MSU LIBRARIES n. RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. ‘FINES wi11 be charged if book is returned after the date stamped below. - ‘ .. -,l . I ' .I ‘0‘ cl Al ‘— |’ 71' THE DEVELOPIENT OF HETEROATOI-SUBSTITUTED FURYL SYNTHONS‘ IN A GENERAL IETHODOLOGY TOWARD THE SYNTHESIS OF TERPENOIDS I. THE PREPARATION OF (i)-LACTARAL II. SYNTHESIS AND REACTIONS OF HETEROATOI-SUBSTITUTED FURYLIETHYL ORGANOIETALLICS III. AN APPROACH TO THE SYNTHESIS OF (t)-APHIDICOLIN By David Brent Head A DISSERTATION Submitted to Iichigan State University in partial fulfillment of the requirements for the degree of ‘ DOCTOR OF PHILOSOPHY Department of Chemistry 1985 ABSTRACT THE DEVELOPIENT OF HETEROATOI-SUBSTITUTED FURYL SYNTHONS IN A GENERAL IETHODOLOGY TO'ARD THE SYNTHESIS OF TERPENOIDS I. THE PREPARATION OF (i)-LACTARAL II. SYNTHESIS AND REACTIONS OF HETEROATOI-SUBSTITUTEI FURYLIETHYL ORGANOIETALLICS III. AN APPROACH To 1111: SYNTHESIS or (fl-APHIDICOLIN By David Brent Head Many syntheses of bio-active natural products contain- ing five-membered oxygen-substituted heterocycles have been approached by construction of a parent carbocycle to which the heteroaromatic ring is then appended. This study is directed toward the deve10pment of a general methodology in which the heterocyclic component is incorporated as an integral part of the molecule. An additional consideration is that access to all possible mono- or di-substitution patterns of furan-, butenolide-, butyrolactone- and tetra- hydrofuran-containing structures (Figure 1) must be possible from common intermediates. We envisioned employing highly functionalized furyl synthons as the operational equivalents of furan, butenolide, butyrolactone and tetrahydrofuran mono- and di-anions in alkylative and annulative processes, respectively. To demon- strate the viability of such an approach, the Grignard reagent of 3-chloromethyl-4-tetrahydropyranyloxymethylfuran 24 was produced and coupled with chloride 26 in the total synthesis of the mushroom metabolite (i)-lactara1 15. A variety of alpha heteroatom substituted, beta furylmethyl organometallic reagents (Figure 6) were investigated in a previously estab- lished alkylation-cyclization sequence (Figure 4). The furyl substituents were selected to regiospecifically direct the cyclization sequence as well as to facilitate the place- ment of oxygen on the five-membered heterocycle in conversions to other oxidation states (Figure 5). These groups included -OTMS, -OTBDMS, -SME, -SPh and -TMS. Precursors to the -SMe (46), -SPh (54) and -TMS (64, 72) substituted organo- metallic reagents were successfully produced and alkylated; however, none of these groups proved to be useful in the annulation sequences studied. However, the -TMS variant provided an efficient means for the production of various 3- and 4-alkyl-2(5H)-furanones (Figure 9). .A synthetic approach 1x3 the antiviral substance (i)- aphidicolin was also undertaken. Furyl-Grignard coupling product 91 was converted to the tricyclic aphidicolin precur- sor 92 via a Lewis acid catalyzed biomimetic cyclization. Completion of a formal total synthesis of (t)-aphidicolin requires the conversion of 92 to the McMurry intermediate cyclopentenone. For my parents, Harvey and Naomi Head, whose love and support made this endeavor possible. 11 ACKNO'LEDGEIENTS The author wishes to express his sincere appreciation to Dr. Steven P. Tanis for his guidance, encouragement and advice throughout the course of this work. The unique learning Opportunity which was provided is also appreciated. Thanks are extended to my colleagues, the faculty and staff of the Michigan State University Chemistry Depart- ment who all contributed to a memorable experience. Financial support from Michigan State University in the form of a graduate assistantship is gratefully acknow- ledged. iii TABLE OF CONTENTS LIST OF FIGURES LIST OF EQUATIONS INTRODUCTION. . . . . I. THE PREPARATION OF (i)-LACTARAL . II. SYNTHESIS AND REACTIONS OF HETEROATOM- SUBSTITUTED FURYLMETHYL ORGANOMETALLICS A. Alkoxy- and Siloxy-substituted Furans B. Sulfur-substituted Furans C. Silicon-substituted Furans. III. AN APPROACH TO THE SYNTHESIS OF (i)-APHIDICOLIN . . . . EXPERIMENTAL. General LIST OF REFERENCES. iv 12 20 21 31 45 60 72 72 120 V r-m——’ Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10 11 LIST OF FIGURES Oxidation states 7-10 and substitution patterns A-C of five-membered oxygen- containing heterocyclic ring systems. Alpha-position carbomethoxy blocked, beta furyl cyclization. . . . . . Conversion of furans 7A to .7B and 7C. 3-Fury1methy1 Grignard reagent formation, alkylation and cyclization sequence X-Group blocked in alkylation, 3-furylmethyl cyclization sequence synthons Heteroatom containing X-Group substituted 3-furylmethyl organometallic reagents Proposed cyclization of 3-alkyl-5-alkoxy substituted furans. . . . . . . . . . Proposed synthetic pathway to alpha-siloxy, beta-hydroxymethyl furans 2-Trimethylsilyl-3- and 4-alky1 furans and their 2(5H)-furanone oxidation products Proposed retrosynthetic scheme to (t)- aphidicolin Furan terminated, epoxide initiated olefin cyclization . . . . . . . . . . . . . 10 20 21 26 59 61 63 10. 11. 12. 12a. 13. 14. 15. LIST OF EQUATIONS Retrosynthetic disconnection of (t)-lactaral. Model study of Grignard coupling in (+)- lactaral synthesis. . . Preparation of 3-chloromethy1-4-tetrahydro- pyranyl-oxymethyl-furan 24 and conversion to Grignard reagent 25. Preparation of (t)-1actaral 15 from Grignard reagent 25. Attempted O-alkylation of 2-(5H)-furanone Preparation of 2- trimethylsilyl- 3, 5- di- methyl- furan 34 . . . Reaction of 34 with an electrophile Attempted 0-silylation of 4-bromomethy1- 2(5H)-furanone. . . . . . . . . . . Attempted stannyl-lithium displacement of 4-bromoethy1-2(5H)-furanone. Preparation of 3-bromo-2(5H)-furanone 36. Attempted 0-silylation of 36. Preparation of 2-trimethylsiloxy-S-bromo- furan 37 from 36. . . . . . . . . . . . Preparation of 2-tbuty1dimethylsiloxy-3- bromo-furan 38 from 36. . . . Benzaldehyde addition product 39 of lithiated 38. . . . . . Attempted stannylation of lithiated 38. vi 16 17 22 24 24 25 26 28 28 28 28 29 30 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33/34. Preparation of 2-nbutyl-furan 40. Preparation of 2- methylthio- 5- nbutyl- furan 41 from 40 . . . . . . . . . . . . . . . Attempted hydrolysis of 41. Production of ethoxylated butyrolactone 42 in hydrolysis of 41 . . . . . . . . Attempted hydrolysis of 43 (sulfoxide of 41). Preparation of alpha-methylthio-beta- hydroxymethyl-furans 44 and 45. Preparation of 2-methylthio-3-chloromethy1- furan 46 from 44. . . . . . . . . . Alkylation product 47 from Grignard reagent of 46 . . . . . . . . . . .,. Stannyl lithium displacement of chloride 46 'providing stannane 49 . 50. Benzaldehyde addition product 51 of lithiated 49. . . . . . . . . . . . Attempted alkylation of lithio-cuprate of 49. Preparation of alpha-phenylthio-beta- hydroxymethyl-furans 52 and 53. Preparation of 2-phenylthio-3-chloromethy1— furan 54 from 52. . . . . . . . . . . . Preparation of 2-pheny1thio-3- tri- nbutyl- stannylmethyl- furan 55. . . . . . Benzaldehyde addition products 56 and 57 of lithiated 55 . . . . . . . . Preparation of 2-phenylthio-3-decy1- furan 58 from 55. . . . . . . . Preparation of 2,6-dimethy1-9-(2-pheny1- thio-3-furyl)-2,6-nonadiene 59 from 55. vii Attempted alkylation of lithiated 49 providing 32 32 33 33 33 34 35 35 36 37 37 38 38 39 39 4O 41 41 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Preparation of 3- tetrahydropyranyloxy- methyl- furan 60 . . Sulfenylation of 60 . Sulfenylation of 2- -trimethylsilyl- 3- hydroxy- methyl- furan 63 . . Preparation of 2-phenylthio-4-chloromethyl- furan 62 from 53. . . . . . . . . . . . . Attempted stannylation of 62. 2-Silylation of 3-furan-carboxylic acid Reduction of 2-trimethylsilyl-3-furan- carboxylic acid to alcohol 63 . . Preparation of 2-trimethylsilyl-3-chloro- methyl-furan 64 from 63 . . . . . . . . Preparation of 2, 6- dimethyl- 9- (2— —trimethy1- silyl- 3- furyl)- nona-2, 6- ~diene- 2, 3- epoxide 65 from 64.. Cyclization of 65; Preparation of 66. Silylation of 3- hydroxymethyl furan; Prep- aration of 2, 5-bis-trimethylsily1-3- hydroxymethyl- furan 67. . . . . . Silylation of 3- trimethylsiloxy furan; Preparation of 2- trimethylsi1yl-4-hydroxy- methyl- furan 68 . . . . Silylation of 3-tbutyldimethylsiloxy-furan. Preparation of 2-pheny1thio-3-tbuty1dimethyl- siloxy- furan 69 from 52 . . Silylation of 69; Preparation of 2-pheny1- thio-5- trimethylsilyl- 3- tbutyldimethylsiloxy- furan 70.. . Desilylation of 70; Preparation of 2- phenylthio-5-trimethylsi1yl-3-hydroxymethy1- furan 71. . . . . Desulfurization of 71; Preparation of 2- phenylthio-S- trimethylsilyl- 3- hydroxymethyl- furan 71. . . Silylation and reduction of 2-bromo-4-furan carboxylic acid; Preparation of 68. viii 42 42 43 44 45 46 47 47 47 48 49 50 51 52 52 52 53 54 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. Preparation of 2-trimethylsi1yl-4-chloro- methyl-furan 72 from 68 . . . . . . . . . Preparation of 2, 6- dimethyl- 9- (2- -trimethyl- silyl- -4- furyl)- -nona-2, 6- -diene- 2, 3- epoxide 73 from 72. . . . . . Preparation of 2-trimethylsily1-3-decy1- furan 74 from 64 and subsequent oxidation to furanones 80 and 81. Preparation of 2-trimethylsilyl-4-decyl- furan 77 from 72 and subsequent oxidation to furanone 86. Preparation of 6, 7- epoxy- 8- -hydroxygeranyl- acetate 89. . . . . . . Preparation of 6,7-epoxy-8-tbutyldimethyl- siloxygeranyl chloride 90 from 89 . . Coupling of 90 with 3-furylmethyl magnesium chloride; Preparation of 1- tbutyldimethyl- siloxymethy1-6-methy1- 9- (3-fury1)- nona-2,6- diene- 2, 3- -epoxide 91. . . . . . Cyclization of 91; Preparation of 92. Conversion of 92 to ketone 93 . Conversion of 93 to diol 94 . Conversion of 94 to acetonide 95. Conversion of 94 to bis-MEM ether 96. ix 55 55 57 58 64 64 65 66 67 67 68 70 I NTRODUCT I ON INTRODUCTION In recent years, a major emphasis of synthetic organic chemistry has been placed on the synthesis of naturally occurring compounds, particularly those with unusual or medicinally useful biological activities. These activities include antibiotic, antifungal, antileukemic, cytotoxic and antitumor properties. The diversity of the structural features that are associated with these compounds. is fascinating' as 'well as bewildering; however, upon closer inspection, a number of recurring themes appear. Many biologically active terpenoids incorporate five-membered oxygen-containing heterocyclic rings within their structures.1 This feature is an. integral part of such molecules as the insect antifeedant ent-neo-clerodane ajugarin I 1,2 the antileukemic pseudoguaianolide rudmollin 2,3 the witchweed germination promotor strigol 3,4 the fish antifeedant nakafuran-B 4,5 and the cytotoxic conferti- folin 5,6 which might also be a synthetic intermediate in the synthesis of the cytotoxic and insect antifeedant drimane sesquiterpene warburganal 6.7 i g E Terpenoids 1-5 exhibit two of the four common oxidation States of the five-membered oxygen containing heterocyclic system. These range from the fully aromatic furan 7 to tetrahydrofuran 10 (Figure 1). The terpenoids 1-5 also display the three (A-C) substitution patterns commonly observed about this ring system. AJugarin I 1.1ncorporates the 3-substituted substructure A, while rudmollin 2, strigol 3, and nakafuran-8 4 provide examples of the 2,3-fused substructure B and confertifolin 5, the 3,4-fused substructure C. [1) O 2 i3, 2 19 o .6. E. Figure 1. Oxidation states,Z;19 and substitution patterns ,A;§,of five-membered oxygen containing hetero- cyclic ring systems. Structural classification of these terpenoids according to the ring systems displayed in Figure 1 suggests that a synthetic, methodology centered around the incorporation of the heterocyclic component into a cyclic or acyclic framework would be a powerful and versatile technique in the synthesis of various terpenoids. The relative similarities of 5 and the non-heterocyclic 6 also suggests that this methodology could be applied to the synthesis of non-heterocycle containing terpenoids as well. As a result of the interesting biological activities associated with compounds such as 1-5, this class of molecules has attracted, considerable attention as targets for total synthesis. Most of the reported syntheses of such compounds have centered upon the preparation of a parent carbocycle upon which the five-membered hetero- cycle is then appended. These schemes have generally not considered the utilization of the five-membered ring nucleus as a unit upon which the remainder of the molecule might be assembled. A truly general approach to the synthesis of molecules 1-5 should provide access to the various states of oxidation (7-10) as well as the different patterns of substitution (A-C) about the heterocyclic nucleus. A synthetic program which has the potential to meet these requirements has been under investigation in our laboratories. This strategy revolves around the preparation of ring systems A-C from common intermediates. If one considers that A-C, in the various oxidation states shown by 7-10, can be obtained from furanoid precursors via standard oxidation or reduction techniquesll, a common intermediate becomes evident. Thus, the fully aromatic furan 7 can function as the precursor to oxidation states 8-10. This suggests that a. common furanoid intermediate might provide access to all structural combinations implied by Figure 1 if substitution patterns A-C could also be readily established. . It is well known that furan suffers electrophilic aromatic substitution at unsubstituted positions alpha to the ring oxygen.10 Therefore, this predisposition can be utilized to introduce various substitution patterns around a furyl nucleus in an annulation sequence. This could be accomplished if, for example, a simple 3-substituted furan (7A, Figure 1) contains a benign electrophilic center in its side chain. Unmasking of the electron-poor center and subsequent electrophilic substitution would provide 78 (Figure 1). In order to prepare the'substitution pattern represented in C (Figure l) by this protocol, the favored cyclization at the alpha-position of 7A must be blocked. Therefore, a functional group X must be introduced at the alpha-furyl position which will prevent the electronically favored closure, directing ring formation to the unsubstituted beta-position. Oishi12 has examined the furan-terminated cyclization of such an X-group blocked substrate and has observed the formation of type 7C (Figure 1) products as shown in Figure 2. Unfortunately, the difficulties associated with the manipulation of the a-COZMe group render this approach less useful. Figure 2. Alpha-position carbomethoxy blocked, beta furyl cyclization. The analysis presented above suggests that the more complicated B and C substitution patterns (Figure 1) should be readily constructed from the' much less complex 3-substituted furans 7A as shown in Figure 3. ./\ O ‘? ,«wr”””’lld" F! 7;! Z/\S ' O a R /\ 4/\ 0 x0 79 Figure 3. Conversion of furans Z5 to lg and 1g. 7 We have previously described a simple route for the preparation of 3-substituted furans8 as well as their cyclizations9 to compounds of type 7B (Figure 1) as illustrated in Figure 4. The readily available isoprenoid Cl _'__Mq_> Cit R. MCPBA 0% R [6“ 2 RX 0" fihfaa—‘q 99:); I 3 LECUQ: .. R3 . (CW’nl/OLRZ . . 43f . “mm; @131. g. A..." R; l . Figure 4. 3-Fury1methyl Grignard reagent formation. alkylation and cyclization. 3-chloromethyl furan is smoothly converted to the corresponding‘ Grignard reagent which, in the jpresence of LiZCuCl4,42 readily couples with a wide variety of primary and secondary alkyl halides as well as allylic halides. In this process, remote electrophilic centers, such as epoxides can be incorporated without destruction. Alternatively, as shown in Figure 4, these centers can be introduced after the Grignard coupling step. 8 With an efficient route to a variety of 3-substituted furans available, the cyclization sequences (Figure 4) leading to bicyclic products were examined.9 In the initial studies, the epoxide function was selected as the initiator for the ring forming sequence. This selection was based upon the ease of epoxide preparation and the relative mildness of the Lewis acidic conditions previously utilized in polyolefin cyclizations. However, the success of these studies was complicated by the poorly nucleophilic character of the furyl terminator coupled with the acid lability of the starting materials and products. These problems were manifested in poor material balance or in the production of unwanted, uncyclized epoxide-opening products. To address these concerns, several Lewis acids ‘were investigated to determine their ability to promote epoxy-furan cyclization. The choice of Lewis acids selected was dictated by the following factors: (1) the ability to readily modify the potency of a group of Lewis acids with a common metal center; (ii) the possibility of moderating the Bronsted acidity of the medium through the choice of Lewis acid, i.e. adventitious protic acid might be scavenged by a Lewis acid possessing a carbon-metal bond releasing an alkane; and (iii) moderation of the Lewis acid-product alcohol complex acidity. Although considerable success was enjoyed in these cyclization studies, the reactions tended to be quite substrate dependent. Exact experimental conditions often 9 had to be worked out carefully with regard to substrate functionality and Lewis acid. These studies have shown, however, that a variety of fused, spirocyclic and bridged furan-containing molecules can be readily obtained. Although these experimental techniques have proven to be quite useful, they cannot provide access to all of the substructures of Figure 1. As shown in Figure 4, only compounds of type 7A and 73 have been produced; however, in principle, oxidation states 8-10 should be accessible from them. In order to produce compounds of the type 70, utilizing the experimental techniques above, it becomes necessary to consider a fundamental modification of the reaction substrates. This modification would consist of the attachment of some group X to the 2-position of furans 7A to direct the cyclization step (Figures 2 and 3). In addition -to cyclization direction, attachment of a group X to a furan 7A would also facilitate the transformation of it into the desired oxidation states 8 and 9. This is particularly appealing in light of the lack of regioselectivity in the oxidation of furans to butenolides. The group X could be converted to oxygen which would regiospecifically produce the desired butenolide 8 which could furthermore be reduced to butyrolactone 9. If 10 ‘was desired, the X-group could simply be removed to provide the parent C followed by reduction. A scheme which illustrates this proposed reaction sequence utilizing X-group appended furyl intermediates 10 is shown in Figure 5. Cases I and II illustrate the two possibilities for 2-position, X-group appended furans of type 7A. In both cases, the reaction sequence is identical to that established previously (Figure 4), i.e., formation of a 3-furylmethyl Grignard reagent followed by alkylation to attach a side chain containing a latent electrophile and cyclization. However, we must now consider the nature and preparation of suitable precursors to X-group appended furylmethyl Grignard reagents. In Case I, the X-group would not exert control over the cyclization but would facilitate conversion of the cyclization products to butenolides of type 8B (Figure 1). In Case II, the X-group serves to block the electronically favored cyclization process, affording the 3,4-disubstitution pattern of furans of type 70. In addition, the X-group can serve to direct the introduction of oxygen in an otherwise relatively symmetrical molecule to afford butenolides of type 8C. - m 1 oxid 4, + Ca; / \ + _ I. /\ Xcoupling + /O\ X mx A 0 a n. /\ )‘<——'—" /\ )-(——" /\ 7 Figure 5. X-Group blocked 3-fury1methy1 synthons in alkylation, cyclization sequence. 11 The goal of the research efforts described herein was to develop new synthetic methods which might allow the transformations illustrated above (Figure 5) to be realized in the laboratory. I. The Preparation of (i)-Lactara113 I. The Preparation of (:)-Lactara113 The Basidiomycotina subdivision of fhngi have provided a myriad of terpenoid metabolites. Notable among these structurally diverse natural products are the hirsutane sesquiterpenoids such as hirsutic acid 11,14 the marasmanes illustrated by marasmic acid 12,15 the fomannosanes such as fomannosin 13,16 the lactaranes represented by lactaroru- fin A 14,17 and the» secolactarane lactaral 15.13a,18 As a result of the unique structures and the promising antibac- terial and antitumor activities exhibited by several com- pounds of these types, there has been considerable effort directed toward their total synthesis. These efforts have culminated in elegant syntheses of 11, 12, 13, and related compounds. However, the lactaranes and seco-lactaranes have not been as thoroughly studied. 12 13 As part of our program directed toward the total synthe- sis of furan containing natural products, the synthesis of (1)-lactaral 15 was undertaken. Central to this program is the development of highly functionalized furans as viable synthons. :Ecau e 0 b e - 2 ____..0 .___.O 0 ~. 1’ ~ 2’ ‘ CHO CHO L?- (1) 14 Initially, attention was focused upon the formation of the C(3)-C(4) bond (lactarane numbering) as the crucial step in the synthesis of (:)-15 (eq. 1). This bond can be constructed in either of two polar senses. Path a uti- lizes an allylic nucleophile while path b employs a benzylic-type nucleophile ix: the bond-forming process. Based upon earlier work by Tanis,8 the reverse polarity (relative to the normal use of substituted furylmethyl derivatives) path b route was chosen as the method of choice. This approach has a distinct regiochemical advantage over the path a construction in which an allylic carbanion is added to an electrophilic furylmethyl residue. Whereas an allyl organometallic may undergo rearrangement, the furylmethyl carbanion approach illustrated in eq. 1 allows the regiochemical integrity of the trisubstituted double bond to be retained. The synthesis of (1)-lactaral 15 via a path b approach requires that the displacement by the furylmethyl carbanion upon the requisite allylic halide proceed in an 8N2 fashion and not SN2'. Previously, the tendency of related furyl organometallics to yield mixtures of products resulting from 8N2 and SNZ' displacement has been 'noted8 when the halogen-bearing carbon and the terminal olefinic carbon are sterically similar. To determine if mixtures of products might be obtained in the preparation of 15, the model system illustrated in eq. 2 was examined. 15 The readily available bromide 17, prepared from the precursor alcoholzo:21 (CBr4, Bu3P)19 and chloride 18 (CCl4, Ph3P),19 were selected as coupling partners for Grignard reagent 16.8 The reaction of Grignard reagent 16 with allylic bromide 17, in the presence of LigCuC14, J“?! "AOLM‘ id. :0. ‘2. ,2, x-Br lg (60%,64) zo L9. mm (47%, nod-0) afforded a 60% yield of a mixture 19 and 20 in a ratio of 60:40 as determined by 250 MHz proton NMR.22a However, the coupling of 16 with the corresponding chloride 18 provided the desired 19 as the sole product in 47% yield. It was gratifying to observe only the desired 8N2 substitution product from the reaction of 16 and 18 although the reasons for the observed selectivity are not immedi- ately obvious.22b 16 \ R2 \ Cl Mg,THF \ MgCl 0::[ _.__.. O 0 1’ R. z’ z’ (3) amp '0. 01' HP .23.. RI‘ R2' “125' 2:, 77% 23 1007. 2‘21 R" R28 CHLOH ( 77%) Q RI-CHZOTHR Reset-50mm.) With this crucial information in hand, a precursor to the nucleophilic furyl residue outlined in path b of eq. 1 was prepared (eq. 3). The formerly inexpensive diethyl-furan-3,4-dicarboxylate 21 was reduced with lithium aluminum hydride to provide alcohol 22 in 77% yield. Protection as a mono-THP ether was accomplished by the procedure of Griec023 affording' a statistical mixture of un-, mono-, and bis-protected diol from which 23 was isolated in 50% yield after chromatography. Numerous attempts to perform the chlorination of 23 by previously used methods24 afforded only trace quantities of 24. However, when the procedure of Meyers25 was employed, chloride 24 was obtained in 77% yield. Initial attempts to convert 24 to Grignard reagent 25 at room temperature provided only low yields of 25 as 17 determined by titration.26 Earlier work27 investigating the formation of Grignard reagents in the presence of acet- alic functions suggested attempting the conversion at a lower temperature. At 10°C (internal), 25 was formed over a period of 2 h in quantitative yield as determined by titration.26 / * / (UHF OR 33 gg 2; R: THP(75°/o) 1§(96°/.) a R = H(IOO°/o) The reaction of 25»‘with chloride 26 (prepared from the corresponding alcohol;180 CC14, Bu3P19) afforded lactarol-THP ether 27 in 75% yield as the sole product. Deprotection23 of 27 gave lactarol 28 in quantitative yield. Alcohol 28 was then oxidized with activated manganese dioxide28 to (:)-lactaral 1518 in 96% yield. Synthetic 18 15 was identical in all respects to a sample of authentic lactaral kindly provided by Professor G. Magnusson. With the successful completion of the synthesis of lactaral accomplished, studies were undertaken toward its cyclization into the lactarane skeletal system (see 14). A cyclization might occur if the cyclopentene double bond could be induced to act in a nucleophilic sense and attack the Lewis acid complexed aldehyde of lactaral or other functional groups such as the corresponding cyclic acetal 30 alcohol 28, or acid chloride 33.29:30 29 R -- o 11’ R=CH20CHO a3 Racoon 93 R=COCl Initially, the direct cyclization of lactaral 15 was examined. A variety of Lewis acids and reaction conditions were employed including SnCl4, ZnIz, BF3¢vOEt2. Et2A1Cl, and TFA/TFAA; however, none of these techniques jprovided cyclized product. The cyclization of lactarol 28 was also attempted with trifluoroacetic acid (TFA) and with formic acid. In the latter case, only the formate 31 was isolated. 19 Finally, lactaral 15 was oxidized via the Corey proce- dure (MnOZ, NaCN, MeOH)31 to the corresponding methyl ester, followed by hydrolysis to afford lactaroic acid 32 (72% yield overall). Acid 32 was converted to the corresponding acid chloride 33 (oxalyl chloride, benzene) in 80% yield. Treatment of 33 with aluminum chloride afforded only trace amounts of cyclized products. It is likely that the fragility of the system results in destruction under the relatively harsh conditions required for closure. II. Synthesis and Reactions of Heteroatom-substituted Furylmethyl Organometallics 11. Synthesis and Reactions of Heteroatom-substituted Furylmethyl Organometallics In order to evaluate the strategies discussed in the introduction which were designed to provide access to 3,4- disubstituted furams as well as to control regioselectivity in butenolide preparation, it was necessary to prepare a variety of Xégroup substituted furylmethyl organometallic intermediates. In the course of the study, a number of novel-and previously unknown furyl compounds were prepared. For those cases in which the desired organometallic was obtained, alkylation and. cyclization studies *were carried out. Additionally, the ease with which these products could be converted to the corresponding butenolides was examined. In order to ensure facile conversion of the furyl precursors into their corresponding butenolides, only heteroatom containing X-groups were considered. - + _ + M M (I \§ ,4“ x 0 2,3-substitution 2,4-substitntion x . -0TMS, -0TBDMS, -sue, -SPh, -TMS Figure 6. Heteroatom containing X-group substituted 3-fury1methy1 organometallic reagents. 20 21 As shown in Figure 6, the X-groups investigated were: O-trimethylsilyl; O-t-butyl-dimethylsilyl; methylsulfenyl; phenylsulfenyl; and trimethylsilyl. In all of these cases, the target organometallic was the Grignard reagent due to the convenience of formation and previously demonstra- ted8:9 utility in alkylation sequences. II. A. Alkoxy- and Siloxy-substituted Furans Since butenolide- and butyrolactone-containing molecules are the ultimate targets of this methodology, the first X-groups to be examined were those in which the heteroatom was oxygen, i.e., X = OR where R = alkyl or trialkylsilyl. The ease of hydrolysis of alkoxy and siloxyfurans3232333:b and their reactivity when treated with exogenous electro- philes33b suggested that X = OR, OSiR3 would be an excellent first choice. One additional benefit of this approach is the possible direct production of a butenolide from the cyclization of a 3-alkyl-5-0R-substituted furan (Figure 7) in a manner analogous to the work done with exogenous electrophiles.33 /0\ 9R __ ““ f; I? R = alkyl, trialkylsilyl . . Fi ure 7. . Proposed cyclization of dialkyl-S-alkoxy substituted furans. 22 Initially, X = OR for R = alkyl was considered. Unfor- tunately, only a few members of this class of compounds are known, i.e., X = methoxy,32a t-butoxy,32b and acyloxy.320 Of these compounds, a methoxy-substituted derivative appeared to be the most suitable for this study since it should be less inclined to R-O bond cleavage than the t-butoxy analog; however, the preparation of the umthoxy-substituted derivative is tedious. Therefore, an attempt was made to prepare it via 'the procedure of Kraus32c which was successfully used to prepare 2-acyloxy- and 2-t-butyldimethylsiloxy-furans. As shown in eq. 5, the attempted O-alkylation of the enolate anion of 2-(5H)-furanone with methyl iodide led to a low yield of only double-bond shifted product. More reactive methylating reagents (Me303F4, MeOSOZCF3) were not investigated due to their expense and difficulty in handling. 1) 1m, HMPA, THF [1 -78°C, 10 min / o 0 2) Mel, -78'—-o'c 1 (13% (5) low yield 23 Attention was then directed to the silicon analogs of the alkoxyfurans, which have been more thoroughly inves- tigated. To date, siloxyfurans have been prepared via 0-silylation of the corresponding butenolides with trimethyl- silyl chloride or t-butyldimethylsilyl chloride by two general methods. Kraus320 has accomplished this conversion via the lithium enolate of a butenolide; however, in this study, the milder Silylation conditions of Asaoka,33a Cazeau,34 and Wiesner35 have been found to be more useful. Typically, a weak base such as triethylamine, often in the presence of a coordinating metal cation, was used to catalyze the Silylation, analogous to 'the manner in which silyl enol ethers of ketones have been formed.36 3,5-Dimethyl-2(5H)-furanone37 was converted to the corresponding siloxy derivative, 2-trimethylsiloxy-3,5-di- methylfuran 3433a via the two methods shown33a:34 (eqs. 6 and 7) to afford 14% and 56% yields of 34, respectively. With siloxyfuran 34 in hand, a reaction designed to give an indication of the ability of O-trimethylsilyl to function as a cyclization blocking group was carried out. Compound 34 was subjected to electrophilic attack via the procedure of Asaoka33b to afford a high yield of alkylated product 35 (eq. 8). Although Asaoka33b has observed beta(3)- position electrophilic attack in reactions of alpha(2)-trimethylsiloxy-alpha'(5)-alkyl furans, the analogous reactions of alpha(2)-trimethylsiloxy-beta(3), 24 O __ 1) TMSCl, Et3N, oi / \ (6) 0 2) 25'c, overnight 0 :35 147. TMSCl, Et3N, Nal, CH3CN n ’ ’33 56% (7) 25'c, 2.5 h I): HC(OEt)3, Snob, __ OTMS CH Cl , -40°->10'C, (8) O 21,2 2 (131:0)2 0 H 1}}; 93 937. alpha'(5)-dialkyl furans (such as 34) have not until now been examined. Treatment of compound 34 with an electrophile resulted in exclusive alpha(5)-position attack even though this position was already methyl substituted. Although these results suggest that electrophilic attack at the beta'(4)-position is not electronically favorable, the possibility of intramolecular cyclization to this position is not ruled out. In the event of an intramolecular cycliza- tion, steric constraints would make attack at the remote 25 alpha'-position to form a 3,5-bridged bicycle very unfavor- able. After having examined the properties of simple siloxy- furans, it was necessary to consider the introduction of a functionality which would facilitate the appendage of a side chain. The placement and nature of the reactive site must be reconciled with the chemistry required .to develop the siloxy furans. One of the most direct and obvious solutions to this problem is illustrated in eq. 9. Using the methodology Br ._ msc1 Et3N 4 H (9) 0 0 OSiMeE5 low yield of Asaoka33 to establish a siloxy X-group, the 3-bromo- methyl-butenolide shown34 was successfully silylated. Unfortunately, the only product obtained, which was isolated in low yield, resulted from the expected halide elimination. In an attempt to overcome problems associated with the halide, an effort was made to transform it into the corres- ponding organotin compound via the procedure of Still;38 however, as expected, only decomposition resulted (eq. 10). 26 Br _ Bu3SnLi O O 4 decomposition (10) THF, -22'c The difficulty in the preparation of a halomethylsiloxy-furan presumably arises due to the relative sensitivity of both functionalities as well as the furan itself. An alternative which addresses the problem of functional compatability is illustrated in Figure 8. f' (lg—£03m; E / \ “’5‘”: O 0 0 Case I co,R 00:“ _ HO . 21 g m . =2 hoses ' Figure 8. Proposed synthetic pathway to alpha-siloxy, beta-hydroxymethyl furans. 27 In this scheme, the organometallic precursor is left in the relatively insensitive carboxylic acid oxidation state in order to establish the sensitive siloxy functionality. However, in path I, the final structure is perfectly set up for decomposition and probably would not survive the relatively acidic conditions required to convert it to the corresponding halide. In path. II, the regioisomeric 2-siloxy-4-hydroxymethyl furan has an electronic structure that promises greater stability. However, the lack of availability of the appropriate precursors to these materials prompted us to abandon this approach in favor of the follow- ing one. Utilizing the "umpolung"39 capabilities afforded by organotin compounds, a one carbon extension of a carbanion can be performed via its alkylation with iodomethy-tri-n-butylstannane and subsequent . tin-lithium exchange (nBuLi).38 The two requisite, regioisomeric siloxy-furylmethyl organometallic reagents (see Figure 5) should, in principle, be obtainable by alkylation of the corresponding siloxy-furylanions with Bu3SnCH21. The carbanionic nature of this mode of functional elaboration should be compatable with the siloxy group whereas the carbocationic nature of the halogenation discussed previously is not. To regioselectively produce :1 beta-furylanion, it is necessary to perform a halogen-lithium exchange on the corresponding furyl halide since the most acidic proton 28 (7% 1) Brz, CC14, 2h 0 heat, 3h 1) TMSCI. ECBN. N81 2) pyridine, benzene, Br 29 66% ‘ ‘7 19 —> product too labile (12) CH3CN, 25°C, 15 min for aqueous workup 2) aqueous workup ' ' Br' 1) TMSCI, Et3N, ZnClz THF 19 . ’ > 1/ \S (12a) 25 C, 1.5h 2) distill directly 0 OTMS 22 37% Br 22 TBDMSCI, EtéN, ch12, i m (13) m, zs'c, 4.51: O orepms 38 49% N 29 resides at the alpha position. Although there are far fewer reports of the additions of beta-furylanions to electrophiles than of the alpha counterparts,46 Wiesner has reported the high yield addition of the lithium derivatives of 3-bromo- and 4-bromo-2-trimethylsilyloxyfuran to steroidal ketones.32 These results suggest that the alkylation of these anions with ICHZSnBug would provide access to the desired organometallic reagents of Figure 5. Therefore, it became necessary to prepare the precursor halo-butenolides. A simple procedure40 for the preparation of the 2-bromobutenolide 36 from 2-(5H)-furanone41 is 11- lustrated in eq. 11. Compound 36 was then subjected to the Silylation condi- tions illustrated in eq. 12, 12a, and 13. The trimethylsilyl derivative, 2-trimethylsilyl-3-bromofuran 37 was successfully prepared; however, it was found to 'be extremely labile. The t-butyldimethylsiloxy derivatiVe 38 was successfully prepared (eq. 14) and found to be reasonably stable. With compound 38 in hand, its conversion to the corres- ponding halogen-lithium exchange product and reaction with electrophiles was examined. Treatment of 38 with nBuLi followed by the addition of benzaldehyde provided the addition product 39 in 90% yield. OH 1) nBuLi, Et 0, -78°C ‘ 38 2 —> I \ (14) 2) 0°C, 30 min 0 dreams 3) PhCHO 33907. 30 Several methods were then examined in an attempt to alkylate the anion derived from 38 with ICHZSnBu3. Direct alkylation of the corresponding organolithium was unsuccessful as was alkylation of the Grignard reagent, produced by treatment of the organolithium derivative with MgBrz. Introduction of Copper I and II catlysts42 was also unsuccessful in promoting alkylation as was the mixed cyano-cuprate. A very low yield of the stannylmethylated butenolide was obtained, however, by alkylation of the organolithium reagent in the presence of HMPA, followed by desilylation as shown in eq. 15. It was concluded that the anion of 38 was not sufficiently nucleophilic to react with ICHZSnBu3 in good yield. $0303 1) nBuLi, EtZO, -78’c .— §§ 4)- (15) 2) O'C, 30 min 3) HMPA 4) inverse addition to <5% ICHZSnBug, 25'c 5)1MIE1 Due to our inability to prepare precursors to the siloxy-furylmethyl organometallics in acceptable yield, this approach to butenolide equivalents was abandoned in favor of sulfur-substituted furans as described in Section II .B. 31 II. B. Sulfur-substituted Furans At this point, our attention was directed toward furans substituted at the 2-position with sulfur-containing X-groups, e.g., X = SR, R = alkyl as potential butenolide dianion equivalents. This functional group appeared to have several advantages over the oxygen-containing analog. First, it was assumed that the poorer pi-system overlap of the sulfur containing group would result in lower furyl electron density imparting more stability to the alkylthio- furans than was observed for the labile siloxy and alkoxy-furans. Additionally, since sulfides in general are more easily prepared than ethers, it was assumed that unlike the siloxyfuryl analogs, the appropriate precursors ‘UD the desired sulfenyl furyl organometallic reagents would In; more readily accessible. Another positive aspect concerning the use of sulfur is the variety of oxidation states in which it can exist. This property might be useful in subsequent cyclization studies especially in light of the electron-withdrawing nature of the furyl substituent employed by Oishi in Figure 4. Along with the greater stability of the sulfenylfurans, however, comes the added complication of removing' sulfur with conversion of the furan to the corresponding butenolide. This, of course, should in principle be achievable via the appropriate hydrolytic conditions similar to the hydrolyses of dithiane alkylation products in the well-known 32 aldehyde and ketone synthetic methodologies reported by Corey,43 Seebach,44 Schlessinger,45 and others. Although alkylthio-furans are known compounds,46 there have been few literature reports of their hydrolyses/173vb II. B. 1. lethylsulfenyl as X-group The first sulfur containing X-group to be investigated was methylsulfenyl, X = SMe. Our initial studies, which examined the introduction of -S-CH3 and its removal by hydrolysis, are described below. The volatile furan was first converted to 2-n-buty1- furan 40 (eq. 16) by a modification of the procedures of Chadwick48 and Levine49. In a similar sequence, 40 ‘was metalated and treated with dimethyl disulfide yielding 41. 1) add to nBuLi, TMEDA, ofc fur“ 4 m (16) 2) 25 c, 30 min 0 3) THF, n-butyl- Br, 0-925 C overnight 40 56% 1 add to nBuLi, TMEDA, EtO 0°C (:2 ) “L WSW (17) 2) 25 C, 30 min Me 3) 88MB, 0-v25 C, 2h 3} 73% 33 With methylthio-furan 41 in hand, hydrolysis studies were carried out. A variety of conditions were explored including some of the nfllder variants which have been used successfully for the hydrolysis of various hemithio- and dithio-acetals and. ketals. These include: HgClz, H20;44 HgClz, HCl, H20, Me0H,47b and HgO, HBF4, THF.50 Unfortunately, these attempts provided unreacted starting material. The high stability of the methylthio-furans to hydroly- sis requires the use of a strong acid such as H2804. How- ever, it is necessary to carefully adjust conditions in order to avoid complete decomposition of the furan. These 41 H2504 cone. (2 eq.) ~ I > uniaolable od t H20 (1 '39-), DME. heat, 20h Pr uc (18) 081: 41 H2809 (10%) a (19) ~ EtOH, heat, 12h O {.3 707. {L1, MCPBA, 0112012 m conc. 1101,1312 41 33% (20) 0’0. 3h ‘IS' 0 25‘0, 24h "’ O 5.3 717. 34 conditions were examined as described in eqs. 18 and 19. Although the conditions of eq. 18 did not provide unreacted starting material, we were unable to isolate or purify the product. However, the milder conditions of eq. 19 did afford a good yield of hydrolyzed material 42 as .the ethanol addition product. This observation may explain the enhanced stability of the product of eq. 19 relative to that of eq. 18. In light of the difficulties encountered in these hydrolysis attempts, we next examined the hydrolysis of the derived sulfoxide 43. Using milder hydrolysis conditions than before, 43, unfortunately, gave only a low yield of reduction product. During the course of the hydrolysis studies, a synthetic sequence to the methylsulfenyl-substituted organometallic precursor was begun. As shown in eq. 21, 3-hydroxymethyl- furan was metalated in a fashion similar to that of eqs. 16 and 17 (2 eq. nBuLi). The dianion. was then reacted with dimethyldisulfide to give a mixture of two regioisomeric sulfenylated 3-hydroxymethyl-furans 44 and 45 in 53% and ca. 5% yields, respectively. The major regioisomer was 0“ 1) add to nBuLi (Zeqo). H OH TMEDA(23 0): Et 0’ / \ 2- (7C * ’ \ an O 8149 "93 O 0 2) 25‘0, 1h . . 3) MESSMe 0-e25 C . 45 5% overnight £5 537 " 35 purified and halogenated in the usual way25 to afford 46 in 79% crude yield (eq. 22). Cl MsCl, LiCl, collidine, DMF I \ (22) 35 u- () Eflme £9 79% crude The chloride 46 ‘was not thermally stable and could not be purified; however, the crude material was clean enough to allow the alkylation studies described below to be carried out. When Grignard formation with chloride 46 could not be induced by conventional methods, an "entrainment" tech- (23) 1) M3, BrCHz CH Br,m (entrainment?$ 9.6. lo \ 2)M ./ :sMe/O 01M . , 47 217. LiZCuC14, OC Wurtz 4product. 36 nique was successfully utilized (eq. 23). The Grignard reagent thus prepared 'was coupled with epoxygeranyl chloride51:52 in the presence of LigCuCl4 catalyst42 to provide a mixture of products from which the SNZ' product 47 could be isolated in disappointingly low yield (21%). A 64% yield of the furan dimer 48 was also isolated. It was hoped that the formation of the Grignard by a different method might reduce the amount of dimerization and also produce more of the desired 8N2 displacement product rather than the SN2' product 47 observed. Unfortunately, when the coupling ,was attempted after Grignard formation by the Rieke53 technique, only the furan dimer 48 was observed. It is thought that the dimerization products are produced during the Grignard formation. In an attempt to circumvent this problem, the chloride 46 was converted into the corresponding tri-n-butylstannane 4938 followed by tin-lithium exchange and treatment with magnesium bromide to form the Grignard reagent (eqs. 24 and 25). Although the stannane 49 was not produced in very high yield (31%, eq. 24), a sufficient quantity was obtained to allow the alkylation of the corresponding Grignard reagent with epoxygeranyl chloride51v52 (eq. 25). . SnBu3 46 1) mix Bu3Snl-I, IDA, 0 C, 15 min / \ (24) ”V 2) add to‘$§, THF, -25‘C 3) -25=.25 0, 1h 0 SMe 93 317. 37 1) nBuLi, THF, -78‘c, 30 min \ ‘33 , fi" / \ (25) 2) 0 C, MgBrz, LiZCuC14 O SMe 3) , 01925'C, 1h so 97, I ~ Cl é The alkylation product 50 was produced in quite low yield (9%); however, no dimerized material was produced and no recovered stannane or its proton exchange product were observed. This suggested that the low yield of the reaction was not due to problems in the anion formation or its alkylation but in the stability of the anion itself. To verify this assumption, the stannane 49 was subjected to tin-lithium exchange at low temperature (-78°C) as before, then reacted directly with benzaldehyde at -78°C (eq. 26). A high yield of addition productwas produced suggesting that the anion was stable at low temperature. 1 mm, THF, 48°C, 30 min ' )n , .. / \ on (26) "' .2) PhCHO, m, -78‘--25 0, 2h 0 m 2 80% 38 Since the Cu(II) catalyzed Grignard coupling process, which is the usual method for alkylation of furylmethyl anions,8 does not work well with alkyl halides at low temper- ature, a mixed organo-cuprate was. utilized (eq. 27). Unfortunately, only the furan dimer 48 was again obtained. 1) nBuLi, tar, ~78'c, 30 min 32 . fi—Jb 8 27 2) CuCN, -78 C, 45 min mgjor ( ) product 3) 01M. -7s'c, 30 min Concurrent with these investigations into methylsulfenyl as X-group, a similar effort was directed toward phenylsul- fenyl as X-group. II. B. 2. Phenylsulfenyl as X-group In a reaction sequence identical to that for the methylsulfenyl analog, 3-hydroxymethyl-furan was elaborated into the chloro-precursor of the corresponding phenylsulfenyl 1) add to nBuLi(2eq.), a... TMEDA(2eq.), Et20, O'C H H . . 2) 25'c. 30 min [Kw (“f0 + (28) (if 3) PhSSPh, ’ 4’s /0\ l \ 34) 0 01-725. C , overnight 0 52 (1:4) §§ "’ 50-80% 39 C1 D 52 14301, LiCl, collidine, MF 1- [\i (29) 2§94% organometallic reagent (eqs. 28 and 29). Unlike the methylsulfenyl derivative 46, phenylthio-furylmethyl-chloride 54 is quite stable and can be purified by distillation. Although the Grignard reagent frOm chloride 54 could be directly formed by standard methods, several alkylation attempts provided the same .results .as those for -the methylsulfenyl analog. As with chloride 46, the phenylthio-furylmethyl-chloride 54 was transformed into the corresponding tri-n-butyl stannane 55 as shown in eq. 30. Initial alkylation studies 1 o SnBu3 54 ) mix Bu3SnH, LDA, 0 C, 1331:; m (30) "' 2) add to g9, um, -25 c 0 84> 3) -25=-25 c, 1h 33 667. of the tin-lithiunl exchange product of 55» gave the same results as those observed for the 4O methylthio-furylmethyl-stannane 49. Therefore, a similar low-temperature reaction with benzaldehyde was attempted 1) nBuLi, THF, —78 C, 30min, 55 A (31) ’" 2) PhCHO, THF, -78-025? C, 331nm. 232nm (eq. 31). The anion of 55 was successfully produced .and added at low temperature to PhCHO. However, in this case, along with the expected product 56, an unusual rearranged product 57 was obtained. However, as is shown in eq. 32 and 33, we found that the phenylthio-substituted organolithium reagent could be successfully alkylated by some of the more reactive alkyl halides in the presence of HMPA at temperatures no higher than -25°C. If the temperature was raised above this, decomposition resulted; and if lowered much below -25°C, the rate of alkylation slowed to an unacceptable level. 41 1) nBuLi, THF, -78‘c, 30mi a l \ (32) 2) HMPA <3 5n) 3)‘\~/,\~,/\\’,~\\//o\ , 58 39% I, -25 c, 1h cv 1) nBuLi THF -78‘c- 30 min ‘~ “ a; a a a ’ I \ (33) 2) HMPA C) 5%! 3) “ “ g3 29% c1, -25’c, 1h This process was improved further by adding the preformed-organolithium reagent, via a dry ice Jacketed syringe, to a solution of the alkyl halide in TH]? and HMPA at -25°C (eq. 34) giving 59 in a more acceptable yield (59%). 55 1) nBuLi, THF, -78‘c, 30 min A. —> 34 2) add '53 ( ) to ‘~ ‘~ 1, 59% .HMPA, -25'c 3) -25 c, 1h Upon successful preparation of 2-pheny1thio-3-a1kyl furans, the problem of efficiently constructing an organometallic precursor of the 2,4-substitution pattern 42 was addressed. Both the 2,3-substituted methylthio and phenylthio organometallic precursors were prepared via an oxygen-directed metalation of 3-hydroxymethy1 furan. However, in both cases, small amounts of the 2,4-substituted regioisomers were produced. We, therefore, assumed that blocking of the directing oxygen, for example, as an ether, might lessen the directing effect and result in a greater proportion of the 2,4-substituted regioisomer. Additionally, it was thought that if this ether was particularly large and bulky, a steric preference for metalation leading to the 2,4-substituted regioisomer might be observed. This hypothesis was tested as shown in eq. 35 and 36. PPTS, DHP, CHQCIZ. lr"j§f‘brfir (35) / \ 25'c, 1.5h / \ C) 99957. 1) add to_nBuLi,TMEDA,Etz0,0'C 52 2) 25'C, 30 min 3) PhSSPh, 0’0 70% 4) EtOH, PPTS, heat, 2h (36) 43 In the event, blockage of the hydroxylic oxygen of 3-hydroxymethyl furan as its THP ether 60 did not provide a sufficient barrier to metalation at the 2-position as the subsequent sulfenylation reaction (eq. 36) afforded only the 2,3-substituted regioisomer 52. In a further attempt to jprepare the 2,4-substituted regioisomer, a different strategy was adopted. Instead of blocking the directing hydroxylic oxygen, a substituent which could be selectively removed at a later time was placed at the favored alpha metalation position. As shown in eq. 37, 2-trimethylsilyl-3-hydroxymethyl furan 63 (prepared from 3-hydroxymethyl furan) was subjected -to sulfenylation conditions as before. 1) add to nBuLi(2eq. ), 0" o'c /\ /o\ 2,0 \ST: 0 TMSZ)25'C, 30 min Q3 3) PhSSPh, 0 C 23% 69% #3 1¢s H O . TMS 25'c, 4h 0 E1751 44 Surprisingly, the major product was not the desired 2-phenylthio-S-trimethylsilyl-4-hydroxymethyl furan (23%) but an unexpected isomer resulting from beta-position metalation (69%). Apparently, the directing effect of the hydroxylic oxygen is strong enough to override the furyl ring electronic factors that favor alpha-position metalation. The 'beta-position. metalation product was subjected to desilylation conditions to provide the novel 3-phenylthio-4-furan methanol 61. Although the major product of this procedure did not give the desired 2,4-substitution pattern, these results suggest a possible solution to this problem. If, in addition to placing a substituent at the favored metalation position, the directing oxygen was also blocked, one would predict that _the desired substitution pattern would be produced. However, this was not pursued further. The minor product of eq. 37 was subjected to desilylation conditions and through this procedure an adequate amount of the 2,4-substituted regioisomer 53 was produced for subsequent alkylation studies. The alcohol 53 was submitted to the usual chlorination conditions to produce the corresponding chloride 62 (eq. 38). As with l 29 MsCl, LiCl, collidilge, W (38) M $5 0 §3€W% 45 compounds 46 and 54, when the conversion of the chloride 62 to the Grignard reagent was attempted, the predominant product was found to be the furan dimer. Unfortunately, we were unable to convert chloride 62 to the corresponding tri-n-butylstannane in greater than 13% yield. A variety of techniques were examined in attempts to overcome this problem (see eqs. 24 and 30); however, the major isolated species was the reduction product shown in eq. 39. N b ¢S /O\ (39) Despite the fact that we had some success in the alkylation of the sulfenyl organometallics, it became apparent that this X-group would not be satisfactory. The methylthio-furans were hydrolyzed only under extremely vigorous conditions and in similar studies , the phenylsulfenyl analog was found to be even more resistant to hydrolysis. II. C. Silicon-substituted Furans At the outset, a silyl-substituted furyl organometallic appeared to have a good chance for success because of the 46 precedent for the preparation of silyl furans and their conversion into butenolides.59 Although silyl furans have been. prepared from furan or alkyl furans via standard metalation techniques,46 a newer ‘procedure allows direct preparation of at precursor to the 2,3-substituted silyl-furylmethyl organometallic. This was accomplished via the procedure of Knight54 in the preparation of 2-trimethylsilyl-3-furan carboxylic acid in 87% yield (eq. 40). The initially formed coon 1) LDA (Zeq.), THF, OOH -78‘c 30 min ' fl \ ’ + / \ W» 2) TMSCl (2eq.) 0 3) -78‘->25°C, 2h 0 ”5 87% carboxylate oxyanion directs abstraction of the adjacent alpha furyl proton by coordination to a second equivalent of base. The dianion is quenched with trimethylsilyl chloride to give, exclusively, the 2,3-substituted regioisomer after hydrolysis of the silyl ester. The trimethylsilyl-furancarboxylic acid was reduced to the corresponding alcohol 63 (76%), followed by conversion to the chloride 64, (77%) (eq. 41 and 42). 47 OH (7": 1) add to ma, Bit—LO, 0‘s» Q—‘g: (41) 93 767. 1 93 MsCl, LiCl, a» (if (42) collidine, DMF ' O TMS 93 m. Unlike the sulfur-substituted analogs, silyl chloride 64, was readily transformed into the corresponding Grignard reagent and alkylated in high yield with epoxygeranyl chlo— ride51:52 to afford 65 in 93% yield (eq. 43). IO , / 64 1) M8. THF. 50 C, 45min. (777°) * I TMS (43) ~ 2) QM , L120uC14 O 02.25‘ c, 1h Q 937. 48 The cyclization of 65 was examined with several differ- ent Lewis acids including ZnIZ Et20, Ti(O-ipr)3Cl, BF3-Et20, and MgBrz EtZO, Et3N. In all cases the solvent used was methylene chloride. Unfortunately, for all the Lewis acids employed, the predominant products were those of silicon loss. However, the magnesium bromide catalyst gave the best results as shown in eq. 44. 92 W (3eq-). Et3N (1.594.; CH2012, o‘c 2) zs'c, 1h 1" TMS | (44) 30% The majority of the cyclization product was accounted for by cyclized, desilylated product (10%) and uncyclized ketone (30%). The silicon-containing cyclization product 66, isolated in 20% yield, is not the expected regioisomer however. By some as yet undetermined pathway, cyclization to the initially silicon-bearing position of the furyl ring, along with migration of the trimethylsilyl group to the remote alpha position, has occurred. 49 While these studies were underway, some potential pathways to a precursor of the regioisomeric 2,4-substituted silyl-furylmethyl organometallic were also being investigat- ed. The most direct way of preparing beta-position functionalized silyl-furans is via Silylation of the corresponding alpha-furylanions as shown in eq. 40. Hence, the major obstacle in the establishment of the 2,4-substitution pattern lies in the placement of the silyl moiety at a position remote from the strongly directing beta-position functionality. As shown in eq. 40, the directing strength of the carboxylic acid group is such that none of the remote alpha silylated product is produced. Submission of 3-hydroxy- methyl-furan to these reaction conditions led to the produc- tion of a mixture of products (eq. 45) including a bis-silyl furan 67, but none of the desired 2-trimethylsilyl-4-hydroxy- methyl-furan was observed. W“ 1) 1m, THF, 0'0, 45 mill. 0 2) M01, o'-o25'c, 2h 3) MeONa, MeOI-I, 25°C, 15 min °" °“ Jim /0\ * /o\ I'MS +TM lo ms (45) 2E% 922W1 gzzum 50 To further reduce this directing effect, 3-hydroxymethyl-furan was protected as its trimethylsilyl ether and then submitted to the same Silylation conditions. As is shown in eq. 46, the desired 2,4-disubstituted furan TMS 'TMs gg um. gg:ua; 22151 a) LDA, THE, 0 c, 10 min b) TMSCl, ozezs’c, overnight c) meoue, neon, 25' c, 15 min 68 was produced in low yield but in sufficient quanti- ties to allow further experimentation. To further decrease the coordinating effect of the 3-hydroxymethyl group beyond that of the trimethylsilyl ether above, the t-butyldimethylsilyl ether 'was jprepared. Additionally, different metalation conditions were utilized in an attempt to decrease the amount of bis-silylated ma- terial produced. As shown in eq. 47, silylation48 of the t-butyldimethyl- silyl ether afforded twice as much of the desired 2,4-substi- tuted regioisomer as before (eq. 46) and none of the bis-silylated material. 51 OTBDMS TBDMS WED!“ a,b,c 1+ \ / \ ,. / \ S M IO (4,) 70% 26% a) add to nBuLi, TMEDA, o‘c b) 25'c, 30 min c) msc1, 0‘425'c, 1h The measures taken above afforded the desired 2,4-substituted regioisomer only as a minor product. To overcome this problem, a previously discussed strategy (see eq. 37) was adopted in which the favored site of reaction is blocked by a removable group. Therefore, Silylation was, at this point, attempted on a furyl sub- strate in which a substituent is placed at the favored alpha metalation position. The t-butyldimethylsilyl ether of 2-phenylsulfenyl-3-hydroxymethyl-furan 69, prepared from 52 as shown in eq. 48, was submitted to standard metalation conditions followed by quenching with trimethylsilyl chloride (eq. 49). A good yield (76%) of the desired silylated material 70 was obtained. The next step toward the 2,4- substituted silyl hydroxymethyl furan is removal of the phenylsulfenyl substituent by desulfurization with Rancy Nickel. However, in practice, it was found necessary to 52 W TBDMSCI, imidazole, / \ TBDMSB) 5¢ DMF, 25‘c, 1b a 4. 52 93 81% 1) add to nBuLi, Etzoe 0.C TBDMS 69 2) 25 C, 2.5h . fl. / \ (49) "' 3) TMSCl, 01.25 0, 1h TMS 0 34. 29 76% cleave the t-butyldimethylsilyl ether before the desulfur- ization could efficiently be carried out. Carefully control- led conditions were required to avoid concurrent C-desilyla- tion. Attempted fluoride desilylation was unsuccessful; however, cleavage with aqueous HOAc in THF did provide 71 in reasonable (52-73%) yield (eq. 50). 29___, HOAc, THP NH+ (C + recovered (50) H 0, heat, TMS o S¢ 0 starting 125h material 21 52-73% 11% 9% 53 The silyl ether cleavage product 71 was then subjected to Raney Nickel55 desulfurization (eq. 51) to afford a 56% yield of the desired 2-trimethylsilyl-4-hydroxymethyl furan 68. This reaction also suffered some reproducibility problems and was extremely sensitive with regard to the time of exposure to Haney-Nickel. 71 Ra-Ni(W-2, 20-fold excess) 3’. W61) EtOH, heat, 6h TMS o 99 567. Although this procedure does provide the desired 2,4-substituted organometallic precursor 68, the reaction sequence is relatively . long (5 steps from 3-hydroxymethyl-furan) and the overall yield is low (16%). This strategy is essentially identical to that reported by Goldsmith and Liotta; however, in our hands, the yields were lower than reported.56 A more efficient process for the preparation of 68 was devised as shown in eq. 52. 2-Bromo-4-furoic acid, prepared by the direct bromination of furoic acid with 54 1) nBULi(20 26¢]. ) a Et 0, 43¢, 30min / \ 2 o a Br 0 2) TMSCI, -78 c 3) 01.25 c 1h (52) 02H 02" C0114 1) add to IAH / \ + / \ .. M “2°22“; 93 TMS o TMS 0 TMS 2) 25 c, 2h 81% 63% 157. 157. pyridinium hydro-perbromide in 60% yield,5'7 was submitted to lithium-halogen exchange conditions followed by quenching with trimethylsilyl chloride. A 63% 'yield. of the corresponding trimethylsilyl furoic acid was obtained along with some bis-silylated and reduction products. This crude mixture was directly reduced (LAH) to afford an 81% yield of the desired 2,4-substituted furan 68; 31% overall yield (3 steps from furoic acid). With a precursor to the silyl-appended 2,4-substituted organometallic reagent readily available, alkylation and cyclization studies were undertaken. 2-Trimethylsilyl-4-hydroxymethyl-furan 68 was smoothly 55 transformed to the corresponding chloride as shown in eq. 53. Like its regioisomer 64, chloride 72 was readily converted into the Grignard reagent and alkylated in high l 93 MsCl, LiCl, collidine, DMF If (53) TMS 0 23 78% yield with epoxygeranyl chloride51 as shown in eq. 54. /o O 72 1) Mg, THF, so 0, 45min, (6470’ | (54) 2) QM, 1.1 CuCl4ih 0 02.35%, 13 907. 56 As before (see eq. 44), similar cyclization studies were undertaken with 73, and as previously noted, the same cyclization conditions (MgBrZOEtZO (3 eq.)), Eth (1.5 eq.) gave the best results. Analysis of the cyclization products demonstrated that the materials isolated were identical to those previously obtained (eq. 44) and the overall yield was similar. The desired cyclization product 66, ‘which previously' was the ‘unexpected regioisomer, was also produced here, albeit in only 20% yield. We must conclude that the trimethylsilyl-furans are too acid labile to be efficiently utilized in Lewis acid catalyzed. cyclization reactions. As an alternative, some variants on trialkylsilyl, such as triethylsilyl or methyldi- isopropylsilyl, which should be less acid sensitive, have been considered and may be investigated in the future. Although considerable success was enjoyed in alkylation reactions of silyl-appended organometallics, it is apparent that the trimethylsilyl variant will not meet all of the requirements for a general methodology as set forth in the introduction. However, theser silyl-substituted. furans should serve admirably as precursors to 3- and 4-alkyl-2- (5H)-furanones. As shown in eqs. 55 and 56, the chlorides 64 and 72 were converted to the corresponding Grignard reagents and coupled in high yield with nonyl-iodide to afford the alkyla- tion products 74 and 77, respectively. 57 In principle, the direct oxidation of a 3-substituted furan will provide the corresponding butenolide;583vb however, regio-chemical ambiguities usually render this approach impractical for general synthetic applications. A more suitable solution to this problem is afforded via oxidation of the corresponding 2- or 5-sily1ated, 3-alkylfurans. As has been demonstrated by Kuwajima59 and recently employed by Schultz,60 Goldsmith and Liotta,56 a trimethylsilyl group can serve to regio-specifically direct the introduction of oxygen providing the corresponding 3- or 4-alkyl-2(5H)-furanones. Utilizing the procedure 'reported 'by Kuwajima,59' the alkylation products 74 and 77 were smoothly converted into their corresponding butenolides 80, 81 and 86. (fl (Lb / \ cwwLfisHls [fig-"9 (55) .___.. + 0 TMS 0 TMS 0 0 fit 1582‘]. ’83 (7870,!‘” ’81; (”Mg ,THF ,b)nony|-l ,L‘gCuCl ,QCHscog-‘l ,NaOAc,CH2Cl2 58 l 0,b '9 1c __. CQHH) / \ ——> / \ ———+ (56) TMS O TMS O O 73 21(7 7 ‘70) g: (9 'O/O) o-c) as in equation 55 Other representative examples designed to examine the relative rate of furan vs. remote olefin oxidation as a function of the degree of alkene substitutions are presented in Figure 9. As shown, good to excellent yields of coupled silyl-furans are realized in all cases; and if the alkene is less than trisubstituted, the major or exclusive oxidation product is the result of attack at the furyl residue. These methods provide access‘ to a wide variety of 2- and 3-substituted butenolides depending on the alkyl halide used. According to the organizational scheme presented in the introduction, these products can be considered as type A (Figure 1) furans with access to all of the oxidation states 7-10. 59 \I\ «an a ..some 2 Jean C O ngk \3I\mW\)I\MWHWXJWI(XJ\)I\MMHV\“ I. z \ .I()I“Y(uo 3: R $2. 9.x 0 0 ms: $8 mm snow“ map 0 . oz» 0 i E 45?... Jéwn 3%. ermu mm. Jrhm..wM o 0 ms: 0 .1. \l\/\/Uw.LV I I gdo $5 an «a: mm o i me: o x _\ . n«)(\ltem \\ \\ 8:225 .88... lmmu 32 coeam Figure 9. -3- and 4-a1ky1 furans and ducts. their 2(5H)-furanone oxidation pro 2-Trimethylsily1 111. An Approach to the Synthesis of (1)-Aphidicolin III. An Approach to the Synthesis of (1)-Aphidicolin The novel, tetracyclic diterpenoid aphidicolin (see Figure 10) was isolated61 in 1972 from cultures of the fungus cephalosporium aphidicola (Petch). Subsequently, intensive biological activity studies were undertaken62 when it was discovered that aphidicolin was an antibiotic that possessed strong in vitro activity against Herpes simplex type 1 and 2 viruses. Because relatively few antiviral substances are known and due to the novel structure of the molecule, there has been intense interest among synthetic organic chemists in aphidicolin. As a result, several total syntheses of aphidicolin have been reported.63a»br°:d The Corey synthesisG3° utilizes a polyene cyclization to establish the A and B rings of aphidicolin with the appropriate stereochemistry. This strategy very efficiently affords access to the basic carbocyclic skeleton upon which most of the published approaches then work to append the sterically congested spirocycle. The construction of the D-ring is very neatly accomplished by McMurry.63a He effi- ciently attacks the problem via\ an intermediate cyclopen- tenone (see Figure 10). With this in mind, a study directed towards a formal, total synthesis of aphidicolin, utilizing 60 61 ovaaocousn ammunu eafiooaoasoo ouuavoauouau sneeze: «it .o ”V Figure 10. Proposed retrosynthetic scheme to (t)-aphidicolin. 62 the methodology explored in this thesis, -was 'undertaken. This study would exploit; 1) a furan-terminated polyolefin cyclization to rapidly establish the carbocyclic nucleus of aphidicolin, followed by ii) a furan to butenolide to cyclopentenone conversion yielding the McMurry intermediate. A retrosynthetic scheme of this strategy is shown in Figure 10. Cyclization of 91 would afford 92 in which the A and B rings of the target have been established. With a furyl moiety integrated into the cyclization product, an efficient transformation of it into the McMurry intermediate should be possible via the target butenolide. As discussed in the introduction, it would have been desirable to have a substituent appended to the furyl ring which could be carried through the cyclization to enable facile conversion to a butenolide. Unfortunately, a suitable substituent which could survive cyclization conditions has not yet been found. However, (since in this case, the desired orientation of the cyclization is produced without a directing furyl substituent, it should be possible to append the substituent after the cyclization has occurred. Although less efficient than introduction as part of an X-substituted alkyl furan, this strategy should serve well in the present case. An analysis of the substituents located at C(3) and C(4) of aphidicolin suggests, as in the Corey synthesis, that an w-oxygenated geranyl halide, after alkylation and 63 cyclization would yield the proper relative stereochemistry at C(4). In addition, should the cyclization substrate possess a 6,7-epoxide, then the cyclized product would have an oxygen at C(3), however opposite in stereochemistry to that desired (see Figure 11). Another desirable feature of this construction is the potential for an asymmetric synthesis should an optically pure geranyl synthon be employed.64 Lewis acid HO 0 w Figure 11. Furan terminated, epoxide initiated olefin cyclization. (E)-8-Hydroxygeranylacetate, prepared from geranyl acetate65 when submitted to Henbest epoxidation, afforded 89 in 94% yield. After protection of the hydroxyl as a TBDMS ether, the acetate functionality was cleaved with 64 potassium carbonate in methanol followed directly by chlori- nation via the procedure of Stork52 to afford chloride 90 in 65% yield overall from 89. 1) MCPBA CH C1 O‘C ' 2 2’ \ \ \ 0A6 o —’ OA‘ (57) H 2) 25 c, 2h 0H 93 947. 1) TBDMSCI, BEEN, HHAE, 93 __C.HzEI_2: 25 '1‘“ (9‘5? \ Ct (58) 2) K2C03, MeOH, o‘c, 30 min 3) 1. nBuLi, Et 0, HMPA, o'c I, ii. 25.0, 10 gin 03+ 29 65% overall 111. 'I‘sCl, L1c1, zs’c, | overnight, (77%) The modified epoxygeranyl chloride 90 was coupled with the Grignard reagent produced from 3-furylmethyl chloride in the usual way8 to afford a 60% yield of coupling product 91 (eq. 59). Several Lewis acids were examined 65 (59) 1) M8: THP ’ /o\ 2) 29, L16CUC14 O-25' , 1h (see eq. 44) in the attempted cyclization of 91, but only two were found to give a significant amount of desired product 92. As is shown in eq. 60, boron trifluoride etherate catalyst in the presence of triethylamine gave the best results as a 28% yield of desired cyclization product 92. Although related cyclizations are usually conducted solely in methylene chloride solvent, in this case, it was discovered that the addition of benzene and pentane to the reaction solvent almost completely inhibited the formation of a mono-cyclic side product which created difficulties in the purification of 92. Unfortunately, the major product of this reaction appeared to be an uncyclized, epoxide-opened ketone as shown in eq. 60. Although considerable effort was expended 66 in an attempt to increase the yield, we were unable to obtain greater than a 30% yield of 92. Titanium tri-isopropoxy chloride was the only other Lewis acid tried that afforded a significant amount of 92. However, the yield was only about half of that obtained with BF3oOEt2 and much more of the troublesome monocyclic product was produced. BFa'OEt2(3eq.), Etlbl(1.5eq.), CH C1 , benzene, pentane, -60°C, 10 min 23. As shown in Figure 10, the configuration of the hydroxyl at C(3) of the cyclization product 92 is opposite to that of aphidicolin. Several techniques for the direct inversion of the hydroxyl in 92 were examined at this stage of the synthesis including the Mitsunobu reaction 66 and nitrite displacement67 of the mesylate of 92. Unfortunately, the highly hindered nature of the C(3) hydroxyl precluded any success with these measures. 67 The inversion was accomplished in a manner identical to that utilized in the Corey630 and McMurry63a syntheses of aphidicolin as illustrated in eqs. 61 and 62. The C(3) hydroxyl of 92 *was oxidized to the corresponding ketone 1) FCC, NaOAc, cu2012, 92 ’" O'C, 30 min 2) 25'C, 1.5h 1) add to 93 BuiNF, THF, Li(sec-Bu)3EH (62) 2' o c, 2h THF, -78'c, 1h ', 0 2) 0°C. tho /. Ho” .. H0 94% 25 79% overall from.92 93 in 91% yield (eq. 61). Desilylation of 93 with tetra-n-butyl ammonium fluoride (94%) and immediate reduction with L-selectride, to avoid decomposition of the intermediate aldol, afforded the diol 94 in a 79% overall yield from 92. 68 With the desired. A-ring functionality of aphidicolin now intact, the conversion of 94 to the McMurry intermediate was undertaken. As discussed above, this was to be approached via the appendage of a substituent to the furyl moiety and subsequent elaboration of the target (see Figure 10). With diol 94 in hand, the most direct route to the McMurry intermediate appeared to lie in protection of the diol system as the acetonide followed by metalation and Silylation of the furyl moiety. As shown in eq. 63, the acetonide 95 was smoothly produced in 90% yield. acetone, oxalic acid, CHZCIZ, zs‘c, overnight However, the metalation and Silylation of 95 turned out to be quite problematic. Initially, the procedure of Schultz6O was applied to the problem. The metalation was carried out by adding 1 equivalent of nBuLi to a solution of 95 in THF at 0°C. The solution was stirred for 2h at 0°C followed by addition of TMSCl and gradual warming to 69 25°C overnight. Upon workup and analysis, only recovered starting material was obtained. The reasons for the failure of this procedure are not clear inasmuch as the substrate Schultz successfully silylated is rather closely related, to 95. Additionally, these reaction conditions have worked in this lab when applied to a different substrate. It was assumed that the problem was not in the alkyla- tion step, but in difficulty with the metalation. Attempt- ing to overcome this, a series of progressively more vigorous metalation procedures were examined. When metalation was attempted by stirring of 95 with 2 equivalents of n-BuLi in THF at 25°C for Zl‘th solvent decomposition became a problem. The same conditions were then applied with Et20 as the solvent with added TMEDA but attempted Silylation yielded none of the desired silylated product. The material that was recovered was slightly different from the starting material and spectral analysis suggested that the acetonide moiety had been altered. Some simple studies of the Silylation of furyl-lithium derivatives revealed that alkylation with TMSCl is consider- ably slower in Et20 than in THF. Hence, metalation of 95 was attempted as before with 2 equivalents of n-BuLi and added TMEDA but with only hexane solvent. After stirring for 2 h at 25°C, the reaction was cooled to 0°C and THF was added. After the addition of TMSCl, the reaction was lanalyzed indicating that much decomposition had occurred. 70 At this point, it was concluded that the acetonide moiety was consuming base. We hOped that a different protecting group here might circumvent the above difficulties and so the bis-methoxyethoxymethyl ether68 96 was prepared from the diol 94 (eq. 64). Our initial attempts to metalate 96 again indicated that the protecting group was not stable to the reaction conditions. 94 TMEMC1(6eq.), (ipr)2NEt(g:q.) "' cnzc12, 25’c, 2h MEMO' MEMO ’ 2§£B% As a last resort, Silylation of the diol 94 was attempt- ed. This was carried out by stirring 94 with 4 equivalents of nBuLi and TMEDA in hexane at 25°C for 30 min. Subsequent- ly, the solution was cooled to 0°C and THF was added followed by TMSCl. Analysis of the reaction mixture showed no furyl Silylation; however, O-silylation at the A-ring hydrox- yls was observed. 71 In conclusion, we have observed that the furyl moiety of this molecule is unusually difficult to metalate. A solution to this problem might lie in the proper selection of the A-ring diol protecting functionality. Alternatively, the furyl nucleus might be selectively and carefully bromi- nated. A simple metal-halogen exchange would lead to the silylated equivalent of 96 after treatment with TMSCl. These possibilities are currently being examined in our laboratories. EXPERIMENTAL EXPERIMENTAL General: Tetrahydrofuran (THF) ‘was dried by idistillation under nitrogen from sodium benzophenone ketyl; methylene chlo- ride *was dried by distillation under nitrogen .from calcium hydride; 3:3 dimethylformamide (DMF) was dried by distillation at reduced pressure from phosphorous tentoxide; hexamethylphos- phoramide (HMPA) was dried by distillation at reduced pressure from calcium hydride; pyridine was dried by distillation under nitrogen from calcium hydride; diisopropyl amine was dried by distillation under nitrogen from calcium hydride. Petroleum ether refers to the 30-60°C boiling point fraction of petrol- eum benzin. Diethyl ether was purchased from Mallinkrodt, Inc., St. Louis, Missouri, and used as received. n-Butyl lithium in hexane was purchased from Aldrich Chemical Company , Milwaukee, Wisconsin and titrated by the method of Watson and Eastham.26 All reagents were used as received unless otherwise stated; all reactions were performed under argon with the rigid exclusion of moisture from all reagents and glassware unless otherwise mentioned. Melting points were determined on a Thomas-Hoover capillary melting point apparatus and are uncorrected. Infrared spectra were recorded. on a lee-Unicanl SP-lOOO infrared. spectrometer or a Perkin-Elmer Model 167 spectrometer with polystyrene 72 73 as standard. Proton magnetic resonance spectra (lH-NMR) were recorded on .a Varian T-60 at 60 MHz, a Varian OFT-20 at 80 MHz, or a Bruker WM-250 spectrometer at 250 MHz as mentioned in deuteriochloroform unless otherwise indicated. Chemical shifts are reported in parts per million (5 scale) from internal standard tetramethylsilane. Data are reported as follows: chemical shifts (multiplicity: s = singlet, brs = broad sing- let, d = doublet, t = triplet, m = multiplet, coupling constant (Hz), integration). Electron impact (El/MS, 70eV) mass spectra were recorded on a Finnigan 4000 with an INCOS 4021 data system. Flash chromatography was performed according to the procedure -of Still et.a1,69 using Whatman silica gel and eluted with the solvents mentioned. Analytical thin-layer chromatography was run on either Macherey-Nagel Polygram SIL G/UV254 pre-coated plastic sheets or Brinkman Instruments SIL G/UV pre-coated glass plates. Spots were visualized by either dipping into a solution of Vanillin (1.5g) in absolute ethanol (100 mL) and concentrated sulfuric acid (0.5 mL) and heating with a heat gun or spraying with a 5% solution of molybdophosphoric acid in absolute ethanol and heating to 120°C. l-(l-Bromoethyl)-cyclohexene 17. To a mechanically stirred 20,21 solution of cyclohexenyl ethanol (3.0g, 24mmol) in dry Et 0 (100 mL) was added carbon tetrabromide (19.9g, 60 mol) 2 followed by triphenylphosphine (14.8g, 60 mmol).19 After 30 minutes, the solution became bright yellow and warm. Upon termination of exothermicity, the solution was refluxed for 8h. After cooling to room temperature, pentane (50 mL) was 74 added, followed by stirring for 10 minutes. The solution was then cooled to 0°C and the resulting precipitate was removed by filtration through celite. After removal of solvent lip .ygggg, the crude product was purified by chromatography on a column of silica gel (60-230 mesh, 25g, hexane-EtZO, 2:1) followed by distillation through a 20 cm vigreux column, BPZ.0mm = 67°C to provide 4.07g (90%) of bromide 17 as a water white 1 liquid. H NMR (60 MHZ) 6 5.75 (brs, 1H), 4.60 (q, J=7Hz, 1H) 2.10 (brm, 4H), 1.75 (d, J=7Hz, 3H), 1.60 (brm, 4H); IR (neat) 2960, 2880, 1660, 1450, 1185, 915, 740 cm‘l; CI-MS (CH4) 189 (M++1, 50.6), 187 (M++l, 46.4), 145 (7.5), 125 (17.6), 109 (base). l-(-chloroethyl)-cyclohexene 18. To a magnetically stirred 20’21 (1.26g, 10 mmol) in solution of cyclohexenyl ethanol dry carbon tetrachloride (20 mL) cooled to 0° (ice 'water) was added tri-n-butyl phosphine (3.04g, 15 mmol).19 After addition, the reaction mixture was reluxed for 2.5h then pentane (20 mL) was added and stirring was continued for 5 minutes. A heavy bottom layer separated out from which the top layer was decanted. Removal of solvent and distillation of the residue through a vigreux column, Bp5mm = 60°C, afforded 0.68g (47%) of 18 as a viscous liquid. 1H NMR (250 MHz) 5 5.77 (brs, 1H), 4.54 (q, J=7Hz, 1H), 2.05 (m, 4H), 1.60 (m, 4H), 1.59 (d, J=7Hz, 3H); IR (neat) 2960, 2890, 1670, 1445, 1380, 1230, 1 1040, 675 cm”; cx-us (CH4) 145 (M++l, 3.2), 119 (9.9), 109 (base), 63 (19.8). 75 Coupling of Grignard Reagent 16 with Bromide 17 . To a solution of the furyl-methyl Grignard reagent 16 (3.9 mmol). (titer determined by titration26) in dry THF (5 mL) at 0°C (ice water) was added bromide 17 (0.57g, 3.0 mmol) in dry THF (1 mL) dropwise followed by Li CuCl catalyst (0.1M in 2 4 THF, 0.12 mL, 0.012 mmol). The solution was stirred at 0°C for 10 minutes then quenched with saturated aqueous NH Cl 4 (2 mL). The reaction mixture was cast into EtZO/hexane 1:1 (50 mL) and washed with saturated aqueous NaHCO3 (50 mL), H20 (50 mL) and saturated aqueous NaCl (50 mL). The organic phase was dried (Na2S04). Concentration in vacuo provided the crude product as a pale yellow liquid.which was purified by chromatography on a column of silica gel (60-230 mesh, 50g, EtZO-hexane, 1:99) to provide 0.34g (60%) of coupling product Rf 0.6 (EtZO-hexane, 5:95). Spectral analysis (integra- tion of vinyl protons in proton NMR below) showed this to be an approximately 3:2 mixture of SN2 vs. SN2' like products, 1H NMR (250 MHz) as 7.30 (s, 211), 7.18 (s, 2H), respectively. 6.37 (s, 1H), 6.23 (s, 1H), 5.40 (brs, 0.6H, 8N2), 5.18 (dq, J=2,7Hz, 0.4H, SNZ'), 1.43 (dd, J=2,7Hz, 1.2H, SNZ'), 1.00 (d, J=7Hz, 1.8H, SNZ); IR (neat) 2940, 2885, 1505, 1455, 1385, 1; EI/MS (70eV) 190 (11*, 35.8), 175 (8.9), 109 1030, 785 cm’ (base). Coupling of Grignard Reagent 16 with Chloride 18. As in the coupling of 16 and 17, 16 (3.9 mmol) and 18 (0.43g, 3.0 mmol) provided 0.27g (47%) of 19 after purification as the sole product, with the following spectral data. 1H NMR (60 76 MHz) 6 7.15 (d, J=2Hz, 1H), 7.00 (S, 1H), 6.05 (d, J=2Hz, 1H), 5.15 (brs, 1H), 2.35 (m, 2H), 1.85 (m, 4H), 1.50 (m, 4H), 1.00 (d, J36Hz, 3H); IR (neat) 2940, 2860, 1670, 1505, 1455, 1 1385, 1030, 880 cm‘ ; EI-MS (70eV) 190 (M+, 37.0), 175 (8.1), 109 (base). 18c Furan-3,4-dimethanol 22. A solution of diethyl-furan- 3,4-dicarboxylate 21 (18.0g, 0.085 mmol) in dry Et 0 (150 2 mL) was added dropwise to LiAlH (7.60g, 0.20 mmol) in dry 4 Et 0 (300 mL) at 0°C (ice water) over lh. After addition, 2 the solution was refluxed for 1h and allowed to stand at 25°C overnight. The reaction was carefully quenched with 20% aqueous Rochelle Salt (200 mL). The ether layer was, separated, the aqueous phase saturated with NaCl, and then extracted with EtZO/EtOAc, 9:1 (3 x 150 mL). The conbined organic phases were dried (Na2804), concentrated in vacuo and distilled _ o 180 _ , (3P0.15mm - 110 C, lit. BP2mm - 130 C) to afford 8.16g (77%) of 22 as a viscous water white fluid. Spectral data was consistent with the literature.180 18c Furan-3,4-dimethanol,mono-THP-ether 23. Furan-3,4- dimethanol 22 (6.10g, 47 mmol) was added to a solution of dihydropyran (3.95g, 47 mmol) and pyridinium paratoluenesul- fonatez3 (1.18g, 4.7 mmol) in dry CH C1 (200 mL) and stirred 2 2 for 4h at 25°C. The solution was then diluted with Et 0 (200 2 mL) and washed with half saturated aqueous NaCl (100 mL). The organic phase was dried (NaZSO4) and concentrated in vacuo to afford 10.86g of crude product which contained three compo- nents by TLC. The desired mono-THP ether was separated from 77 the bis-THP ether side product and unreacted diol by flash chromatography on a column of silica gel (60-230 mesh, 300g, CHZClZ/EtOAc,‘ 6:1) as the second compound to be eluted. This was performed in two runs on the same bed of silica gel pro- viding 4.90g (50%) of mono-THP ether 23, BPO.l5mm = 125°C (no literature value was reported18c), 3.40g (25%) of bis-THP ether and 1.50g (25%) of unreacted furan dimethanol. Spectral data of the mono-THP ether was consistent. with the literature.18c 3-Chloromethy1-4-tetrahydropyranyloxymethyl furan 24 . To the mono-THP-ether 23 (1.0g, 4.8 mmol) was added collidine (0.70 mL, 5.3 mmol) followed by a solution of dry LiCl (0.41g, 9.6 mmol) in dry DMF (10 mL). The solution was cooled to 0°C (ice water) and methanesulfonyl chloride (0.41 mL), 5.3 25 The resulting solution mmol) was added over 15 minutes. became a‘ bright yellow suspension which was stirred for 2h at 0°C, and at 25°C for 1h. The reaction mixture was diluted with EtZO-pentane, 1:1 (150 mL) and washed with saturated aqueous NaHCO3 (150 mL) and saturated aqueous Cu(N03)2 (150 mL). The organic phase was dried (Na2304), concentrated _i__n_ vacuo and purified by bulb-to-bulb (Kugelrohr) distillation, oven 140°C (0.02 mm) to provide 0.85g (77%) of chloride 24 1H NMR (250 MHz) 6 7.46 (d, J=Hz, as a slightly yellow fluid. 1H), 7.40 (d, Jsznz, 1H), 4.72 (t, J=4Hz, 2H), 4.64 (s, 1H), 4.58 (s, 1H), 4.56 (s, 2H), 3.92 (m, 1H), 3.56 (m, 1H), 1.60 (brs, 6H); IR (neat) 2940, 2860, 1560, 1450, 1270, 1140, 1035 cm'l; EI-MS (70eV) 230 (11*, 1.7), 146 (9.4), 129 (98.4), 85 (base). 78 Lactarol-THP-ether 27. A solution of 3,-chloromethyl-4- tetrahydropynanyloxymethyl furan 24 (0.50g, 2.2 mmol) and dibromoethane (2 drops) in dry THF (1 mL) was added over 5 minutes to activated magnesium turnings (0.060g, 2.5 mmol) covered with dry THF (4 mL) at 25°C. After addition was com- plete, the internally measured temperature of the reaction began to rise and at 30°C a cooling bath was introduced so that the temperature was maintained at 10°-15°C as the magnesium was consumed. This occurred over 2h at which time titration26 of an aliquot of the Grignard solution showed formation to be quantitative. A portion of the Grignard solution (0.38 'M, 3 mL), 1.1 mmol) was cooled to 0°C (ice water) and 4,4- 18° (0.16g, 1.0 mmol) dimethyl-1-(l-chloroethyl)-cyclopentene in dry THF (1 mL) was added via syringe followed immediately by LiZCuCl4 catalyst (0.1 M in THF, .04 mL, .004 mmol). The solution was stirred at 0°C for 30 minutes then allowed to warm to 25°C and worked up as before (see coupling of 16 and 17) to afford 0.48g crude product. Purification was accom- plished by flash chromatography on a column of silica gel O-PE, 5:95) to afford 0.24g (75%) 1 (230-400 mesh, 100g, Et2 of lactaral-THP-ether 27, Rf 0.46 (EtZO-PE, 1:9). (250 MHz) 6 7.36 (S, 1H), 7.17 (8, 1H), 5.22 (brs, 1H), 4.64 H NMR (s, 1H), 4.59 (s, 1H), 4.33 (dd, J=12,4Hz, 1H), 3.90 (m, 1H), 3.55 (m, 1H), 2.62 (dd, J=14,6Hz, 1H), 2.47 (ddt, J=6,7,7Hz, 1H), 2.33 (dd, J=l4,7Hz, 1H), 2.09 (brs, 4H), 1.60 (brm, 6H), 1.06 (s, 3H), 1.05 (s, 3H), 1.00 (d, J=7Hz, 3H); IR (neat) 2970, 2890, 1550, 1465, 1368, 1125, 1030 cm-1; EI-MS (709V) 79 318 (11", 1.7), 300 (1.2), 233 (5.0), 216 (19.0), 123 (18.8), 85 (base). Lactarol 28. A solution of lactarol-THP-ether 27 (2.68g, 8.4 mmol) in absolute ethanol (800 mL) and pyridinium toluene- sulfonatez3 (2.68g, 10.6 mmol) was heated to 55°C for 3h. After cooling to 25°C, the ethanol was removed in 2&9 and the residue diluted with Et 0 (200 mL). After washing with 2 half saturated brine (200 mL), the organic phase was separated and dried (Na2804). Removal of solvent afforded 1.87g (95%) 1H of lactarol 28 which required no further purification. NMR (250 MHZ) 6 7.35 (S, 1H), 7.16 (S, 1H), 5.22 (brs, 1H), 4.51 (S, 2H), 2.62 (dd, J=14,6HZ, 1H), 2.47 (ddt, J=6,7,7HZ, 1H), 2.33 (dd, J=14,7HZ, 1H), 2.09 (brs, 4H), 1.57 (brs, 1H), 1.06 (8, 3H), 1.04 (8, 3H), 1.00 (d, J=7Hz, 3H); IR (neat) 3380, 2970, 2890, 1545, 1465, 1365, 1150, 1055, 805 cm-1; EI-MS (70eV) 234 (M+, 11.1), 216 (14.8), 201 (17.1), 160 (15.0), 123 (base). Lactaral 15. To a solution of lactarol 28 (1.0g, 4.3 mmol) in methylene chloride-pentane, 1:1 (100 mL), was added "acti- vated" Mn0228 (23g). The suspension was stirred overnight and the solid was removed by filtration through celite. Concen- tration _i_n vacuo provided the crude product which was purified by flash chromatography on silica gel (60-230 mesh, 70g, Et O-PE, 1:4) to afford 0.96g (96%) of lactaral 15, Rf 0.62 2 (EtZO-PE, 1:4). as a slightly yellow liquid. Spectral data was identical to that kindly provided by Dr. G. Magnusson.18c In NMR (60 MHz) 6 9.76 (s, 1H), 7.80 (d, J=2Hz, 1H), 7.07 (brs, 80 1H), 5.10 (brs, 1H), 2.60 (m, 2H), 2.40 (m, 1H), 2.02 (brs, 4H), 1.06 (s, 3H), 1.04 (s, 3H), 1.00 (d, J=7Hz, 3H); IR (neat). 2990, 2870, 2770, 1700, 1545, 1480, 1155, 1055, 825, 762 om‘l; EI-MS (70eV) 232 (M+, 9.1), 214 (12.8), 199 (19.1), 123 (base), 81 (84.5). Lactaral ethylene acetal 30. To a biphasic mixture of anhydrous ethylene glycol (0.80g, 12.9 mmol) and lactaral 15 (0.30g, 1.3 mmol) in dry benzene (10 mL) was added para- toluenesulfonic acid (0.02g, 0.1 mmol). The mixture was warmed to 60°C for 3h, then after cooling the 25°C, was cast into ether (20 mL), washed with water (20 mL), saturated aqueous NaHCO3 (20 mL), and dried (Na2804). Removal of solvent'.in_ Leuq afforded 0.30g (84%) of lactaral ethylene acetal 30, (Rf identical to 29) which required no further purification. 1H NMR (60 MHz) 6 7.32 (d, J=2Hz, 1H), 7.05 (brs, 1H), 5.72 (s, 1H), 5.13 (brs, 1H), 3.92 (m, 4H), 2.50 (m, 3H), 2.09 (brs, 4H), 1.06 (s, 3H), 1.04 (s, 3H), 1.00 (d, J=7Hz, 3H); IR (neat) 2970, 2890, 1550, 1465, 1368, 1125, 1030 cm-1; EI-MS (70eV) 276 (M+, 1.5), 258 (1.3), 214 (15.3), 123 (base). Attempted cyclization of Lactaral 28. A solution of lactarol 28 (0.09g, 0.43 mmol), formic acid (99%, 4.3 mL) and cyclohexane (4.3 mL) was stirred at 25°C for 1h. The organic phase was separated, washed with saturated aqueous NaHCO3 (10 mL), brine (10 mL), and dried (Na2S04). Removal of solvent afforded 0.08g (71%) of lactarol formate 31. 1 H NMR (60 MHZ) 6 7.95 (S, 1H), 7.31 (S, 1H), 7.10 (S, 1H), 5.15 (brs, 1H), 3.98 (s, 2H), 2.39 (m, 3H), 2.05 (brs, 4H), 1.05 (brs, 9H); IR 81 (neat) 2980, 2865, 1735, 1555, 1470, 1373, 1170, 1060 om‘l; EI-MS (70eV) 262 (M+, 6.6), 247 (0.7), 216 (23.8), 201 (13.4), 123 (base). Lactaroic Acid 32. Lactaral 15 (0.18g, 0.8 mmol) was stirred for 12b in a solution containing sodium cyanide (0.25g, 5.0 28 (5.0g), acetic acid mmol), "activated" manganese dioxide (0.1 mL), 1.5 mmol) and methanol (10 mL). The methanol was then removed _i_n_ vacuo and the residue diluted with Et20 (50 mL). The organic phase was dried (Na2S04) and removal of solvent afforded 0.16g of crude methyl lactarate. The ester was hydrolyzed by refluxing for 12h in a solution of THF (10 mL) containing 5 equivalents of 20% aqueous NaOH solution. Acidification of the reaction medium followed by extraction with EtZO (50 mL) provided an organic phase which was dried (Na2S04). Removal of solvent afforded 0.14g lactaroic acid 1H NMR (60 32 (71% overall) which was not purified further. MHZ) 6 12.34 (hrs, 1H), 7.99 (8, 1H), 7.16 (S, 111), 5.16 (brs, 1H), 2.61 (m, 3H), 2.02 (brs, .4H), 1.06 (S, 311), 1.04 (S, 3H), 1.00 (d, J=7Hz, 3H); IR (neat) 3100, 2980, 2640, 1695, 1; EI-MS (70eV) 248 (M+, 26.4), 1540, 1435, 1310, 1245, 1150 cm- 233 (21.4), 215 (3.6), 149 (5.2), 123 (base). Lactaroyl Chloride 33. A solution of lactaroic acid 32 (0.38g, 1.5 mmol) and oxalyl chloride (0.30g, 3.5 mmol) in dry benzene (10 mL) was stirred at 25°C for 12h. Removal of solvent in vacuo afforded 0.32g (80%) of lactaroyl chloride 33 as a brown oil which was not purified further. 1H NMR (60 MHZ) 6 7.99 (S, 1H), 7.16 (s, 1H), 5.16 (brs, 1H), 2.61 82 (m, 3H), 2.02 (brs, 4H), 1.06 (S, 3H), 1.04 (S, 3H), 1.00 (d, J=7Hz, 3H); IR (neat) 2980, 2890, 1775, 1535, 1470, 1390, 1; EI-MS (70eV) 266 (11", 5.1), 251 (4.5), 1160, 818, 687 cm- 231 (4.7), 215 (2.6), 123 (base). Reaction of 34 with (Et0)3CH/SnCl4. Preparation _of 35. To a solution of 2-trimethy1siloxy-3,5-dimethy1-furan 34 (0.20g, 1.1 mmol) and triethyl orthoformate (0.16g, 1.1 mmol) in dry methylene chloride (2 mL) cooled to -40°C was added a catalytic amount of stannic chloride (four drops). The reaction temper- ature was gradually raised to 10°C over a period of 2h. The mixture was cast into Et20 (25 mL) and washed with saturated aqueous Na'HCO3 (2 x 25 mL) and' brine (25 mL). The organic phase was dried (Na2804) and removal of solvent afforded 0.22g (93%) of 35. 1 H NMR (60 MHZ) 6 6.99 (q, J=2Hz, 1H), 4.30 (s, 1H), 3.65 (m, 4H), 1.90 (d, J=2Hz, 3H), 1.41 (S, 3H), 1.23 (m, 6H); IR (neat) 2985, 2900, 1765, 1665, 1455, 1375, 1125, 1080 cm"1 ; CI-MS (CH4) 215 (M++1, 1.0), 169 (8.6), 141 (4.9), 103 (base). 3-Bromo-2(5H)-furanone 36. To a solution of 2-(5H)-furanone (11.2g, 0.13 mol) in dry carbon tetrachloride (40 mL) was added bromine (21.9g, 0.137 mol) at 25°C. The solution was stirred until most of the bromine had been consumed and became light orange in color (2h). The solvent and excess bromine were then removed _i_n_ vacuo to afford 32.4g (100%) crude yield of the 2,3-dibromobutyrolactone as a red fluid. The material was subjected to dehydrohalogenation without further purifica- tion. To a solution of the dibromide (32.4g, 0.134 mol) in 83 dry benzene (250 mL) was added pyridine (51.8 mL, 0.64 mol). The solution was heated under reflux for 3h, then after removal of most of the solvent _i_g vacuo, the residue was diluted with methylene chloride (200 mL). After washing with 1N HCl (2 x 200 mL), the organic phase was dried (NaZSO4) and removal of solvent afforded 20g of an orange solid. The crude product was purified by filtration through a column of silica gel (60-230 mesh, 100g, methylene chloride-hexanes, 1:1) followed by recrystallization from CCl -CH C1 to give 13.9g (66%) 4 2 2 1 of 36 as yellow-green flakes (MP = 59°C). H NMR (60 MHz) 6 7.76 (s, 1H), 4.97 (S, 2H); IR (CHC13) 3065, 1785, 1620, 1; EI-MS (70eV) 164 (M+, 30 0), -1360, 1230, 1170, 1068, 1003 cm” 162 (M+, 33.6), 135 (70.0), 133 (77.6), 107 (40.9), 105 (42.0), 39 (base). 2-Trimethy1siloxy-3-bromofuran 37. To the Z-bromobutenolide 36 (1.50g, 9.3 mmol) in dry THF (70 mL) was added a catalytic amount of zinc chloride (10 mg) followed by triethylamine (1.6 mL, 11.0 mmol). The solution was stirred at 25°C for 1.5h. After the suspended solids had settled, the solution was transferred via syringe to a dry flask, and the solvent was removed by distillation. Continued distillation under reduced pressure afforded 0.82g (37%) of 37 as a cloudy liquid 1 (BP10mm = 65 C). (d, J=3Hz, 1H), 0.33 (S, 9H); IR (neat) 2980, 1645, 1522, 1 H NMR (60 MHz) 6 6.78 (d, J=3Hz, 1H), 6.23 1380, 1260, 1065, 865 cm- ; EI-MS (70eV) 164 (M+, 15.4), 162 (M+, 15.9), 135 (41.9), 133 (42.2), 107 (21.2), 105 (23.9), 85 (base). 0( SC 84 2-tButyldimethylsiloxy-3-bromo-furan 38. To a solution of the 2-bromobutenolide 36 (3.0g, 18.5 mmol) in dry THF (14 mL), under argon was added a catalytic amount of zinc chloride (10 mg) followed by tbutyldimethylsilyl chloride (2.88g, 19.1 mmol) and triethylamine (2.76 mL, 19.1 mmol). The solution was stirred at 25°C for 4.5h. After removal of most of the solvent Q mug, the residue was diluted with EtZO-pentane, 1:1 (150 mL) and washed with saturated aqueous NaHCO (2 x 3 150 mL). The discolored organic phase was shaken with a gener- ous amount of activated charcoal followed by filtration through celite. The solution was dried (Na2804) and removal of solvent gave a red fluid which was purified by bulb-to-bulb (Kugelrohr) distillation, oven temp. = 100°C (1 mm), to afford 2.5g (49%) of 38. 1 H NMR (60 MHZ) 6 6.76 (d, J=3Hz, 1H), 6.23 (d, J=3Hz, 1H), 1.00 (m, 9H), 0.27 (s, 3H), 0.50 (s, 3H); IR (neat) 2990, 2950, 2890, 1640, 1520, 1480, 1380, 1255, 1065, 1010, 860 om'1; EI-MS (70eV) 278 (11*, 6.7), 276 (11", 7.9), 193 (3.5), 191 (3.9), 139 (5.1), 137 (5.0), 73 (base). Reaction of Lithiated 38 with Benzaldehyde. Preparation of 39. To a solution of 2-tbutyldimethylsiloxy-3-bromo-furan 38 (0.28g, 1.0 mmol) in dry Et 0 (4 mL) was added n-butyl 2 lithium (1.3 M in hexane, 0.77 mL, 1.0 mmol) dropwise at -78°C. The solution was warmed to 0°C and stirred for 30 minutes, then benzaldehyde (0.1 mL, 1.0 mm) in dry Et 0 (1 mL) was 2 added over 5 minutes. The reaction was stirred for 1h at 0°C then quenched with saturated aqueous NaHCO:3 (1 mL). The solution was diluted with Et20-pentane, 1:1 (50 mL) and washed 85 with saturated aqueous NaHCO3 (2 x 50 mL). The organic phase was dried (Na 804) and removal of solvent 11; vacuo afforded ‘ 1 0.27g (90%) of 39 as an orange-yellow liquid. H NMR (60 MHZ) 6 7.26 (m, 5H), 6.69 (d, J=2Hz, 1H), 6.09 (d, J=2Hz, 1H), 5.68 (S, 1H), 2.10 (m, 1H), 0.98 (m, 9H), 0.24 (m, 6H); IR (neat) 3410, 2950, 2890, 1645, 1530, 1420, 1260, 1010, 912, 1 860, 796, 707 om'; EI-MS (70eV) 304 (11*, 4.5), 287 (1.1), 247 (11.4), 189 (0.4), 172 (26.1), 75 (base). 2-Methylthio-5-nbutyl-furan 41. To a solution of n-butyl lithium (2.6 M in hexane, 15.3 mL, 40 mmol) and tetramethyl- ethylene diamine (6.03 mL, 40 mmol) in dry Et 0 (50 mL) at 2 0°C (ice water) was added 2-n-butyl furan-40 (5.0g, 40 mmol) via syringe. The resulting solution was warmed to 25°C and stirred for 30 minutes followed by cooling to 0°C (ice water) and addition of dimethyl disulfide (3.6 mL, 40 mmol) via sy- ringe. The resulting solution was allowed to warm to 25°C with stirring overnight. The reaction was then diluted with Et 0 (300 mL) and washed with water (300 mL), 5% aqueous NaOH 2 (300 mL), and saturated aqueous NH Cl (300 mL). The organic 4 phase was dried (Na2S04) and removal of solvent in vacuo fol- lowed by distillation, BP5mm = 90°C, afforded 5.0g (73%) of 1 41 as a yellow liquid. H NMR (60 MHz) 6 6.25 (d, J=3Hz, 1H), 5.85 (m, 1H), 2.55 (t, J=7Hz, 2H), 2.29 (S, 3H), 1.44 (m, 4H), 0.90 (m, 3H); IR (Heat) 2980, 2910, 1600, 1510, 1465, l 1130, 1020, 985, 805 cm’; EI-MS (70eV) 170 (11*, 27.5), 127 (base), 113 (5.0), 99 (12.0). 0v an pt p1 86 Attempted Hydrolysis of 41. Preparation of 42. A solution of 10% sulfuric acid (5 mL), 95% ethanol (5 mL), and 2-methy1- thio-5-nbuty1-furan 41 (0.5g, 2.9 mmol) was heated under reflux overnight, cooled to 25°C, then diluted ‘with Et 0* (50 mL) 2 and washed with water (50 mL). After drying, the .organic phase (Na2S04) and removal of solvent provided 0.44g of crude product. Purification by chromatography on a column of silica gel (60-230 mesh, 50g, EtZO-hexane, 1:9) provided 0.40g (75%) 1 of 42 as a colorless liquid. H NMR (60 MHz) 6 4.11 (q, J=6Hz, 2H), 2.50 (m, 2H), 1.40 (m, 8H), 1.00 (m, 6H); IR (neat) 3000, 1; EI-MS (70eV) 186 (M+, 2910, 1750, 1380, 1200, 1040, 830 cm- 2.8), 171 (0.35), 156 (2.6), 141 (26.6), 101 (52.2), 85 (61.5), 57 (base). 2-Methy1thio-3-hydroxymethy1-furan 44 and 2-Methy1thio-4- hydroxymethyl-furan 45. To a solution of n-butyl lithium (2.6 M in hexane, 48.0 mL, 125 mmol) and tetramethylene diamine (18.9 mL, 125 mmol) in dry Et 0 (150 (mL) at 0°C (ice water) 2 was added 3-hydroxymethy1-furan -(5.9g, 60 mmol) in dry Et20 (5 mL) via syringe. The solution was warmed to 25°C and stirred for 1h, during which time a white precipitate formed. The mixture was cooled to 0°C and dimethyl disulfide (5.37 mL, 60 mmol) was added dropwise via syringe. The solution was slowly warmed to 25°C while stirring overnight. The reaction mixture was diluted with Et 0 (150 mL) and washed with water 2 (250 mL), 10% aqueous NaOH (250 mL), 1 N aqueous HCl (250 mL). The organic phase 'was dried (Na2804) and removal of solvent in vacuo afforded 6.0g crude product which was purified 87 by distillation through a vigreux column, BPO.2mm = 75°C, to yield 4.64g (53%) of 2-methy1thio-3-hydroxymethy1-furan 44 as a colorless liquid. A second fraction from the distill- ation, BPO.2mm = 85°C yielded less than 5% of the regioisomeric 2-methylthio-4-hydroxymethy1-furan 45. The following is spec- 1H NMR (60 MHz) 6 7.35 (d, J=2Hz, 1H), troscopic data for 44. 6.40 (d, J=2Hz, 1H), 4.50 (bds, 2H), 2.55 (m, 1H), 2.35 (s, 3H); IR (neat) 3400, 2955, 2910, 1495, 1435, 1320, 1150, 1055, 1010, 900, 775, 745 om‘1; EI-MS (70eV) 144 (11", 50.1), 128 (base). 2-Methylthio-3-chlormethyl-furan 46. On the same scale and utilizing a procedure identical to that used in the prepar- ation of chloride 24, 46 was prepared from alcohol 44 in 79% crude yield as a slightly yellow liquid. The chloride could not be purified by distillation due to its thermal instability; however, 1the crude material was found to be sufficiently pure 1H NMR (60 MHz) 6 7.34 to be utilized in subsequent reactions. ((1, J=2Hz, 1H), 6.42 (d, J=2Hz, 1H), 4.50 (s, 2H), 2.33 (s, 3H); IR (neat) 3000, 2970, 2920, 1500, 1450, 1290, 1155, 1065, 730 om'1; EI-MS (70eV) 164 (11*, 20.7), 162 (11", 44.0), 149 (4.6), 127 (base). Reaction of Grignard Reagent of 46 with Epoxygeranyl chlo- ride;51’52 Preparation of 2,6-Dimethy1-6-vinyl-7-(2-methyl- thio—3-furyl)-hept-2-ene-2,3-epoxide 47. In a Grignard entrain- ment procedure, to magnesium powder (0.15g, 6.2 mmol) in dry THF (10 mL) was added a solution of 2-methylthio-3-chloro- methyl-furan 46 (0.50g, 3.1 mmol) and ethylene dibromide (0.27 88 mL, 3.1 mmol) in dry THF (20 mL) over 1h so that reflux was maintained. ‘After the addition was complete, the solution was stirred at 60°C until most of the magnesium had been con- sumed (30 min.). The solution was then cooled to 0°C (ice catalyst (0.1 M in THF, 0.3 mL, 0.03 mmol) 51,52 4 was added followed by epoxygeranyl chloride water) and LiZCuC1 (0.58g, 3.1 mmol) in dry THF (2 mL). The resulting solution was warmed to 25°C over the course of 1h. After most of the solvent was removed .1_!_1_ M112: the residue was diluted with EtZO/pentane, 1:1 (100 mL) and washed with water (100 mL), saturated aqueous NH Cl (100 mL) and saturated aqueous NaHCO (100 mL). The 4 3 'organic phase was dried (Na2804) and removal of solvent afforded 0.90g of crude product. The crude product was purified by chromatography on a column of silica gel (60-230 mesh, 100g, EtZO-hexane, 2.5:7.5) to provide 0.18g (21%) of 47, R 0.51 f (EtZO-hexane, 1:9). Isolation of a lower-lying component provided 0.50g (64%) of dimerized 46, the Wurtz product 48. 1H NMR (60 MHZ) 6 7.36 (m, 1H), 6.36 (m, 1H), 5.81 (dd, J=18,10Hz, 1H), 5.14 (dd, J=14,2Hz, 1H), 4.92 (dd, J=8,2Hz, 1H), 2.71 (bdt, J=8Hz, 1H), 2.33 (m, 3H); IR (neat) 2990, 2950, 1615, 1495, 1450, 1390, 1375, 1320, 1250, 1190, 1150, 1120, 1080, 1030, 1005, 980, 925, 900, 710 cm_1; EI-MS (70eV) 264 (M+-16, 4.1), 127 (97.2). 2-Methylthio-3-tri-n-butylstannylmethyl-furan 49. According to the procedure of Still38 to LDA prepared in the usual way (3.5 mmol), in dry THF (5 mL) chilled to 0°C (ice water) was added tri-n-butyl stannane (0.82 mL, 3.1 mmol) and the resulting SO tr to re IC 111 111 Ne re CI 31' 89 solution was stirred for 15 minutes at 0°C. The resulting tri-n-butyl stannyl lithium solution was then added via syringe to 2-methylthio-3-chloromethyl-furan 46 (0.50g, 3.1 mmol) in dry THF (5 mL) cooled to -25°C (dry ice-isopropanol). The resulting solution was stirred at -25°C for 1h, then. warmed to 25°C over the course of 1h. The solution was then diluted with Et20-pentane, 1:1 (150 mL) and washed with water (100 mL), saturated aqueous NH4C1 (100 mL) and saturated aqueous NaHCO3 (100 mL). The organic phase was dried (Na2S04) and removal of solvent Q 13.9.1.9. afforded 1.3g crude product. The crude product was purified by chromatography on a column of silica gel (60-230 mesh, 50g, hexanes) to provide 0.38g (31%) l of 49, Rf 0.23 (pentane). 1H), 6.03 (d, J=2Hz, 1H), 2.25 (S, 3H), 2.03 (S, 2H), 1.33 H NMR (60 MHz) 6 7.20 (d, J=2Hz, (m, 18H), 0.91 (m, 9H); EI-MS (709V) 363 (5.1), 362 (4.7), 361 (36.8), 360 (12.8), 359 (26.5), 358 (9.8), 357 (14.5), all M+-Bu; 127 (base, M+-SnBu3). 2,6-Dimethyl-9-(2-methylthio-3-furyl)-non-6-ene-2,3-oxirane 50. To a solution of 2-methy1thio-3-tri-n-buty1-stannylmethyl- furan 49 (0.32g, 0.8 mmol) in dry THF (3 mL) cooled to -78°C (dry ice-isopropanol) was added nBuLi (2.6 M in hexane, 0.31 mL, 0.8 mmol). The solution was stirred for 30 minutes at -78°C and after warming to 0°C (ice water) a solution of anhy- drous magnesium bromide (0.18g, 1.0 mmol) in dry THF was added CuCl followed by Li 4 catalyst (0.1 M in THF, 0.1 mL, 0.01 51,52 2 mmol). Epoxygeranyl chloride (0.15g, 0.8 mmol) in dry THF (1 mL) was then added via syringe and the solution was 90 warmed to 25°C over the course of 1h. After most of the THF had been removed in vacuo, the residue was diluted with Et20- pentane, 1:1 (50 mL) and washed with water (50 mL), saturated aqueous NH Cl (50 mL) and saturated NaHCO (50 mL). Drying 4 3 of the organic phase (Na2S04) and removal of solvent afforded 0.40g crude product. The crude material was purified by chroma- tography on a column of silica gel (60-230 mesh, 50g, EtZO- hexanes, 2.5:97.5) to afford 0.02g (9%) of 50, R 0.31 (Et20- 1 f hexanes, 1:9). H NMR (60 MHz) 6 7.35 (d, J=2Hz, 1H), 6.24 (d, J=2Hz, 1H), 5.03 (bdm, 1H), 2.66 (m, 1H), 2.20 (m, 3H), 1.97 (m, 2H), 1.60 (m, 9H), 1.24 (m, 611); IR (neat) 2960, 2890, 1490, 1450, 1385, 1320, 1255, 1150,’ 1115, 1080, 1060, 1; EI-MS (70eV) 264 (M+-16, 9.9), 207 (17.2), 980, 900, 755 om' 127 (base). Reaction of Lithiated 49 'with Benzaldehyde. Preparation of 51. To a solution of 2-methy1thio-3-tri-n-butylstannylmethyl furan 49 (0.33g, 0.8 mmol) in dry THF (5 mL) was added nBuLi (2.6 M in hexane, 0.31 mL, 0.8 mmol) at -78°C (dry ice-isoprop- anol). The solution was stirred at -78°C for 30 minutes during which time it became dark orange in color. Benzaldehyde (0.085g, 0.8 mmol) in dry THF (1 mL) was then added via syringe and the solution changed to a light yellow color. Stirring was continued and the temperature 'was raised to 25°C over the course of 1h. The solution was then diluted with Et 0- 2 pentane, 1:1 (100 mL) and washed with H 0 (100 mL), saturated 2 aqueous NH4C1 (100 mL) and brine (100 mL). The organic phase was dried (NaZSO) and removal of solvent _i_n vacuo afforded 4 ing The fit di 118 in Th by tl De 11' tc 52 91 a two-phase solution. The upper (colorless) phase (tin contain- ing’ products) was separated from the lower (orange) phase. The lower layer provided 0.15g (80%) of 51 which was not puri- fied further. 1H NMR (60 MHz) 6 7.25 (bds, 6H), 6.15 (m, 1H), 4.80 (m, 1H), 3.40 (m, 1H), 2.80 (m, 2H), 2.20 (s, 3H); EI-MS (70eV) 233 (M+, 32.4), 217 (12.1), 127 (base). V Preparation of 2-Pheny1thio-3-hydroxymethyl-furan 52 and 2-Phenylthio-4-hydroxymethyl-furan 53. To a solution of nBuLi (10.2 M in hexane, 30.3 mL, 0.31 mmol) and tetramethylethylene- diamine (47.1 mL, 0.31 mmol) in dry Et 0 (125 mL) at 0°C (ice 2 water) was added 3-hydroxymethyl-furan (14.7g, 0.15 mmol) in dry Et20 (30 mL) via syringe over a period of 15 minutes. The solution was then warmed to 25°C and stirred for 1h followed by cooling to 0°C. A solution of diphenyl disulfide (34.3g, 0.16 mmol) in dry Et20 (125 mL) was then added over the course of 1h. The resulting solution was gradually warmed to 25°C and stirred overnight. The reaction was diluted. with Et20 (1000 mL) and washed with water (750 mL), 1N HCl (750 mL), 1N NaOH (750 mL), water (750 mL) and brine (750 mL). The organic phase was dried (Na2804) and removal of solvent _in mug gave a cloudy, viscous orange oil which was filtered through a column of silica gel (60-230 mesh, 250g, EtOAc- petroleum ether, 3:7) to provide 25.9g of a golden yellow liquid. The product mixture was purified by preparative HPLC (Waters Prep 500, 2-columns, EtOAc-hexanes, 3:7, 250 mL/min.) to afford 20.7g (67%) of 2-pheny1thio-3-hydroxymethyl-furan 52 and 4.6g (15%) of 2-pheny1thio-4-hydroxymethyl-furan 53 In In 92 1 as colorless liquids. Spectral data for 52: H NMR (60 MHz) 6 7.42 (d, Jf-ZHZ, 1H), 7.06 (S, 5H), 6.48 (d, J=2Hz, 1H), 4.52 (S, 2H), 2.28 (bds, 1H); IR (neat) 3400, 2960, 2900, 1585, 1510, 1480, 1445, 1395, 1165, 1150, 1075, 1055, 1030, 900, 880, 805, 750, 700 om‘1; MS (El/70eV) 206 (11", base), 189 (9.5), 176 (6.6), 160 (22.0), 115 (33.9), 69 (46.2). Spectral data for 53: 1 H NMR (60 MHZ) 6 7.37 (s, 1H), 7.12 (s, 5H), 6.63 (s, 1H), 4.42 (s, 2H), 2.65 (bds, 1H); IR (neat) 3400, 2970, 2910, 1585, 1505, 1480, 1445, 1390, 1175, 1130, 1075, 1030, 985, 925, 880, 750, 700 cm-1; EI-MS (70eV) 206 (M1, base), 188 (11.9), 176 (11.6), 160 (28.1), 115 (49.5), 69 (57.4). 2-Phenylthio—3-chloromethyl-furan 54. On the same scale and utilizing a procedure identical to that used in the prepar- ation of chloride 24, 54 was prepared from alcohol 52 in 94% 1 yield (BP 85°C). H NMR (60 MHz) 6 7.43 (d, J=2Hz, .007mm 1H), 7.12 (S, 5H), 6.52 (d, J=2Hz, 1H), 4.53 (s, 2H); IR (neat) 3900, 2890, 1590, 1490, 1485, 1445, 1285, 1265, 1150, 1120, 1; EI-MS (70ev) 226 (M+, 36.3), 224 1060, 1035, 892, 770 cm- (M+, base), 189 (87.5), 161 (47.3), 128 (63.0). 2-Pheny1thio-3-tri-n-butylstannylmethyl-furan 55. On the same scale and utilizing a procedure identical to that used in the preparation of stannane 49, 55 was prepared from chloride 54 in 66% yield. 1 H NMR (60 MHZ) 6 7.38 (d, J=2Hz, 1H), 7.06 (m, 5H), 6.19 (d, J=2Hz, 1H), 2.09 (S, 2H), 1.09-1.48 (m, 12H), 0.80-0.96 (m, 15H); IR (neat) 2980, 2950, 2880, 1575, 1480, 1465, 1445, 1380, 1265, 1165, 1100, 1030, 1000, 895, 93 1; EI-MS (70ev) 421 (M+-Bu, 85.2), 367 (5.1), 760, 745, 695 cm- 309 (6.0), 291 (21.3), 235 (48.2), 179 (base). Reaction of Lithiated 55 with Benzaldehyde. Preparation of 56 and 57. On the same scale and utilizing a procedure identical to that used in the reaction of stannane 49 with benzaldehyde, 2-phenylthio-3-tri-n-butylstanny1methy1 furan 55 was lithiated and alkylated with benzaldehyde to provide the crude product as a pale yellow oil. Purification of the crude produce was accomplished by chromatography on silica gel to provide 56 (38% yield) and 57 (37% yield). Spectral 1H NMR (60 MHz) 6 7.43 (d, J=2Hz, 1H), 7.24 (s, data for 56: 5H), 7.05 (m, 5H), 6.29 ('d, J=2Hz, 1H), 4.77 (t, J=6Hz, Ph-Cfi-OH), 3.89 (s, -0H), 2.96 (d, J=6Hz, f-QHZCHOH); EI-MS (70eV) 296 (111-, 20.2), 190 (base), 161 (21.0), 129 (37.9), 107 (43.2). SpectrLl data for 57: 1H NMR (60 MHZ) 6 7.23 (m, 11H), 6.27’ (d, J=2Hz, 1H), 5.75 (brs, 1H, f-Cfl(0H)Ph), 3.86 (S, 2H, f-Cfiz-SPh), 2.23 (brs, 1H, -O_Ii); EI-MS (706V) 296 (M+, 5.8), 278 (2.7), 186 (base), 109 (99.4). 2-Pheny1thio—3-decyl-furan 58. To a solution of 2-phenyl- thio-3-tri-n-buty1stannylmethyl-furan 55 (0.96g, 2.0 mmol) in dry THF (5 mL) was added nBuLi (1.8 M in hexane, 1.1 mL, 2.0 mol) at -78°C (dry ice-isopropanol). The solution was stirred at -78°C for 30 minutes and then HMPA (0.35 mL, 2.0 mmol) was added via syringe. The solution was warmed to -25°C (dry ice-isopropanol) and nonyl-iodide (0.5lg, 2.0 mmol) in dry THF (1 mL) was added via syringe. The resulting solution was stirred for 1h at -25°C then diluted with Et20 (50 mL) 94 and washed with saturated aqueous NH4C1 (50 mL), saturated aqueous NaHCO (50 mL) and brine (50 mL). The organic phase was dried (Na2804) and removal of solvent afforded 1.0g of crude product. The desired product was separated from tin by-products and unreacted nonyl-iodide by chromatography on a column of silica gel (60-230 mesh, 100g, hexane) to provide 0.23g (39%) of 58. 1 H NMR (60 MHz) 6 7.33 (d, J=2Hz, 1H), 6.98 (m, 5H), 6.23 (d, J=2Hz, 1H), 2.47 (bdt, J=6Hz, 2H), 1.23 (bds, 16H), 0.88 (m, 3H); EI-MS (70eV) 316 (11*, base), 302 (11.8), 189 (28.5), 161 (30.0), 128 (30.4). 2,6—Dimethy1-9-(2-phenylthio-3-furyl)-2,6-nonadiene 59. To a solution of 2-phenylthio-3-tri-n-butylstannylmethyl- furan 55 (0.96g, 2.0 mmol) in dry THF (3 mL) was added nBuLi (1.8 M in hexane, 1.1 mL, 2.0 mmol) at -78°C (dry ice-iSOpropa- n01). The solution was stirred at -78°C for 30 minutes and then was added dropwise via a syringe jacketed with dry ice held in place by aluminum foil to a solution of geranyl chlor- ide (0.35g, 2.0 mmol) and HMPA (0.35 mL, 2.0 mmol) in dry THF (3 mL) at -25°C. The resulting solution was stirred at -25°C for 1h then diluted with Et 0 (50 mL) and washed 2 with saturated aqueous NH Cl (50 mL), saturated aqueous NaHCO 4 3 (50 mL) and brine (50 mL). The organic phase was dried (Na2804) and removal of solvent in M afforded 1.0g of crude product which was purified by flash chromatography on a column of silica gel (230-400 mesh, 100g, hexane) to I afford 0.39g (59%) of 59. H NMR (60 MHZ) 6 7.34 (d, J=2Hz, 1H), 7.03 (m, 5H), 6.28 (d, J=2Hz, 1H), 5.04 (bdm, 2H), 1.92- Ix‘! 95 2.17 (m, 8H), 1.50-1.64 (m, 9H); EI-MS (70eV) 326 (M+, 29.8), 217 (2.6), 202 (2.9), 189 (base), 161 (30.9), 128 (29.3). 3-Tetrahydropyranyloxymethy1-furan 60. To a solution of 3-hydroxymethyl-furan (2.5g, 25.5 mmol) in dry methylene chloride (100 mL) was added dihydropyran (3.2g, 38.5_ mmol) and pyridinium toluenesulfonate (0.64g, 2.55 mmol). The resulting solution was stirred at 25°C for 1.5h. The majority of the solvent was then removed _i_n v_a_c_1_l_9_ and the residue was diluted with Et 0 (150 mL), washed with water (150 mL), 2 brine (150 mL) and dried (Na2S04). Removal of solvent in vacuo afforded a colorless liquid which was purified by distil- - o ' ' 1 lation, Bp0.17mm - 65 C, to yield 4.42g (95%) of 60. NMR (60 MHZ) 6 7.25 (m, 2H), 6.30 (d, J=2Hz, 1H), 4.64 (m, H 1H), 4.59 (d, J=llHZ, 1H, AB), 4.30 (d, J=11HZ, 1H, AB), 3.30-4.00 (bdm, 2H), 1.40-1.70 (bdm, 6H); IR (neat) 2940, 2860, 1560, 1450, 1270, 1140, 1035 cm-1; EI-MS (70eV) 182 (M+, 0.6), 161 (0.6), 98 (9.6), 85 (38.2), 81 (base). 3-Phenylthio—4-hydroxymethyl-furan 61. On the same scale and utilizing a procedure identical to that used in the preparation of 52, 2-trimethylsilyl-3-hydroxymethy1-furan 63 was sulfenylated to give a crude product that was purified by chromatography on a column of silica gel (230-400 mesh, EtZO-petroleum ether, 2:8). Two major components were iso- lated: 2-trimethylsilyl-4-pheny1thio-3-hydroxymethyl-furan (69%), (Rf 0.60, EtZO-petroleum ether, 1:1) and 2-phenylthio- 5-trimethylsilyl-4-hydroxymethyl-furan (23%), (Rf 0.32, Et 0- 2 petroleum ether, 1:1). The nature of the 2,3,4-substituted 96 regioisomer was demonstrated upon analysis of material that was desilylated as follows. The major product above was stirred with stOH (1 eq.) in dry CH CN at 25°C for 4h. After 3 most of the solvent was removed Q vacug, the residue was diluted with Et20 and washed with saturated aqueous NaHCOB. The organic phase was dried (Na2804), followed by removal of solvent in vacuo to provide 3-phenylthio-4-hydroxymethy1- furan 61 (75%) with the following spectral data. 1H NMR (60 MHZ) 6 7.53 (m, 1H), 7.42 (m, 1H), 7.11 (S, 5H), 4.34 (brs, 2H), 2.09 (brs, 1H); IR (neat) 3420, 2970, 2910, 1590, 1520, 1484, 1445, 1145, 1095, 1035, 885, 820, 755, 705, 645 om’1; EI-MS (70ev) 206 (11*, base), 187 (21.8), 159 (9.8), 147 (8.4), 134 (7.3), 128 (65.8). 2-Phenylthio—4-chloromethy1-furan 62. On the same scale and utilizing a procedure identical to that used in the prepar- ation of chloride 24, 62 was prepared from alcohol 53 in 1 67% yield after distillation, BP '= 115°C. H NMR (60 .005 MHZ) 6 7.48 (S, 1H), 7.19 (s, .5H), 6.72 (s, 1H), 4.37 (S, 2H); IR (neat) 3900, 2890, 1580, 1495, 1485, 1450, 1280, 1 1150, 1050, 890, 760 cm”; EI-MS (79eV) 226 (11*, 9.9), 224 (11", 32.3), 177 (6.2), 161 (13.3), 149 (18.8), 128 (18.5), 115 (9.5), 40 (base). 2-Trimethylsilyl-3-hydroxymethyl-furan 63. To a suspen- sion of lithium aluminum hydride (0.11g, 3.0 mmol) in dry Et 0 (10 mL) at 0°C (ice water) was added 2-trimethylsilyl- 2 3-furoic acid (0.36g, 2.0 mmol) in dry Et 0 (5 mL) over 15 2 minutes. The mixture was allowed to warm to 25°C and was 97 stirred for 2h. The reaction mixture was cooled to 0°C and quenched with water (10 mL), followed by dilution with Et 0 2 (25 mL) and washing with water (2 x 25 mL). The organic phase was dried (Na2804) and removal of solvent afforded a slightly yellow oil which was distilled, BP5mm =, 90°C, to afford 0.26g (76%) of 63 as a colorless liquid. 1H NMR (60 MHZ) 6 7.46 (d, J=2Hz, 1H), 6.37 (d, J=2Hz, 1H), 4.51 (S, 2H), 2.28-2.40 (bdm, 1H), 0.29 (bds, 9H); IR (neat) 3390, 2995, 2935, 1580, 1490, 1425, 1265, 1120, 1060, 1010, 860, l 775 cm- ; EI-MS (70eV) 170 (M+, 28.3), 153 (base), 137 (40.2), 77 (86.1). 2-Trimethylsilyl-3-chloronethyl-furan 64. . 0n the same scale and utilizing a procedure identical to that used in the preparation of chloride 24, 64 was prepared from alcohol 1 63 in 69% yield after distillation, BP8mm = 75°C. (60 MHZ) 6 7.43 (d, J=2Hz, 1H), 6.34 (d, J=2Hz, 1H), 4.51 H NMR (S, 2H), 0.29 (bds, 9H); IR (neat) 2960, 2900, 1565, 1485, 1450, 1380, 1255, 1175, 1105, 1055, 920, 860, 735 cm-1; EI-MS (706V) 190 (3.4), 188 (9.0), 173 (64.7), 153 (base), 137 (17.2), 95 (39.4). 2,6-Dimethyl-9-(2-trimethylsilyl-3-furyl)-nona-2,6-diene- 2,3-epoxide 65. To a solution of dry 2-trimethylsilyl-3- chloromethyl-furan 64 (4.08g, 22.0 mmol) in dry THF (50 mL) was added magnesium powder (0.53g, 22.0 mmol), with stirring at 25°C. The solution was heated to 50°C and Grignard forma- tion soon became initiated. After 45 minutes at 50°C, most of the magnesium had been consumed producing a slightly turbid shc [0 111 ch re of re by 111 wa 3E 98 blackish-brown colored solution. Titration26 of an aliquot showed Grignard formation to be 77%. The solution was cooled to 0°C (ice water) and Li 2CuCl4 catalyst (0.1 M in THF, 1.0 mL, 0.1 mmol) was added followed by a solution of epoxygeranyl chloride51’52 (3.14g, 16.7 mmol) in dry THF (10 mL). The resulting solution was slowly warmed to 25°C over the course of 1h. Most of the solvent was removed _i__n_ vacuo and the residue was diluted with EtZO-pentane, 1:1 (200 mL) followed by washing with water (200 mL), saturated aqueous NH Cl (200 4 mL) and saturated aqueous NaHC03 (200 mL). The organic phase was dried (Na2804) and removal of solvent afforded 5.5g of .an orange oil which was distilled, BP = 110°C, to afford 007mm 4.76g (93%, based on epoxygeranyl chloride) of 65 as a slightly 1H NMR (60 MHz) 6 7.44 (d, J=2Hz, 1H), 6.21 yellow liquid. (d, J=2Hz, 1H), 5.18 (m, 1H), 2.00-2.80 (bdm, 5H), 1.64 (bds, 7H), 1.27 (s, 6H), 0.27 (m, 9H); IR (neat) 2900, 2890, 1575, 1465, 1390, 1260, 1110, 1065, 860, 770 cm-1; EI-MS (70eV) 306 (M+, 2 9), 291 (1.4), 233 (3.6), 207 (15.3), 153 (base). Attempted Cyclization of 65. Preparation of 66. To a solution of 65 (0.10g, 0.33 mmol) and triethylamine (0.07 mL, 0.50 mmol) in dry methylene chloride (10 mL) at 0°C (ice water) was added freshly prepared magnesium bromide etherate (0.23g, 0.99 mmol). The solution was stirred at 0°C for 20 minutes then warmed to 25°C. The solution was then diluted with Et 0 (50 mL), washed with water (50 mL) and saturated 2 aqueous NH4C1 (50 mL). The organic phase was dried (Na2804) and removal of solvent _ip vacuo afforded 0.1g crude product 99 which was purified by chromatography on a column of silica gel (60-230 mesh, 15g, EtZO-petroleum ether, 3:7) to afford 30 mg (30%)- of uncyclized ketone, Rf 0.64 (EtZO-petroleum ether, 1:1) and a mixture of two unseparated components which were separated by analytical HPLC. The mixture was found to consist of 10 mg (10%) of cyclized desilylated material, R 0.17, and 20 mg (20%) of 66, R l f 0.24, with the following H NMR (250 MHZ) 6 6.39 (s, 1H), 3.30 (dd, f spectral data . J=11.3,5.0Hz, 1H), 2.47‘ (dd, J=12.5,l6.3Hz), 2.20 (dt, J=12.5,3.8Hz, 1H), 1.50-1.90 (bdm, 7H), 1.19 (s, 3H), 1.05 (s, 3H), 0.87 (s, 3H), 0.21 (s, 9H); IR (neat) 3440, 2990, 2890, 1505, 1465, 1385, 1260, 1100, 1035, 940, 855, 765 cm-1; EI-MS (70eV) 306 (M+, 21.6), 291 (49.5), 273 (27.5), 219 (13.7), 201 (16.9), 73 (base). Silylation of 3-Hydroxymethyl-furan. Preparation of 2 , 5-Bis-trimethylsi1yl-3-hydroxymethyl-furan 67 . To LDA (52.0 mmol), prepared in the usual way, in dry THF (50 mL) at 0°C (ice water) was added 3-hydroxymethyl-furan (2.5g, 26.0 mmol) in dry THF (10 mL) via syringe. The solution was stirred for 45 minutes at which time the solution became a thick, pale green suspension. Trimethylsilyl chloride (8.0 mL, 63.0 mmol) neat was then added via syringe and the resulting solution was warmed to 25°C over 2h. Most of the solvent was removed _i__n_ 339112 and the residue was diluted with Et 0 (100 mL), followed by washing with water (100 mL) 2 and saturated aqueous NH4C1 (100 mL). The organic phase was dried (Na2S04) and removal of solvent in vacuo afforded wi an th m t( (1' 100 6.33g of crude product which was directly submitted to desilyl- ation conditions, as follows, without further purification. A solution of the crude product and sodium methoxide (2.80g, 52.0 mmol) in methanol (100 mL) was stirred at 25°C for 15 minutes. After careful removal of most of the solvent it; M, the residue was diluted with Et20 (100 mL) and washed with water (2 x 100 mL). The organic phase was dried (Na2S04) and removal of solvent i_n_ vacuo followed by distillation through a vigreux column afforded 0.61g (24%) of recovered 3-hydroxymethyl-furan, 1.64g (37%) of 2-trimethylsilyl-3- hydroxymethyl-furan 63, and 0.76g (12%) of 67, BP 1 5mm = 108 C, with the following spectral data.- H NMR (60 MHz) 6 6.58 (s, 1H), 4.52 (S, 2H), 2.07 (bds, 1H), 0.29 (bds, 18H); IR (neat) 3400, 2990, 2890, 1580, 1420, 1255, 1090, 1015, 940, 855, 765, 705 cm'1 ; EI-MS (70eV) 242 (M1, 26.4), 227 (73.9), 147 (39.4), 133 (28.9), 75 (base). Silylation of 3-Trimethylsiloxymethyl-furan. Preparation of 2-Trimethylsily1-4-hydroxymethy1-furan 68. To LDA (39.0 mmol), prepared in the usual way, in dry THF (100 mL) was added 3-trimethy1siloxymethyl-furan (6.6g, 39.0 mmol) in dry THF (10 mL) via syringe. The resulting solution was stirred for 10 minutes at 0°C during which time it took on a dark golden brown color. Trimethylsilyl chloride (4.95 mL, 39 mmol) was added followed by gradual warming to 25°C overnight. The reaction mixture was worked up and subjected to silyl ether cleavage as before (see the procedure for the preparation of 67) to afford 5.59g of crude product. 101 The crude material was purified by chromatography on a column of silica gel (230-400 mesh, 400g, EtZO-petroleum ether, 15:85) to afford 2.27g (34%) of 2-trimethylsilyl-3-hydroxymethyl-furan 63, Rf 0.39 (EtZO-petroleum ether, 1:1); 1.47g (16%) of 2,5-bis-trimethylsi1y1-3-hydroxymethyl-furan 67, Rf 0.50; and 0.65g (10%) of 68, Rf 0.28, with the following spectral data. 1H NMR (60 MHz) 6 7.28 (s, 1H), 6.58 (s, 1H), 4.45 (s, 2H), 2.36 (bds, 1H), 0.29 (bdm, 9H); IR (neat) 3320, 2955, 2880, 1595, 1465, 1405, 1250, 1155, 1070, 1010, 975, 1 905, 835, 755 cm’ ; EI-MS (70eV) 170 (M+, 58.3), 155 (96.7), 127 (42.4), 97 (45.2), 75 (base). 2-Phenylthio—3-tbutyldimethylsiloxymethyl-furan 69. To a solution of 2-pheny1thio-3-hydroxymethy1-furan 52 (5.65g, 27.4 mmol) and imidazole (4.66g, 68.5'mmol) in dry DMF (25 mL) was added t-butyldimethylsilyl chloride (4.97g, 32.9 mmol). The resulting solution was stirred at 25°C for lb and then diluted with Et 0 (150 mL) followed by washing with 2 water (100 mL) and saturated aqueous Cu(N03)2 (100 mL). The organic phase was dried (Na2S04) and removal of solvent in vacuo afforded a yellow liquid which was purified by filtra- tion through a column of silica gel (60-230 mesh, 35g, EtZO- petroleum ether, 1:9) to afford 7.20g (81%) of 69 as a slightly 1 yellow liquid. H NMR (60 MHZ) 6 7.53 (d, J=2Hz, 1H), 7.17 (bds, 5H), 6.60 (d, J=2Hz, 1H), 4.68 (s, 2H), 1.00 (S, 9H), 0.14 (m, 6H); IR (neat) 2970, 2880, 1590, 1485, 1445, 1265, 1 1105, 1070, 860, 790, 750 cm” ; EI-MS (70eV) 320 (11", 2.4), 3C 102 305 (1.3), 263 (96.5), 189 (base), 167 (21.8), 161 (29.0), 128 (34.0). 2-Pheny1th1o-5-trimethylsi1y1-3-tbuty1dimethylsiloxymethy1- furan 70. To a solution of nBuLi (2.08 M in hexane, 0.48 mL, 1.0 mmol) in dry Et 0 (3 mL) at 0°C (ice water) was added 2 2-pheny1thio-3-tbuty1dimethylsiloxymethyl-furan (0.32g, 1.0 mmol) in dry EtZO (2 mL) via syringe. The solution was then warmed to 25°C and stirred for 2.5h or until a deep reddish- brown color was observed. Upon cooling to 0°C, trimethylsilyl chloride (0.15 mL, 1.25 ‘mmol) was added and the solution was gradually warmed to 25°C over the course of 1h. A precipi- tate was observed soon after the‘addition of trimethylsilyl chloride. The reaction mixture was diluted with Et 0 (25 2 mL), washed with water (25 mL), and saturated aqueous NH4C1 (25 mL). The organic phase was dried (Na2804) and removal of solvent afforded 0.35g of a dark brown oil. The crude product was purified by filtration through a column of silica gel (60-230 mesh, 10g, EtZO-petroleum ether, 1:99) to afford 1H NMR (60 MHz) 0.30g (76%) of 70 as a colorless liquid. 6 7.18 (bds, 5H), 6.83 (S, 1H), 4.67 (S, 2H), 0.98 (8, 9H), 0.35 (S, 9H), 0.10 (m, 6H); IR (neat) 2980, 2880, 1590, 1485, 1260, 1095, 1060, 940, 855, 785, 745 cm-1; EI-MS (70eV) 392 (3+, 4.0), 377 (1.6), 335 (17.1), 261 (27.8), 73 (base). 2-Phenylthio-5-trimethylsi1yl-3-hydroxymethyl-furan 71. A solution of 2-phenylthio-5-trimethy1silyl-3-tbutyldi- methylsiloxymethyl-furan 70 (6.65g, 17.0 mmol), glacial acetic 103 acid (50 mL), water (20 mL) and THF (50 mL) was heated to reflux while being monitored by TLC analysis (EtZO-petroleum ether, 1:1). After 70 minutes, most of the solvent was removed 1; vacuo and the residue was diluted with water (50 mL) fol- , lowed by addition of solid NaHCO3 until- gas evolution ceased. The aqueous solution was then extracted with Et201(2 x 100 mL) and the combined ether extracts were dried (Na2S04). Removal of solvent in vacuo provided 6.1g of crude product which was purified by chromatography on a column of silica gel (60-230 mesh, 300g, Et20-petroleum ether, 1:1) to afford 2.46g (52%) of 71 along with 0.39g (11%) of 52 and 0.60g (9%) of recovered starting material. The cleavage of the 1H NMR and it was silyl ether to give 71 was verified by subjected to desulfurization below without further characteri- zation. Desulfurization of 71. Preparation of 2-Trimethylsilyl- 4-hydroxymethy1-i'uran 68. 2-Pheny1thio-5-trimethy1silyl-3- hydroxymethyl-furan (2.9g, 10.4 mmol) and a suspension of W-2 grade raney-nickel catalyst (58.0g) in absolute ethanol (150 mL) was heated to reflux for 6h. The solution was then filtered through celite followed by rinsing of the catalyst with ethanol (3 x 50 mL) and filtration of the combined rinse solvent. Rinse and filtrate were combined and removal of solvent in vacuo followed by distillation, BP = 90°C, 3mm afforded 1.0g (56%) of 68, identical in all respects to that obtained previously. 104 Silylation of 2-Bromo-4-furoic acid. Preparation of 2-Trimethylsilyl-4-hydroxymethyl-furan 68. To a solution 57 (0.43g, 2.25 mmol) in dry Et20 (10 mL) cooled to -78°C was added nBuLi (2.9 M in hexane, of 2-bromo-4-furoic acid . 1.71 mL, 4.95 mmol), after which a precipitate formed. The Solution was stirred at -78°C for 30 minutes and then tri- methylsilyl chloride (0.63 mL, 5.0 mmol) was added via syringe. The solution was then warmed to 0°C (ice water) and gradually warmed to 25°C over 1h. 'After this time, the solution was diluted with Et 0 (50 mL) and washed with 0.1 M aqueous HCl 2 (2 x 50 mL). The organic phase was dried (Na2804) and removal of solvent in vacuo afforded 0.45g crude product. The crude product was purified by chromatography on a column of silica gel (230-400 mesh, 50g, Et O-petroleum ether-MeOH, 5:94.5z0.5) 2 to afford 0.09g (15%) of 2,5-bis-trimethylsilyl-3-furoic acid, Rf 0.24 (EtZO-petroleum ether-MeOH, 10:89.5:0.5) and 0.30g of a mixture of two components which was directly submit- ted to hydride reduction conditions as before (see preparation of 63). The crude material from this reduction was purified by chromatography on a column of silica gel (60-230 mesh, 30g, EtZO-petroleum ether, 15:85) to afford 0.23g (81%) of 2-trimethylsi1yl-4-hydroxymethy1-furan 68 (51% overall from 2-bromo-4-furoic acid) which was identical in all respects to that obtained previously. 2-Tr1methylsilyl-4-chloromethyl-furan 72. On the same scale and utilizing a procedure identical to that used in the preparation of chloride 24, 72 was prepared from alcohol 105 1 68 in 78% yield as a colorless liquid, B = 90°C. H P10mm NMR (60 MHz) 6 7.57 (s, 1H), 6.61 (s, 1H), 4.42 (s, 2H), 0.27 (bds, 9H); IR (neat) 2980, 2925, 1685, 1600, 1455, 1385, 1; EI-MS (70eV) 188 (M+, 1260, 1185, 1140, 1085, 855, 715 cm" 28.3), 173 (89.2), 153 (18.2), 93 (base). 2,6-Dimethy1-9-(2-trimethylsilyl-4-furyl)-nona-2,6-diene- 2,3-epoxide 73. On the same scale and utilizing a procedure identical to that used previously (see the procedure for the preparation of 65), 2-trimethylsilyl-4-ch1oromethy1-furan 72 was converted to the corresponding Grignard reagent (64% 3 51,52 by titration) and coupled with epoxygeranyl chloride to afford a 90% yield (based on epoxygeranyl chloride) of 73 after purification by chromatography on a column of silica gel (60-230 mesh, EtZO-petroleum ether, 5:95), Rf 0.58 (Et20- 111 NMR (60 MHz) 6 7.29 (s, 1H), 6.45 petroleum ether , 1: l) . (S, 1H), 5.18 (m, 1H), 2.00-2.80 (bdm, 5H), 1.64 (bds, 7H), 1.27 (S, 6H), 0.27 (m, 9H); IR (neat) 2990, 2880, 1465, 1385, 1; EI-MS (70eV) 306 (M+, 1255, 1130, 1090, 920, 855, 770 cm" 2.3), 291 (1.9), 233 (2.8), 220 (4.3), 207 (13.6), 153 (96.7), 73 (base). General Procedure for the Coupling of Silyl-substituted Furylmethyl Grignard Reagents with Alkyl Halides. Preparation of 2-Trimethylsilyl-3-decyl-furan 74 and 2-Trimethylsily1-4- decyl-furan 77. To a solution of dry 2-trimethylsilyl-3- chloromethyl-furan 64 or regioisomer 2-trimethylsi1yl-4-chloro- methyl-furan 72 (1.0g, 5.3 mmol) in dry THF (15 mL) was added magnesium powder (0.13g, 5.3 mmol) with stirring at 25°C. 106 The solution was heated to 50°C and Grignard formation soon became initiated. After 45 minutes at 50°C, most of the ‘ magnesium had been consumed producing a slightly turbid blackish-brown colored solution. Titration26 of an aliquot .showed Grignard formation to be 77% and 64% from chlorides 64 and 72, respectively. The solution was cooled to 0°C CuCl (ice water) and Li catalyst (0.1 M in THF, 0.27 mL, 2 4 0.027 mmol) was added followed by a solution of nonyl-iodide (1.2g, 4.8 mmol or 0.9g, 3.4 mmol, respectively based on the Grignard conversion of 64 or 72) in dry THF (2 mL). The resulting solution was slowly warmed to 25°C over the course of lh. Most of the solvent was removed _i_n_ 3M9. and the, residue was diluted with EtZO-pentane, 1:1 (50 mL) followed by washing with water (50 mL), saturated aqueous NH4C1 (50 mL), saturated aqueous NaHCO3 (50 mL) and drying (Na2S04). Removal of solvent _ip vacuo afforded the crude product which was purified by chromatography on a column of silica gel (60-230 mesh, 70g, petroleum ether) to afford 1.1g (82% of 74, Rf 0.51 (petroleum ether) and 0.7g (77%) of 77, R], 0.51 (petroleum ether) with yields based on the Grignard conversions of 64 and 72, respectively. Spectral data for 74: 1H NMR (60 MHz) 6 7.45 (d, J=2Hz, 1H), 6.20 (d, J=2Hz, 1H), 2.25 (bdt, J=6Hz, 2H), 1.35 (bds, 16H), 0.90 (m, 3H), 0.28 (s, 9H); IR (neat) 2980, 2950, 2880, 1575, 1470, 1400, 1260, l 850, 770 cm"; EI-MS (70eV) 280 (11", 15.8), 265 (5.1), 195 (39.0), 181 (31.6), 154 (43.1), 73 (base). Spectral data 1 for 77: H NMR (60 MHz) 6 7.34 (s, 1H), 6.46 (s, 1H), 2.41 (it t] 107 (bdt, J=6Hz, 2H), 1.34 (bds, 16H), 0.92 (m, 3H), 0.29 (s, 9H); IR (neat) 2960, 2920, 1465, 1250, 1075, 710, 840, 750 cm-1; MS (EII70eV) 280 (M+, 3.3), 265 (2.2), 251 (1.8), 237 (1.9), 196 (2.5), 167 (12.6), 154 (base). 2-Trimethylsilyl-3-(6-hepteny1)-furan 75. According to the general procedure above, on an identical scale, 2- trimethylsilyl-3-chloromethyl-furan 64 afforded a Grignard reagent which was coupled with 6-bromo-1-hexene and purified 1 to afford a 74% yield of 75, R 0.43 (petroleum ether). R f NMR (60 MHz) 6 7.43 (d, J=2Hz, 1H), 6.20 (d, J=2Hz, 1H), 5.46-6.00 (m, 1H), 4.80-5.18 (m, 2H), 2.46 (bdt, J=6Hz, 2H), 2.04 (m, 2H), 1.39 (m, 6H), 0.29- (s, 9H); IR (neat) 2980, 29.50, 2880, 1645, 1575, 1460, 1495, 1255, 1155, 1000, 920, 1; EI-MS (70eV) 236 (11", 13.5), 221 (25.0), 207 860, 770 cm’ (12.8), 195 (15.9), 163 (10.1), 153 (54.7), 111 (base). 2—Trimethylsilyl-3-(3-hepteny1)-furan 76. According to the general procedure above, on an identical scale, 2- trimethylsi1y1-3-chloromethyl-furan 64 afforded a Grignard reagent which was coupled with 1-chloro-2-hexene and purified to afford a 67% yield of 76, Rf 0.38 (petroleum ether). 1H NMR (60 MHz) 6 7.40 (s, 1H), 6.18 (s, 1H), 5.37 (bds, 2H), 1.89-2.52 (m, 6H), 1.30 (m, 2H), 0.90 (m, 3H), 0.29 (s, 9H); IR (neat) 2955, 2920, 2880, 1675, 1570, 1450, 1390, 1250, 1 1090, 965, 835, 755 cm”; EI-MS (70eV) 236 (M+, 4.4), 221 (3.3), 208 (9.3), 193 (16.6), 163 (8.8), 153 (60.4), 73 (base). 108 2-Trimethylsilyl-4-(6-hepteny1)-furan 78. According to the general procedure above, on an identical scale, 2—- trimethylsilyl-4-chloromethyl-furan 72 afforded a Grignard reagent which was coupled with 6-bromo-l-hexene and purified to afford a 70% yield of 78, Rf 0.36 (petroleum ether). 1H NMR (60 MHz) 6 7.32 (s, 1H), 6.44 (s, 1H), 5.46-6.06 (bdm, 1H), 4.79-5.07 (m, 2H), 2.37 (bdt, J=6Hz, 2H), 2.00 (m, 2H), 1.40 (m, 6H), 0.29 (s, 9H); IR (neat) 2960, 2930, 2860, 1640, 1; EI-MS (70eV) 236 (3*, 1450, 1250, 1075, 910, 840, 755 cm' 8.1), 221 (4.8), 193 (6.2), 167 (11.3), 154 (63.6), 73 (base). 2-Trimethylsily1-4-(3-hepteny1)-furan 79. According to the general procedure abOVe, on an identical scale, 2-- trimethylsilyl-4-chloromethy1-furan 72 afforded a Grignard reagent which was coupled with l-chloro-Z-hexene and purified to afford a 60% yield of 79, Rf 0.31 (petroleum ether). 1H NMR (60 MHz) 6 7.34 (s, 1H), 6.43 (s, 1H), 5.40 (m, 2H), 1.84-2.48 (m, 6H), 1.32 (m, 2H), 0.90 (m, 3H), 0.29 (s, 9H); IR (neat) 2960, 2920, 1450, 1250, 1075, 910, 840, 750 cm-1; EI-MS (70eV) 236 (M+, 11.2), 221 (5.9), 207 (3.9), 193 (9.7), 154 (98.5), 73 (base). General Procedure for the Oxidation of Silyl-substituted Furans to Butenolides: Oxidation of 74. Preparation of 3-Decyl-2(5H)-furanone 80 and 3-Decyl-2(3H)-furanone 81. According to the procedure of Kuwajima,59 to a stirred suspen- sion of 40% peracetic acid (1.9g, 10.0 mmol) and sodium acetate trihydrate (0.41g, 3.0 mmol) in methylene chloride (10 mL) was added 2-trimethy1silyl-B-decyl-furan 74 (0.70g, 2.5 mmol) 109 in methylene chloride (1 mL) at 0°C (ice water). The reaction mixture was ‘stirred for 3h then diluted with Et 0 (50 mL) 2 and washed with water (50 mL), saturated aqueous NaHCO3 (2 x. 25 mL), saturated aqueous sodium thiosulfate (2 x 25 mL) . and saturated aqueous NaHCO3 (25 ml) followed by drying (Na2804). Removal of solvent in vacuo afforded 0.60g of crude product which was purified by chromatography on a column of silica gel (60-230 mesh, 60g, EtZO-Petroleum ether, 1:3) to afford 0.22g (39%) of 80, R 0.33 (EtZO-petroleum ether, 1’ 1:1) and 0.22g (39%) of 81, Rf 0.22. Spectral dgta for 80: 1H NMR (60 MHz) 6 7.15 (m, 1H), 4.77 (m, 2H), 2.27 (m, 2H), 1.30 (m, 16H), 0.89 (m,-3H); IR (neat) 2920, 2850, 1750, 1655, 1455, 1350, 1230, 1190, 1110, 1060, 1000, 825 cm'1; EI-MS (70ev) 224 (M+, 15 0), 195 (2.5), 179 (5.3), 167 (4.9), 153 (8.4), 139 (16.4), 125 (20.9), 98 (base). Spectral data 1H NMR (60 MHz) 6 6.80 (m, 1H), 5.57 (m, 1H), 3.17 for 81: (bdm, 1H), 1.30 (m, 18H), 0.90 (m, 3H); IR (neat) 3350, 2925, 2850, 1760, 1660, 1465, 1335, 1190, 1100, 1050, 1000, 825 om'1; EI-MS (70eV) 224 (M+, 2.0), 195 (2.1), 179 (3.0), 167 (3.5), 153 (1.2), 98 (base). Oxidation of 75. Preparation of 3-(6-Hepteny1)-2(5H)- furanone 82. According to the general procedure above, on an identical scale, 2-trimethylsilyl-3(-6-heptenyl)-2(5H)- furanone 75 was oxidized to 82 in 67% yield after purification by chromatography on silica gel, Rf 0.29 (EtZO-petroleum 1 ether, 1:1). H NMR (60 MHZ) 7.19 (m, 1H), 5.49-6.15 (bdm, 110 1H), 4.73-5.10 (m, 2H), 4.80 (m, 2H), 2.03-2.36 (bdm, 4H), 1.43 (bds, 6H); IR (neat) 2950, 2880, 1765, 1645, 1460, 1355, 1; EI-MS (70eV) 180 (M+, 0.5), 1210, 1075, 1010, 925, 845 cm’ 165 (1.3), 151 (3.2), 135 (7.0), 123 (6.4), 98 (base). Oxidation of 76. Preparation of 3-(3-Heptenyl)-2(5H)- furanone 83 and 3-(3-Heptenyl-3,4-epoxy)-2(5H)-furanone 84. According to the general procedure above, on an identical , scale, 2-trimethylsilyl-3-(3-heptenyl)-furan 76 ‘was oxidized F to a mixture of 83 and 84 which was purified by chromatography on silica gel to provide a 52% yield of 83 and a 36% yield L of 84, R 0.28 and 0.14 (ERZO-petroleum ether, 1:1), respec- l f tively. Spectral data for 83: H NMR (60 MHz) 6 7.09 (ni, 1H), 5.40 (m, 2H), 4.77 (m, 2H), 2.36 (bds, 4H), 1.84-2.10 (m, 2H), 1.29 (m, 2H), 0.91 (m, 3H); IR (neat) 2980, 2955, 2880, 1755, 1660, 1460, 1355, 1210, 1080, 980, 925, 840, 1; MS (El/70eV) 180 (M+, 10.3), 165 (1.7), 151 (4.1), 135 (6.6), 123 (7.3), 98 (base). Spectral data for 84: 1H 740 cm- NMR (60 MHZ) 6 7.23 (m, 1H), 4.80 (m, 2H), 2.74 (m, 2H), 2.43 (m, 2H), 1.78-2.04 (m, 2H), 1.47-1.59 (m, 4H), 1.00 (m, 3H); IR (neat) 3430, 2995, 2960, 2900, 2280, 1765, 1660, 1460, 1; MS (El/70eV) 196 (M+, 1355, 1215, 1085, 925, 845, 745 cm" 0.4), 178 (0.6), 167 (3.9), 139 (7.5), 124 (61.4), 85 (base). Oxidation of 65. Preparation of 2,6-Dimethy1-9-(2- trimethylsilyl-3-furyl)-nona-2,6-diene-2,3-6,7-dioxirane 85. According to the general procedure above, on an identical scale, 2,6-dimethyl-9-(2-trimethy1silyl-3-furyl)-nona-2,6- diene-2,3-epoxide 65 was oxidized to 85 in 93% yield after purlf petrc 91) . 111( 281 86. 502 86 ge 111 purification by chromatography on silica gel, Rf 0.50 (Et20— petroleum ether, 1:1). 1H NMR (60 MHz) 6 7.48 (s, 1H), 6.25 (S, 1H), 2.54-2.86 (m, 4H), 1.56-2.01 (bds, 6H), 1.26 (bds, 9H), 0.31 (S, 9H); IR (neat) 2990, 2890, 1460, 1390, 1260, 1110, 860, 770 cm‘1 ; EI-MS (70ev) 322 (M1, 2.1), 306 (3.3), 281 (1.2), 233 (4.7), 153 (34.2), 73 (base). Oxidation of 77 . Preparation of 4-Decy1-2(5H)-furanone 86. According to the general procedure above, on an identical scale, 2-trimethylsilyl-4-decy1-furan 77 was oxidized to 86 in 91% yield after purification by chromatography on silica 1H NMR (60 MHz) gel, Rf 0.16 (EtZO-petroleum ether, 1:1). 6 5.86 (m, 1H), 4.78 (bds, 2H), 2.45 (bdt, J=7Hz, 2H), 1.31 (m, 16H), 0.90 (m, 3H); IR (neat) 3400, 2960, 2880, 1785, 1755, 1650, 1475, 1345, 1185, 1145, 1040, 950, 900 cm-l; EI-MS (70eV) 224 (M+, 2.3), 195 (10.1), 181 (3.6), 164 (26.8), 98 (19.1), 85 (base). Oxidation of 78. Preparation of 4-(6-Hepteny1)-2(5H)- furanone 87. According to the general procedure above, on an identical scale, 2-trimethy1silyl-4-(6-heptenyl)-furan 78 was oxidized to 87 in 77% yield after purification by chromatography on silica gel, Rf 0.15 (EtZO-petroleum ether, 1:1). 1H NMR (60 MHz) 6 5.39-6.00 (m, 1H), 5.77 (m, 1H), 4.80-5.08 (m, 2H), 4.73 (S, 2H), 2.34 (m, 2H), 2.04 (m, 2H), 1.42 (bds, 6H); IR (neat) 2955, 2885, 1785, 1755, 1645, 1455, 1; EI-MS (70ev) 181 (M+, 1.1), 1380, 1140, 1035, 920, 745 cm' 165 (3.9), 151 (22.4), 139 (12.8), 121 (26.4), 111 (39.8), 98 (73.3), 85 (base). fura an 79 chr 1:1 3H3 11! of me 112 Oxidation of 79. Preparation of 4-(3-Hepteny1)-2(5H)- furanone 88. According to the general procedure above, on ‘ an identical scale, 2-trimethylsilyl-4-(3-hepteny1)-furan 79 was oxidized to 88 in 51% yield after purification by chromatography on silica gel, Rf 0.22 (EtZO-petroleum ether, 1 1:1). 3 NMR (60 MHz) 6 5.87 (m, 13), 5.45 (m,23), 4.75 (m, 2H), 2.46 (m, 4H), 1.99 (m, 2H), 1.33 (m, 2H), 0.98 (m, 3H); IR (neat) 2990, 2960, 2910, 1785, 1755, 1645, 1455, 1 1185, 1145, 1045, 985, 900, 745 cm“ ; MS (El/70eV) 180 (3+, 2.4), 165 (0.4), 151 (12.1), 137 (3.1), 123 (10.0), 98 (base). 6,7-Bpoxy-8-hydroxygeranyl-acetate 89. To a solution 65 of (§)-8-hydroxygeranyl acetate '(3.0g, 14.1 mmol) in dry methylene chloride (50 mL) was added metachloroperbenzoic acid (85%, 3.0g, 15.0 mmol) at 0°C (ice water). The solution was warmed to 25°C and stirred. for» 2h. After this time, the mixture was diluted with Et 0 (150 mL) and washed with 2 saturated aqueous Nazszo3 (2 x 150 mL) and saturated aqueous NaHCO (2 x 150 mL). The organic phase was dried (Na2S04) 3 and removal of solvent _i_n vacuo afforded 3.0g (94%) of 89 13 NMR (60 MHz) 6 5.37 (m, which was not purified further. 13), 4.57 (d, J=7Hz, 23), 3.58 (s, 23), 3.18 (bds, 13, -03), 2.97 (t, J=5Hz, 13), 2.00-2.30 (m, 43), 2.05 (s, 33), 1.73 (s, 33), 1.30 (s, 33); IR (neat) 3480, 2965, 1740, 1675, 1450, 1385, 1260, 1135, 1080, 1040, 965, 910, 880, 760, 690 cm'1; EI-MS (70eV) 213 (M+-0, 0.5), 169 (1.0), 153 (2.2), 137 (3.9), 126 (10.8), 111 (11.9), 43 (base). 113 6,7-Epoxy-8-tbutyldimethylsiloxygeranyl chloride 90. To a solution of 6,7-epoxy-8-hydroxygeranyl-acetate 89 (3.0g, 13.0 mmol) 13 dry methylene chloride (25 mL) was added tri- ethylamine (2.23 mL, 15.0 mmol), dimethylaminopyridine (65 _mg, catalytic) and tbutyldimethylsilyl chloride (2.21g, 14.0 mmol). The resulting solution was stirred at 25°C for 1h and then diluted with Et 0 (200 mL) and washed with 0.01 2 N aqueous HCl (200 mL) and saturated aqueous NaHCO (200 3 mL). The organic phase was dried (NaZSO4) and removal of solvent Q vacuo afforded 4.1g of the crude TBDMS ether which was not purified further, R 0.61 (EtZO-petroleum ether, f 6:4). The crude material was.directly submitted to deacetyla- tion conditions as follows: a solution of the TBDMS ether (4.1g, 12.0 mmol) in methanol (30 mL) was cooled to 0°C (ice water) and potassium carbonate (5g) was added. The resulting solution was stirred at 0°C for 30 minutes and then diluted with Et 0 (200 mL) followed by washing with water (2 x 200 2 mL). The organic phase was dried (Na2S04) and removal of solvent _i_p vacuo afforded 3.3g (91%) of crude deacetylated alcohol, Rf 0.24 (EtZO-petroleum ether, 6:4) which was readily transformed into the chloride below without further purifica- tion. To a solution of the alcohol (3.3g, 11.0 mmol) in dry Et 0 (10 mL) and HMPA (4.5 mL) at 0°C (ice water) was 2 added nBuLi (2.08 M in hexane, 5.2 mL, 11.0 mmol) via syringe. The resulting solution was warmed to 25°C and stirred for 10 minutes. Toluenesulfonyl chloride (2.09g, 11.0 mmol) in dry Et20 (5 mL) was then added followed by anhydrous lithium 114 chloride (0.63g, 15 mmol). The resulting solution was stirred at 25°C overnight followed by dilution 'with EtZO-petroleum ether, 1:1 (100 mL) and washing with water (100 mL), saturated aqueous NaCl (100 mL), and drying (NaZSO4). Removal of solvent _i__n_ vacuo afforded the crude product which was purified by bulb-to-bulb (Kugelrohr) distillation, oven 160°C (0.04: mm) to afford 2.7g (77%, 65% overall) of 90 as a slightly yellow 1 liquid, R 0.69 (EtZO-petroleum ether, 6:4). H NMR (60 f MHZ) 6 5.55 (bdt, J=8HZ,'1H), 4.13 (bdd, J=8Hz, 2H), 3.65 (m, 2H), 2.75 (m, 1H), 2.10-2.30 (m, 4H), 1.82 (bds, 3H), 1.35 (bds, 3H), 1.00 (m, 9H), 0.15 (m, 6H); IR (neat) 2920, 2855, 1660, 1460, 1380, .1350, 1250, 1160, 1085, 1035, 775 cm'1; EI-MS (70eV) 261 (M+-tBu, 1.1), 225 (2.9), 198 (16.1), 157 (30.4), 93 (base). l-tButyldimethylsiloxymethyl-G-methyl-9-(3-fury1)-nona- 2,6-diene-2,3-epoxide 91. In the usual procedure,8 3-chloro- methylfuran (0.49g, 4.2 mmol) was converted to the correspond- ing Grignard reagent and coupled with 6,7-epoxy-8-tbutyldi- methylsiloxygeranyl chloride 90 (1.3g, 4.2 mmol) to jyield, after purification by chromatography on a column of silica gel (60-230 mesh, 100g, EtZO-petroleum (ether, 5:95), 0.91g 1 (60%) of 91, R 0.24 (EtZO-petroleum ether, 5:95). H NMR f (80 MHz, d acetone) 6 7.37 (t, J=2.5Hz, 1H), 7.27 (bds, 1H), 6 6.30 (bds, 1H), 5.21 (bdt, J=6.2Hz, 1H), 3.65 (d, J=11HZ, 1H), 3.49 (d, J=11Hz, 1H), 2.76 (t, J=6.3Hz, 1H), 2.38 (bd, t, J=8Hz, 1H), 2.15 (m, 2H), 1.61 (m, 7H), 1.23 (S, 3H), 0.90 (s, 9H), 0.07 (d, 6H); IR (neat) 2970, 2880, 1470, 1390, 115 1; EI-MS (70eV) 307 (M+-tBu, 1265, 1100, 850, 785, 690 cm- 0.6), 249 (1.4), 225 (6.0), 215 (8.1), 175 (18.8), 131 (39.4), 81 (base). Cyclization of 91. Preparation of 92. To a solution ‘ of l-tbutyldimethylsiloxymethy1-6-methyl-9-(3-furyl)-nona-2,6- diene-2,3-epoxide 91 (3.55g, 8.24 mmol), in dry methylene chloride (30 mL), benzene (15 mL) and pentane (15 mL) was added triethylamine (2.04 mL, 12.36 mmol). The resulting solution was cooled to -60°C (dry ice-isopropanol) and boron trifluoride etherate (3.60 mL, 24.72 mmol) was added via syringe. The resulting solution was stirred at -60°C for 10 minutes taking on a deep golden yellow' color, followed by quenching with Et20 (10 mL) and saturated aqueous NaHCO3 (20 mL). The reaction mixture was then diluted with Etzo (200 mL) and washed with saturated aqueous NaHCO (200 mL), 3 saturated aqueous NH Cl (200 mL) and water (200 mL). The 4 organic phase was dried (Na2S04) and removal of solvent af- forded 3.5g crude product. This material was purified by chromatography on a column of silica gel (60-230 mesh, 300g, EtZO-petroleum, 15:85) to afford 1.95g (55%) of uncyclized ketone product, Rf 0.43 (EtZO-petroleum ether, 15:85) and 1.00g (28%) of 92 as a slightly yellow liquid, R 1 f 0.24 (EtZO- petroleum ether, 15:85). H NMR (80 MHz, d acetone) 6 7.18 1 6 (d, J=2Hz, 1H), 6.08 (d, J=2Hz, 1H), 3.70 (d, J=10Hz, 1H), 3.40 (dd, J-14,7HZ, 1H), 3.33 (d, J=10Hz, 1H), 2.30-2.48 (m, 2H), 1.50-1.78 (m, 7H), 1.19 (S, 3H), 0.88 (S, 9H), 0.75 (S, 3H), 0.08 (S, 6H); IR (neat) 3500, 2970, 2890, 1510, 147 SC 01 n-u 116 1475, 1390, 1255, 1095, 1030, 855, 785, 735 cm-1; EI-MS (706V) 307 (M+-tBu, 8.8), 289 (1.3), 215 (base), 187 (13.0), 175 (58.8). 1 Preparation of Ketone 93. To a solution of 92 (0.10g, . 0.28 mmol) in dry methylene chloride (1 mL) was added anhydrous Sodium acetate (23mg, 0.05 mmol) followed by pyridinium chloro-chromate (0.12g, 0.55 mol) at 0°C (ice water). After 30 minutes, the mixture was warmed to 25°C and stirred to 1.5h. Saturated aquecus NaHCO (5 mL) was added followed 3 by dilution with Et 0 (25 mL), and washing with saturated 2 aqueous NaHCO3 (2 x 25 mL), 1 N aquecus HCl (25 mL) and saturated aqueous NaCl (25 mL). The organic phase was dried. (Na2804) and removal of solvent g m afforded 0.95g crude product which was purified by chromatography on a column of silica gel (60-230 mesh, 50g, Et O-petroleum ether, 3:97) 2 to provide 0.91g (91%) of ketone 93 as a slightly yellow liquid, Rf 0.57 (EtZO-petroleum ether,‘ 2:8). 13 NMR (80 MHZ, d6 acetone) 6 7.26 (d, J=1.7Hz, 1H), 6.14 (d, J=1.7Hz, 13), 3.65 (d, J=9.0Hz, 13), 3.46 (d, J=9.0Hz, 13), 2.20- 2.73 (m, 6H), 1.59-1.84 (m, 3H), 1.19 (S, 3H), 0.96 (S, 3H), 0.86 (s, 9H), 0.06 (s, 3H), 0.03 (s, 3H); IR (neat) 2960, 2880, 1705, 1510, 1475, 1390, 1260, 1100, 850,1785 cm-l; EI-MS (70eV) 305 (M+-tBu, 55.0), 223 (3.8), 213 (33.2), 171 (23.5), 157 (40.0), 147 (base). Preparation of D101 94. To a solution of ketone 93 (0.417g, 1.15 mmol) in dry THF (0.5 mL) at 0°C (ice water) was added tetra-butyl ammonium fluoride (1 M in THF, 2.3 117 mL, 2.30 mmol) via syringe. The solution was stirred for 2h at 0°C and then diluted with Et20 (10 mL) and washed with saturated aqueous NaHCO3 (10 mL) followed by saturated aqueous NaCl (10 mL). The organic phase was dried (Na2S04) and removal of solvent ip 32.2119 afforded 0.268g (94%) of crude product which rapidly turned brown in color' at .25°C. Due to ‘the instability of the compound, it was directly submitted to subsequent reduction conditions without further purification. To a solution of lithium tri-sec-butyl borohydride (1 M in THF, 2.16 mL, 2.16 mmol) in dry THF (3 mL) at -78°C was added the crude keto-alcohol (0.268g, 1.08 mmol) in dry THF (1.5 mL) via syringe. The resulting solution was stirred at -78°C for 1h then warmed to 0°C (ice water) and stirred for 1h. The reaction was quenched by carefully adding' methanol (2 mL) followed by 20% aqueous NaOH (3 mL) and 30% aqueous hydro- gen peroxide (6 mL). The solution was stirred overnight and then diluted with Et 0 (100 mL), washed with water (100 2 mL), saturated aqueous NH4C1 (100 mL) and saturated aqueous NaHCO (100 mL). The organic phase was dried (Na2804) and 3 removal of solvent Q vacuo followed with purification by chromatography on a column of silica gel (60-230 mesh, 50g, EtOAc-methylene chloride, 1:9) afforded 0.212g (79% overall from 93) of diol 94 as a white solid, R 1 f 0.13 (EtOAc-methylene chloride, 1:9). H NMR (80 MHz, d acetone with D20) 6 7.20 6 (d, J=1.7HZ, 1H), 6.08 (d, J=1.7HZ, 1H), 3.69 (bdt, J=3.3Hz, 1H), 3.59 (d, J=14.8Hz, 1H), 3.33 (d, J=14.8Hz, 1H), 2.31-2.47 (m, 2H), 1.52-1.89 (m, 7H), 1.19 (S, 3H), 0.78 (S, 3H); IR 118 (neat) 3350, 2950, 1505, 1480, 1455, 1385, 1270, 1210, 1165, 1 1135, 1055, 1000, 890, 745 cm" ; EI-MS (70eV) 250 (3*, 29.5), 235 (14.4), 217 (base), 199 (19.6), 159 (32.4), 149 (40.1). Acetonide 95. To a solution of diol 94 (25 mg, 0.10 mmol) in dry methylene chloride (1 mL) and acetone (99.5%, 0.5 mL) at 25°C was added a few crystals of oxalic acid fol- lowed by stirring overnight. The solution was diluted with Et 0 (20 mL), washed with saturated aqueous NaHCO (2 x 20 2 3 mL) and dried (Na2804). Removal of solvent followed with purification by chromatography on a column of silica gel (60-230 mesh, 10g, Et O-petroleum ether, 2:98) afforded 26 2 mg (90%) of acetonide 95 as.a white solid, Rf 0.47 (Et2 1 acetone) 6 7.20 O- petroleum ether, 2:8). H NMR (80 MHz, d 6 (d, J=1.7Hz, 1H), 6.09 (d, J=1.7HZ, 1H), 3.74 (m, 1H), 3.69 (d, J=12.2HZ, 1H), 3.33 (d, J-12.2HZ, 1H), 2.33-2.52 (m, 2H), 1.61-1.89 (m, 7H), 1.37 (S, 3H), 1.25 (S, 3H), 1.20 (s, 3H), 0.82 (S, 3H); IR (KBr) 3010, 2970, 2890, 1505, 1480, 1380, 1205, 1165, 1095, 1010, 860, 765 cm_1; EI-MS (706V) 290 (3*, 41.5), 275 (25.8), 232 (4.5), 217 (base), 149 (47.6). Bis-MEI Ether 96. To a solution of diol 94 (20 mg, 0.08 mmol) in dry methylene chloride (1.0 mL) was added diis0propy1, ethyl amine (0.11 mL, 0.64 mmol) and methoxy- ethoxymethyl chloride (0.073 mL, 0.64 mmol) at 25°C. The resulting solution was stirred for 2h. The reaction mixture was diluted with Et 0 (25 mL) followed by washing with satu- 2 rated aqueous NH4C1 (25mL) and saturated aqueous NaHCO3 (25 mL). The organic phase was dried (NaZSO4) and removal of 501 on 91) 119 solvent i_n vacuo followed with purification by chromatography on a column of silica gel (60-230 mesh, 5g, EtZO-petroleum ether, 5:95) afforded 29 mg (85%) of bis-MEM ether 96 as a colorless liquid. 13 NMR (80 332) 6 7.13 (d, J=1.7Hz, 13), _ 6.07 (d, J=1.7Hz, 13), 4.76 (d, J=5.0Hz, 23), 4.69 (d, J=5.0Hz, 23), 3.53-3.72 (m, 113), 3.38 (s, 33), 3.39 (s, 3H), 2.33-2.48 (m, 23), 1.55—1.86 (m, 73), 1.22 (s, 33), 1.00 (s, 33); IR (neat) 2980, 2880, 1510, 1455, 1370, 1290, 1250, 1205, 1000- 1; EI-MS (70eV) 335 (31-2 x 303, 0 1), 1200, 855, 740, 705 cm" 274 (14.2), 261 (15.6), 247 (M+-2_x MEM, 9.5), 163 (68.7), 121 (base). LIST OF REFERENCES 10. 11. 12. 13. LIST OF REFERENCES Nakanishi, K. et. al. eds., "Natural Products Chemistry", Kodansha Ltd., Tokyo, 1974. Kubo, 1.; Lee, Y.-W.; Balogh-Nair, V.; Nakanishi, K.; Chapya, A. J. Chem. Soc. Chem. Commun. 1976, 949. Herz, W.; Kumar, N.; Blount, J. F. J. 03. Chem. 1981, 46, 1356. Cook, C. E.; Whichard, L. P.; Turner, B.; Wall, M. E.; Egley, G. H. Science 1966, 154, 1189. Schulte, G.; Scheuer, P. J.; McConnel, O. J. Helv. Chim. Acta 1980, 63, 2159. ' Appel, 3. 3.; Connolly, J. D.; 0verton, K. 3.; Bond, R. P. M. J. Chepl. Soc. 1960, 4685; Nakanishi, K.; Kubo, 1. Israel J. Chem; 1977, 16, 28. Tanis, S. P.; Nakanishi, K. J. Am. Chem. Spc. 1979, 101, 4398. Tanis, S. P. Tetrahedron Lett. 1982, 23, 3115. Tanis, S. P.; Herrinton, P. M. J. Org. Chem_._ 1983, 48, 4572. . Dunlop, A. P.; Peters, F. N., "The Furans", Reinhold, New York, 1953. Sargent , M. V. ; Cresp, T. M. , "Comprehensive Organic Chemistry", Barton, D.; 01113, W. D.; Eds., Pergamon Press, Oxford, 1979, Vol. 4, pp. 693-744 and references cited therein. Akita, H.; Naito, T.; Oishi, T. Chem. Pharm. Bull. 1980, _2_8_, 2166 . a) Tanis, S. P.; Head, D. B. Tetrahgdron Lett. 1982, _2_3_, 5509. b) Tanis, S. P.; Head, D. B. Tetrahedron Lett. 1984, _2_5, 4451. 120 14. l5. 16. 17. 18. 19. 20. 21. 22. 121 Trost, B. M.; Shey, C. D.; DiNinno, F. Jr. J. Am. Chem. Soc. 1979, 191, 1284; Hashimoto, H.; Tsuzuki, T.; Sakan, F.; Shirahama, H.; Matsumoto, T. Tetrahedron Lett. 1974, 3745; and references cited therein. Boeckman, R. K. Jr.; Ko, S. S. J. Am. Chem. Soc. 1980, 102, 7146; Greenlee, W. J.; Woodward, R. B. Tetrahedron 1980, _3_6_, 3367; Greenlee, W. J.; Woodward, R. B. Tetrahedron 1980, 36, 3361; Greenlee, W. J., Woodward, R. B. J. Amr. Chem; Soc. 1976, _98, 6075; and references cited therein. Semmelhack, M. F. ; Tomoda, S. ; Nagota, H. ; Boettger, S. D. ; Hurst, K. M. ,1. Am. Chem. Soca 1982, 194, 747; Miyano, K. ; Ohfune, F. ; Azuma, S. ; Matsumoto, T. Tetrphedron Lett. 1974, 1545; and references cited therein. a) Vidari, G.; DeBernardi, M.; Vita-Finzi, P.; Fronze, G. Phytgchemigtry 1976, _]._§, 1953; b) Froborg, J.; Magnusson, G.; Thoren, S. J. Org. 'Chgm, 1975, 4_0, 1595; - c) Fex, T.; Froborg, J.; Magnusson, G.; Thoren, S. J. Org. Chem. 1976, 41, 3518; d) Froborg, J.; Magnusson, G. Teprahgdron 1978, a, 2027; e) Ayer, W. A.; Brown, L. M. Tetrghggrpp 1981, fl, 2199. a) Magnusson, G.; Thoren, S. Tetrghedrop 1974, 3_0, 1430; b) Battaglia, R.; DeBernardi, M.; Fronza, G.; Mullerio, G.; Vidari, G.; Vita-Finzi, P. _L Nat. Prod, 1980, 4_3, 319; c) Froborg, J.; Magnusson, G.; Thoren, S. Act; Chem. chd. Ser. B 1974, 28, 265. Hooz, J.; Giliani, S. Q. J. Cheap 1968, 46, 86; Hooz, J. Org, Syp. 1976, 54, 63. Royals, E. E.; Hendry, C. M. J. Org. Chem. 1950, 15, 1147. Gemal, A. L.; Luche, J. L. J. Am. Chem. Soc. 1981, 103, 5454. a) The ratio of 19:20 was determined by a comparison of the integrated areas of their respective olefinic protons. An examination of the spectrum of pure 19 prepared from chloride 18 corroborated this analysis. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 122 b) For other examples of the reaction of organometallic compounds with allylic electrophiles, see: Stork, G.; Grieco, P. A.; Gregson, M. Tetrahedron Lett. 1969, 1393; Anderson, R. J.; Henrick, C. A.; Siddall, J. B. J. Am. Chemé Soc. 1970, _9_3_, 735; Anderson, R. J.; Henrick, C. A.; Siddall, J. B.; Zurfluh, R. J. Amé Chem. Soc. 1972, g, 5379; Fouquet, G.; Schlosser, M. Angew. Chem. Int. Ed. 1974, 1_3, 82; Goering, H. L.; Singleton, V. D.', Jr. J. Am. Chem. Soc. 1976, 98, 7854; Rossi, R.; Giorgi, R. Gazz. Chim. Ital. 1976, 10_6_, 215; Ziegler, F. E.; Tam, C. C. J. Org. Chem. 1979, 44, 3428; Yamamoto, Y.; Yamamoto, S.; Yatagai, H.; Maruyama, K. J. Am. Chem. S09. 1980, 102, 2318; and references cited therein. Miyashita, N.; Yoshikoshi, A.; Grieco, P. A. L Org. Chem. 1977, 4_2_, 3772. Kirner, W. R. J. Amt. Chem. Soc. 1928, _5_Q, 1955; Sherman, E.; Amstutz, E. D. J. Am. C m Soc. 1950, 1_2_, 2195. .‘ Collington, E. W.; Meyers, A. I. J. Org. Chem. 1971, 36, 3044. Watson, S. L.; Eastham, J. F. J. Orgganet. Chem. 1967, g, 165. Stowell, J. C. J. Org. Chem. 1976, g, 560. Attenburrow, J.; Cameron, A. F. B.; Chapman, J. H.; Evans, R. M.; Hems, B. A.; Jansen, A. B. A.; Walker, T. _J. Chemfi. Soc. 1952, 1094. Paquette, L. A.; Han, Y.-K. J. Am. ChegnL. Soc. 1981, 103, 1838. Baker, R.; Sims, R. J. J. Chem. Soc. Perkin I 1981, 3087. Corey, E. J.; Gilman, N. W.; Ganem, B. E. J. Am. Chem, 890. 1968, ,9_Q, 5616. a) Wiesner, K.; Marini-Bettolo, R.; Flecker, P.; Tsai, T. Y. R. Can, J. Chem. 1981, ,5_9_, 1403. b) Sornay, R.; Meunier, J. M.; Fournari, P. Bull. 390. Chim. France 1971, 990. c) Kraus, G. A.; Roth, B. J. Org. Chem, 1978, 4_3, 2072. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 123 a) Asaoka, M.; Miyake, K.; Takei, H. Chem. ILett. 1977, 167. b) Asaoka, M.; Sugimura, N.; Takei, H. Bull. Chem. M 1979. 52. 1953. Cazeau, P.; Moulines, F.; Laporte, 0.; Duboudin, F. J. Organomet. Chem. 1980, 9g, 201. Wiesner, K.; Bettolo, R. M.; Tsai, C. S. J.;,Tsai, T. Y. R. Heterocycles, 1981, 15, 305. A House, H. O.; Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org. Chem. 1969, 34, 2324. 0111s, W. D.; Hussain, S. A. T.; et. a1., J. Chem. Soc. Perkin I 1975, 1480. Still, W. C. J. Am. Chem, Soc. 1978, 100, 1481. Seebach, D. Angew. Chem. Internat. Edit. 1979, 18, 239. Owen, L. N.; Sultanbawa, M. U. S. J. Chemu Soc. 1949, 3105. Price, C. C.; Judge, J. M. Org, Syn. 1965, 45, 22. Tamura, M.; Kochi, J. Synthesis 1971, 303. Corey, E. J.; Seebach, D. Angew. Chem. Internat. Edit. 1965, 4, 1075, 1077. Seebach, D.; Beck, A. K. Org. Syn. 1971, 5;” 76. Richman, J. E.; Herrmann, J. L.; Schlessinger, R. H. Tet, Lett. 1973, 3267, 3271. Gschwend, H. W.; Rodriguez, H. R. Org. Reagt., 1979, .26, 151. a) Okazaki, R. J. Chem. Soc. Chem. Commun. 1982, 1055. b) deGroot, A. E.; et. a1. Tet. Lett. 1982, .23, 4831. Chadwick, D. J.; Willbe, C. J. Chem. Soc. Perkin .1 1977, 887. Levine, R.; Ramanathan, V. J. Org, Chem. 1962, 1216. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 124 Trost, B. M. ,ngm1_figy1 1978, 363. Epoxygeranyl chloride52 was often used in the studies presented herein due to its eff iency' in coupling with furylmethyl grignard reagents and cyclizability of the resulting furyl substrates as demonstrated by: Nasipuri, D.; Das, G. J. Chem. Soc. Perkin .1 1979, 2776. Stork, G. Org. Syn. 1978, Q4, 68. Rieke, R. D.; Bales, S. E. J. Am. Chem. Soc. 1974, gg, 1775. Knight, D. W.; Nott, A. P. J. Chem. Soc. Perkin ,1 1981, 1125. Org. Syn., coll. vol. III, 176. Goldsmith, D.; Liotta, D.; Saindane, M.; Waykole, L.; Bowen, P. Tet. Lett. 1983, 24, 5835. Ferraz, J. P.; do Amaral, ,L. J. Org. Chem. 1976, .41, 2350. a) see: Katritzky, A. R.; et. al., Eds. Adv. Heterocycl. Chemm 1982, .39, 167 and references cited therein. b) Foote, C. S.; et. a1., Tetrahedron 1967, .23, 2583. KuwaJima, I.; Urabe, H. Tet. Lett. 1981, 22, 5191. Schultz, A. G.; Motyka, L. A. J. Am. Chem, Soc. 1982, 104, 5800. Dalziel, W.; Hesp, B.; Stevenson, K. M.; Jarvis, J. A. J. J. Chem. Soc. Perkin I 1973, 2841. Bucknall, R. A.; Moores, H.; Simms, R.; Hesp, B. Antimicrob. Agents Chemother. 1973, g, 294. a) McMurry, J. E.; Andrus, A.; Ksander, G. M.; Musser, J. M.; Johnson, M. A. J. Am. Chem. Soc. 1979, .191, 1330; Tetrahedron 1981, .31” suppl. no. 1, pp. 319. b) Trost, B. M.; Nishimura, Y.; Yamamoto, K. .11 Am. Chem, Soc. 1979,,191, 1328. c) Corey, E. J.; Tius, M. A.; Das, J. J. Am, Chemm Soc. 1980, 192” 1742. d) Ireland, R. E.; Dow, W. C.; Jollie, D. G.; Thaisrivongs, S. J. Org. Chem. 1984, 4_9,, 1001; Ireland, R. E.; Godfrey, J. D.; Thaisrivongs, S. J. Am. Chem. Soc. 1981, 1Q§, 2446. 64. 65. 66. 67. 68. 69. 125 Sharpless, K. B.; Katsuki, T. J. Am. Chem. Soc. 1980, 102, 5974. Sharpless, K. B.; Umbreit, M. A. J. Am. Chem. Soc. 1977,199, 5526. Mitsunobu, 0. Synthesis, 1981, 1. a) Brooks, D. W.; Grothaus, P. G.; Palmer, J. T. Tet. Lett. 1982, 13, 4187. b) Raduchel, B. Synthesis 1980, 292. Corey, E. T.; Gras, J. L.; Ulrich, P. Tet. Lett. 1976, 11, 809. Still, W. C.; Kahn, M.; Mitraw A. J. Org; Chqmm 1978, 43, 2923.