.' I- I I b‘II I'.’ I 'III .. H ’l .H‘ IL'II‘I " "III II‘II-I' I . I‘" .I ._I .I '[m '. I'.'II)’ I'IL I". ‘ II'I‘IIIIEIQIII JI‘I‘.\‘ 56%,.“ "2‘. I .I' . .. I'I'II ' ..1‘ I III I IIQ I" NI ..~ -.‘- . ‘.. .. '._."II MI;- '1 I GYIIIIIII. ' II‘HII'II'I}:I I H' I, -' ' ’IIVI‘ II 'I' I" ‘Hil' .I ‘MIII‘I; IIIII'I "' IIIII.I.II IUI' ""IIm IIII 31:13. """I' III' I; «JIIIIII'I I .I' q“ :IAIS’IIIII fill I'v‘IIIIII' :{IIIHIi I‘MII'I: .EIII: W‘ '.Il ,'A'I F' I’l I. {‘.I':l'1'v"“';.' I’M I I N I I 1 III II'LtI JVIWISII‘J’I} w!" ’.V'fi I,I ‘IIK' a I . ‘|, "’ #03fil 'I'l:? "‘v 3. l|"'I I I .III a “I! I ‘5': IYI.' (”'\\‘ I~II.. -1;II- . 'II." I‘M :-’,I “If-III I.. '.I- 5’ I I I ‘.IIIIII: (10)." ‘I I’ \ I l “’ IvoIQIIIIi'II ‘21:?) ~ t )1 .4 ‘ InIII “’3? :55 - I'II' kai‘R-Vm‘ III-III:”IIg'II'III'IIIII tIII‘h hi 24%} IZiI'II figné “III E ’Iq‘ a. .I.-I~ ..~ " 'I‘L-IILI' W- Af'ldlt‘lIIII1|I I '| ll. ‘l': u . | -;”I,.'I. man-I, II' wu‘ III “U I" '. I, H ‘IIl' ")RIIIII‘I’IIIII'J" -\ I iII,'I:I‘II‘~-I‘-I 'I'CIII‘IJ‘I'III; ‘I ‘IOI’II' ‘I III IV‘I'II'III'I "Inn; «III: C I" ‘ I I III IIIIIJII-flz»: ”:95 I I * - ' , ' III‘ IIII I‘II;II III.” . 'I IIH I.. ‘ :. IIII 1‘ t I' ' II . ‘..; II‘ . hfl':}:g¢w ‘5' ‘I 1- . 9 . . “.7: Isa-fl ‘.‘zyi . g, ' ' I'I ... : II It ‘ ‘-‘F. u - If I-RIIIIII ‘ I. I.IIJI hg‘}~j§ ‘an'hh-x‘a..-‘: , 2‘ ‘I Q‘I'r“r"’v“(7$“:"7‘ I“ ‘ I I III‘ | I ‘I I V i by?- ‘- . If III‘I 5r fa? III ‘f h I.- - DI II; . I I I 'I I III II I II. . . .I , ”.1 'IYI}"-.;IIIIIIIII:- $235,: I I III II' .I ~‘1'lr1 '\ 1:“ III I I —.I I,“ (“32: H.915" §§&).:IN és I , . '.Ir 'I‘ITV. ' y " W‘I" :sf‘f t. IJ II I I .. I ‘ I". (3&3 ‘1, ‘ :1 . H I I I I . “Va ”I’ .IlIn {ah ".i & )p IE'III' III-5:1; “w; .'II I‘L‘JJ: . ' 'I I I“: I It I III . ’I I I ..’lll‘l.v‘)l"l“~'r II'III'IIQ_ ‘7‘.I‘I"Ll ' I‘ II .. I 'Hfi, .0 , I- I - I I IIIITIIII III II'I '1" ‘ .-I I TI; III, -II . I .- .‘.-.-"I- my. I * ‘ '. . I III I. ,- | IV I Z}. .—- t 3 I - . I II .II II I I I I - \’. 1 . IIII I y‘ I .l I ' ' ':I ‘ ' - I ‘ I I I . ‘ a, I. I , , Ii : ‘ ll~ I ’5‘- )- ‘* heats: LIBRARY Michigan State University I' fl. This is to certify that the dissertation entitled Studies Directed Toward the Synthesis of Euphane Triterpenes presented by Lawrence Ko lac zkowski has been accepted towards fulfillment of the requirements for Ph .D . , Chemistry degree in WM H. W Major professor Date 1 ‘3 , Cs ‘3 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 MS 85minsmaeuts: 1.] Place in book drop to LIBRARJES remove this checkout from w your Y‘QCOY‘d. FINEE will V be Charged if Book is returned after the date stamped below. STUDIES DIRECTED TOWARD THE SYNTHESIS OF EUPHANE TRITERPENES BY Lawrence Kolaczkowski A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1984 DEDICATION Dedications are, by nature, usually rather short. However in the journey leading to this degree, several people have had a profound influence on my life, and I dedicate this work to them. To my parents, Eugene and Esther Kolaczkowski, who gave me a warm and loving environment to grow up in. The only advice they gave was to do the best I could. Their only wish for me was to be happy. To my brothers and sisters, Gene, Marcia, Jim, Sandy, Dan, and Chris, who along with mom and dad have made my family one of life's true treasures. To Susan Kauzlarich, whose love and encouragement has made the difference in my life. To Dave Odelson, a special friend and colleague, whose help and companionship made the ups and downs of life in graduate school a lot easier to take. To Brian Sumner, one of my oldest and dearest friends, for always being there. To Dr. Robert M. Coates, who gave me my first opportun- ity to do research. To Dr. Caetan Vaz, who taught me the art of chemistry on a millimole scale. To Dr. William Reusch, who taught me never to speculate until all of the data is in. And last, but certainly not least, to Al and Fran Price, Dave and Emily Johnson and all the Johnson Scholars, who reminded me that there is a world outside of chemistry. -iii- ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to Dr. William Reusch for his guidance, support, and rigorous professional example during the course of this work. Many thanks to the organic chemistry faculty at Michigan State, especially Dr. Stephen P. Tanis, for several helpful discussions. Thanks also to the staff and my fellow graduate students for their friendship and sharing. I will truly miss you all. Finally, the author wishes to thank Michigan State University for a Teaching Assistantship, and also the Amway Corporation and Union Carbide, for a Summer term Research Fellowship. -iv- Kolaczkowski, Lawrence DIBAL, followed by mild acid treatment gave the desired ring A enone as a mixture of C-17 hydroxyl epimers Si and 83. Protection of these alcohols as MEM-ethers, followed by methylation with excess potassium t-butoxide and methyl iodide produced the C-4 gem-dimethyl compounds 2} and 23. Deprotection with TiCl4 followed by PDC oxidation gave diketone 2g. Finally catalytic reduction of the A1 double bond, followed by selective reduction of the C-3 carbonyl function with NaBH4 yielded 23. The structure of I} has been confirmed by X-ray crystallography. Preliminary work on a modification of this synthesis to give the AB-trans product ll is also presented. ABSTRACT STUDIES DIRECTED TOWARD THE SYNTHESIS OF EUPHANE TRITERPENES BY Lawrence Kolaczkowski Studies directed toward the synthesis of the euphane triterpenes are presented. The key steps in the synthesis are a regiocontrolled, Lewis acid-catalyzed Diels-Alder reaction, and a selective photochemical epimerization at a quaternary carbon center. Thus, the Diels—Alder reaction between quinone zg and diene 25 in the presence of BF3°OEt2 gave the tetracyclic adduct 22 in good yield. Irradiation of a mono-enol acetate derivative of 22 (SI) with light filtered to block wavelengths less than 365 nm yielded a 5.5:1 mixture of the C-10 methyl epimer S3 and recovered SI respectively. In this manner our key tetracyclic intermedi- ate was converted from a lanostane configuration to a euphane configuration in one step. Surprisingly, solvolysis of enol acetate Q3 in basic methanol yielded the AB-cis fused enedione 29 rather than the expected AB-trans isomer fié. The C-4 carbonyl oxygen in 29 was removed by selective reduction with zinc and acetic acid, conversion of the resultant alcohol 22 to mesylate SQ, followed by reduction with zinc and sodium iodide. Reduction of dienone 83 with TABLE OF CONTENTS PAGE INTRODUCTION. 0 C O O O O O O O O O O O O O O O O O O O O 1 RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . . 17 Preparation of Diane 24. . . . . . . . . . . . . . 17 Dials-Alder Reactions of 24. . . . . . . . . . . 19 Lewis Acid Catalysis of Enone- Like Dienophiles in Diels-Alder Reactions. . . . . . . . . . . . . .26 Stereochemical Considerations. . . . . . . . . . . 35 Functional Group Transformations Leading to the C-5 epi-Butylospermol Ring System. . . . . . . 50 EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . . 68 General. . . . . . . . . . . . . . . . . . . . . 68 General procedure for the evaluation of Lewis Acids in the Diels-Alder reaction of 28 and 32. . .69 Preparation of quinol 36. . . . . . . . . . . . . .70 Diels-Alder reaction of~ quinol 26 and 32. . . . . .71 Diels-Alder reaction of quinone 28 and 24. . . . . 72 Separation of 29 and 30.. . . . . . . . . . . . . 73 Preparation of enol—acetate 61.7. . . . . . . . . .75 Photoenol—acetate 62. . . . . . . . . . . . . . . .76 Preparation of trikEtone 79. . . . . . . . . . . . 77 Preparation of alcohol 72. . . . . . . . . . . . . 78 Preparation of mesylate 8Q. . . . . . . . . . . . 79 Preparation of enedione 83. . . . . . . . . . . . .80 Preparation of enones 81 and 82. . . . . . . . . . 81 Preparation of MEM-ether 8g. . . . . . . . . . . . 83 Preparation of 91 and 92 . . . . . . . . . . . . . 84 Preparation of alcohol 93. . . . . . . . . . . . . 86 Preparation of diketone 9§. . . . . . . . . . . . .87 Preparation of saturated diketone 96. . . . . . . .88 Preparation of 73. . . . . . . . . T . . . . . . . 89 Preparation of 8§. . . . . . . . . . . . . . . . . 90 Preparation of 88. . . . . . . . . . . . . . . . . 91 APPENDIX. C O O O O O O O O O O O O O O O O O O O O O O 93 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . .169 _V— LIST OF TABLES PAGE Catalyzed Diels-Alder reactions of 24 and 28. . .19 Effect of several catalysts on the Diels- Alder reaction of 28 and 32. . . . . . . . . . . 3O Attemptedcleavage of 8§ and 86. . . . . . . . . 57 Attempted olefin isomerization of 96. . . . . . .63 Attempted olefin isomerization of 92. . . . . . .64 ‘Vi" ORTEP drawing ORTEP drawing Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass Mass spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum LIST OF FIGURES of of of of of of of of of of of of of of of of of of of of of of of Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared Infrared spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum spectrum 3 ”23232323323253: 7 L0 2 ooqxoxoxoxooooooo 1m fifififififigfiGEEESESEK'. ~vii- PAGE . 93 .94 . 95 .96 .98 . 99 100 .101 102 .103 104 .105 106 .107 108 .109 110 .111 112 .113 114 .115 116 .117 118 .119 120 Kai 122 Lu3 124 .125 126 .127 128 .129 130 .131 132 FIGURE 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 PAGE Infrared spectrum of 93. . . .133 Infrared spectrum of 95 . . . . . . . 134 Infrared spectrum of ag. . . . .135 Infrared spectrum of 73 . . . . 136 Infrared spectrum of 85. . . . . . . . .137 Infrared spectrum of 88 . 138 PMR spectrum of 35. . . . . 139 PMR spectrum of 35 . . . . .140 PMR spectrum of 3g. . . . . . 141 PMR spectrum of 3] . . . . . .142 PMR spectrum of 29. . . . . 143 PMR spectrum of 31 . . . .144 PMR spectrum of 81. . . . . . 145 PMR spectrum of 63 . . . . .146 PMR spectrum of 79. . . 147 PMR spectrum of 72 . . . 148 PMR spectrum of 89. . . . . 149 PMR spectrum of 83 . . . .150 PMR spectrum of 81. . . . . 151 PMR spectrum of 83 . . . .152 PMR spectrum of 82. . . . 153 PMR spectrum of 91 . . . .154 PMR spectrum of 92. . . . 155 PMR spectrum of 93 . . .156 PMR spectrum of 95. . . . 157 PMR spectrum of 96 . . . . 158 PMR spectrum of 73. . . . . . 159 PMR spectrum of 8g . . . 160 PMR spectrum of 88. . . 161 CMR spectrum of 22 . .162 CMR spectrum of 79. . . 163 CMR spectrum of 63 . . . .164 CMR spectrum of 72. . . . 165 CMR spectrum of a; . . .166 CMR spectrum of 81. . . . 167 CMR spectrum of 33 . . .168 -viii- INTRODUCTION Terpenoid compounds are ubiquitous in the plant kingdom. Their presence in plants in such great numbers and structural diversity is enigmatic. Whereas many terpenoid compounds accumulate in concentrations greater than those normally associated with active metabolism, their exact function in plants is largely unknown. Triterpenes represent the largest class of terpenoid compounds. Many triterpenes, especially the more highly functionalized members, exhibit a wide variety of physiological activity. Tetracyclic triterpenes, a small yet structurally diverse sub-group, are composed primarily of two families, the euphanes and the lanostanes. Constitutionally both the euphanes and lanostanes are very similar to the steroids. All have a common origin; enzyme mediated cyclization of squalene.l The conformational constraints imposed by interaction with a specific enzyme determine into which manifold the cyclizations fall. As demonstrated in Scheme I, an all chair folding prior to cyclization leads to the euphanes, whereas a chair- boat-chair folding leads to the lanostanes and eventually the steroids. 2.2385 H mzmmum 0522 5.20 =4 2.20:3 I -3- The euphanes are important compounds, not only in their own right, but as precursors to a variety of oxidized and rearranged natural products such as the limonoids, meliacins, and simaroubalides. quassin veprisone _ . a Simaroubalide a limonoid kivol a meliacin Euphol, one of the best known euphane triterpenes was first isolated by Newbold and Spring in 19442 from several plants of the genus Euphorbia. The correct structural assignment however, was not reported until 1955. Many chemical and physical tests suggested a structure for euphol which was identical to the previously identified lanosterol. There were, nonetheless, subtle differences in the reactivity of the two compounds, the most conspicuous being their behavior under acidic conditions. Treatment of lanosterol with protic acid results in 8,9 7,8 isomerization of the A double bond to the A position. -4- However, when euphol is treated under identical conditions, it undergoes rearrangement to yield the isoeuphane ring system. The structural and/or stereochemical differences responsible for such drastically different reactivity were unclear. Finally in 1955 Barton et a1.3 determined conclusively the structure of euphol and isoeuphol. Barton reasoned that the structural relationship of the C and D rings in euphol and lanosterol was enantiomeric (although this initial report did not fix the stereochemistry at C—17). This configuration forces the B and C rings of euphol into strained "half-boat" conformations. Protonation and subsequent sigmatropic 1,2 methyl shifts establish a BC trans-ring fusion, which allows both rings to assume more stable chair conformations. Barton termed this relief of steric strain a "conformational driving force." isoeuphol -5- In the lanostanes, where the B and C rings are already in the chair-chair conformation, the rearrangement does not occur, as this would lead to the higher energy cis fused system. lanosterol R A lanosterol Despite the structural similarity between tetracyclic triterpenes and steroids there is a a great disparity in the amount of work published in each area. While over 100 steroid syntheses have been reported, to date only two tetracyclic triterpene syntheses have appeared in the literature. The great structural similarities belie the significantly greater synthetic challenge of the tetra- cyclic triterpenes. Both tetracyclic triterpene syntheses reported have been in the lanostane family, and this work represents our efforts toward the first euphane triterpene synthesis. -6- The two 1anostane syntheses reported represent substantially different strategic approaches. The first was reported by Woodward and co-workers in 1954.4 In it they take advantage of the constitutionally and configura- tionally steroid-like character of lanosterol. Starting with cholesterol, the conversion to lanosterol was accomplished in 17 steps, as outlined in Scheme II. Dihydrolonosterol This approach emphasizes one of the most challenging problems in tetracyclic triterpene synthesis; construction of the bis-angular methyl CD ring fusion. Most of the steps in the WOOdward synthesis addressed this problem. -7- The synthesis reported by the van Tamelen group5 used a biomimetic strategy. The polyene precursor was synthe— sized in 23 steps from (-) limonene. Lewis acid-initiated cyclization of the epoxide gave a 30% yield of tetracyclic adducts (see Scheme III). It is of interest to note that all the isomers formed from a chair-boat-chair folding are isoeuphane in nature (presumably resulting from the initially-formed euphane ring system). This demonstrates not only the sensitivity of these molecules to rearrangement, but also the inapprOpriate nature of this approach to the synthesis of the euphane triterpenes. Previous work in our research group6 has provided a simple means of preparing multigram quantities of diketone 1. This diketone contains the requisite trans dimethyl functionality found in the CD rings of many tetracyclic triterpenes. Use of 1 as a starting material for the synthesis of such compounds handily avoids the problem of angular methyl group introduction encountered by the Woodward group in their synthesis of lanosterol. H H H mimmvm 3. n3 .32.ou3=8_ T. 1.x. an. 3:23:08. . e at \\ [’a\\ S..\ 2.932333 -9- The synthesis of 1 is outlined in Scheme IV. If the (N1 0 CH, H. CH, 5%RI'I /Al 2 3 KOH <¢ 5 SCHEME IV aldol cyclization is effected with enantiomerically-pure proline as the catalyst, good yields of optically active Wieland-Miescher ketone i can be obtained.7 This enables us to synthesize either optical antipode of 1, one enantiomer for euphane synthesis, and the other for the lanostanes. Since 1 is the only chiral component in our strategy, this provides an easy access to an optically active synthesis of the euphanes. Our approach to the synthesis of the euphane triterpenes was envisioned as two steps: fusing of the AB ring system to the six-membered ring and attaching the side chain to the five-membered ring. -lO- Several strategies are possible for construction of the carbocyclic framework from diketone 1. Using terminolo- gy borrowed from steroid chemistry, they fall into three categories, each which will be discussed in turn. I. A +CD ——> ACD ——> ABCD In this approach an A-ring component is added to the CD ring system, and then the B-ring is closed to complete the tetracyclic system. Several examples of this strategy, where an A-ring anion is added to a C-ring carbonyl have been accomplished by W.S. Johnson et al.8 A direct application of this strategy using diketone 1 has been reported recently by Bull and Bischofberger9 in their l4a-methyl-l9-norsteroid synthesis (see Scheme V). Addition of the diethyl(phenylsulfinyl)methyl- phosphonate anion to 1 gave the expected unsaturated sulfinate 6. Double bond isomerization followed by sulfinate-sulfoxide rearrangement yielded allylic alcohol 1. Oxidation followed by conjugate addition -11.. SCHEME V 10 9 of 3-methoxybenzylmagnesium chloride to 8 joins the A and CD rings in compound 9. Treatment of 9 with para- toluene sulfonic acid closes the B-ring to give 10 as a 8,9 9 11 mixture of A and A ’ isomers. II. B-+CD -—> BCD -—> ABCD A majority of the applications of this strategy are again found in the work of W.S. Johnson. The hydrochrysene approach10 has been used in the synthesis of several steroids, as outlined in Scheme VI. _12_ SCHEME VI III. A-+CD ——> ABCD Of the three approaches, the A-+CD cycloaddition is by far the most efficient for construction of the tetra- cyclic framework. Because of the high degree of stereo- selectivity in the Diels-Alder reaction, the stereochemistry at the newly generated centers should be cleanly fixed. Despite these advantages, very little attention has been given to this approach. Lora—Tomayo et al.11 have reported the thermal cyclo- addition of quinone 11 to diene 13 to yield tetracyclic adduct 13 as the only product. Reaction of -13- methoxy-p-benzoquinone 11 with diene 14 reportedly yields 0 I I + fi cu,o’ : o / 11 12 cycloadduct 15,12 again as a single regioisomer. Compound 15 o 9. ° + cmo —-> 0 / cup 0 11 14 15 cup C . 16 was then converted in several steps to the steroid-like 16. Both of these reports, however, are suspect. The regio- chemistry of the cycloadditions is contrary to that expected, based on results obtained both in our laboratory13 and elsewhere.14 A study undertaken by Inouye and Kakisawa15 confirms the questionable nature of these reports. -14_ Lora-Tomayo and co-workers have reported16 the cyclo- addition of quinone 11 and styrene 12 to yield 18 as a o o l \ + / ——-> C1130 O / c1130 0 18 11 17 single isomer. Inouye and Kakisawa report a completely different result for this same reaction. Careful chromatographic analysis of the product shows it to 18 19 be a mixture of two adducts 18 and 19, the latter predominating by a 12:1 margin. Denmark17 has pr0posed an A-+CD Diels-Alder approach to steroid synthesis as outlined in Scheme VII, but to date the only report in this area has been on the synthesis of the trans-hydrindenone 21. _15_ 20 23 SCHEME VII In this dissertation we present our efforts toward the first synthesis of a euphane triterpene ring system. Our strategy is based on the two key reactions, a Lewis acid-catalyzed regiocontrolled Diels-Alder reaction, and a unique photochemical epimerization at a quaternary carbon center. A summary of our approach is outlined in Scheme VIII. In the course of this synthesis a study of 'the effect of various Lewis acids on the regioselectivity of Diels-Alder reactions with 2-methoxy-5-methyl-p- benzoquinone was conducted, and will be discussed. 71 -l6- SCHEME VIII RESULTS AND DISCUSSION Preparation of Diene 24_ Synthesis of the carbocyclic framework of the euphane triterpenes by an A-+CD ——> ABCD Diels-Alder strategy requires an efficient route to cisoid diene 24 from diketone 1. A straightforward approach to 24 involves addition of a vinyl anion to the six-membered ring carbonyl, followed by dehydration of the resultant tertiary allylic carbinol. Of central importance to this approach is the ability to carry out reactions selectively at the cyclohexanone carbonyl. A greater rate of nucleOphilic addition reactions for cyclohexanones versus cyclopentanones has been documented.18’19 This selectivity is reflected in many reactions of diketone 1. Thus Martin, Tou, and Reusch20 have reported selective addition of a series of nucleophiles to the six-membered ring carbonyl of 1. For example, the reaction of excess vinyl magnesium bromide _17- -13_ with 1 gives a good yield of alcohol 25. Changing the solvent from tetrahydrofuran (THF) to toluene enhances mg Br 3, THF ——> 60% C.H,CH,—> 75% PM? / \ \ ‘\ the 1,2 addition of Grignard reagents to easily enolizable ketones, and improves the yield of 25 from 60% to 75%. Care must be exercised in the dehydration of alcohol 25 to diene 24. In the presence of strong Bronsted acids the only products obtained are the transoid dienes 25 and 27, resulting presumably from isomerization of the initially formed diene 24. Since the transoid dienes 25 and 22 do not react with dienophiles, and isomerization to 24 is sluggish, formation of these transoid dienes must be avoided. Earlier work has shown that a solution of boron trifluoride etherate (BFB-OEtZ), in benzene/THF21 serves to dehydrate 25 to 24 without subsequent isomerization. Furthermore, recent studies have shown that copper(II) -19- sulfate induced dehydration22 of 25 on a small scale gives a near quantitative yield of 24. Further studies on the use of this method are currently underway. Diels-Alder Reactions of 24_ Tou and Reusch23 have demonstrated the feasibility of using an A-+CD Diels-Alder approach for synthesis of the lanostane ring system. Under thermal conditions diene 24 reacted with 2-methoxy-5-methy1-p-benzoquinone 28 to give a poor yield of tetracyclic adducts 28 and 38. Clearly this is not a suitable reaction for synthetic applications; however it is possible to improve both the yield and the I selectivity of this Diels-Alder reaction by Lewis acid catalysis. The effect of two of the most selective catalysts (BF3-OEt and SnCl4) is shown in Table 1. In 2 01-130 29 30 Table 1 CATALYST RATIO Q29:30) 121312 Heat 2:1 34% SnCl4 1:99 74% BF3-03t2 10:1 55% -20- each case examined, the products exhibited an a-endo configuration (i.e. had the same relative configuration at C-10, C-13, and C-14 found in the lanostanes). In planning a synthesis of the euphane skeleton there were two fundamental questions about this Diels-Alder reaction we wanted to answer. First, was it possible to increase the regioselectivity favoring isomer 28? While in principle both isomer 28 and isomer 88 could be converted to the A-ring system found in many tetracyclic triterpenes, conversion of isomer 28 would clearly be more straightforward. Secondly, could the stereochemical course of the reaction be altered to provide products having a euphane-like configuration at the C-10, C-l3, and C-14 centers? (a-exo transition state geometry in the Diels-Alder reaction). From a study of the vast body of data on Diels-Alder reactions several conclusions concerning regiochemistry 23-25 and stereochemistry have been made. 1) Reactions between simple dienes and simple dienOphiles. A) 1-substituted dienes favor products in which alkyl substituents are on adjacent sites (ortho alkyl effect). t: >150 major minor -21_ B) 2-substituted dienes favor para products. t i no major minor C) With 1,2 disubstituted dienes, the l-substi- tuent influences regiochemistry more than the 2-substituent. T + :1 +1115} major minor 2) Quinone dienophiles: A) Electron donating groups on a quinone deactivate the double bond to which they are bound towards reaction. 0 j + 1’ cmzf© / cup 0 B) In monoalkoxyquinones the alkoxy group is able to donate electron density to one of the two carbonyl groups, making it ester—like. In the Diels-Alder reaction between monoalkoxyquinones and 1—substituted dienes, the alkoxy group directs the regiochemistry in such a way that the substi- tuent on the diene is proximal to the ketone- like carbonyl in the product. -22- 61-130 cu, Note that in the last example the directing effects observed in 1A) and 2B) complement each other. However when these influences oppose one another as they do for quinone 21 mixtures of products are observed with l-substituted dienes.26 In order to achieve good 0 cmo C) 40% regioselectivity in Diels—Alder reactions of quinone 28 one or the other of these effects must be enhanced or moderated vis-a-vis the other. One means of enhancing the ortho alkyl effect is to 1 place an electron withdrawing group on C-1 of the diene. COzEt ' 0 ca, + \ ——> cup / cap 0 -23- In a similar way,an electron donating substituent in this O ac...3 O CHZOH CHZOH + \ HF CH,O / C" g 0 o OCH, 93% 27 position favors the other adduct. An alternative means of achieving greater selectivity in the Diels-Alder reaction of quinones is to alter the electrophilic character of one of the two carbonyl groups by selective activation or deactivation. By deactivating one of the carbonyl functions of a quinone we obtain a cross-conjugated dienone system which should resemble a cyclohexenone in its function as a dienophile. Although cycloalkenones are generally reluctant dienOphiles, their reactivity may be enhanced by Lewis acid catalysis. (A discussion of the Lewis acid— catalyzed Diels-Alder reaction appears in the following section). For the present, we will focus on deactivation by selective addition to one quinone carbonyl group. 28’29 have achieved selective Liotta and co—workers addition of nucleophiles to substituted quinones to give good yields of mono-addition products. We have obtained similar results in the reaction of methyllithium with _24- CH; U > TMEDA C1130 o HO CH3 CH ,0 quinone 28. This addition took place at the more reactive ketone-like carbonyl function, giving quinol 38 in 57% yield. 0 cu, cu, emu cup mm“ cmo o HO CH3 28 35 We reasoned that if quinol 38 served as a dienophile, the regioselectivity with or without Lewis acid catalysis would be improved and would lead to the desired product. As anticipated, both SnCl4 and BF3°OEt2 catalysis in the reaction of 38 with piperylene led exclusively to adduct 32. CF! 3 \ Snot, or 4- CH30 / BFa-OEt HO CH3 36 32 -25- A similar result has been reported for the related quinol 98. Unfortunately, when we attempted a Diels-Alder CH ’ + \ Snot, cu, / HO CH: 33 32 39 treaction between quinol 38 and diene 24, no cycloaddition products were obtained. Apparently the reduced activation and increased steric hinderence of dienOphile 36 combined with the bulkiness of diene 24 prohibit this cyclo- addition. A similar 1ack of reactivity had been observed 26 earlier for dienone 40 and 24. NR 40 Selective activation of one quinone carbonyl group presents an alternative means for controlling regiochemis- try in the Diels-Alder reaction. This can be accomplished with the use of Lewis acid catalysis, as was demonstrated in the Diels-Alder reaction between quinone 28 and diene 24 (see Table l). The influence of SnCl4 and BF3-OEt2 on the distribution of products may be explained in terms of the complexes they form with the quinone dieno- phile. It is proposed that tin, which is able to expand its -26- coordination sphere to hexacoordinate, is stabilized by chelation as in 41. Such coordination increases the electrophilic character of the more substituted position on the quinone double bond. Boron, which is normally Of'LA 10 CH3 CH3 LA‘0 0 41 . 42 tetracoordinate, prefers complexation at the more basic ester-like carbonyl group as in 48. This activates the less-substituted position on the dienOphilic double bond. A wide variety of Lewis acids are available as potential electrophilic complexing agents and should provide a large degree of flexibility in control of the Diels-Alder reaction. Lewis Acid Catalysis of Enone-like Dienophiles in Diels- Alder Reactions In 196031 Eaton and Yates reported a dramatic rate acceleration for the Diels-Alder reaction between anthracene and maleic anhydride in the presence of A1C13. Subsequent reports by other groups have shown that Lewis acids can markedly effect the regiochemistry of the cycloaddition as well. 33 As first reported by Lutz and Bailey in 1964, the -27- :+ f“ TON/{)0K A 9 71% 29% SnCl 4% 93% 7% presence of a Lewis acid catalyst tends to enhance formation of the isomer favored in the thermal reaction. In more complex systems (most notably with quinone dienophiles) the regiochemistry of the catalyzed reaction may be completely reversed from that of the thermal reaction. Such examples were first reported by Valenta and co- workers in 1972,33 followed by a more elaborate report in 1975.34 The following examples are from the latter work. + o o CH, (:113 \ A CM3 - + + z’ 01 TO 0' CH 81-3-0612 ’ > o -28- / \ o 5'3 In their study of the regioselectivity of Diels-Alder reactions involving quinone 28 Tou and Reuschl3 were the first to suggest that differences in complexation are important in determining product structure. To gain further understanding of this subject, a study of different Lewis acid complexes with 28 was undertaken. We hOped that one result of this study would be improved regio- selectivity and yield in the preparation of adduct 28. Although the list of possible Lewis acids is extensive, aluminum chloride (AlCl3), tin(IV) chloride (SnCl4), and boron trifluoride etherate (BF3-OEt2) are by far the most widely used catalysts. The effectiveness of these catalysts in improving the yield and controlling -29_ regiochemistry in the reaction of quinone 28 with diene 24 is well documented.13 We therefore chose to study the effect of altering the electronic environment on the acceptor atoms of these three catalysts by varying the ligands. The results of these experiments are summarized in Table 2. (Standard reaction conditions may be found in the Experimental section.) In place of diene 24 (expensive), we used piperylene as a reference diene. This simple l-substituted diene has been shown to exhibit the same regiochemistry in these Diels-Alder reactions as does 24.26 Because previous studies have shown BF3-OEt2 to be the most effective catalyst for synthesis of 28, the boron based catalysts were the first investigated. Boron trifluoride, when introduced as the diethyl etherate, must undergo exchange and recomplexation with a basic oxygen to be effective. To determine the effect of slowing down this exchange, the more basic ether THF was substituted for diethyl ether in this reaction system. Although the reaction with BF3-THF was noticably slower than with BF3-OEt2, the distribution of products proved to be the same. Alternatively, the use of BF gas would 3 provide a more active catalyst, but measurement and addition of exact quantities of BF3 gas is difficult. Thus, when piperylene was added to a methylene chloride solution of quinone 28 saturated with BF3 gas, only adduct 34 was -30_ Table 2 Diels-Alder reaction of 28 and 32. (The ratio ' of catalyst to 28 is 1:1 unless otherwise indicated.) Yield of cyclo- Catalyst Ratio(33:34) adducts Heat 1:1 80% SnCl4 1:20 85% BF3-OEt2 4:1 85% BF3°THF 4:1 70% BF3 gas 34 only 43% B(Et)3 0% B(Ph)3 0% SbCl5 0% AlCl3 (0.5 eq) 3:2 64% (1.0 eq) 7:6 76% (2.0 eq) 1:2 AlCl3/Al(i-Pr0)3 (0.5 eq/0.5 eq) 2.5:1 35% (1.0 eq/0.5 eq) 1:1 50% Al(i-Pr0)3 0% AlClZEt 10% AlClEt2 0% AlEt3 0% TiCl4 34 only 87% TiC14/Ti(OEt)4 (1.0 eq/1.0 eq) 34 only 40% Ti(OEt)4 0% n-BuSnCl3 0% (n-Bu)ZSnC12 0% (n-Bu)4Sn 0% -31- formed. The ambiguity in measuring the amount of BF3 gas introduced in such reactions makes it difficult to draw any conclusions about this result. If the electron withdrawing halogen substituents on boron are replaced with electron donating alkoxy groups the catalytic activity of the resulting borate esters vanishes. Based on their Lewis acidity, trialkyl boranes should exhibit intermediate catalytic activity. However, they undergo facile addition reactions with quinone 28. Thus, reaction of 28 with piperylene and BEt3 did not give cycloaddition products. The only product isolated and identified from this reaction was 48, which presumably results from initial 1,4 addition to 28, followed by 0 H: 01.0 611,04, 0 43 oxidation back to the quinone state. Similar addition reactions of several alkyl boranes with p-benzoquinone 44 giving the corresponding 2-alkyl-hydroquinones were reported by Hawthorne and Reinte35 in 1965. Interestingly, these authors also noted that quinones other than 44 were relatively unreactive towards alkyl boranes. Furthermore, when the alkyl groups attached to boron were replaced with phenyl substituents (BPh3), only starting materials were recovered. -32- We have observed a similar pattern of behavior for aluminum catalysts. Thus, aluminum chloride was an effective catalyst, but aluminum isoprOpoxide was ineffective in promoting the Diels-Alder reaction between quinone 28 and piperylene. The use of mixed aluminum chloride/aluminum isopropoxide catalysts did not improve either the yield of cycloaddition products, or the regioselectivity compared with AlCl3 alone. When the halogen atoms on AlCl3 were replaced with alkyl groups, the yield of cycloaddition products drOpped dramatically. For example, when added to a solution of 28 and piperylene, ethyl aluminum dichloride (AlEtClz) yielded less than 10% of the expected cycloaddition product after 6 hr. Diethyl aluminum chloride (AlEt2C1) and triethylaluminum (AlEt3) did not catalyze this reaction at all. An unusual observation was made during the AlEtCl2 and AlEt2C1 experiments. While, in general, addition of Lewis acids to a solution of quinone 28 produced orange— yellow solutions, the two alkyl aluminum chlorides gave a deep blue color. This behavior is reminiscent of the classical quinhydrone reaction. Thus, mixing equimolar amounts of p-benzoquinone 44 and hydroquinone 45 in basic solution gives a dark green (almost black) solid called quinhydrone. This reaction proceeds by an electron transfer from the hydroquinone ion to the quinone molecule, giving resonance-stabilized semiquinone ions 48. -33- 0 OH _ 6: , '1 (DH 4' ’ + 2 H20 l OI ("1 *' -q?: ‘ 44 45 46 Solutions of this complex have an intense green coloration. Although direct physical evidence is lacking in this case, a similar charge transfer complex, involving hydroquinone species generated by 1,4 alkyl addition to 28 may be responsible for the blue coloration observed in the ethyl aluminum chloride reactions. Similar studies were conducted with tin based reagents. While SnCl4 gave good yields of Diels-Alder adducts, n-butyl substituted tin chlorides as catalysts were ineffective. Ironically, the most effective catalyst found in our survey proved to be selective not for 38 but rather for 34. Titanium(IV) chloride (TiC14) gave cycloaddition yields and regioselectivity which surpassed even SnCl4 in the synthesis of 34. Interestingly, this result conflicts with the report of Henderickson and Singh36 that TiCl4 catalyzed Diels-Alder reactions with a series of quinones and dienes gave regioselectivity identical to that of the corresponding thermal reaction. It was hoped that a catalyst-regioselectivity relationship might emerge from our study of Diels-Alder _34- reactions of quinone 28. Unfortunately, this did not happen. Furthermore, no such relationship has appeared in the literature. Consequently catalyst selection has been in many respects a trial and error process. Childs, Mulholland, and Nixon37 have published a scale of relative Lewis acid strengths based on induced proton NMR chemical shifts for various Lewis acid-carbonyl compound complexes. These values vary by less than an order of magnitude from the strongest acid (BBrB) to the weakest (AlEt3), and do not correlate well with the observed effectiveness of these catalysts in the Diels- Alder reactions of quinone 28. Some catalysts such as BBr3 and SbCl5 which are listed as "stronger" Lewis acids than BF3-OEt2 and SnCl4 do not catalyze the Diels- Alder reaction of quinone 28 with piperylene. It is not clear, however, whether this is due to reactions of the catalyst with the quinone (as was clearly the case with BEt3) of just a failure of the catalyst to form the type of complex necessary to activate the quinone towards reaction. It is apparent from this brief study that much more work needs to be done before any conclusions can be drawn concerning the catalyst regioselectivity question for quinone 28. It remains to be seen whether a more effective catalyst than BF3-OEt2 exists for the synthesis of 28. -35- Stereochemical Considerations Control of stereochemistry in our key Diels-Alder reaction was the second problem posed by a euphane synthe- sis. Four stereoisomers are possible from the reaction of diene 24 with quinone 28. Dienophile 28 can approach from the a or 8 face of diene 24 and in each case in an endo or exo orientation to lead to adducts 42, 48, 48 and 28. Two of these stereoisomers, 48 and 48, would be useful for a euphane triterpene synthesis, as they have the required relative stereochemistry at C-lO, C-13 and C-l4. The stereochemistry at the two other newly-generated centers should be amenable to change at a later stage. Thus the C-5 configuration may be changed if necessary by -35- epimerization, and the C-9 stereocenter will eventually be 7’8 double bond to the A8'9 lost by migration of the A position. 7 In practice the only stereoisomer obtained from Diels- Alder reactions between diene 24 and quinone 28 is the B-endo adduct 28. A Dreiding model of diene 24 shows that the C-13 methyl group is oriented over the endocyclic double bond of the diene, effectively blocking the bottom face of the diene from attack. DienOphile approach is therefore favored at the 8 face of the diene, in spite of the fact that this yields a product in which the C—ring is forced into a twist boat conformation. An effort was next made to alter the stereochemical course of the Diels-Alder reaction. In the transition state leading to isomer 28 the methoxy group of the dienophile lies above the plane of the CD ring system, very close to the C-14 methyl group. By changing the methoxy group to a bulkier group (e.g. trityl), we hoped to increase the steric congestion enough to favor a B-exo transition state. However, initial attempts to convert quinone 28 to a bulkier ether by simple exchange were unsuccessful. The prospect of a multi-step quinone synthesis, coupled with early success in an alternative approach to the stereochemical control problem led us to curtail our studies in this area. A solution to the problem of stereochemical control at C-10 was finally achieved by a photochemical _37_ isomerization following the Diels—Alder reaction. The desired transformation is, in fact, similar to the classical example of usnic acid racemization. The structure of usnic acid 50, a constituent in several genera of lichens, was the subject of debate for over 100 years following its isolation by Rochleder and Heldt in 1843. Although optically active, usnic acid racemized when heated in acetic acid, or on acetylation with acetic anhydride in the presence of strong acids. .This racemization represented a rare example of an isomerization at a quaternary carbon center. In 1955 Stork39 proposed a' diradical/ketene pathway for this racemization (see Scheme IX). Bond cleavage gives diradical 51, and ll 50 51 - 52 SCHEME IX -38.. one resonance form which can be drawn is the conjugated diene ketene 58. Ring closure then leads to racemic usnic acid. Beginning in 1957, Barton and co-workers published a series of papers under the unassuming title "Photo- chemical Transformations". In the sixth paper of the series Barton and Quinkert40 reported their investigations into the photochemistry of cyclohexadienones, including the photochemical racemization of usnic acid. Quinkert and co-workers have since continued an extensive study of fully conjugated cyclohexadienone systems.4]‘a-e In 1979 Quinkert et a1.4le reported a detailed study of solvent effects and excitation wavelength effects on the isomeriza- tion of 2,4 androstadien-l-one 53. Irradiation of 53 in the absence of external nucleophiles gave a mixture of 58 and 54. Irradiation of 54 gave the same relative -39- ratio of 58 and 54, indicating a true photoequilibrium established between the two tetracyclic compounds. Photo- stationary ratios of 58 and 54 varied from 6.5:1 to 3.7:1, depending on the reaction conditions. A ketene intermediate, 55, was supported by spectroscopic studies,42 and by trapping experiments with cyclohexylamine. That the naturally occurring C-lO B-methyl isomer 58 predominates at equilibrium is not surprising. The anti relationship of the C-10 methyl and the C-9 proton allows the B-ring to assume a stable chair conformation. The C-9, C-10 cis relationship in isomer 54 forces the B-ring into an energetically less favorable boat conforma4 tion. These results suggest a solution to our problem of configurational control at C-10 for purposes of affecting a euphane synthesis. By introducing a A4’5 double bond into the Diels-Alder adduct 28 we generate a fully conjugated cyclohexadienone system analogous to the minor isomer 54 in the Quinkert study. If the substituents at C-3 and C~4 and the A7’8 double bond do not seriously perturb the photochemical reaction, it should be possible to use this same transformation to convert 28 to the desired C-10 epimer. Relief of steric interactions in the B-ring twist boat conformation, resulting from the B-endo transition state in the Diels-Alder, provides the driving force for this conversion. -40- Several methods are available for converting ene- diones such as 28 to a 2,4 cyclohexadienone system. Treatment of triketone 28 with two equivalents of lithium diisopropylamide (LDA) followed by quenching the resultant bis-enolate with two equivalents of trimethylsilyl chloride (TMSCl) should give the bis-silyl enol ether 58. Not surprisingly the bis-enolate was insoluble in THF 0 OSEE 9 t 29 56 at -78°C and quenching the enolate without use of a co- solvent gave poor yields of 58. Large amounts of hexa- methylphosphoric triamide (HMPA) achieved solution of the bis-enolate, but this procedure was undesirable on a large scale. Furthermore, 58 proved to be moisture sensitive, which made subsequent manipulations unnecessarily cumbersome. ' An alternative route to the A-ring dienone involved addition of a methyl group to the more reactive C-4 carbonyl function. This was attempted by addition of a methyl organometallic reagent, followed by a dehydration to give 58. Wittig olefination to give 58 followed by isomerization to 58 was another possibility. -41.. The same procedure which had been applied successfully to the synthesis of quinol 38 failed to give selective addition to the C-4 carbonyl of 28 in reasonable yield. Furthermore dehydration of 52 under a variety of conditions failed to yield dienone 58. Similarly, Wittig olefination suffered from poor yields, and these approaches to the 2,4 cyclohexadienone system were ultimately abandoned. It was from seemingly unrelated work in the lanostane synthesis that a source of an apprOpriate linearly conjugated dienone system was found. During the course of studies on the lanostane system, attempts were made to isomerize the C-5 position in 28 with base to give the AB-trans ring fusion product 58. -42.. Under a variety of conditions only starting material was recovered. That enolization had in fact occurred was demonstrated by trapping with acetic anhydride to give enol acetate Q}. Treatment of d} with mild base (K2C03 in methanol) gave back triketone l2, indicating that the AB-cis configuration was more stable. Clearly enol acetate Q} fulfills all requirements as a substrate in the photochemical reaction. It is an easily made, stable, non-hygroscopic crystalline solid, and the acetate group is easily hydrolyzed by treatment with mild base. Of concern, however, was the possible effect of the oxygen functionality at C-3 and C-4 on the course of the photoisomerization. -43- In the event, irradiation of a solution of enol- acetate 5; in dry acetonitrile with a 400 watt medium- pressure Hanovia lamp gave a 5.5:l mixture of photoenol acetate isomer fig and starting material d} respectively. The structure of 62 has been confirmed by X-ray crystallography43 (see Figure 1). Conditions for the reaction of d} are more critical than those in the Quinkert work, because of the presence of a C-17 carbonyl function. Previous work in our laboratories45 has demonstrated that irradiation of keto- acetate 63 in ether solution with a pyrex filter causes isomerization to the cis-fused keto-acetate 63. Since this Norrish type I cleavage-recombination reaction 63 64 -44- requires higher energy light than the photoepimerization of fil, this undesired epimerization is avoided by using a saturated copper(II) sulfate filter solution, which blocks wavelengths below 365 nm. Although Q} and 03 have very different molecular shapes, a mixture of these two epimers was not easily separable by routine chromatOgraphic techniques. Small amounts of pure 03 were obtained by thick layer chroma- tography using a repeated development technique (silica, 2000 microns, 50% ether/hexanes). Fortunately, the two C—lO isomers may be separated by flash chromatography45 after solvolysis. Treatment of the photoequilibrium mixture with potassium carbonate (K2C03) in methanol gave isomeric enediones 29 and 65. E o ‘_\ cup cmo ococn, 51 A o(ococu,)2 NaOCOCH, DMAP K,co, cup" 0 -45- Our assignment of the AB—trans ring fusion to isomer SE was based on an overwhelming body of evidence for a number of related systems. Bohlmann, Mathar, and Schwarz14 have studied the Diels-Alder reactions of several quinones, including 2g, with a number of simple dienes. In each case the initially formed cis-fused bicyclic adducts were readily converted to the correspond- ing trans-fused products on mild base treatment. In fact, we have observed that solid samples of 33 on standing will eventually isomerize to the trans compound 35. *0 cmo O 34 35 With a more complex system Valenta et al.34 reported the synthesis of Diels-Alder products 09 and 02. Although fig was converted smoothly to 0g, under a variety of conditions 02 failed to give 02. As noted earlier, the anti-trans relationship at C-8, C-13 and C-14 in fig is a particularly stable configuration, whereas the cis relationship in 02 is more stable than the related syn- anti configuration 03. A similar observation was made for -45- O O O 66 O O O 67 69 the regioisomeric Diels-Alder adducts 22 and 39.26 -47- From these results, since an anti-relationship at C-9 and C-lO exists in 63, it followed that the trans ring fusion should result after hydrolysis. This conclusion also agreed with the work of W.S. Johnson46 on the relative stability of perhydrophenanthrene isomers. Johnson's analysis uses the B-ring of the tricyclic system The Penanthrene Ring System as a reference, and counts the number of eclipsing and skew interactions between the fused rings. After assigning a numerical value to each interaction (3.6 kcals/eclipsing interaction and .8 kcal/skew interaction) the relative stability of each isomer may be predicted. A simple mneumonic for this analysis is "the greater number of equitorial ring bonds to the B-ring, the more stable the configuration." While the B-ring in 65 and 20 is not a true chair because of the presence of the A7'8 double bond, the distortion is not a serious one. Dreiding models show that the three ring bonds in 6§ are equitorial, while in 19 two ring bonds are equitorial and one is axial. This -48- analysis also predicts the formation of 65 from 62. 0' .9 90 CH ,0 65 70 At this stage of our synthesis, the critical task of fixing the relative configuration of the angular methyl groups was completed. The remaining task, modification of the A-ring, required the introduction of geminal methyl groups at C-4, introduction of an equitorial hydroxyl at C-3, and removal of the C-1 carbonyl to give 71. 71 As a definitive structure proof of 71 it was our intent to convert it to 72, a compound recently obtained by Levisalles and Audouin47 by chemical degradation of the side chain of euphol. -49- 72 For this comparison we needed to effect isomerization 7’8 double bond to the A8’9 of the A position, and methylation of the C-3 hydroxyl. Compound Tl proved to be resistant to all attempts at migration of the double bond. Since this is a well known reaction in the euphane ring system we sought to confirm our structural assignment on ll by X-ray crystallography.44 It was at this point we discovered the actual structure of our product was not the trans-fused 71 but rather the cis-fused 73. 9...... 73 -50- Functional Group Transformations Leading to the C-5 epi-Butyrospermol Ring System Our plans for converting the A-ring of 79 to the A-ring structure found in triterpenes centered on intermediate enone 74. Thus the A-ring system in 74 appeared ideally 70 74 suited for introduction of the geminal methyl groups at C-4. Alkylation at C-l6 is prevented by reduction of the C-l7 carbonyl, and alkylation at C-2 is blocked by the A1 double bond. Our initial approach to 73 was based on a reaction sequence developed by Woodward and co-workers for their synthesis of cholesterol.4 This conversion has been studied by a member of our research group working with lanosterol intermediate 33. Yields for the reactions in this sequence were generally good to excellent, with the exception of the final reductive deacetoxylation. -51- 1)DIBAL 2)H,o‘ 1) Aczo/ny . coon, 2)2n/HOAc <35% OVERALL YIELD 76 Since a low yield sequence such as this was deemed unacceptable, an alterative procedure was sought. One such alternative was found in the work of Speziale, Stevens, and Thompson, involving a modification of the 48 Woodward cholesterol synthesis. Thus, treatment of Diels-Alder adduct 72 with zinc dust and aqueous acetic Zn liOAq/Hgo H5 H 77 78 acid gave an excellent yield of the C-4 alcohol 7g as a single isomer. Selective reduction of one carbonyl function in this case is not surprising, since the C-1 -52- carbonyl is ester-like in character, due to donation of electron density from the C-3 methoxy group, and is consequently less reactive than the C-4 carbonyl. We anticipated that this reaction could be applied to the euphane synthesis without significant modification, and that it would reduce only the C-4 carbonyl. The presence of an additional ketone function at C-l7 was not a concern, as experience has demonstrated this to be the least reactive of the three carbonyl groups. In the event, treatment of 79 with zinc dust in aqueous acetic acid for 1.5 hr gave an excellent yield of alcohol 72 as a colorless crystalline solid. In order to complete the conversion of 79 to enone 73 in a concise fashion, we planned to transform the hydroxy function at C-4 to a good leaving group, and then effect simultaneous 0 o 99..... p 70 79 R=H 80 R=Ms displacement and reduction of the two remaining carbonyl groups with an appropriate hydride reducing agent. Work 49 by Holder and Matturro suggested that lithium triethyl- borohydride (Super-Hydride), which is a source of -53- exceptionally nucleophilic hydride, was the reagent of choice for this transformation. To this end alcohol 73 was smoothly converted to the corresponding mesylate 89 in 94% yield. Attempts to prepare the corresponding tosylate gave less than 50% conversion after 24 hours. Unfortunately reduction of 30 with three equivalents of Super-Hydride produced a complicated mixture of products. The problem with this 9. 81 82 reaction may lie in the work-up. Triethyl boron (a pyrophoric liquid) is a product in the reaction and must be destroyed before it is exposed to the atmosphere. How well the reduction product survives the basic hydrogen peroxide treatment normally used to destroy the -54- triethyl boron is not known. The crude reduction product obtained from subsequent extraction followed by a mild acid wash yielded small amounts of the desired enones Si and 63. Attempts to effect the same reductions using solu— bilized lithium aluminum hydride (LAH) in THF, which Brown claims50 is superior to an LAH/THF slurry for the reductive removal of tosylates and mesylates, gave similar results. In light of these observations, a stepwise removal of the mesylate followed by reduction of the remaining carbonyl functions was explored as an alternative route to Bl and 83. Several methods for the selective removal of mesylates in the presence of other sensitive functionality have been reported.51a-C. One of the 51° involves mildest, the method of Fujimoto and Tsatsuno, treatment of mesylate SQ with sodium iodide and zinc dust in refluxing glyme. This gave a very good yield of enedione 33 as a crystalline solid. Subsequent reduction A Zn 80 83 of diketone 83 with two equivalents of diisobutylaluminum hydride (DIBAL) in methylene chloride, followed by -55.. treatment with dilute HCl in 4:1 THF/water, then gave an 86% yield of 81 and 82 in a 2:1 ratio. 1) DIBAL 83 2)H* 81 82 The overall yield for our four step conversion of 79 to a mixture of 8} and 83 was greater than 69%, which more than doubles the yield from the Woodward procedure. Introduction of the geminal dimethyl substituents at C-4 presented many unexpected problems. Initial alkylation studies were conducted using lithium diisopropylamide (LDA) as a base. It was hoped that the oxaphilic character of lithium, coupled with the hindered nature of the C-17 oxygen would serve to minimize O—alkylation at that site. However, the alkoxyenolate generated by treatment of 8} with two equivalents of LDA proved to be insoluble in THF. Addition of HMPA as a co-solvent was necessary to achieve a homogeneous system. Reaction of this solution with one equivalent of methyl iodide followed by one equivalent of a proton source gave a mixture of C-4 methylated, and O-methylated products, together with recovered starting material. -56- O-methylation at C-l7 was not viewed as a serious problem at this stage, due to the variety of reagents available for the cleavage of methyl ethers. The most reasonable course at this point seemed to be exhaustive methylation, followed by cleavage of the C—l7 O-methyl ether. Unfortunately attempts at exhaustive methylation using LDA as the base gave mixtures, and the yield of 'the tri-methylated product was never maximized to a satisfactory level. Remarkably, treatment of 8} with a large excess of potassium t-butoxide in THF followed by excess methyl iodide gave only two products, 83 and 85. Extended reactions times led to 85 as the sole identifiable product; however the yield of 85 was only fair (45%). Although the sensitivity of the euphane triterpenes to protic acids limited the choice of reagents for ether cleavage,52 several well-studied methods seemed attractive. The most promising of these were boron tribromide 53a,b 54a-c (BBr3) and trimethylsilyl iodide (TMSI). -57- Table 3 Attempted cleavage of 85 and 86. Substrate Reagent Solvent Time Temperature Product Reference 85 TMSCl/NaI CH3CN 24 h RT nr 54c 85 TMSCl/NaI CH3CN 25 h 82° nr 54c 85 TMSI CHCl3 12 h RT nr 54a 85 TMSI CHCl3 12 h 62° nr 54a 85 BBr3 CH2C12 16 h O-rRT multiple 53a ‘~ products 85 BBr3. KI CH C1 12 h -46° tar 53b ,~ 2 2 18-crown-6 86 TMSCl/NaI CH3CN 16 h 82° 87 54c 86 TMSI CHCl 12 h 62° 87 54a Unfortunately, under a variety of conditions, these reagents failed to yield any products which could be identified as the C-17 hydroxyl compound. The results of these experiments are summarized in Table 3. Dreiding models show that the C-17 methyl ether is extremely hindered by surrounding groups, making approach of these relatively large demethylating reagents difficult. Comparative studies with the steriod derivative 17- methoxy-androst-4-en-3-one 86 bear out this argument. Although BBr3 failed to give cleavage products, the silicon reagents gave variable yields of testosterone 82 as seen in Table 3. -58.. To ensure that the A1 double bond in 85 was not causing unwanted side reactions in the experiments with BBr it was removed by catalytic hydrogenation with 3! . . 5 . . Wilkinson's catalyst. 5 However reaction of 88 With BBr3 under a variety of conditions still gave only complex (Ph,P),RhCI H2 85 88 product mixtures. A different approach to ether cleavage was reported recently by Olah et al.56 Treatment of simple secondary methyl ethers with cerric ammonium nitrate and sodium bromate in aqueous acetonitrile gave the corresponding ketone as the chief cleavage product. However, in our hands, this method failed to give any of the desired methyl cleavage product. The resistance of our C-17 methyl ether derivatives to a variety of ether cleavage reactions required that a different protective group be used for the C-17 hydroxyl function. The hindered nature of this group was once again evident in the protection step. Common reagents for the introduction of trimethylsilyl (TMS), t- butyldimethylsilyl (TBDMS), benzyl, benzoyl, and -59- pivaloyl protecting groups all failed to react, or did so in poor yield. Fortunately, B-methoxyethoxymethyl chloride (MEM-Cl), a reagent developed by E.J. Corey et al.,57 reacted smoothly with a mixture of 81 and 82 to give the corresponding MEM ethers in 87% yield. Dimethylation of 89 90 the protected alcohols with potassium t-butoxide and methyl iodide in THF as previously described gave a 57% yield of the gem-dimethyl compounds 94 and 95. As 91 92 expected, the MEM protecting groups proved to be easily removed by treatment with TiCl4 in methylene chloride. It should be noted, however, that zinc bromide, which is effective for deprotection in most instances, failed in this case to give cleavage products cleanly. -60- 91 +92 93 94 Finally, oxidation of the resulting epimeric alcohol mixture with pyridinium dichromate (PDC)58 in dimethyl- formamide gave diketone 95. 93 + 94 PDC > [NWF 95 The last two steps of this synthesis were straight- forward. Catalytic hydrogenation of the A1 double 55 bond with Adam's or Wilkinson's catalyst gave the saturated ketone 96 in near quantitative yield. -51- g 0 0f (Ph3P)3RhCI ® H, o 95 96 Finally, taking advantage of the low reactivity of the C-17 carbonyl function, treatment of 96 with sodium borohydride gave selective reduction of the C-3 carbonyl, yielding the axial alcohol 73 in 91% yield. 99.... 95 73 At this stage we were unaware of the AB-cis nature of the ring fusion in 73. It was our intention to convert the final product (which we believed to be 71) into 72 in order to achieve a final structure proof. An optically active antipode of 72 was obtained recently by -62- . . 4 . . . LeVisalles and Audouin 7 through a Side chain degradation of euphol. A direct comparison of our synthetic product with this degradation compound would require an isomeriza- tion of the A7’8 double bond, followed by methylation of the C-3 hydroxyl group in 71. Butyrospermol has been converted to euphol by two different methods,59'60 and these served as a basis for the double bond isomerization experiments. The first method involves an acid-catalyzed double bond shift. Although treatment of butyrospermol with strong acid leads to isoeuphane PIOdUCtS, a simple A7'8 to A8'9 double bond migration results from mild acid treatment (HCl gas in chloroform). In the second method, treatment of butyro- spermol with a hydrogen saturated Adam's catalyst (PtOz) induces olefin isomerization to give euphol. Since 4,4- dimethyl steroids containing an equitorial hydroxyl at C-3 undergo a ring contraction rearrangement on treatment with strong acid, our acid catalyzed isomerization reac- tions were carried out on the 3-keto derivative. -63- Diketone 96 was treated with a series of acid and transition metal catalysts in an attempt to isomerize the A7,8 summarized in Table 4. double bond. The results of these experiments are Much to our surprise in all cases Table 4 Attempted olefin isomerization of 96. Catalyst Solvent Temperature Time Result collidine-HCl CH2C12 40° 4 hr nr HCl CHCl3 0° 2 hr nr HCl CHCl3 25° 1.5 hr nr HCl CHCl3 25° 8 hr nr PtOz/H2 HOAc/CGH12 25° 1 hr nr RhC13-3HZO EtOH/CHCl3 70° 24 hr formation of 95 p-TSA-HZO THF 67° 12 hr nr C6H6 25° 1 week nr C6H6 80° 4 hr nr 70% HClO CH CN 25° 10 hr nr 4 3 but one starting material was recovered unchanged. -64- Even treatment with perchloric acid yielded only starting material, which seemed remarkable in light of the euphanes proclivity toward rearrangement under these conditions. Isomerization was next tried with 92. 97 The results of these experiments are summarized in Table 5. Once again no Table 5 Attempted olefin isomerization of 9]. Catalyst Solvent Temperature Time Result HCl gas CHCl3 25° 2 hr cleavage of MEM ether PPTS C6H 25° 24 hr nr PPTS C6H6 48° 24 hr nr PPTS C6H6 80° 24 hr nr P-TSA C H 80° 24 hr several 6 5 products PdC12-2C6H5CH2CN C6H5 80° 12 hr giggizis PdClz-HCl HOAc/HZO reflux 24 hr several products -55- products corresponding to the A8'9 double bond isomer were identified. The failure of these efforts to effect double bond isomerization was disturbing. Consequently, a sample of keto-alcohol 73 was submitted for X-ray crystallographic analysis. The results of this analysis was that 73 proved to be the unexpected AB-cis isomer (see Figure II). We can now rationalize the reluctance of the A7’8 double bond to migrate to the A8'9 position. Models of 98 show severe interactions between either the C-4 a-methyl O 0“ H and the C-13 angular methyl, or the C-4 B-methyl and the C-10 angular methyl. Only in 96 can these methyl interactions be avoided without invoking non-chair forms in the A and B rings. Once we discovered that the previous synthesis had produced a 5-epi-butyrospermol derivative (AB-cis), modifications were sought that would redirect it to the desired AB-trans ring fusion. The most attractive of these modifications was based on work by Pike, Summers and Klyne.61 Thus lithium-ammonia reduction of the -66- 4,4-dimethy1-1umista-5,7-diene-3-one derivative 99 was reported to give the 4,4-dimethy1-5B—lumistan-B-one 100 in good yield. In our system an analogous B-ring homoannular diene may be generated first by introduction 1lJ th 2)E°Il o 99 100 of a A4'5 double bond in 89 and 99, followed by bis- methylation at C-4 as previously described (Scheme X). Lithium-ammonia reduction of 102 should then lead to the butyrospermol derivative 103. Modification of 103, using SCHEME X -57- methodology developed in the synthesis of 73, would lead to 73. Preliminary work in this area is encouraging. Indeed, initial results indicate that the alkylation step (to form 102) proceeds in higher yield than that observed for the formation of the cis—fused 4,4-dimethyl adducts 93 and 92. ~ EXPERIMENTAL General Except where otherwise indicated, all reactions were conducted under a dry argon or nitrogen atmosphere, using solvents distilled from apprOpriate drying agents. All glassware was oven dried at 180-200°C for four hours previous to use, or flame dried (three times) under a continuous stream of dry argon or nitrogen. Reactions were monitored by thin layer chromatography (Merck Silica Gel- 60, F-254, .2 mm) with visualization by ultraviolet fluorescence, or spray reagent (30% H2804 or 5% p- anisaldehyde in ethanol) with subsequent heating. Small scale chromatographic separations were accomplished with the use of 2 mm silica plates (Merck F-254, 20 X20 cm). Larger scale separations were effected by flash chromatography (40-63 millimicron silica gel, Merck 9385). Melting points were determined on either a Thomas-Hoover capillary melting point apparatus, or a Reichert hot-stage microscopic, and are uncorrected. Infrared (IR) spectra were recorded on a Perkin-Elmer 237B grating spectrophotometer. Mass spectra (MS) were obtained with a Finnigan 4000 GC/MS spectrometer. Proton magnetic resonance spectra (PMR) -68- ~69- were taken in deuterochloroform and recorded on either a Varian T-6O or a Bruker WM 250 spectrometer at 250 MHz, and are calibrated in parts per million (6) downfield from tetramethylsilane (TMS) as an internal standard. Carbon magnetic resonance spectra (CMR) were recorded on a Bruker WM 250 Spectometer at 69.8 MHz using deutero- chloroform as solvent and are calibrated in parts per million (6) downfield from TMS as internal standard. Microanalyses were performed by Spang Microanalytical Labs, Eagle Harbor, MI. General Procedure for the Evaluation of Lewis Acids in the Diels-Alder Reaction of Quinone 28 and Piperylene 33 To a solution of quinone 23 (152 mg, 1 mmol) in 5 mL of methylene chloride, which has been cooled to 0°C, is added 1 mL of a l M solution of Lewis acid in methylene chloride. After stirring for.5 hr, the solution is cooled to -15°C and 1 mL (681 mg =10 mmol of a 1:1 cis/trans mixture) of piperylene 33 in 4 mL of methylene chloride is added slowly over.5 hr. Reaction progress is monitored by GLC. The reaction is quenched by addition of 1:1 methanol/water and warming to room temperature. The mixture is diluted with methylene chloride, washed with saturated sodium bicarbonate, water, and brine, and dried over sodium sulfate. Removal of solvent yields the crude product as an orange oil. -70.. Preparation of quinol 36 Quinone 26 (152 mg, 1.0 mmol) was dissolved in 50 mL of dry THF in a 100 mL, three-neck flask. To this solution is added 0.9 mL (696 mg, 6.0 mmol) of tetra- methylethylenediamine (TMEDA) and this mixture was cooled to -78°C. Following slow addition of .72 mL of 1.4 M methyllithium (1.01 mmol) in ether, the reaction mixture was stirred for 7 hr, quenched with methanol, and warmed to room temperature. Removal of solvents under vacuum gave a brown oil, which was dissolved in methylene chloride, and washed with water, brine, and dried over sodium sulfate. Evaporation of this solution yielded crude 36 as a brown oil. Further purification was effected by passage through a short column of silica gel, followed by crystallization from chloroform/hexanes to give 95 mg (57%) of 36 as a white powder. Characteristic properties of 36 are: M.P. 103-103.5°C Mass spectra (70 eV) m/e (rel. intens.) 168(40), 153(100), 125(34), 108(14). 69(33), 53(18), 43(42), 39(20). Infrared (CDC13) 3560(s), 3400(br), 2970, 2230, 1750, 1675, and 1640 cm‘l. PMR (CDC13) 250 MHz, 5 1.49 (d, 3H, J =5.8 Hz), 1.86 (d, 3H, J'=1.5 Hz), 3.04 (s, 1H), 3.77 (d, 3H, 6.1 Hz), 5.44 (S, 1H),6.45 (br s, 1H). -71- Diels-Alder reaction of quinol 36 and piperylene 33 Quinol 36, (76 mg, .45 mmol) was dissolved in 7 mL of methylene chloride and cooled to -46°C (COz/cyclohexa- none). To this solution was added 0.5 mL of a 1.0 M solution of tin (IV) chloride in methylene chloride (0.5 mmol). After stirring for 15 min, a solution of 0.5 mL of piperylene (50% cis-trans isomers, 0.34 g, 5.0 mmol) in 5 mL methylene chloride was added slowly over 1 hr. After stirring for 4 hr at -46°C, there was no evidence of adduct formation by GLG analysis so the temperature was increased to 0°C. After stirring for 6 hr at 0°C, the reaction mixture was quenched with water. The organic phase was diluted with methylene chloride, washed with saturated sodium bicarbonate, water, brine, and then dried over sodium sulfate. Analysis by GLC showed formation of a single adduct, plus a small amount of unreacted quinol. Column chromatography (silica-ether) yielded 50 mg (47%) of 33 as a white solid. An analytical sample was prepared by recrystallization from ether/hexanes. When this Diels-Alder reaction was conducted using boron trifluoride etherate as the catalyst, an adduct which proved identical to 33 by mixed melting point and 250 MHz PMR was isolated in 41% yield. Characteristic properties of 33 are: M.P. 130.5-132°C Mass spectra (70 eV) m/e (rel. intens.) 236(12), 203(14), 175(13), 169(63), 152(27), 137(98), 128(67), 127(30), -72- 109(44), 100(58), 91(30), 70(100), 68(54), 43(80). Infrared (c0013) 3550(br), 2960, 2240, 1710, 1670, 1640, 1610, and 1260 cm"1 PMR (c0c13) 250 MHz, 5 0.99 (d, 3H, J =7.5 Hz), 1.35 (s, 3H), 1.46 (s, 3H), 2.32 (s, 1H), 1.10-2.50 (m, 4H), 3.75 (s, 3H), 5.43 (s, 1H), 5.68 (d, 1H, J =9.6 Hz), 5.79 (m, 1H). Diels-Alder reaction of quinone 28 and diene 24 Quinone 23 (15.2 g, .10 mol) was dissolved in 500 mL of methylene chloride in a two-liter, three-neck, round bottomed flask equipped with a mechanical stirrer, and a pressure equalizing addition funnel. After cooling this solution to 0°C, 14.7 mL (16.96 g, .12 mol) of boron trifluoride etherate was added in one portion. The resulting orange solution was stirred for 15 min., cooled to -15°C (COZ/ethylene glycol) and a solution of diene 23 (11.8 g, .062 mol) in 250 mL of methylene chloride was added slowly over a 2 hr period. The temperature of the now dark green solution was maintained at -15°C for 24 hours. After addition of 100 mL of 50% methanol/water, the reaction mixture was warmed to room temperature. The organic phase was separated and concentrated to ca. 150 mL. Unreacted quinone was removed by reduction with 200 mL of 10% sodium bisulfite. The disappearance of quinone was monitored by TLC (silica— ether). When the quinone was no longer present the -73- organic phase was separated, washed with water, brine, and dried over sodium sulfate. Evaporation of the solvent gave an orange oil, which when triturated with ether yielded a creme colored solid. Recrystallization from methylene chloride/cyclohexane gave 10.9 g of 23 as an off-white solid. The mother liquor proved to be a 1:1 mixture of 23 and regioisomer 33. This mixture was easily separated according to the following procedure. §§paration of 23_§gg;}3 A mixture of regioisomers 23 and 33 (1.9 g, 5.6 mmol) was added to a solution of sodium bicarbonate (.48 g, 5.71 mmol) in 150 mL of methanol and heated under reflux for 6 hr. After chilling in ice, the mixture was filtered and the collected solid was dissolved in methylene chloride, washed with water, brine, and dried over sodium sulfate. Removal of the solvent gave 1.0 g of an off-white solid, which by TLC and 250 MHz PMR proved to be pure 23. Regioisomer 33, contaminated with a small amount of 23, may be recovered from the mother liquor by removing the solvent under vacuum, and repeating the previous process on the resulting solid. The combined yield of 23 from the initial crystallization and subsequent treatment of the mother liquor is 12.47 g (59% based on starting diene). -74- Characteristic properties of 23 are: M.P. 260-261°C Mass spectrum (70 eV) m/e (rel. intens.) 342(1), 314(15), 299(9), 189(11), 155(10), 154(100), 145(12), 119(15), 105(17), 69(10). Infrared (CDC13) 2960, 2150, 1720, 1700, 1660, 1605, 1150 cm'l. PMR (CDC13) 250 MHz, 6 1.01 (s, 3H), 1.15 (d, 3H, J =0.9 Hz), 1.39 (s, 3H), 1.40-2.80 (m, 11H), 2.99 (dd, 1H, J'=7.9 and 10.1 Hz), 3.78 (s, 3H), 5.29 (dd, 1H, J =3.1 and 6.4 Hz), 5.67 (s, 1H). CMR (CDC13) 69.8 MHz,6 219.54, 201.54, 196.02, 159.11, 145.15, 115.74, 109.54, 56.31, 56.10, 50.90, 50.66, 46.75, 42.31, 34.20, 30.76, 27.58, 25.94, 25.35, 24.55, 23.38, 17.58. Characteristic properties of 33 are: M.P. 218-220°C Mass Spectrum (70 eV) m/e (rel. intens.) 342(1), 327(1), 314(33), 299(27), 145(23), 131(31), 126(22), 125(26), 119(25), 105(46), 91(48), 79(22), 77(27), 69(100), 55(36), 41(41). Infrared (CDC13) 2920, 2240, 1760, 1725, 1665, and 1610 cm‘l. PMR (CDC13) 250 MHz, 6 1.06 (s, 3H), 1.24 (d, 3H, J’=0.9 Hz), 1.39 (s, 3H), 1.49-2.95 (m, 11H), 2.51 (ddd, 1H, J =1.8, 9.8 and 19.2 Hz), 3.77 (s, 3H), 5.27 (dd, 1H, J'=2.7 and 7.0 Hz), 5.78 (s, 1H). -75- Preparation of enol-acetate 63 A mixture of the triketone 23 (8.4 g, 24.6 mmol), anhydrous sodium acetate (4.03 g, 49.2 mmol), and 4- (dimethylamino)pyridine (DMAP) (20 mg, .16 mmol) was added to 300 mL of 2:1 (v/v) benzene and acetic anhydride. After heating under reflux for four days, the mixture was cooled to room temperature and filtered to remove sodium acetate. The solvent was removed under vacuum, and the resulting solid dissolved in methylene chloride, washed with water and brine, and dried over sodium sulfate. Removal of solvent yielded a solid for which TLC (silica-ether) showed two spots Rf==.34, and Rf =.21. Separation by flash chromatography (silica-40% methylene chloride/ether) yielded 8.7 g (92%) of 63 as the lower Rf component, and .4 g (5%) of unreacted 23 as the higher Rf component. An analytical sample of enol-acetate was prepared by recrystallization from ethyl acetate/petroleum ether. Characteristic prOperties of 63 are: M.P. l95-197°C Mass spectrum (70 eV) m/e (rel. intens.) 384(3), 342(24), 327(15), 324(22), 299(11), 129(11), 128(10), 119(11), 115(11), 105(18), 91(16), 69(20), 55(13), 43(100), 41(13). Infrared (CDC13) 2950, 1760, 1740, 1675, and 1605 cm-1. PMR (CDC13) 250 MHz, 6 0.97 (s, 3H), 1.26 (s, 3H), 1.50 (s, 3H), 2.26 (s, 3H), 1.60-3.10 (m, 11H), 3.75 (s, 3H), 5.43 (s, 19), 5.56 (q, 1H). -76- Photoenol-acetate 62 ~ A solution of 63 (1.5 g, 3.9 mmol) in one liter of dry acetonitrile was deoxygenated by bubbling a stream of dry argon through the solution for 15 min. The solution was then irradiated for 1.5 hr using a Hanovia medium-pressure mercury lamp filtered by pyrex and a saturated copper(II) sulfate solution. The filter solution -was cooled by an ice-calcium chloride bath, and circulated through a jacketed vessel surrounding the lamp. Removal of the solvent yielded 1.5 g of a gummy yellow solid. Separation of products was difficult, but was accomplished on a small scale. Thus 150 mg of crude product was subjected to preparative TLC (silica-ether, 4 passes). Three bands were noted: Rf==.58 (8.4 mg, not identified), Rf =.50 (109 mg) 63, and R ==.44 (19 mg) 63. f Characteristic properties of 63 are: M.P. 176-177.5°C Mass spectrum (70 eV) m/e (rel. intens.) 384(8), 342(42), 207(16), 204(14), 181(20), 167(35), 166(83), 161(18), 119(20), 105(17), 69(21), 43(100). Infrared (c0c13) 2990, 1785, 1755, 1645, and 1600 cm'l. PMR (CDC13) 250 MHz, 6 0.97 (s, 3H), 1.06 (s, 3H), 1.27 (s, 3H), 2.27 (s, 3H), 0.80-3.30 (m, 11H), 3.77 (s, 3H), 5.43 (d, 1H, J =3.9 Hz), 5.49 (s, 1H). CMR (CDC13) 69.8 MHz, 6 218.87, 200.72, 199.10, 168.67, 164.85, 143.50, 142.41, 117.59, 99.92, 56.25, 51.14, -77- 50.25, 46.73, 43.11, 34.14, 31.08, 26.88, 26.23, 24.17, 23.25, 20.23, 17.35, 15.09. Preparation of triketone 73 The mixture of enol-acetates 63 and 63 (2.0 g, 5.2 mmol) was dissolved in 100 mL of dry methanol. Potas- sium carbonate (0.5 g, 3.6 mmol) was then added in one portion and the mixture was stirred for 10 min. Filtration followed by solvent removal yielded a brown solid which was dissolved in methylene chloride, and washed with 10% aqueous acetic acid, water, and brine. The aqueous phase was back-extracted with methylene chloride. The combined organic phases were dried over sodium sulfate. Evaporation of the solvent yielded a brown solid, which after trituration with ether, gave 1.5 g of a light yellow solid. Flash chromatography (silica gel—5% methylene chloride/ether) yielded two products, 23 (Rf==.39, .25 g), and 73 (Rf==.28, 1.25 9). Characteristic properties of 73 are: M.P. 229-232°C Mass spectrum (70 eV) m/e (rel. intens.) 342(19), 299(10), 211(15), 171(16), 138(12), 119(16), 114(100), 105(29), 91(19), 86(32), 69(28), 55(12), 44(15). Infrared (CH2C12) 2990, 1750, 1725, 1680, and 1625 cm‘l. PMR (CDC13) 250 MHz, 6 0.91 (s, 3H), 0.97 (s, 3H), 1.34 (s, 3H), 1.30—3.15 (m, 12H), 3.81 (s, 3H), 5.46 (q, 1H), 5.87 (s, 1H). -73- CMR (CDC13) 69.8 MHZ, 6 218.5, 201.5, 193.3, 162.1, 141.8, 117.9, 109.5, 56.3, 53.5, 52.1, 50.5, 46.6, 40.1, 34.1, 30.7, 26.9, 24.5, 23.7, 21.1, 19.2, 14.3. Preparation of alcohol 73 Triketone 73 (1.5 g, 4.39 mmol) and zinc dust (.5 g, 7.65 mmol) were added to 60 mL of 2:1 (v/v) acetic acid/ water. After stirring at room temperature for 1.5 hr the unreacted zinc was removed by filtration, washed several times with hot methanol, and discarded. The combined organic and aqueous phases were washed with ether (5 times), and the combined ether phases were washed with saturated sodium bicarbonate, water, and brine. After drying over sodium sulfate removal of solvent by rotary evaporation yielded 1.43 g (95%) of 73 as a white solid. An analytical sample was prepared by recrystallization from methanol (colorless plates). Characteristic properties of 73 are: M.P. 256.5-257°C Mass spectrum (70 eV) m/e (rel. intens.) 344(10), 155(17), 154(100), 123(13), 56(8). Infrared (CDC13) 3585(5), 3545(br), 2240, 1724, and 1613 cm'l. PMR (CDC13) 250 MHz, 6 1.00 (s, 3H), 1.02 (s, 3H), 1.08 (s, 3H), 1.20-2.75 (m, 12H), 2.99 (d, 1H, J =2.8 Hz), 3.78 (s, 3H), 4.34 (dd, 1H, J =2.4 and 10.0 Hz), 5.31 (s, 1H), 5.49 (q, 1H, J =3.1 Hz). -79- CMR (CDC13) 69.8 MHz, 6 202.0, 189.0, 173.7, 143.4, 118.3, 100.2, 68.1, 56.1, 50.5, 47.6, 46.8, 46.2, 36.5, 34.2, 30.9, 27.2, 24.6, 24.1, 24.0, 18.6, 14.3. Analysis: Calculated for C21H2804: C, 73.23; H, 8.19 Found: 73.39 8.16 Preparation of mesylate 80 The alcohol 73 (1.36 g, 3.95 mmol) was dissolved in 50 mL dry pyridine and cooled to 0°C. Mesyl chloride (0.62 mL, 0.91 g, 6.3 mmol) was added via syringe, and the mixture was placed in the freezer overnight. Pyridine hydrochloride was removed by filtration and discarded. After removal of the solvent under vacuum, the residue was dissolved in methylene chloride, and washed with water, brine, and dried over sodium sulfate. Removal of solvent by rotary evaporation yielded 1.56 g (94%) of 83 as a yellowish solid. Recrystallization from methylene chloride/heptane gave pure 83 as a white crystalline solid. Characteristic properties of 83 are: M.P. 196-198°C Mass spectrum (70 eV) m/e (rel. intens.) 422(8), 343(48), 326(19), 187(27), 154(57), 153(81), 150(56), 119(59), 113(100), 105(62), 98(60), 69(28), 55(35), 41(34). Infrared (CDC13) 2950, 1725, 1650, 1620, 1350, and 1175 cm-1. -80- PMR (CDC13) 250 MHz, 6 1.01 (s, 3H), 1.05 (s, 3H), 1.12 (s, 3H), 1.25-2.65 (m, 12H), 3.17 (s, 3H), 3.82 (s, 3H), 5.34 (d, 1H, J'=9.5 Hz), 5.41 (d, 1H, J’=1.2 Hz), 5.52 (m, 1H). CMR (CDC13) 69.8 MHz, 6 200.3, 189.4, 168.8, 143.5, 117.6, 102.3, 78.1, 56.4, 50.4, 48.4, 44.8, 39.1, 36.6, 34.1, 30.9, 27.2, 24.6, 24.1, 18.6, 14.0. Analysis: Calculated for C H O S: C, 62.54; H, 7.16; 22 30 6 S, 7.59; Found: C, 62.71; H, 7.14; S, 7.51. Preparation of enedione 83 The mesylate 83 (1.0 g, 2.37 mmol), sodium iodide (1.78 g, 11.8 mmol), and zinc dust (1.54 g, 23.6 mmol) were added to 20 mL of dry glyme and heated under reflux for 3 hr. After cooling to room temperature and filtering, the solution was diluted with water and extracted with ether. The combined ether extracts were washed with brine and dried over sodium sulfate. Evaporation of the solvent yielded 0.698 g (89%) of 83 as a white powdery solid. An analytical sample of 83 was prepared by recrystallization from ether (white needles). Characteristic properties of 83 are: M.P. 215-217°C Mass spectrum (70 eV) m/e (rel. intens.) 328(47), 313(14), 230(17), 215(19), 213(23), 139(95), 138(100), 119(37), 105(38), 99(19), 91(19), 40(20). Infrared (c0013) 2975, 1740, 1620, 1390, and 900 cm'l. -31- PMR (CDC13) 250 MHz, 6 1.01 (s, 3H), 1.05 (s, 3H), 1.06 (s, 3H), 1.35-2.80 (m, 14H), 3.70 (s, 3H), 5.28 (d, 1H, J =1.4 Hz), 5.44 (dd, 1H, J =2.8, and 6.5 Hz). CMR (CDC13) 69.8 MHz, 6 202.94, 189.26, 175.42, 143.87, 117.50, 100.41, 55.53, 50.57, 46.84, 46.13, 38.25, 34.67, 34.14, 32.99, 30.87, 28.70, 27.11, 24.64, 23.94, 18.41, 14.20. Analysis: Calculated for C21H2803: C, 76.79; H, 8.59 Found: 76.89 8.44 Preparation of enones 83 and 82 Diketone 83 (1.10 g, 3.35 mmol) was dissolved in 20 mL of dry methylene chloride, and cooled to 0°C. To this solution was added 10 mL of 1 E diisobutylaluminum hydride (DIBAL) in hexane (10 mmol). After stirring for 1.25 hr excess hydride was destroyed by the addition of 10 mL of saturated sodium potassium tartrate. The reaction mixture was then extracted with ether. A solid which appeared in the aqueous phase was collected, boiled with methanol, and filtered. The filtrate was combined with the ether extracts, and the solvent was removed to yield 1.10 g (99%) of a white solid, which showed a single spot on TLC (silica-ether). This product was dissolved in 100 mL of 4:1 (v/v) THF/water along with four drops of concentrated HCl. The resulting solution was stirred at room temperature overnight, diluted with water, and then extracted with ether. The combined -82.. ether extracts were washed with saturated sodium bicar- bonate, water, brine, and dried over sodium sulfate. Evaporation of solvent yielded a light yellow solid which by TLC (silica-ether) proved to be a mixture of three products. Flash chromatography (silica-ether) gave 83 (Rf==.43, 578 mg, 58%), 83 (R.f =.35, 278 mg, 28%), and a small amount of material which was not identified (not UV—active on silica TLC). The total yield for 83 and 83 was 86%. Characteristic properties of 83 are: M.P. l81-183°C Mass spectra (70 eV) m/e (rel. intens.) 300(1), 285(7), 267(4), 192(67), 174(49), 159(96), 109(100), 105(58), 91(49), 79(39), 41(46). Infrared (CH2C12) 3625(sh), 3490(br), 2975, 2910, 1700, and 1325 cm’1 PMR (CDC13) 250 MHz, 6 0.77 (s, 3H), 1.06 (s, 3H), 1.22 (s, 3H), 1.20-2.80 (m, 15H), 3.78 (dd, 1H, J =l.7 and 7.5 Hz), 5.31 (q, 1H, J'=3.1 Hz), 5.96 (d, 1H, J'=10.l Hz), 6.97 (d, 1H, J =10.1 Hz). CMR (CDC13) 69.8 MHz, 6 200.42, 162.72, 146.08, 128.15, 116.57, 79.44, 49.74, 46.64, 41.81, 39.93, 39.75, 37.76, 35.22, 33.22, 29.22, 28.34, 27.99, 25.24, 19.64, 18.99. Analysis: Calculated for C H O - C, 79.96; H, 9.39 20 28 2' Found: 79.95 9.31 -33- Characteristic prOperties of 83 are: M.P. 168-169°C Mass spectrum (70 eV) m/e (rel. intens.) 301(3), 285(9), 192(56), 177(18), 159(25), 133(32), 119(37), 109(100), 91(58), 79(52), 41(48). Infrared (CH2C12) 3610(5), 3465(br), 2965, 2890, 1680, 1665, and 1480 cm'l. PMR (CDC13) 250 MHz, 6 0.88 (s, 3H), 1.04 (s, 6H), 1.40- 2.80 (m, 15H), 4.10 (dd, 1H, J'=6.9 and 8.7 Hz), 5.28 (q, 1H, J =3.1 Hz), 5.96 (d, 1H, J =10.1 Hz), 6.96 (d, 1H, J'=10.1 Hz). CMR (CDC13) 69.8 MHz, 6 201.12, 162.38, 145.20, 127.91, 116.40, 80.20, 48.54, 43.92, 41.49, 39.40, 39.31, 37.43, 33.30, 30.16, 29.81, 28.87, 27.05, 20.67, 19.40, 18.37. Analysis: Calculated for C 79.96; H, 9.39 20H2802‘ C' Found: 79.92 9.29 Preparation of MEM-ether derivative 89 ~ This procedure will be illustrated for the single C-l7 epimer 83. Normally mixtures of 83 and 83 are protected. Alcohol 83 (700 mg, 2.33 mmol) was dissolved in 3 mL of methylene chloride in a dry 10 mL pear shaped flask with a side arm. To this solution was added 8- methoxyethoxymethyl chloride (MEM-Cl) (400 microliters, 435 mg, 3.5 mmol), followed by diisopropylethylamine (610 microliters, 452 mg, 3.5 mmol). After stirring at -84- room temperature for 3 hr the starting material was no longer evident by TLC analysis (silica—ether) and the reaction was halted. After addition of 5 mL of water, the organic phase was separated, and the aqueous phase was extracted twice with methylene chloride. The combined organic phases were washed with brine, and dried over sodium sulfate. Removal of the solvent by rotary evapora- tion yielded 783 mg of 83 (87%) as a yellow oil. Characteristic properties of 83 are: Mass spectrum (70 eV) m/e (rel. intens.) 389(2), 283(28), 175(25), 137(9), 109(17), and 89(100). Infrared (cpc13) 2900, 2240, 1660, and 1050 cm'l. PMR (CDC13) 250 MHZ, 6 0.80 (8, 3H), 1.05 (5, 3H), 1.16 (s, 3H), 0.80-2.80 (m, 15H), 3.39 (s, 3H), 3.56 (m, 2H), 3.67 (m, 2H), 4.61 (d, 1H, J =6.7 Hz), 4.72 (d, 1H, J'=6.7 Hz), 5.30 (dd, 1H, J =3.1 and 6.4 Hz), 5.95 (d, 1H, J =10.l Hz), 6.95 (d, 1H, J =10.1 Hz). Preparation of 91 and 92 To a 100 mL pear shaped flask containing 1.22 g (10 mmol) of potassium t-butoxide (Aldrich) was added a mixture of 83 and 93 (905 mg, 2.33 mmol) as a solution in 50 mL of THF. The yellow-orange solution turned brown on contact with the base. After stirring this mixture for 10 min, methyl iodide (.62 mL, 1.42 g, 10 mmol) was added in one portion. The solution immediately turned a milky white, and the flask became warm to the touch. Stirring -35- was continued at room temperature for two days. An additional portion of methyl iodide was added, and following an additional two days, the reaction mixture was diluted with 100 mL of water and 100 mL of ether. The aqueous phase was separated, extracted twice with ether, and the combined ether extracts were washed with water (twice), brine, and dried over sodium sulfate. Removal of the solvent yielded 946 mg of product as a dark yellow oil. Preparative TLC (silica-50% ether/hexanes) yielded two products, 93 (Rf==.62, 362 mg), and 93 (Rf==.53, 180 mg). The total yield for 93 and 93 was 56%. Characteristic properties of 93 are: Mass spectrum (70 eV) m/e (rel. intens.) 416(4), 372(4), 340(4), 320(4), 311(1), 174(7), 159(7), 145(4), 137(15), 89(77), 59(100). ' Infrared (neat) 2932, 2880, 1660, 1460, 1370, 1111, and 1035 cm‘l. PMR (CDC13) 250 MHz, 6 0.76 (s, 3H), 1.06 (s, 3H), 1.08 (s, 3H), 1.15 (s, 3H), 1.17 (s, 3H), 0.80-2.80 (m, 13H), 3.39 (s, 3H), 3.56 (m, 2H), 3.67 (m, 2H), 4.62 (d, 1H, J'=6.7 Hz), 4.73 (d, 1H, J =6.7 Hz), 5.38 (q, 1H, J'=3.1 Hz), 5.93 (d, 1H, J =10.1 Hz), 6.89 (d, 1H, J =10.1 Hz). Characteristic prOperties of 93 are: Mass spectrum (70 eV) m/e (rel. intens.) 372(1), 340(1), 311(1), 280(1), 174(7), 159(7), 137(15), 89(77), and 59(100). -86- Infrared (neat) 2946, 2884, 1664, 1457, 1350, 1105, and 1057 cm'l. PMR (CDC13) 250 MHz, 6 0.84 (s, 3H), 1.04 (s, 3H), 1.05 (s, 3H), 1.10 (s, 3H), 1.15 (s, 3H), 0.75-2.80 (m, 13H), 3.40 (2, 3H), 3.55 (m, 2H), 3.68 (m, 2H), 3.99 (dd, 1H, J =6.4 and 9.2 Hz), 4.70 (s, 2H), 5.35 (q, 1H, J =3.1 Hz), 5.94 (d, 1H, J =10.1 Hz» 6.88 (d, 1H, J =10.l Hz). Preparation of alcohol 93 This procedure will be illustrated for the single C-17 epimer 93. Normally mixtures of 93 and 93 are deprotected. A solution of 93 (271 mg, .65 mmol) in 3 mL methylene chloride was cooled to 0°C. Pyridine (40 microliters, 39.5 mg, .5 mmol) was added, followed by titanium (IV) chloride (215 microliters, 371 mg, 1.95 mmol). The solution turned dark brown. After 30 min the starting material was no longer evident by TLC analysis (silica- ether). The reaction was quenched by the addition of 2 mL concentrated ammonium hydroxide solution and then diluted with 5 mL of water. After separation of the organic phase, the aqueous phase was extracted twice with methylene chloride. The combined organic phases were washed with water until the aqueous phase was neutral, then with brine, and finally dried over sodium sulfate. Removal of the solvent gave a brown oil. Passage through a short silica column yielded 171 mg (80%) of 93 as a yellowish solid, which was recrystallized from methylene chloride/ hexanes. -87- Characteristic properties of 93 are: M.P. 221-224°C Mass spectra (70 eV) m/e (rel. intens.) 328(1), 313(1), 192(29), 159(34), 137(100), 105(24), 93(18), 91(17), 43(19), 41(21). Infrared (cuc13) 3600, 2950, and 1660 cm‘l. PMR (CDC13) 250 MHz, 6 0.73 (s, 3H), 1.06 (s, 3H), 1.09 (s, 3H), 1.16 (s, 3H), 1.21 (d, 3H, J'=0.7 Hz), 0.80-2.80 (m, 13H), 3.77 (dd, 1H, J =1.5 and 7.6 Hz), 5.39 (q, 1H, J =3.l Hz), 5.94 (d, 1H, J =10.1 Hz), 6.90 (d, 1H, J =10.1 Hz). Preparation of diketone 96 A mixture of epimeric alcohols 93 and 93 (40 mg, .12 mmol) was added to a stirred solution of pyridinium dichromate (PDC) in 2 mL of dimethylformamide (DMF). The initial bright orange solution turned brown as the reaction proceeded. After stirring overnight (16 hr) at room temperature, the solution was diluted with water and extracted (three times) with ether. The combined ether extracts were washed with water, brine, and dried over sodium sulfate. Evaporation of the solvent yielded 32 mg (81%) of 96 as a brown oil. Passage through a short silica column gave pure 96 as a colorless oil. Characteristic prOperties of 96 are: Mass spectrum (70 eV) m/e (rel. intens.) 327(1), 311(1), 190(1), 175(14), 137(97), 133(24), 44(43), 40(100). -88.. Infrared (neat) 2960, 2880, 1735, 1665, 1455, and 1370 cm-1. PMR (CDC13) 250 MHz, 6 0.98 (5, 3H), 1.03 (s, 3H), 1.08 (s, 3H), 1.10 (s, 3H), 1.17 (s, 3H), 0.70-2.40 (m, 10H), 2.42-2.56 (ddd, 1H, J'=1.8, 9.5 and 19.2 Hz), 2.66-2.80 (m, 1H), 5.53 (q, 1H, J'=3.1 Hz), 5.96 (d, 1H, J'=10.1 Hz), 6.88 (d, 1H, J'=10.1 Hz). Preparation of saturated diketone 96 A 25 mL pear shaped flask containing 10 mg (.044 mmol) of platinum oxide (Adam's catalyst) suspended in 2 mL of benzene was connected to an atmospheric pressure hydrogena- tion apparatus. After three cycles of evacuating the system under aspirator vacuum then filling with hydrogen gas, a solution of 96 (105 mg, .32 mmol) dissolved in absolute ethanol was added with a syringe, and the hydrogen gas uptake was monitored. After 6 hr hydrogen uptake ceased and the reaction was stopped. The catalyst was removed by filtration, and the remaining solution was concentrated to yield 102 mg (97%) of 96 as a colorless oil. Characteristic properties of 96 are: Mass spectra (70 eV) m/e (rel. intens.) Infrared (neat) 2960, 1745, 1700, 1465, 1375, and 730 cm'l. PMR (CDC13) 250 MHz 6 0.92 (s, 3H), 1.02 (d, 3H, J =0.85 Hz), 1.06 (s, 3H), 1.08 (s, 3H), 1.10 (s, 3H), 0.80-2.40 (m, 13H), 2.43-2.58 (ddd, 1H, J =1.8, 9.5 and 19.2 Hz), -89- 2.59-2.75 (ddd, 1H, J'=5.2, 14.8 and 9.6 Hz), 2.93-3.08 (m, 1H, 5.51 (dd, 1H, J'=3.4 and 6.7 Hz). Preparation of 73 Ketone 96 (35 mg, .11 mmol) was dissolved in 3 mL of 95% aqueous ethanol, and cooled to 0°C. To this pale yellow solution was added 1 mL of .1 g sodium borohydride in 3 3 aqueous sodium hydroxide. This mixture was stirred at 0°C, and the reaction progress was monitored by TLC (silica-ether). After 3 hr the starting material was no longer present, and the solution was diluted with 5 mL of water and extracted three times with ether. The combined ether extracts were washed with water, brine, and dried over sodium sulfate. Removal of the solvent by rotary evaporation produced 32 mg (91%) of 73 as an oil which crystallized on standing. An analytical sample was prepared by recrystallization (methylene chloride/petroleum ether) to yield 73 as a colorless crystalline solid. Characteristic properties of 73 are: M.P. 192.5-194.5°C Mass spectra (70 eV) m/e (rel. intens.) 330(4), 297(29), 230(21), 133(22), 119(37), 107(24), 105(39), 91(41), 79(26), 57(36), 55(62), 43(92), 41(100). Infrared (CDC13)3670, 3600, 2920, 2220, and 1725 cm‘l. PMR (CDC13) 250 MHz, 6 0.82 (s, 3H), 0.89 (s, 3H), 0.97 (d, 3H, J'=0.92 Hz), 1.04 (s, 3H), 1.05 (s, 3H), -90- 0.80-2.40 (m, 15H), 2.41-2.55 (ddd, 1H, J’=148, 9.5 and 19.2 Hz), 2.74-2.87 (m, 1H), 3.31 (dd, 1H, J'=6.4 and 9.5 Hz), 5.43 (q, 1H, 3.2 Hz). CMR (CDC13) 69.8 MHz, 6 219.82, 143.28, 119.70, 79.33, 50.65, 50.24, 46.63, 39.86, 35.63, 34.65, 34.60, 34.37, 30.95, 29.51, 27.21, 27.08, 24.99, 24.70, 23.75, 23.52, 17.59, 14.81. Preparation of 85 A mixture of 83 (100 mg, 0.33 mmol), potassium t- butoxide (244 mg, 2.0 mmol), and 5 mL dry THF was stirred at room temperature for 10 min. To this Opaque orange mixture was added methyl iodide (200 microliters, 3.2 mmol) in one portion. The reaction mixture immediately turned milky white. After stirring for 93 hr the reaction was halted by diluting with 50 mL of water, and extracting the resulting clear solution three times with ether. The combined ether phases were washed with water, brine, and dried over sodium sulfate. Removal of the solvent yielded a yellow oil, which when subjected to preparative TLC (silica-ether) gave 53 mg (46%) of 86 as a yellow oil, plus unidentified polymeric material. Characteristics of 86 are: Mass spectra (70 eV) m/e (rel. intens.) 342(1), 327(29), 312(15), 284(15), 269(17), 257(18), 256(38), 160(29), 129(21), 128(42), 105(82), 73(100), 55(63), 43(57), 41(72). -91- Infrared (cuc13) 2950, 2250, 1680, 1125, and 910 cm_l. PMR (c0c13) 250 MHz 6 0.75 (s, 3H), 1.05 (s, 3H), 1.09 (s, 3H), 1.15 (br s, 6H), 0.60-2.80 (m, 12H), 3.18 (dd, 1H, J =1.8 and 7.3 Hz), 3.25 (s, 3H), 5.37 (q, 1H, J'=3.3 Hz), 5.94 (d, 1H, J'=10.4 Hz), 6.89 (d, 1H, J =10.4 Hz). Preparation of 83 A 25 mL pear shaped flask equipped with a side arm containing 5.0 mg (0.02 mmol) of platinum oxide (Adam's catalyst) and 2 mL of benzene was connected to an atmospheric pressure hydrogenation apparatus. After three cycles of evacuation under aspirator vacuum, followed by filling with hydrogen gas, a solution of 35 mg (0.10 mmol) of 86 in 3 mL absolute ethanol was added to the flask with a syringe. After approximately 6 hr hydrogen uptake ceased, and the reaction was stopped. The catalyst was removed by filtration, and the remaining solution was concentrated to yield 35 mg (100%) of 83 as a brown oil. Characteristics of 83 are: Mass spectra (70 eV) m/e (rel. intens.) Infrared (neat) 2950, 1725, 1480, 1390, and 1140 cm-1 PMR (CDC13) 250 MHz, 6 0.83 (s, 3H), 0.89 (s, 3H), 1.06 (s, 3H),1.08 (s, 3H), 1.14 (s, 3H), 0.80-2.40 (m, 14H), 2.69 (ddd, 1H, J =5.2, 15.0 and 14.7 Hz), -92- 2.87-3.04 (m, 1H), 3.21 (dd, 1H, J'=1.8 and 7.3 Hz), 3.26 (s, 3H), 5.34 (q, 1H, J =3.4 Hz). APPENDIX -93_ ‘ L’: ' xl r \ (it Figure 1. ORTEP drawing of 63. -94- Figure 2. ORTEP drawing of 73. -95- ccu am. an: o: m.”- a: 9- mm- . R- Na..- 36 2 do 2.!”- om: ..m “33 .ww mo Ennuommm mmmz .m musmflm on. v— ~— 8— 8 ~; 5 .m k 8 O? u .2 mm 8 z. . r u e8. -:.3. ~33 com-w «many-cm at: + 8532 8x33— .Smhuumm mm.)- -96- saw -u 66- em- jfi a: . m2 3- a: 9 mn— - ma EN 823 ; 2.§U:#a Av ma. e-vm “a-ao o: ,- m— Na .mw mo Esnuommm mmmz sm— t- no” 8- mm .8 wusmflh 8 am 9. m .2 mm ... mm mm 3 R .v .m r 0.9.... r mw .... 0.8- mm. ... mm. ~33 Sea .33 «IN + gamma: 82?: Eda—omum mmd: -97- . mm mo guuommm mmmz .m mun-mam E- a: eu- . 8 8 . 2.. ma- . m u u . . u r v P ..l r n - . b FLI- . r . _ d;= _ r? m . ... - . _ -e- an 8 - an 2- 2. 0.8 :o .20. 1 guru . P- o 8- , .. . n3 .. no. .10 --m-a-~uq§- 8.9 .393 .e-a... 5:- 2 .3 :6 m:- + 8622 32.-8 mm; "m: mmrm 0 mm. mmnv 2:3 ...-mama mm;- -93- .finv ...-u 0.16 0U. ma #5 :3 m; «8.. .53 _ .... -_ «w. e- m...— .2. .MW mo Ednuommm mmmz .o musmfim ..,.— . on ma.- - _ ”mt... 8- mm . . . an a . a... -m . no.9... .03 . me T .r g [008“ a. u a: 83 + 3&3: .9:me grown—m mmol -99.. omv 90v can own .mw mo Esnuommm mmmz .5 ouamflm ova wan can can own ova uxr 33 L 1 0.20 QQNN em. a. Ace «.4co Au mg. was.. .c—c: F. b -Fbpbt.’ _P:5L.h’b _b’h. _.”b hl’b" -.’..‘ -E. hE.-ch .' —E. .’-\."-b.. D ..‘P.’ -br..r.’ -’... ...... a mvm a .ma A~m~ ~n~ mum 1 8“ r v.0 -...3 -98. we -..8 r .988. m. 1 at: «.9328 2.: -98. no I an. ~xmxax~ucgoxuoo—_ .mdmtcm omuN + oo.~_um. voxmo\mo tampoumm mmct -lOO- .Hm mo Eduuommm mmmz Gav 3' can own rrr—.»r......—::.»:L_rp:..>.._.>r.... L .4 J vav 1 can 8N 8. 8— 0: ON— ..2 8 3 PN Z 3 m. _ m2 m: m: r 8; m. 8 m: a sh mm 1 m3 5 $02. - 69 OnIU am no do 33 mm. was: 3.23 9:28.23. 36%; n8: «wan-8m 38 + 8668. 53630 tam—own...“ mm»... 3. .m musmflm th -lOl- umuu “ _.... .. .. 11mww;uqmflmuus ‘ “l.\ c—J I. D. . .5 H «an .:090. guru ma 4:0 ”—410 vva n mv .c.¢o .mw mo Esnuommm mmmz .m mnsmfim own can eon awn own oqu w z . ....u 33%.. . w..u__u.o...... mmu a.“ . -aam rode a: 8. 3. 8 S 2. w .... n u p n q - . ..7 av K. 3. __ mm mm.k _ .. a” on . .m , . _ .. . 3 roam . f a. -92: R. . 3.. .xisflugi~m$.w:;a .m: .mazmfi_=¥mxa zfiswamw: -102- omv can own ova «8” NW van . awn mmmmw Q emu OQN can Owa ova nth- _mmm: ~. . N N m: . ...: . ~n «nu — a J u 60 o ...... 8mm~ no 4C0 “—ACU Av um. “—On .Gbca uxmzaxnucaax «don .udmz¢m vouu + oou~0.n. ~oxm~xo~ tampouam mmcx .mw mo Esuuommm mmmz .oa mnsmfim can 11. can on“ so“ mx: .l.‘ .1: “NW NW" IV . N I 1. ..o.on 1 ..o.oa_ ou— 11— so ow av th w . _ n g no as . .. .m on . n . .. 8 r ..o.on r v - 98. on. a mu. -103- .mN mo Efiuuowmm mmmz .HH musmflm mac 93. gm com 03 can 3m 8m emu av“ wxr . Pbrp..4F4rr—.1p:.:.—L:_:_>_.br_:>...p:_:t... _L[.b::—...r«..t. —r-L4L:_Es.::_.........47... fl an .8 ~21~81 .6” ;~ 1 - v15 :92 .1 r 37. .L [can an emu 03 8. 9.. aa— 8‘ 3 3 2. mx... ..-...EL..F.. .-..pr..._r: . . 1...; ..51. . ._ __n? ~fl GE «2 mm. #2 ma mm: mm— _ o t am i 1 1. . v: r . . m: on r 1 03 to a. nu— ..IU . ..Nvu vm— loéw 3. n 8: ~33: van» a; m. do SS 8:.” + 839N— maxeko 0 no; vane “Eco 5&5me mm? -105- .mw mo Esuuommm mmwz .ma musmflm 8.. .8 m3. . r....Amfi..b>...s_.....,p...&»p..u.rLL_ . «w 5 v r n . a . - 95 -23 2.1.1., - no.8. a...“ on. to or. ox: . mum .na .va 8n . m a ... E n: . . r . E 3 . 1 a a B 3 . 1 -98 a _ om r . fi 9 . .6 s. a...» n . -98. . 8. n 8: .339. k: .323 2 .8 58 «as . 8.8": 83?: 8: k: «Eco 558mm mmat -106- .mw mo Esuuommm mmmz .va wusmflm an .. 9...... 3m own 6.6 can 30 0mm 8N ova mu... 1E-r_.r.LLL:_.L.L.::_rtp.r[._..:PP>.._»LLLm F»_>»Cbt:_::.pr-t.....L»...7.4:»..7... ‘_ 5N mum mmN «VN fi 0—0 1. u 0mm 10.3 1 an...” 1 r98. .mu com so. 8‘ 03 ON. -PP*b bbtxflbhth —EP_b’ph_PP :4 u . — d 7 nm— Hm: m ‘ N - 10.3 .910 . fi .wumm [0.8— mm. 0 8. mg: 98 E.J.-3m no do 23 ovum + 86:: gxno O 8a 33 ~53 Empowmm mm»! -lO7- .mw mo Esuuommm mmmz .mH musmflm 23 an gs o2. as on u? can ..~ -~ 5 f . , , i ;, on RN ; ~m 4 fi n: # 8. A a~ r 3 - v: .m uoén g: L O _-__ mfi‘ P—rm— L 8‘ lo 8— “ on. .. 6. m .xmzax~u¢4ax ..Nfi .udmzcm u 2 do :3 :a + 863 8392 x. .8. :fi .33 5.5%. m3. -lO8- 3m so pL—bblrhp-h-rhbbrbbnt— .15. DC {—- _3 2 Oman" fl mmnm q .- hm... J ....m “ m. .30 33 um. N76 «33 N»: b: .mw mo Esuuowmm mmmz on. 3— am .oa musmflm mu... 3 or; 3. «m as an. n has u; «Km «Hr-3m $3 + 0323 maxmoxmo 550$ mmcz 10.0w [c.3— -109- sec L vo_m ova l 2.2m .mvwm hum «o_x mm aux: wmam 00v .3 :a u; oo— mud .mw mo Esuuommm mmmz ow— mw— m a mvu ova ov— on. u _ can can . cud oo— m u mm mod mo RWNN m._m a. a who uxmzaxmucdox mowu— «mamxam omuu + oouwvun. voxvmxmo tsupuumm mmcz .ha musmflm war 1.0.om ruo.oa_ mxz .5” Wade -110- am? so? 9mm awn ova mad}; ; cum sow om. ow— ov— wmw mow mm— ww— mm~ meg mm A vu— k a mvo¢m_ 1 no aco u—Jdo 2m: . mm: 9.3 «3.3 .mw mo Esupommm mmmz .mH musmwm own com 0mm own ova w.= :m r f T98. wanna '— [Go-wo— 8. 8. 3 8 3 w: 8 cm , a: . 8. k 8 B =. , .9 . r 19.8 8 r an [@000- omu I omzzzm no. ob mm. uxmzaxmocqoz ovum .uamzcm -u~ + oouumu—u voxmoxoo tamboumm mmcz -lll- .mw mo Esupommm mmmz can L m__mw «:3. o. .10 33 mm. 9.8 .53 fin «a .mH mhdmflh 80 08 SN ovw mu: . vw mmm mmm 5N own I 9.8— wé an I 0.3— QZN + 8&qu 3% 558mm mm? -112- .mw mo Esnuommm mmmz .om musmflm a~v age can own men can can gnu own ova uxz LPDL-FED-DPII—DD’D-FIDb—VpD’.bIDD—D’PD-bhhl-’DDP- I own L n o 1 g t .4 v 1 I 0.9.. .A r 1 fi thm I. goOd L l 008‘ omw owm 8a 02 o: and 8— 8 3 0.. mi $36. :9... . JFK mul: .n ....L 0. .3... u ... r».. a oh. Watt . p . mum o_~ .om _ mm. _.fi mw“ .___ ___Mwu ma 1 mm . vud mv. on. a . . as mm mm 3 In“ . «mu 1 1 mm— . 1 I 0.3 1 fl r Numm ~c. ..o.oo. I. .. Mk. mg ~08. Gan-am no .89 :3 vufi + oofiuum 5x880 «so 803 «58 8.5593 mg -ll3- .mw mo Esuuommm mmwz .HN musmflm 3.. one 8m 3m own can 3m 3a 3a 03 m3. _ ...:L r..._t. ..t _ . .....ILt...Lt.u.....pu....:_t:.::..ILPILIU. _ RN _ _ e um mmN =.N an loam x008 L [008— 8a cum 8. 3. o: 8. 2.. 8 3 2. Mr: :_..:.»..1. 3....Jp4..u_.+..»:.:: «8 m2 ma. . a. om. ~m. ~v. so mm mm mm P: a“ 8. no 02 v. soda 3 9. :98. . 272%. 26. .559fl!£§~ma.qxxm m. Jag ..4cu m.“~ + ooumvuo vox~ox~o 8. $8 “33 538% $3. -ll4- momm 1% 3m 98 9mm 8N ...p. d.Lbs-r_... L—Lm—VTJL. mmw mm 3N nu emu mmm m. 4G0 ..4cu 0 3.3.38 .mn mo Esuuommm mmmz .NN wusmflm om? 8. we 2: mm 3 R a . B. a . {can 9 . ... no.8. a. - 3. a; 08m flan-am avuu + 8830. «QR—No 55.3mm mm! -115- ON? 96? 9mm 0mm 0v” phi»_bPPPhPtP—P->h-thyb—>bthPh>—nD-P-DDP4-.>—“ubbb-r «#0 L .&m L A QNN @DN emu Qw— 9v— 4 ~2 . “Nu _a~ uh— m1. 1 QB ‘ w—a 0N“ m. ACO u—JCU mat ~mmm “Chca .mw mo Esuuommm mmmz can end m _ m _ so. «m no emu .mmN mm on m. : ma. _xm:ox~ucqox aomm .uamtcm «was + oauwmum nmxadxua zacpummm mmcx ova vm .mm mnsmam wax .uo.om ..@.oa. mu: ..o.oo~ -ll6- 8 8 8 TRANSMITTANCE(%) N O 4000 3500 3000 2500 2000 15% momma ‘CA ‘) TRANSMITTANCE(%) 1400 1200 1000 800 "EQUGNC v C M Figure 24. Infrared spectrum of 35. -ll7- 100 80 w w §qu<23mz<5 a. m. m limialin- 0 2 2500 2000 I 500 3000 3500 'V’Illfl'o o 8 w w Aiuuzitzmzip I 400 l 200 1000 800 l I 800 2000 x M rREOUlN‘ I Figure 25. Infrared spectrum of 35. ~ ~118- 100 8 8 8 TRANSMITIANCE(%) M O 4000 3500 3000 2500 2000 1 500 maumcv mm -. 100 8 8 TRANSMIHANCE(%) 5 O 20 2000 1800 1600 1400 1200 1000 800 FREQUENCY CM 1 Figure 26. Infrared spectrum of 36. -ll9- 100 on O 8 40 TRANSMIITANCE(%) 20 4000 3500 3000 2500 2000 1 500 "(00va .CM') 8 8 8 IRANSMITTANCE(%) N O 2000 1 800 1600 1 400 I 200 1 000 800 "(Outlaw ‘.M . Figure 27. Infrared spectrum of 37. ~120- 100 (D O 0 O 40 TRANSMITTANCE(%) 20 £000 3500 3000 2500 2000 1 500 ‘NOUENCV 'CM 1 8 8 8 TRANSMITTANCE(%) M O 2000 1800 1600 1400 1200 1000 800 FthutN; v M Figure 28. Infrared spectrum of 29. 100 0 on O O TRANSMITTANCE(%) A O 6 8 8 TRANSMITIANCE(%) N O 0 2000 1800 Figure 29. -121- 2500 FREOUINL v CM . 1600 1400 1200 F'EGU‘Nx v \._'U\ Infrared spectrum of 5%. 1000 800 -122- 100 8 TRANSMITTANCE(%) 2500 FIEOUENLV ‘(M '1 8 8 8 IRANSMITTANCE(%) M O 0 2000 1800 1600 1400 1200 1000 800 FIEOUENCV CM . Figure 30. Infrared spectrum of 61. 100 0 on O O A O TRANSMITTANCE(%) ~123- 8 8 IRANSMIITANCE(%) & O 20 0 1 2000 1800 Figure 31. 2500 FR‘QUng v (M n 1600 1400 1200 "£00! NC 1 t'. M Infrared spectrum of 53. 800 -124- 8 8 8 TRANSMITTANCE(%) M O 4000 3500 3000 2500 2000 1 500 FIEOutNt v CM . 8 8 TRANSMITTANCE(%) & O 20 o 1 I r 2000 1 800 1600 1400 1200 1000 800 FREQmM v 5M Figure 32. Infrared spectrum of 70. -125- 100 8 8 8 TRANSMITTANCE(%) M O 4ooo ° 3500 3000 2500 2000 1500 F'EQUENQ Y (M 1 0 O IRANSMIITANCE(%) h C 20 2000 1600 1 400 FIEQU‘ Ht 1 L M Figure 33. Infrared spectrum of 22. -126- 0 an O O 1RANSMIITANCE(%) b O 400.0— 3500 3000 2500 2000 1500 "(000va CM . 8 8 8 TRANSMITTANCE(%) N O 2000 1 800 1 600 1 400 1 200 1 000 800 "(Quint I J M Figure 34. Infrared spectrum of 80. ~ -127- 100 8 8 “o” TRANSMITIANCE(%) K) C 4000 3500 3000 2500 2000 1500 ‘REQUtNC 7 CM 1 .... on O TRANSMITTANCE(%) 20 o ‘ i . 2000 1800 1600 1400 1200 1000 800 FREQUENQ V M Figure 35. Infrared spectrum of 33. TRANSMIITANCE(%) 1RANSMIITANCE(%) an O 8 & O 20 Figure 36. -128- 3000 2500 Infrared FIEOUfNLv .CM 1 1400 1200 "£QUEV.‘\ ' 1 1M spectrum of 81. 2000 1500 800 -129- 100 0 an O O IRANSMITTANCEHG) § 0 20 4000 3500 3000 2500 2000 1 500 "600000 .‘CM 1 TRANSMIIIANCE(%) 2000 1800 1600 1400 1200 1000 800 FREQUENC v CM . Figure 37. Infrared spectrum of 82. -130- 1RANSM|TTANCE(%) 2500 "EOUQNC 1! CM -) 8 IRANSMITIANCE(%) ‘ O M O 2000 1800 1600 1400 1200 1000 800 FIEOUENU CM . Figure 38. Infrared Spectrum of 89. -l3l- 100 8 8 8 1RANSMI1’1ANCE(%) N C FOE Qqu:_ r M 0 ' ' .' : : 4000 3500 3000 2500 2000 1 500 00 I..- 80 TRANSMITTANCE(%) 0 1 2000 1 800 1 600 1400 1 200 1000 800 ‘QCOUIN' ' -M Figure 39. Infrared spectrum of 91. 100 b 0 an O O O TRANSMITIANCE(%) N O )H 1RANSMITTANCE(% A c: N O 2000 1800 Figure 40. 1600 Infrared -l32- "(WM Y ICM '1 1400 1200 FIEOUNK V CM " spectrum of 92. 1000 800 -133- 100 b 0 (D O O O TRANSMITIANCE(%) N C 4000 3500 3000 2500 2000 1500 "(QUINCY ICM ') A 0 on O O O TRANSM111ANCE(%) N O 2000 1800 1600 1400 1200 1000 800 'IEQUEFK Y CM '. Figure 41. Infrared spectrum of 93. -l34- 100 A O on O O O IRANbMIIIANCE(°/o) M O 4000 3500 3000 2500 2000 1500 'lEOUtNCY «CM '1 ... It 0 on O O O TRANSMITTANCE(%) N D 2000 1800 1600 1400 1200 1000 800 ‘IEOUENC r :0» 0 Figure 42. Infrared spectrum of 95. -l35- 100 A O on O O O IRANSMIfiANcuo/o) M O 4000 3500 3000 2500 2000 1500 "(QUINCY ICM ') A 0 on O O O IRANSMIUANcem) N O 2000 1800 1600 1400 1200 1000 800 FIEOUIM v iCM '- Figure 43. Infrared spectrum of 9g. -136- 100 A 0 on O C O TRANSMITTANCE(%) M O 4000 3500 3000 2500 2000 1500 IKOUQ NCY ICM '1 A 0 on O O O TRANSMITTANCE(%) N O 0 . . 2000 1800 1600 1400 1200 1000 800 "(001!“ r 'CM -. Figure 44. Infrared spectrum of 23. -l37- A 0 CO 0 O O 1RANSMITTANCE(%) N O “CEOUENCV (CA 'I TRANSMITTANCE(%) 1 000 800 1400 1200 fliOU! NCV .CM ‘1 1800 1600 Figure 45. Infrared spectrum of 8g. 100 A 0 on O O O TRANSMITTANCE(%) N O A O on O O O TRANSMITTANCE(%) N O 0 2000 1800 Figure 46. -l38- 2500 MOUQNCY «CM ‘1 1600 1400 1200 FREQUENU ~(M‘ Infrared spectrum of 88. 1000 800 ~139- .ww mo asuuommm mam .54 musmflm d4 A444444dd<1ddddddejd444dd‘JdAddefidiii—4444441dd#dd‘dqdi4udlddiidd4d11444414 1 j .4- £37 A O of”. .mw mo Esuvummm m2m .mv musmflm a p N . a v m c n -l40- fiJfiuddqquqqfiqu44_qqu4—44—«qqd—d—qq—dqqqdq—uqqqua«qa—«dqdqdudu—dddd1quW—quqfid 17 17 13 . -l4l- _ .mw mo Esnuommm mzm .m¢ musmflm : p N m e m e p a jjfiijqjifiquifijj—fijqaqaj_44q.44444d44Ju‘d«¢44444-‘144444«4444 1 1 1' J 3713i} 4 :0 .10 0:8 -l42- .Mw mo Esuuommm MZm .om musmflm o u N n v . c 1111414 11114141 11. n m c h H q i 14 1141114 d-d -41. 41414.4 1|: d 41.141111114111113 — 1141411141411 #1141141]ij a 311 :1- ,1- C) -l43- .mM «0 Ednuommm mzm .Hm mnsmflm c F N . m e m c h qJJ—AJJJJJA—dJ1—d1rdl-Atdde—J-JJAJJ-j4dddJ—JJWJdJIdA—dJdeAJdJ—dddlqddddededludlleJA—ddd41- 3]: m 41 -_ -l44- .mw mo Esupommm m2m .Nm musmflm o F N n e m a p jA—dAJAu—Jdd—4‘4JJA‘uqdddeJd«Add—qdd4qu44—44ddd4de—du«ddfidedd‘dedd‘ddduqJAJ JD .43; 3 < I 1‘ .m3 m0 Enuuommm mzm .mm musmflm a F N n v m a h -l45- 41A444444ddfiflJHA-JJ4111HJ41—41‘114444dd111—4deJJJdd1‘41441‘1444441—41144d44441—jJJ14444J444444 4111313323531 1% I -l46- 1’ o .3414 -— 4141.14-41J1fl141fl “41424141141414. 4 _ a14141414141|4114 .—14111414141414|«141—1114111J1Jul41_11414..— ‘ u 34 d *4 < 4 .—v a a 1 4 4 q « qJ 4‘ q 4 q ,4 711/ 4 — 3 its N n :2:ng v .m@ we Esnuommm mZm m .vm mnsmflm h 7 A 11 41 -147- .mw mo asunommm mam .mm magmas o F N n e m e h «44:411-41—4144444444fi11111111fi4 <<<<<< 411414111111 «1411—1411111141141141414141— 1-1ql1d-14114.14:d 1114 — 414 141.4.11114 q q. #11 1:1 114%, 2. seas-1141 j 4 '6 O -l48- .mN mo Esuuommw mZm .mm musmflm e _. N n e m o h .l441—qud_dd4d.—4444dajJ.—4444|«441—A—4JJNJIAJJJJ—jd4114fidddddleNAJJ—JJ—AAJNJAJAJ-JJ __ 33% 3 -149- a w d41—d4ddaduud— ' J 1 N _ n «Addqdddde-_Jdd—-dduduqu-du .mw mo Esuuommm mzm .nm mnsmflm v m o n —d414ddddd1—udd_«Judd—ddfiudJuJfi—qqdj a Jew .- -150- -J1—J—Jdld udJJqJ—J J I .mw mo Esuuommm mZm .mm musmflm A N _ n e m o h —Jdd#JJJde—1-JJJJ—Jdd #1— dJJdJJdel—dedJJdd 44 1.44414 44441—444Jdddd41—d1-1-41d w 1? 11 -151- .mm mo ssuuommm mam .mm musmflm a p N n v m e h j—J djj dJJqu-Jd 3+4 4144 —J dJJJJJdJl—J djlqlqud41— deJ—J 41.414.1— 44444 NJJI-I‘J1fldd1—J «444d 4141—4 4 3% 4% pi -152- c p N a .mw mo Esuuowmm mam .oo musmflm v m c h +1—NJJAJAJ1—J _ —NJJJ1-Jjjd1_11444 d .— ddelfiJdeNN‘JJJ-‘AJJ-jdddjlquJA—jAJAJJdejj j 3 4.? j 47 -153- JAA—ddeA #4414441NJNNJJJ141—d4444du44 ’11 e 4 a F E j i N _ n gain] .mw mo Esnuommm m2m .Hm musmflm v m a h 1—44414414aN4—441<11dedJul—JJNdeJJJAJddd3414441—JHNJ1GJ :3 ~154- .mw mo Esuuommm m2m .No musmflm a . a . a .. ... o a. 4J1—d444 NdeAddj-dejdl—Jd «u «A de—JJJddqq NJ‘AJINJjJJd‘ ‘AJJAAJJJJ dejdlqudeJJ-Iq 1 i3 1. -155- .Mm NO 85.30me 53 .mo musmfim a F N . ... e m a N. 441—441—1‘J441-41—441q 44444 deJAjjAJl—JJJAJJJJJI— AJ—J—J JJINJ111_J1H1-d4441— J#j1dde did «411141.44. 011111111 ~156- .mw mo Esuuommm mzm .vm musmwm c — N . a v m o N. dd—AJJAddde—JJJAJANNN—ddd44dd+4—444dd«4444444444d141—dd4d«41411—4‘4‘444‘4—44444 9% N 3% 1: 3 -157- .mw mo Esnuommm mzm .mo musmflh a F N . ... e m o N quq—qduqduu-d—qdqdu—quN—uqduq——ud1—.qqu—adfidd—dqqddqufidddduudddufi—«d1—-qqdu—Nude If - -- j 4:1 -158- .mw mo asuuommm mam .om musmflm c _. N . n v m c h «dd—d4dd‘qduqdc«dad«dddddfidduddfiJuddddddefided«a-«Nd‘fidqqu—-qr—4qq1ded—q41—J1fl ...- 1 N _ -159- p .mN mo Esuuommm mZm .nm mnsmflm N 7 a e ... o N 4.4 —Jq44444 N d u ' ,4 ; _ WAJNuq-_J413d44d4-qJ—dd—alfiddddfldqududddI—Idludu«q‘44—‘44dedddfiqdd 14, Illfl< A -160- J—ld —J—J—l—J NJ 1141—14ij 3441—41-44 414441— # dJJJ-JJNJ 41‘1- de 4J4+1_41 NJ 4141- JAJHAJJJJJ-AJAJJIJJJ 11: I j p N .n .mw mo Esuuommm mzm .mw musmflm v m c h :11 41: Ni -l6l- .mw mo Esuuommm mzm .mm musmflm o — N N v m o N. fiJ—d#dfiqfifijfiqd444