supported The field tests conducted by Cardé and coworkers such a proposal. Mori and coworkers have also carried out a synthesis of (+)-disparlure.17 This route, which made use of the asymmetric centers of 2S,3S (+)-tartaric acid, IA, to give the C-7 and C-8 asymmetric centers in (+)-disparlure, is shown in Figure 3. The precursor of the (+)-disparlure produced by this route was shown by proton nmr studies to have an enantiomeric purity in excess of 98%, and the final product also performed well in field tests.17 Once the potent attractancy of (+)-disparlure was demonstrated, it was seen that the continued extensive field testing of the pheromone as a possible means of population control would require more than the gram quantities of material readily produced by Mori's synthesis. That route, due to its length and complexity,was not judged amenable to commercial development and scale-up. For this reason, we decided to devise a totally new stereospecific synthesis of (+)-disparlure; one which, with only minor modifications, could be adapted for commercial use. Sev- eral ideas were kept in mind during the design and develOp- ment of our synthesis, including the cost and safety of the reagents employed, the possibilities for recycling of reagents or byproducts, and the adaptability of the route to the production of analogues or homologues which 10 .opsapmomeIA+v no mfimocpczm o>HuomHomOOLOpm magma: < .m opswflm .HmHLOme mcfiupwum Some 0 oagumEEzmm monocoax a. 05.59:. T. zonxu 2 :9. 23. ow to: a: an I ’42.: 4 853;" I I‘ll u\ 1.: _ o: 2 A z :37... 2 inf... 3.8-..“ a. ’0: 960...: RAW/4.. :33 a I O x $10 A— :é \sssz/A \\\\\>\/A «.096 rue mp tr 3 ooosz ”zone «to :«8 .28 z full Al. : §S\/\/A ‘m .auuflfl. n: :0 a : gouzouzu :uou 11 might prove to be biologically active. Of particular interest were the preparations of (-)—disparlure, and the episulfide analogues of both (+) and (—)-disparlure. After the publication of our synthesis of El§f(+)‘ disparlure,18 described in the Results and Discussion, Pirkle and coworkers19 reported a synthesis of (+)-dis- parlure which was similar to our own in some respects. Since several of the intermediates in that route are the same as those which we employed, this alternative route will be discussed in conjunction with ours. Pirkle's synthesis, while philosophically different from ours, appears to be an alternative, viable route to fairly large quantities of (+)-disparlure, although it has not been used as such yet. RESULTS AND DISCUSSION Synthetic Strategy Two general approaches to the problem of producing an optically active epoxide with very high Optical purity were considered by our research group. The first was the direct asymmetric epoxidation of the corresponding gi§_olefin, 6. As evidenced by the number of approaches to racemic disparlure which utilize various olefin syntheses, (Z)-2-methyl-7-octadecene was readily available, making such a route very attractive. However, although asymmetric epoxidations have been investigated,20 the major drawback to the methods devised is the relatively low degree of optical purity obtained, which is generally less than 50% enanthiomeric excess. Furthermore, the best results are obtained in cases where the side chains of the olefin to be epoxidized are markedly different, for example, in cases of allylic alcohols. In the case of the olefinic precursor to disparlure, the great similarity of the alkyl groups would require a highly discriminating chiral epoxidation reagent. One method which would bear further investigation, however, is based on the work of Panunzi and Paiaro,21 l2 13 and involves the complexation of a chiral platinum (II) reagent to the gi§_olefin, with subsequent nucleophilic attack by an epoxidizing reagent. The second approach, and the one to which our group turned its attention, was the nucleophilic displacement of a leaving group to form the epoxide in the last step of a synthetic sequence. This was the method of epoxide forma- tion chosen by both Muramo15 and Mori,l7 both of whose routes employed a-hydroxy tosylates which were treated with. base to yield disparlure. The consideration of other potentialleaving groups for that reaction led us to the work of Johnson and co- workers22 which provided the real inspiration for our synthesis. In this work, optically active styrene oxide, 25, was produced from a B-hydroxy sulfoxide, 21, as shown in Figure A. Such a synthesis, making use of the chirality of a sulfoxide to control stereochemistry at the a carbon, followed by reduction to the sulfide and use of the sul- fonium ion as the leaving group for the epoxidation, served as an intriguing model for a synthesis of (+)— disparlure. Such a synthesis, diagrammed in Figure 5, would differ from those previously described in that neither of the chiral centers of the epoxide would be present in the starting material, making this a true asymmetric syn- thesis. Optically active sulfoxides are readily prepared23 1A .oofixomasm O>Huom kHHmOdeo cm Eonm . oofixodo o>Hpom >HHwOHon cm no mfiwozpczm OHLuoEEmmw :4 .: ogsmfim mu 5. it: o omsa oz «zoanu Mm 0 cu m :1 st I Iwfl ‘ I ‘99?— I _3nlc In IUIU==~ AT", SCI: .Im :UIU’ . I ._¢U . :o :o z 2 G acts, NW ‘5.— PN oxu ._ a .. - A n 1 Life 10.. All ._ are I! m ‘8... 02.32 «4 .. _ e 5.5.... _ .. o :o z . o 15 . ”N 0 CH ‘ {@5020 2% C "3_.@._?_ OH 26 3) NC! 17 ~ ”$002 WVBY 12) (I) ' >~fiv~xwhm o 3 m CH3 é‘ocio'flo "’ so rfluli {ms-O 9 09-80“ +5. Hum _ - + H0 m "‘ -‘v/\/AI/ 'mHleV~/\/\z "m 9.3 ’ =2:- + t Hun . pro- trons Hun Pro-cis H 935 Hmu NOAI[OCH2CH20CH3]2H2 H nu Houm THF #511": 34a 35 ~ ‘)(CH3)3 OBF‘ 2)NoOH/CH2CI2 imzo ‘— \\I O ’1 x )8 E H (+) 'dispotluro Figure 5. The synthesis of 7R,88 (+)-disparlure as de- veloped by the Farnum group. 16 by Grignard reactions on 1—menthyl-pftoluenesulfinate, 50. In this reaction the l-menthol is recovered, thus re- cycling the reagent used to generate the optical activity for the entire sequence. The alkyl group of the sulfox- ide and the aldehyde used in the condensation could be varied at will, thus enabling the same general reaction sequence to be used for the production of various homo- logues, and for the production of 7S,8R (-)-disparlure, Al, as shown in Figure 6. A further opportunity to make this sequence economi- cally efficient is afforded at the B-hydroxysulfoxide stage. After the separation of the diastereomeric hydroxy- sulfoxides, 53a and 2QR’ leading to gi§_and trans epoxide, it was hoped that the pro—trans compound, éAb, could be converted into the more useful pro-ci§_hydroxysulfoxide, 53a, thus improving the useful yield of this reaction. The work on the synthesis of (+)-disparlure was done by several members of our research group, whose special contributions are cited in the appropriate sections of the following Discussion. After the viability of the approach was demonstrated, first by model studies conducted by T. Reitz and A. Cardé and then by the initial preparation of a small amount of (+)-disparlure, my work was focused upon several specific aspects of the synthesis. These included the development and scale-up of the synthesis of the early precursors, the proposed interconversion of l7 .OLSHLmomHoIAIV mw.mn mo mfimocucmm ommoeoma one .c opzmfih Z 3 23.59% I: \I\I\ .3... 3S gadolzuoa: :o 1&4 V/\/\_:__ I W W». oco.coa\:OoZ AN 1 go «$qu 2 )\( + 0G” 2 I I :0 cc .2721: :22 UV 1...: I in u I z z ”E Audi .w/m All Sam-‘8 )x _ a nu _ a n A - /\/\/\/\/\/Mk .mui/\/\/\/\/\/\ on 18 "pro—cis" and "pro-trans" hydroxysulfoxides mentioned above, and the preparation of analogues, such as the cis-(-)- epoxide and the episulfides previously mentioned. Synthesis of £-menthyl:p-toluenesulfinate (éQ) Since large quantities of A-menthyl—prtoluenesulfinate, ég, would be required for the preparation of significant quantities of (+)-disparlure, or any of its analogues desired, attention was focused upon the optimization of the conditions for its synthesis. The first reported synthesis of A-menthyl-p-toluene- sulfinate was by Phillips?” and is shown in Figure 7. The reduction of pftoluenesulfonyl chloride, 26, was carried out on a large scale as described in Organic Syntheses.25 The resulting p—toluenesulfinic acid sodium salt was converted to the corresponding acid with hydro- chloric acid, then dried under vacuum and with protection from light. Reaction of the acid with thionyl chloride afforded pftoluenesulfinyl chloride, 28, which was used without purification. Attempts to distill the acid chloride gave extensive decomposition. Initially the menthyl ester was prepared from ethyl- petoluenesulfinate, 29, as in the Phillips procedure?” because it was thought that if the ethyl ester could be purified by distillation, the menthyl ester derived from l9 . . Kw .mum:finasmoco3H0plm|H%cpcoEIAav no mfimocpc>m mawaaficm one 5 mp: we (.59 mm m c. a O o Em... C 32....-. a :ouzonrug m Mm 8 «8: :2 All, 0 A44] 0 O IOIWQ .UI OZOIm\.\ Om OZ .Ulw. . .+ u .. mw o o 20 it would be easier to crystallize. As it happened, this proved unnecessary since direct reaction of A-menthol with freshly prepared pftoluenesulfinyl chloride afforded z-menthyl—petoluenesulfinate, 50, which crystallized upon standing, or in rare cases, upon scratching or seeding.26 This material was recrystallized to constant specific rota- tion, first from acetone; then from an acetone/water mix- ture. The mother liquor from the first recrystallization could be epimerized to a new equilibrium mixture of di- astereomers for further recovery of the crystalline ester by simply allowing it to stand for several weeks. Al- ternatively, the procedure of Herbrandson and Dickenson27 could be employed for the epimerization, using tetraethyl- ammonium chloride and hydrogen chloride in nitrobenzene. This was found to be much less convenient than the first method, and gave no better recovery of the desired di- astereomer. Without the additional material provided by such epimerization, the overall yield of this sequence was approximately 25% when the ethyl ester was employed, and ap- proximately A5% when the menthyl ester was prepared directly from pftOluenesulfinyl chloride. 21 Preparation of_p-tolyleundecylsulfoxide ($7) and,p-tolyl-6-methylheptylsulfoxide (52) The formation of optically active sulfoxides by Grig- nard reaction of alkylmagnesium halides on optically active sulfinate esters was first described by Andersoan3whO observed that the reaction appeared to give inversion of configuration about the sulfur. This was in agreement with the earlier reporg8 that sulfinate esters undergo nucleo- philic substitution on sulfur with such inversion of con- figuration. An elegant study by Mislow and co-workers;26b established the absolute configurations of l-menthyljp- tolylsulfinate and the sulfoxides derived from it, thereby showing that the Grignard reaction does in fact proceed with inversion of configuration at the sulfur. By the use of undecyl magnesium bromide, éé, and 6- methylheptyl magnesium bromide, él, the tolyl sulfoxide precursors for (-) and (+)-disparlure were prepared, as shown in Figure 7. In each case, the menthol by-product was recovered by distillation or sublimation prior to the purification of the sulfoxide. It was established that column chromatography on alumina did not induce racemiza- tion of the sulfoxides, and this was the method used for their purification. 22 oufe >Nvevn~rNhBr :>\~/\~/~:R‘N' _ I "S\o 3) 5\\ + I montho Tel/99 :O‘\’ 32 \w l ——’ 3&5- NWWBr NW \ + l-monthol 37 Tbl ~ Figure 8. The stereochemistry of the formation of the -tolyla1kylsulfoxide precursors to (+) and -)-disparlure. o .XCQSS o R _ 'S'\“\.° I) '- Buli/efher/‘78 ”mg-R ol g-Bu R is undecyl“ or 6-methylheptyl Figure 9. The stereochemical course of the conversion of Retolylalkylsulfoxides to tfbutylalkylsulfoxides. 23 Preparation of t-butylundecylsulfoxide(38) and t-butyl-6-methylheptylsulfoxide(@gl Since it was observed by Gilman and Webb29that the metalation of alkylarylsulfides takes place both a to the sulfide and on the aromatic nucleus, it was thought best to avoid such a side reaction on our sulfoxides by replac- ing the aryl group by another alkyl substituent. However, it was noted by Johnson and co-workerSyDthat, whereas optically active alkyl-arylsulfoxides were readily produced by the method of Andersonffisin the case of unsymmetrical dialkylsulfoxides the reaction was much less satisfactory. So it was decided to employ the method described by JohnsonNDfor the displacement of aryl groups from di- aryl- or alkyl-arylsulfoxides by alkyl lithium reagents. When an excess of tfbutyl lithium was added to a solution of petolylundecylsulfoxide or pgtolyl-6-methylheptylsul- foxide at -78°C, the corresponding tfbutylalkylsulfoxide was formed in excellent yield. This reaction was shown30 to proceed with inversion of configuration at the sul- foxide, as diagrammed in Figure 8. Purification of these trbutylalkylsulfoxides was attempted, but both distilla- tion and chromatography led to extensive decomposition. 2A The Preparation of the Diastereomers of 2-methyle8-hydroxyoctadecan:7:yl—t-butylsulfoxide (55a andgb) and 2-methyl-7-hydroxyoctadecan-8-yl-t- butylsulfoxide (59a and b) The reaction of undecanal31 with the anion generated by treatment of t—butyl—6-methylheptylsulfoxide, 3%, with nebutyl lithium at -78°C resulted, after chromatography (silica gel/ether), in two diastereomeric B-hydroxysul- foxides, 53a and b, as shown in Figure 5. The compound eluting first, éAa, when carried through the remainder of the synthesis, afforded gi§7(+)-disparlure. The later material, égb, when carried through the same sequence, gave the (+)‘EEQEE diastereomer of disparlure. These hydroxysulfoxides could be further purified by medium pressure chromatography, removing about 2% other impuri- ties. However, the disparlure produced from this highly purified material was indistinguishable in physical properties and biological activity from disparlure pro- duced from the less pure precursor.18 Similarly, treatment of the a—sulfoxo anion of £- butylundecylsulfoxide $8 with 6-methyl-heptanal, A2, gave a mixture of diastereomers, one of which led to gi§é(-)- disparlure, 39a, the other, 828’ resulted once again in the production of transe(+)—7,8-epoxy-2-methyloctadecane. Since it was by the reaction of the appropriate aldehyde with the anion generated by nebutyl lithium treatment of 25 the trbutylalkylsulfoxides that the configuration at C—7 of (+)-disparlure or C-8 of (-)-dispar1ure was established, it is important to consider the stereochemical course of the reaction. That configuration is determined by the con- formation of the anion as it reacts with the aldehyde. As the reaction with the electrophile begins to take place, the anion begins to attain a tetrahedral geometry. The pos- sible tetrahedral configurations of the anion are shown in Figure 10, as Newman projections viewed by looking down the C-S bond axis from carbon to sulfur. Conformation la of the configuration l_would be the most stable, since the two bulky alkyl groups are opposite one another and the next largest group, the oxygen of the sulfoxide, is between the hydrogen and the alkyl group. This conforma- tion would probably have the greatest population. If the anion were quenched with D20, the D+ abstraction is rapid, and the product is the one resulting from reaction with the predominant conformation of the anion.12 In configuration 2, the carbon center has been inverted. Conformation ga_of this configuration is the one which allows the least hindrance to the attack of the anion on a carbonyl or alkyl halide. Since that is the slow, rate determining step in the alkylation of the anion, this most reactive conformation would be the one leading to product. Thus the product from the reaction of the anion with an aldehyde has the opposite configuration from that resulting (D Figure 10. Possible conformations of the two configurations of the a—sulfoxoanion of a tfbutylalkylsulfoxide 27 from quenching with D20;32 The reaction, although very stereoselective at carbon a to the sulfoxide, gives a mixture of diastereomers at the 6 carbon. The BB and 8S diastereomers are both formed, and must be separated by column chromatography. In the early stages of the development of this reac— tion, A5% pro-gi§_hydroxysulfoxide was produced, as well 18 However, these yields could not be as 30% pro-trans. achieved with consistency, and the more usual ratio was actually 35% pro-gi§_: A5% pro-trans. This ratio was somewhat sensitive to reaction conditions, especially the temperature at which the reaction was quenched. Work done by H. Brown in our laboratory showed that the addition of aqueous ammonium chloride at -78°C instead of at 0°C did increase the amount of pro-gig compound from 30% to 35%. 13 was that during the An explanation proposed for this warming from —78°C to 0°C prior to the quench, a retro- aldol might occur followed by a recondensation whose stereochemical course differed from that of the condensa- tion which had taken place at —78°C. The production of A5% "pro-trans" hydroxysulfoxide in this reaction was discouraging, since the (+)-trans diastereomer of disparlure appears to be biologically inactive.7 Attention was therefore turned toward the possibility of converting the "pro-trans" hydroxysulfoxide into the much more important "pro-cis" material. 28 One avenue for this interconversion which was explored was oxidation of the "pro:§§§n§" B-hydroxysulfoxide to a B-ketosulfoxide, followed by a stereoselective reduction to give a new mixture of diastereomeric hydroxysulfoxides. Since the approach of a reducing agent from the least hindered side of the keto-sulfoxide would produce the desired hydroxysulfoxide, there seemed to be a good chance of achieving a favorable ratio of "pro-cis" : "prostrans". Unfortunately, the alcohol proved very resistant to oxida- tion under conditions which would not induce recemization of the asymmetric center a to the sulfoxide. Pyridinium chlorochromate3u treatment of the "proftrans" hydroxy- sulfoxide did effect some oxidation, but the yields were very poor. Since an alternative approach was available, this approach was shelved. The method chosen involved the nucleophilic displace- ment of the mesylate derived from "pro-trans" hydroxy- sulfoxide. Corey, 33 al.35 developed a procedure whereby the inexpensive, commercially available potassium super- oxide (K02) could be used as a nucleophilic reagent in organic solvents. By the use of 18-crown—6, the K02 could be made soluble enough in such solvents as dimethyl- sulfoxide, dimethylformamide, or dimethoxyethane to func- tion as a very reactive and effective oxygen nucleophile. It can be used to convert bromides, mesylates and tosylates into the corresponding alcohols in 50-95% yield. For 29 example, the 15R prostenoid (A3) (see Figure 11) was l)McsCl ——-> mxoz lO-crown-b Figure 11. The inversion of configuration of a hydroxyl- bearing carbon via nucleophilic displacement by superoxide ion. epimerized35 to the 158 prostenoid (AA) in 75% yield by treatment of the mesylate of (3Ab) with four equivalents of K02 and A.5 equivalents of l8-crown-6 in a 1:1:1 mix- ture of DMSO, DMF, and DME. Previous attempts to carry out this inversion with other nucleophiles had been un- successful. It was noted by Corey35 that while the use of equi- molar amounts of K02 and l8-crown-6 gave the shortest reaction times, the reaction could be carried out with catalytic amounts of crown. This seemed imperative in our situation, considering the cost of l8-crown-6 and the fact that a large excess of K02 was usually employed. If this were to be a viable method for accomplishing our interconversion, it must be economically, as well as chem- ically, feasible. We found that the reaction proceeded successfully, 30 by use of four equivalents of KO.2 and 0.25 equivalent of 18-crown-6. The superoxide was added as a solid, over a period of 10-25 minutes, at 0°C to the "pro-trans" hydroxy- sulfoxide in a solution of 1:1:1 DMSO, DMF, and DME. The reaction requires more time to go to completion when catalytic amounts of crown are used, so the mixture was allowed to stir at least overnight at room temperature, after the addition of K02 was complete. Care must be exercised during the addition, because the reaction is potentially very exothermic. In the case of the reaction with 6-methylheptylbromide in DMSO, there was a short (10 minute) induction period when the reaction was done on a relatively large scale (18 g K02), after which the reaction became violent. Unfortunately, the extremely slow addition of KO2 (for example one equivalent at a time over a period of several hours), resulted in poor yields of hydroxysulfoxide, even though t.l.c. showed that the starting material had been consumed. This could have been due to the formation of hydroperoxides instead of the desired displacement to form alcohol. An alternative method for converting the "pro-trans" hydroxysulfoxide into useful material would bear further investigation. This procedure involves formation of the inverted acetate from the hydroxysulfoxide,36 fellowed by reduction, which is the usual next step in the synthesis. 31 The acetate would be reduced along with the sulfoxide, resulting in a mixture of B—hydroxysulfide diastereomers. These could probably be separated during the same chroma— tography which is required for the purification of the hydroxysulfide. The Preparation of 7S,8S-2-methyl-8-hydroxyoctadecane- 7-yl-t-butylsulfide (35) and 7S,8S-2-methyl- 7-hydroxy—8-yl-t—butylsulfide (Ag) The reduction of the B-hydroxysulfoxides to B—hydroxy- sulfides was carried out in the early stages of our work by a procedure developed by Dr. T. Veysoglu.l8 This method involved the treatment of the hydroxysulfoxide with stannous chloride-acetyl chloride37 at 0°C in a mixture of pyridine, acetonitrile, and dimethylformamide to give the correspond- ing acetoxysulfide, which was directly cleaved by lithium aluminum hydride to give an 85% yield (after chromatography on silica gel/benzene) of pure hydroxysulfide. This reaction was carried out on a fairly small scale, and proved unwieldly when the problem of scale-up was attacked by H. Brown.33 For that reason, a new method was developed, which was a modification of the procedure of Ho and Wang,38 employing sodiumbi§(methoxyethoxy)aluminum- hydride (Vite, Red-Al). The addition of the reducing agent to the sulfoxide solution at room temperature produced some 32 elimination product, but this problem was overcome by lower- ing the temperature during addition to -78°C, followed by warming to 68°C. The result of this change in reagent was a slight decrease in yield (from 86% to 81%), but greatly improved handling of large reaction mixtures. Other reagents which might also effect this transforma- tion, and should be considered in making future modifications to the synthesis, are sodium cyanoborohydride in conjunc- tion with a catalytic amount of l8—crown—6, as employed by Durst t al.39 and the bromo- or iodotrimethylsilane re— agents developed by Olah gt 1.“0 At this point, it seems appropriate to discuss the synthesis of (+)-disparlure by Pirkle and coworkers19 which was mentioned in the Introduction, since the key intermediate in that synthesis is a B-hydroxysulfide very similar to our own. In this sequence, diagrammed in Figure 12, the hydroxysulfide is prepared in racemic form, and the R-l-(l—naphthyl)ethylisocynate-derived carbamates are used to effect a resolution, by the separation of the diastereomeric carbamates on a high pressure liquid chromatograph. This is, of course, the basic philosophical difference between the Pirkle synthesis and our own; no asymmetric induction is employed, but rather the racemic material is derivatized and the diastereomers separated. This method leaves the undesired carbamate diastereomer as wasted 33 ..I II .3qu E 83.8%? A: co 2852.6 .2 333.1. o H pm m... ..( I 2 S. 9353.1 A+v 1!! 1002 AN :0 . u same: «30.92 c I .092 Am «05: o 3500333? 20502235 Hx {a o > QUEER—.0. . .i 6 Au Zinc: :u qfiuruv :uumroV o.oc>u0m_ 3:3 Orton: I o I — I w A— _ u. mzoI III/Hz JW/“s £9615 5809233 1 + 2 m... we LI‘II/\/vL/AIII saw/girl. All :o/\/\/L/ pim— /0 mzoofiuzuvx 3A material, which is a problem which would be magnified if the sequence were carried out on a large scale. The ex— pense of the optically active isocyanate might also be prohibitive. However, the procedure could be used to advantage on our hydroxysulfide intermediate, for pro- ducing ultra-high purity disparlure in small quantities for testing, or as a standard. Preparation of 7R,8S (+)-disparlure and 7S,8R (-)-dispar1ure The conversion of the B-hydroxysulfide precursors into 18 by (+) and (—)-disparlure were originally carried out a procedure developed by Dr. T. Vesoglu, involving alkyla- tion with trimethyloxoniumfluoborate in CH2C12/CH3NO3 followed by epoxidation with NaOH in CH2C12/H2O. The reaction time and temperature for the alkylation were very critical, and the procedure required the rapid removal of solvent under vacuum at 0°C prior to the addition of base. This problem and the fact that the long (lO-l2 h) reaction time at 0°C for the epoxidation step was both in- convenient and often led to the formation of elimination products, resulted in a reinvestigation of this procedure by H. Brown,33 R. Muenchausen, and Dr. M. Lipton. After such modifications as the use of 100% CH2Cl2 as the solvent for the alkylation, and changing the base from NaOH to NaH were attempted without success, a procedure was 35 developed by Lipton which resulted in much shorter reac- tion times and consistently higher yields. In this method the hydroxysulfide was treated with Meerwein salt at 0°C in CH2C1 then warmed to room tempera- 2, ture for l h, after which time the solvent was removed under vacuum at room temperature. The residual crude sulfonium salt was taken up in pentane, and treated with 0.5DflNaOH, at 0°C. The mixture, after stirring for A.5 h. at 0°C, was warmed to room temperature before work—up. The method gave yields in the range of A5-55% after distil- lation of the product. Preparation of 7§i8R,cis-7,8-epithio-2- methyloctadecane and 7R,88,cis-718- epithio-2-methyloctadecane It was shown by Cardé and coworkersul in 1972 that the presence of the olefinic precursor to racemic disparlure, pipg2-methyl-7-octadecane, effectively inhibited the attractancy of disparlure to male moths. The discovery of this inhibitory phenomenon, which had been previously observed with other insect species, led to the synthesis and field testing of a number of compounds related to disparlure to determine whether they might also function ”2 Some of these compounds as antagonists or attractants. were found to be somewhat attractive, though none was as good an attractant as racemic disparlure, which seemed to 36 indicate that the moths' receptors were not entirely specific for disparlure. One of the compounds tested was gig: 7,8-epithio-2-methyloctadecane. This compound showed practically no attractancy in its racemic form; however, considering the fact that racemic disparlure is much less attractive than (+)-disparlure, it was thought that perhaps the optically active episulfide of the same configuration as (+)-disparlure might prove attractive. In addition, since (-)-disparlure is rather less attractive than racemic material, perhaps its episulfide analogue, being struc- turally very similar, might function as an antagonist. Accordingly, we decided to prepare these optically active compounds for electroantennogram and field testing. Beroza and coworkers!42 prepared the episulfide used in their tests from racemic disparlure by reaction with thiourea.”3 Unfortunately, the reaction gave a yield of only about 5%. Clearly, we could not afford such a low yield if precious (+) and (-)-disparlure were to be used in the episulfide preparations. The chemistry of thiiranes, including their synthesis from epoxides, has been reviewed by Fokin and Kolomiets.uu Such a conversion offered an excellent way to produce the episulfides stereospecifically, since the mechanism of the reaction of epoxides with such reagents as potassium thiocyanate or thiourea had been well studied. The mechanism A5 proposed by Ettlinger, which is shown in Figure 13, was 37 In a 8 .‘Unv .1 my .l H... pa .moofixoao spas coflpowOL zomx pom Emflcmcooz — 2o». «a 32 + o .IITIII I .. .a .:H mpswfim .1 38 rigorously demonstrated by van Tamelenu6 to be correct. The reaction begins with the nucleophilic opening of the epoxide by the thiocyanate anion, followed by the iso- merization of the resulting alkoxide anion, via a cyclic intermediate, into a thiolate anion. This is then converted into the thiirane with displacement of the cyanate anion. By this mechanism, inversion should occur at both carbons of the original epoxide ring. A similar mechanism was proposed by Culvenoru7 for the reaction of epoxides with thiourea. These mechanisms were further confirmed by studies of the reactions of KSCN and thiourea with optically active 2-phenyloxirane, 2-methyl—3-phenyloxirane, and 2,3- d1phenyloxirane.u8 The conversion of these compounds to the corresponding thiiranes resulted in the inversion of the configuration of the epoxide at both asymmetric centers. This could only be explained by the cyclic intermediate or transition state already proposed. The conditions for these reactions appeared to be critical in a number of cases, with yields varying from 30-80%, depending upon solvent, pH, and temperature. After trying numerous combinations of those conditions in reactions of racemic disparlure with KSCN or thiourea, it was evident that no significant improvement in yield ”2 would be achieved with these re- over that of Beroza agents. An alternative method of carrying out this conversion 39 was devised by Chan and Finkenbine,“9 which makes use of triphenylphosphine sulfide. This method gave good yields of several thiiranes not readily produced by the KSCN or thiourea procedures. However, the mechanism of this reaction involves a trigonal bipyramidal intermediate or transition state, which could undergo pseudorotation. Thus the stereochemistry of the reaction has not yet been conclusively established. For this reason, we bypassed that reagent in favor of theprocedure developed by Calo and coworkers50 in which 3- methylbenzothiazole-2-thione in the presence of trifluoro- acetic acid was used to effect the conversion of oxiranes to thiiranes. This procedure was reported to be success- ful in several cases where the conventional reagents had failed, giving high yields of thiiranes. Of importance to us was the fact that the reaction could be explained mechanistically in a manner similar to that demonstrated to be the case for the thiourea and KSCN reactions. This mechanism is outlined in Figure 1A, and led us to believe that reaction would probably be stereospecific. The reaction of racemic disparlure with 3-methylbenzo- thiazole-2-thione and an equivalent of trifluoroacetic acid gave, after purification by column chromatography and preparative t.l.c. (silica gel, toluene), yields of around 35-A0% episulfide. The reaction time was somewhat longer than that reported by Calo,5O but the progress of the A0 IMO”!:‘\ 5 I JG (3 + l H >5 + “>136” ——’ Ran 32 s R R4 3‘ (DH 2 Ir. if ‘9‘:>=° ” 2513c“: + IH~fl“. V Figure 1A. Proposed mechanism for epoxide to eniSilfide interconversion. A1 reaction could be followed with t.l.c. The same procedure was used to convert small quantities of (+) and (-)-dis- parlure to their episulfide analogues for testing, which is incomplete. Determination of Enantiomeric Purity of Disparlure Precursors Because the two alkyl chains are very similar, the optical rotation of pure (+)-disparlure is very small: [djg5 = + 0.A8° (CClu), [aJSS = + 0.23l° (neat). Thus, optical rotation is not a very accurate means for de- termination of the enantiomeric purity of disparlure. Our group, as well as every other group which has devised a synthesis of optically active disparlure, was faced with the difficulty of finding a method for making such a determination. Chiral solvating agents such as those 51 extensively studied by Pirkle do not complex with epoxides well enough to produce significant chemical shift dif- ferences between enantiotopic protons in the n.m.r. spectrum. Unfortunately, the enantiomers of epoxides equipped with no distinguishing characteristics except two slightly dif- ferent alkyl side chains are not distinguished by chiral shift reagents either. We ruled out procedures requiring degradation of the epoxide because any method chosen would need to be applied to the naturally occurring material A2 for comparison, and the difficulty and expense involved in procuring that material in any quantity precludes its use for that purpose. Thus our group, as did Mori and Pirkle, came to rely upon the determination of the enantiomeric purity of the immediate precursor to the epoxide, in our case the B- hydroxy-Erbutylsulfides, 35 and Ag. Knowing the optical purity of that precursor, we must rely upon the stereo- specific nature of the final epoxidation, which has been demonstrated, and handle the material with sufficient care to prevent accidental racemization during purifica- tion. The pfbutyl group proved to be a useful handle for the determination of the enantiomeric purity of the hydroxy- sulfide by NMR making use of the chiral shift reagent Eri§(3-heptafluorobutyryl-d—camphorato)europium III, Eu(hfbc)3. The Rebutylhydroxysulfoxides could also be studied by this method, but the shifts induced in the sulfides were of greater magnitude, and the sulfide was the immediate precursor to the final epoxide. In studies conducted by A. Cardé on a Brucker 180 WH instrument, it was found that by the use of approximately 25 mg Eu(hfbc)3 on a 20 mg sample of hydroxysulfide in d6- benzene, chemical shift differences of approximately 20 Hz between enantiotopic pybutyl protons could be obtained. The hydroxysulfide precursor to (+)-disparlure was found ”3 by this method to be contaminated with A.5-5.5% of its enantiomer. In the case of the precursor to (-)-dispar- lure, the hydroxysulfide was found to be 96.6% optically pure. EXPERIMENTAL General All melting points were determined with a Thomas-Hoover apparatus and are uncorrected; boiling points are also un- corrected. Reagents and solvents were reagent grade, and used as received except as noted in individual cases. Routine proton magnetic resonance spectra were obtained on a Varian T-60 instrument. Infrared spectra were measured on a Perkin-Elmer Model 137 Spectrophotometer; liquids were examined as neat films, and solids as Nujol mulls. Gas chromatographic separations were routinely achieved using an F & M Model 700 Laboratory Chromatograph equipped with a thermal conductivity detector. The standard column used was 10 foot 30% SE-30 on Chromasorb W. Other columns and instruments are noted in the individual procedures. Mass spectra were run on an Hitachi RMU—6 instrument with an ionizing voltage of 70 eV. Preparation ofgp-toluenesulfinic acid (27) To a A L beaker containing 2.5 L H20 was added 600 g (u.76 mol.) Na2803 and A20 g (5.00 mol.) NaHCO3. The mix- ture was heated with stirring to 70-80°C, and maintained AA A5 at this temperature by a regulator on the oil bath or hot plate. To this was added, in 10 g portions over a period of 3 h, A8A g (2.5A mol.) pftoluenesulfonyl chloride. After addition was completed, the mixture was allowed to come to room temperature and stand for a period of 12 h, at which point the beaker was full of white fluffy crystals of the sodium salt of pftoluenesulfinic acid. These crystals were collected and dissolved in hot water, with stirring (70°C). To this solution was added an equivalent of concentrated HCl. After the mixture had cooled enough to handle, the crystals were collected, and the mother liquor evapor- ated to give additional crops. The combined crops were dried for several days ip’vaggp over Drierite, followed by P205, to give a yield of 80-85% for the two steps. This material should be used as quickly as possible after drying is complete, due to its tendency to decom— pose, especially in the presence of light. Preparation of petoluenesulfinyl chloride (28) Freshly distilled (from triphenyl phosphite) SOCl2 (120 mL, 1.65 mol.) and 120 mL dry ether were placed in a 3 neck flask equipped with mechanical stirring and N2 flow. Dry petoluenesulfinic acid, 27, (23A g, 1.5 mol.) was added as a solid, over a period of about 10 minutes. Foaming and gas evolution occurred during the addition and determined the addition rate. Stirring was continued A6 until no further gas evolution was observed (2-3 h). The system was warmed slightly (30-A0°C) toward the end of this period. Excess SOCl2 and ether were then removed under vacuum at room temperature, with care taken to prevent moisture from reaching the acid chloride during this process. The resulting crude petoluenesulfinyl chloride was not purified further, due to its tendency to decompose during distillation, and the esterification re- action was carried out immediately. Preparation of etpyl p-toluenesulfinate (29) Absolute ethanol (86 mL, 1.5 mol.), dry pyridine (133 mL, 1.65 mol.) and 250 mL anhydrous ether were placed in a 1 L, 3 neck flask equipped with mechanical stirring, condenser, addition funnel, and N2 flow. The mixture was cooled to 0°C, and maintained at that temperature throughout the reaction. The acid chloride, 28, prepared in the previous reaction was added dropwise at a rate which kept the tem- perature at 0°C. Stirring was continued at 0°C for 1 h after the addition was complete; then the mixture was poured into water and the phases separated. The aqueous phase was extracted twice with ether. The combined ether extracts were washed twice with H2O, twice with 5% HCl, and once with saturated NaCl solution, then dried over MgSOu. The solvent was removed on the rotary evaporator, and the residue distilled under vacuum. A small forerun u? (30-50°CA3J&mIHg) preceded the ester (95-100°C/0.1nm1Hg). Care had to be taken to assure that the pot temperature did not exceed 120°C, since decomposition was observed at higher temperatures. 160 g of ethyl p—toluenesulfinate, 29, was obtained after distillation, for a yield of 58% over the two steps from the sulfinic acid. 1H NMR (CDC13): 60.90 (3H, t, J=3Hz); 2.0 (3H, s) 3.5 (2H, complex m), 7.15 (AH, dd, J=9, J'-A Hz). Preparation of A-menthyl-p—toluenesulfinate (30) From (29) Ethylep-toluenesulfinate, 29, (100 g,o.5A3 mol.) and l-menthol (8A.7 g,0.5A3 mole) were stirred together at 60°C under partial vacuum (15 mm Hg) for A8 hr. The mixture was then allowed to cool at atmospheric pressure. The stirring was then stopped and the mixture allowed to stand overnight, after which crystallization had occurred. In other runs, if such spontaneous crystallization did not occur, it could be induced by seeding or scratching. The crystalline mass was taken up in a minimum of acetone and recrystallized to yield menthyl—petoluenesulfinate (71.8 g, A5%) as white needles, m.p. 106-107°c. This material was then recrystallized 7-8 times from a 17:3 mixture of acetone/H2O, until a constant specific rotation of [a]g3 = -199.7°i0.3° was obtained in acetone solution. NMR (CDC13) 60.70-1.02 (19H, m); 2.A0 (3H, s); 7.20-7.75 (AH, m); IR (CClu): u6.25, 8.8, 11.7. H8 Preparation of 6-methy1heptyl-p—tolylsulfoxide (3g) Diastereomerically pure grmenthy1-pgtoluenesulfinate, 30, (70.0 g, 0.238 mol.) was dissolved in 625 mL dry ether in a 2 L three—necked round bottom flask equipped with a stirrer, an addition funnel and N2 inlet-outlet tubes. A solution of 6-methylhepty1magnesium bromide, prepared by refluxing 6-methylheptyl bromide (H5.9 g; 0.238 mol.) and magnesium (6.53 g; 0.2“7 mol.) in dry ether for one hour, was transferred via a cannula to the addition funnel. The Grignard reagent was added dropwise over a period of 50-60 minutes to the stirred reaction mixture. The stirring was continued for an additional three hours. The reaction mixture was then quenched with 300 mL saturated ammonium chloride solution. The layers were separated, and the ethereal layer was washed with water. The aqueous por- tions were combined and, in turn, extracted with ether. The ethereal portions were combined, dried over Na2sou, and concentrated. The crude product was then subjected to sublimation (50°C at(JJ2mm Hg) until only a very small amount of menthol was detected in the residue by t.l.c. (silica gel/CHCl3). Finally, removal of the unreacted starting material from the residue by column chromatography (Woelm alumina; l:l hexane/ether, followed by 100% ether) gave the desired sulfoxide in 65% yield. (Slight contamina- tion with menthol could be tolerated in the next step.) “9 1H NMR: 50.80 (6H, d, J=3 Hz) 1.0-1.75, (9H, m); 2.32 (3H, s); 2.65 (2H, t, J=3 Hz); 7.25 (uH, dd, J=8, J'-u Hz). Preparation of t-butyl-6-methylheptylsu1foxide (33) prolyl-6-methylheptylsulfoxide (20.1 g; 0.08 mol.) was dissolved in 2 L of dry ether and added to a flame dried 3 L three-necked flask, equipped with a mechanical stirrer, rubber septum, and N2 flow. The system was cooled to -78°C, and 177 mL tfbutyllithium (1.8 M, 0.32 mol.) was added, slowly at first, to the solution by cannula. The mixture was allowed to stir an additional three hours at -78°C after addition was complete, after which the cold bath was removed and distilled water (300 mL) was added slowly to quench the reaction mixture. Upon warming, the layers were separated, the ether phase extracted with water (150 mL), and the combined aqueous layers extracted with ether (3 x 100 mL). The ether extracts were combined, dried with Na2S0u, concentrated, and dried (0.1 mm Hg for 6 h) to give 98-105% crude tfbutylsulfoxide, which was used in the next step without further purification, due to its instability. 50 Preparation of 7S,8S-2-methyl-8-hydroxyoctadecane:17y1- t-butylsulfoxide(éfla, cis Precursor) and its,JS,8R Diastereomer(3£b,gtrans Precursor) To a flame dried 500 mL three necked flask equipped with a mechanical stirrer, rubber septum, and 125 mL addition funnel with a nitrogen outlet tube, was added crude tfbutyl-6-methylheptylsulfoxide (9.“7 g; 0.0H3H mol.) in 130 mL dry ether. The system was then cooled to -78°C. 23.9 mL nebutyllithium (2.0 M in hexane, 10% excess) was added to the solution. The generated anion (color change from light yellow to dark yellow or light brown) was allowed to stir for 95-60 minutes, after which freshly distilled (5H°C, 0.13 mm Hg) undecanal (8.86 g; 0.521 mol.) was added. The reaction mixture was allowed to stir for an additional N5-60 minutes, after which 100 mL cold saturated ammonium chloride solution was added drop- wise from the addition funnel. After addition was complete, the ice bath was removed. Upon warming to 0—10°C, the phases were separated, the aqueous phase reextracted with ether (2 x 100 mL), and the combined ether extracts dried with Na280u. Concentration of the extracts gave 32.3 g crude hydroxysulfoxide. Separation by column chromatography (silica gel/ether) yielded 10.8 g (35%) of "pro-gig" hydroxysulfoxide, 3Qa,and 13.9 g (“5%) of the "pro-trans" l diastereomer, 33R. For the 7S,8S pro-cis isomer: H NMR: (001“): 60.85 (9H, m); 1.1-1.5 (36H, m); 2.6 (1H, m); 51 3.7 (1H, m); “.1 (1H, br, s) IR (neat film): u3.0, 3.5, 5.8 (w), 6.9, 7.u, 8.6, 8.9, 9.3, 10.0. For the 7S,8R, "pro-Egang" hydroxysulfoxide: 1H NMR (CClu): 60.85 (9H, m); 1.1-1.8 (36H, m); 3.25 (1H, m); 3.u (1H, br s); u.0 (1H, m); IR (neat film): u2.9, 3.4, 5.75 (w), 6.8, 7.3, 7.6, 8.5, 8.8, 9.9, 13.8. Preparation of 7S,8S-2-methyl-8-hydroxyoctadecane-Z-yl- t-butylsulfide (35) To a 500 mL round bottom flask, equipped with a West condenser topped by an addition funnel, was added "pro- gisfl hydroxysulfoxide (20.0 g; 51.“ mmol.) in 150 mL dry tetrahydrofuran. The system was cooled to -78°C, and sodium bis(methoxyethoxy)aluminum hydride (No.0 mL, 80% solution in benzene) in 100 mL dry THF was added dropwise over a period of two hours. The system was allowed to warm to room temperature over a period of one—half hour, and then warmed to 68°C for at least eight hours. The reaction mix- ture was allowed to cool, and then slowly added to a mixture of 500 mL 1.0 N HCl and 300 mL benzene in a 2 L separatory funnel. The aqueous layer and resulting salts were ex- tracted with additional benzene (3 x 300 mL), the combined extracts dried with MgSOu, concentrated, and purified by solumn chromatography (silica gel/benzene) to give 15.5 g (81%) of hydroxysulfide which is normally at least 95% 52 pure by t.l.c. 1H NMR (001“): 60.8 (m, 9H); 1.0-1.6 (m, 3uH); 2.2—2.u (m, 2H); 3.1-3.2 (m, 2H). Preparation of (+)-Disparlure (13) Original procedure: To a 500 mL round bottom flask was added 7S,8S-methyl-8-hydroxyoctadecan-7—yl-tfbutyl- sulfide (10.0 g; 28.6 mmol.) and 100 mL of solvent (CH2C12: CH3NO3; 1:1) under nitrogen. After cooling to 0°C, tri- methyloxonium tetrafluoroborate (7.9“ g; 53.7 mmol.) was added, and the system allowed to stir vigorously for 1 hour and 10 minutes. The solvent was then removed by distillation, with the temperature not rising above 25°C. The total reaction time should not exceed 2 h. CH2C12 (200 mL) and