BEIURNING MATERIALS: 1V1531o} Place in book drop to LJBRARJES remove this checkout from w. your record. FINES will be charged if book is returned after the date stamped below. STUDIES TOWARD THE TOTAL SYNTHESIS OF CUCURBITACINS By BALAN CHENERA A DISSERTATION Submitted to MICHIGAN STATE UNIVERSITY in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1984 r": —-V~‘ ’71 “I: I 'Tj"‘fi mn‘m -' V: ‘TMV "—'\ T (‘7—1 YT v~_ ~"_ 8 UDth LC ARD ugh ltLAL SIUTILS_S c: CUCDfirlTAClTS ‘7 WT T.'Tj RALAI: C:- SARA The studies directed toward the total synthesis of cuCurbitacins, a group of roughly twenty tetracyclic triterpenes, are presented. This research has culminated in the construction of tetracyclic ring structures ideally suited for the cucurbitacin synthesis. A major effort has also been directed toward the preparation of several CD bicyclic synthons for tetracyclic triterpenes. Thermal Diels Alder reaction of diene 4 and p—benzo- quinone yielded the adduct 10 in over 753. The chemistry of the adduct10 was explored as a possible route toward the introduction of 11-oxo group and 95-methyl group, which are unique for cucurbitacins. Brief acid treatment of'10 gave the A-aromatic derivatives11emui12ih.80:20 ratio. Prolonged treatment of 10with acid isomerized it to a mixture of A ’11 and A8’9 isomers(17 and 12) in 80:20 ratio. The facile double bond migration is attributed to the stabilizing influence of C-1 oxygen functionality in17’. The conversion of17’to 16followed by epoxidation and ring opening yielded the 11-oxo derivatives 26 and 27, in which the 93isomer,27was found to be the kinetically BALAT‘T CTTEITEJRA favored product. The stable enol 301*:as isolated when 26 and 27was subjected to base treatment. The alkylationBCD—->ABCD A wide variety of steroids have been prepared by Johnson and coworkers using this route starting from 5-methoxy-2- tetralone.20 2. CD +A——>ACD—>ABCD Johnson's estrone methyl ether synthesis is an example of this approach. 3. CD+A—->A3,CD Scott Denmark has proposed an AJFCD Diels Alder strategy for the construction of an 11-oxo steroid, and a key bicyclic synthon has been synthesised towards this end.22 Valenta's synthesis of a D-homosteroid involves an AB+D-—->ABCD strategy for the construction of the tetracyclic system.23 0 f o s O. o O 0 AB 0c ABCD D-HOMOSTEROID t KL) Since the present study involves the construction of the cucurbitacin skeleton, it is appropriate to consider advances made in the synthesis of steroidal natural products bearing groups that are similar to those of the cucurbitacins. The introduction of an 11-oxo group and the establishment of the unusual 9B-methyl configuration are two such features. The introduction of an 11-oxo group has been acheived ,11 in many ways. Hydroboration-oxidation of A steroidal olefins routinely gives excellent yields of 11-oxo analogues and their alkylations have been studied as a means to introduce alkyl groups at C-9.21+ Epoxidation of a ‘8’11 double bond followed by acid catalyzed ring opening 25 s also gives 11-oxo steroids. secently Stork has reported an elegant synthesis of 11-oxo steroid by an intramolecular 26 Diels Alder reaction, as described in equation 2. 10 In order to introduce the 95-methyl group and 11-hydroxy group simultaneously, several workers have explored the migration of 1OB—methyl group to the 95-position during the acid catalyzed cleavage of 9,11-epoxides. Apsimon and coworkers studied the boron trifluoride catalyzed cleavage of some steroidal 9,11-epoxides.27 In the case of 9a,11a—epoxy androst-4-ene-3,17—dione, cleavage with boron trifluoride yielded two major products. One of these was identified as the 3,11a-hydroxy-93-methylestra-1,3.5(10) trien-17-one. The other proved to be 11a-hydroxy-9B-androst- 4,8(14)-diene-3,17-dione (equation 5). .qJ 11 Iavie and coworkers attempted a lanostane to cucurbitane 1 transformation using Apsim n s strategy. However only elimi- 28 nation products were isolated.(equation 4). '- Edwards and Paryzek have reported extensive studies with tetracyclic triterpenes. Cne of these involved an efficient 1O 9 methyl migration leading to compounds having the cucurbitacin skeleton.15 Then 33-acetoxy-93,J1B-epoxylanost- an-7-one was treated with boron trifluoride in acetic anhy- dride, the 93-methyl isomer was isolated in moderate yield. (equation 5). The course of this rearrangement was found to be very sensitive to substrate configuration and reaction conditions. The corresponding 9a-11a-epoxide isomer reacted sluggishly to give 7,11-diketolanostane product.(equation 6). 12 Since the cucurbitacins possess no functionality at C-7, reductive removal of the 7-cxo group was attempted. This failed and the authors suggest that this difficulty is due to the steric hindrance by the 95-methyl, 15-methyl- ene and 1ha—methyl groups. Unfortunately, the undesired 7-oxo function was necessary for fi-epoxidation. Epoxidation of Eg-acetoxy lanost-9(11)-ene gave only the a-epoxide.29 Preparation of the B-epoxide via the bromohydrin route failed to yield this product. Cnly the reaction of BB-acet- oxylanost-9C11)-ene-7-one with m—chloro perbenzoic acid gave substantial fi-epoxidation.30 (equation 7). A00 MCPBA > a+fi Epoxldu («1.7) O Very recently Edwards and Paryzek have reported another lanostane to cucurbitane transformation.31 They applied a Westphalen type rearrangement to a 9a-hydroxy-11-ketone32 derived from lanosterol in order to induce the migration of C-10 methyl group to C-9. Three products were identified from this reaction. (equation 8 ). Although this work represents the first synthesis of a true cucurbitacin skeleton, the desired product is formed in poor yield. Improvements are needed for an efficient synthesis of cucurbitacins. anp8) 14 As a result of previous work in our laboratory, many transformation of bicyclic dione 1 have been studied and the synthesis of tetracyclic intermediates has been acheived. Because of the difference in reactivity of the two carbonyl functions it is possible to conduct many reactions at the more reactive site without protecting the other. The side chain construction has also been effected on a model compo- und 2 derived from the parent dione 1.311L (equation 9). 1 N8 3H4 2 Base Adel Several Steps (eq- 9) 15 The first part of this dissertation discusses an efficient construction of a tetracyclic intermediate ideally suited for cucurbitacin synthesis. Several useful and novel transformations leading to 1441-methyl estrane derivatives will also be discussed. The second part describes the synthesis of a large class of functionalized bicyclic synthons for the construction of tetracyclic triterpenes, starting from trans- 1,6-dimethylbicyclth,3,O)-nona42,7—dione l, RESULTS AND DISCUSSION As noted earlier, trans-1,6-dimethyl bicyclo(4,3,0)- nonane-2,7-dione, l) undergoes nucleophilic reactions selectively at the six membered carbonyl function. For example, sodium borohydride reduction, ketalization, organe- lithium addition etc. proceed exclusively at this site. The difference in reactivity between five and six membered cyclic ketones was discussed by H.C.Brown almost three decades ago and is attributed to the I-strain effect35’ 36. 16 17 Upon treatment with boron trifluoride etherate in THF- benzene, vinyl carbinol é yields diene 3 which undergoes Diels-Alder reactions with various dienophiles. Both the thermal and Lewis-acid-catalyzed Diels-Alder reactions of diene 3 provide the¢£kendo adducts exclusively. The structures of some of these adducts have been confirmed by X—ray crysta- llography. 18 it was the remarkable¢xyendo selectivity observed in the Diels-Alder reactions of diene 4 that prompted us to invest- ~ igate the use of the homologous diene 2 in a cucurbitacin total synthesis Us 4 (947) 1 7°" '73. C Benzene ’ 857. Schenue 4 The poor to moderate yields from our previously reported preparation of vinyl carbinol‘2,were attributed to extensive enolization of the carbonyl functions by the Grignard reagent in THF solution. By performing the reaction at low temperature in a hydrocarbon solvent with freshly prepared vinyl magnesium bromide, we obtained an excellent yield (80 to 85%) of 2: Next we observed that treatment of’é with copper sulfate in hot benzene gave facile dehydration t0,fl with no trace of the isomerized dienelo. 28 In an attemt to improve the yield of the quinone adduct 19, the Diels Alder was carried out with stannic chloride and boron trifluoride as catalysts. However the desired adduct was obtained only in moderate yield (45% . Isomerization of the diene‘g’to an extent of 256 was also observed during the Lewis-acid—catalyzed reactions. Fortunately slow addition of diene E’to a refluxing solution of benzoquinone in benzene provided the desired adduct in over 70% yield. In this case the thermal Diels Alder reaction gave better results than the Lewis-acid—catalyzed modification. Two isomers were actually obtained in the thermal Diels Alder reaction. The minor isomer was observed to rearrange slowly to the major isomer in solution. Therefore rigorous structural assignment of the minor isomer was not attempted. However the major isomer was tentat- ively assigned thecflbendo configuration based on the known structures of related adducts from other quinone dienophiles. O Benz. 10 29 Brief acid treatment of l9 according to the procedure of Ansell and coworkersl+1 gave the nonconjugated aromatic isomer 1; contaminated with a small amount of the conjugated isomer 1 .(Bquation 14). ~ cuacozn/HCI 1 Hr (eq.14) One notable feature observed for 11 is the large homoallylic coupling constant between the 66¢ and 69 hydrogens and the 99 hydrogen (J6ec ’99 ‘=5.5 Hz, J69 ’9; =8.0 Hz). Large homoallylic coupling constants of this kind have been reported for 1,4—dihydronaphtha1ene systems of rigid conformation?2 30 In compound ll, the large homoallylic coupling constants, together with other coupling constants (J63,7:§.2 Hz, J&¢,6p =22.6 Hz), serve to establish the assigned configuration. The relative insolubility of Ll and 12 hindered their purification, so this mixture was converted in excellent yield (91%) to the dimethyl ether derivative 13. Chromatography of l; on silica gel gave crystalline l; with a small amount of is; It is known that for steroidal analogs having an aromatic A ring, the 59’9 and 25’11 isomers may be equilibrated by acid treatment, and the equilibrium ratio depends upon structural changes in ring D. Iainaut and Bucourt have studied the effect of ring D puckering and angle strain on the equilibrium composition of zg’gand ‘2’11 isomers of estratetraenes and their results are summerized in table 3 43. 31 A8’9 and A9,11 uh ium composition of . Equilibr 7:717 5 -r‘JAJ TA isomers of estratetraenes in acid. _ 11111 . . - J llllll Id """""""""""""""""""""""""""" — . . _ . . . _ . . . _ . _ l l l l l _ — - — n 8 - 8 — 1. .d . _ h. . Ru . o/ . ”U“. C - 6 _ - — - a _ _ . . . ..A . . _ . _ .x: n . . _ . _ - 47. cal— - - - - - _ . . _ _ . -I' 11 II. """""" J """"""""" 4 """""""" J """"""""" - _ _ . . . _ _ l l l l l . v) d. . . . _ .l _ _ — _ - 9 C _ - _ 2 — _ _8A_ a. . . ,6 . 1. _ .2 . . _ O . 22 . . . .,3 n. 4. _ . . . — o... .l — . - - - . . _ . _ _ . . . . . _ - llll % """""""" L- """""""""" ‘- """"""""" L— """"""""" u _ . . . _ .1. . . _ . . . _ . . _ . . . . . _ _ . . _ . . . _ . _ . _ _ . . u u u m.“ — . _ . _ _ 2. _ . . . by. . . . . _ _ _ u u n m n . . . . . . . . _ _ . by. . l l l l — 1 _ _ . . . nn_ - d _ . _ _ _ n _ . . . . u . . . . _ O _ . . . . @ _ . . . _ . m l l l u _ . _ no . . . . nv . . . . n3 .K . . . . . R. . 32 Assuming no serious perturbation by the 14¢cpmethyl group Q,11 we expect a 40:60 ratio of c?’9 and a: isomers of 14, Indeed a recent report by Bull and others show that the equilibrium composition of A to a’ product isomers is 1:1 during the acid catalyzed cyclization of compound 12. He also found that on conversion of the ketone function at 0-17 to a ketal this isomer ratio changed to 40:60 in favor of the 8’9 >4“. isomer (equation 15 (013.15) Treatment of 13 and 18 with p—toluene sulfonic acid gave a 1‘éamixture of conjugated isomers 15,and lé in quantitative yield. The angular methyl groups in lg’and 16 showed 1H NMR signals that correspond closely to those reported from the 3-methoxy-14sc-methylestratetraen-17-ones by Bischofberger and Bulluq. However the C-11 olefinic proton in lé’has shifted downfield due to its proximity to the C-1 methoxy group. The predominance of this isomer in the acid catalyzed equilibrium mixture is unexpected, when one considers long range conform- ational effectsl+3 and the crowding of the 1-methoxy substituent. 3A :ith regard to the possible conversion of these 14cc-methyl e trane derivatives to cucurbitacins, it is clear that the (1') 9(11) double bond in 16 might serve as a latent 11-keto function. However the petential utility of such an approach was diminished by the substantial amount of 13 that accompanied 16. We anticipated that this difficulty might be lessened in the hydroxyl series due to the smaller size of 0H as compared to 0He, and were pleased to find that 1; (or even 10) was : isomerized to the required 9,11-double bond isomer 13 in over 80$ yield by acid treatment. This transformation can now be achieved by dissolving the initial adduct l9.in glacial acetic acid containing a mall amount of concentrated hydrochloric acid, and allowing the solution to remain overnight at room temperature (equation 16). 10 (eq.‘16) \u U1 With previous work in mind, it seems likely that the predominant formation of the c?"‘ isomer in our studies is due to a stabilizing influence of the oxygen function at 0-1. The close proximity of the C-1 hydroxyl group to 0-11 was confirmed by the chemical shift of the vinyl proton at 0-11 (5 7.1 ppm). The 0-1] hydrogen in the corresponding lace-methyl estrane derivatives reported by Bull appeared at 6.2 ppm in the 1H NMR . Attempted dehydrogenation of ll and its derivative led to some novel transformations. When 11 or a mixture of ll and L; was treated with a palladium catalyst (10% Pde), in refluxing xylene, lg and 12 were isolated, their amount and ratio varying with the reaction conditions. If the weight ratio of catalyst to substrate was 0.3 or less the B-aromatic-1,n-diketone lg was obtained in 50% yield. With larger amounts of catalyst (ratio ca 1.0) and a longer reaction time (24 hrs), the naphthaquinone derivative 12 was formed in almost quantitative yield. We explored the possibility of effecting a photochemical oxidation of 12 at 0-11, following the procedure of Rommel and Wirz for 5—methyl-1,4-naphthaquinone ; however this reaction gave complex mixturesh5(equation 17). 36 0 HM: I M XYIQI‘IG.A HOAc —>Complox- («1.17) 19 The formation of 1Q in the palladium-charcoal dehydrogenation was unexpected. The keto-enol tautomerism of 1,4-dihydroxy naphthalenes has been studied by Bruce and others“5 , and these workers have found that additional hydroxyl groups in the 5 and 8 positions of the 1,4-dihydroxynaphthalene system favored the 1,4—diketo form over the enol tautomer. The unexpected stabilization of the carbonyl tautomer in this case was attributed to intramolecular hydrogen bonding. However there 37 is no such stabilization in compound lg. We speculate that the peri interactions in lg may be less than those in the naphthalene diol tautomer thus stabilizing 1g. The four protons on 0-2 and C-3 displayed as a singlet at S 3.0ppm in excellent agreement with that of similar compounds. In an attempt to prepare AB aromatic 1hcc-methylestrane derivatives, similar dehydrogenations were effected with the dimethyl ether derivative 1;. Treatment of 13 with Pd/C in refluxing xylene gave a mixture of ,a?’9 and the desired AB aromatic compound 29. Although the mass recovery was nearly quantitative, this reaction proved to be capricious, and the products were extremely difficult to separate. A clean high yield oxidation of 12 to 21 was acheived by treatment with dicyano dichloro parabenzoquinone, but the 11,12—double bond resisted all hydrogenation efforts. Finally, N-bromosuccinimide reacted with 13 to give 22, a bromo derivative of 21 (scheme 5 ). N N ”I 58 Scheme 5(cont) DEN) 13 44.. Benzene 118$ 13 Since the facile preparation of compound 12 from dione 1, had been demonstrated, our attention was next turned to the transformation of this olefin into the 11-oxo derivative. Again, because 12 proved to be insoluble in most common organic solvents, it was converted in excellent yield (90%) to the bismethoxy derivative 19, by reaction with sodium hydride and methyl iodide. 00% "' 16 ’11 olefins The formation of Hoe-alcohols of steroidal A is well known4?As the stereochemistry at C-11 was not critical to our synthesis, a similar hydroboration was attempted on lg, The protection of ketone was not undertaken because this functional group is known to be very hindered and therefore unreactive. Also Bull et a143 found that upon ketalization of compound 2;, the ,¢_’11 double bond shifted substantially to the A8’9 position. Furthermore, an extremely poor yield (#096) of the ketal was realized in their synthesis (equation 18). do Hydroboration—oxidation of lg surprisingly gave reduction of the C-17 carbonyl. High resolution NMR revealed it to be a mixture of oCand P alcohols, which was readily oxidised to 16 with pyridinium chlorochromateEB(equation19 ). Since the introduction of an oxygen function at C-11 via hydroboration failed we next studied the rearrangement of epoxides derived from lé’ Initial attempts to epoxidize 19 with unbuffered meta chloro perbenzoic acid gave rise to several rearranged products (mainly allylic alcohols and 11-ketones). However, epoxidation of 1Q by the two phase dichloromethane/aqueous potassium carbonate procedure of Anderson and Veygeslu‘.“9 afforded a crystalline, but extremely acid sensitive epoxide mixture. The ratio of these epoxides was determined by 1H mm, to be 70:30 favoring the p-epoxide. The two epoxides exhibited the following relevant resonances in ‘H NMR (250 MHz, CD013). A doublet at S 4.88 (J=5.8 Hz) 1+1 and a triplet at S 5.05 (J=2.14 Hz) for 24 and 25 respectively. The coupling constants observed for these two compounds were in excellent agreement with that predicted from an inspection Dreiding models and an application of Kouplus equation. They were also similar in magnitude to that observed for somewhat related steroidal 9,11-epoxides50. However the large chemical shifts observed for 24 and 25 (4.88 and 5.05 respectively) might be attributed to the influence of C-1 methoxy group. In CPBA CHZC 1,;Nch03 16 24d'9dfiic 25 .-. 99,113 We then examined the rearrangement of epoxides 23 and 2?. Lithium perchlorate was selected as our first Lewis acid because it cleanly catalyzes the rearrangement of certain cyclohexene epoxides51.into the corresponding cyclohexanones. When the epoxides 23 and g; were treated under reflux in dry benzene containing a small amount of lithium perchlorate, the products obtained included the expected 9¢H and 99H 11-ketones 2g and 22 together with a substantial amount of AF: £9’11 diene 28 in 30 to h0% yield.(equation 20) 42 The more potent Lewis acid boron trifluoride etherate in methylene chloride as a catalyst for epoxide rearrangement proved to be too severe, providing diene 22 in poor yield (30%) as the only identifiable product (equation 21). CHHCH 45 Fortunately, the conversion of epoxides 2g and a; to the 11-keto isomers 26 and 22 was improved by conducting the rearrangement of these epoxides in tetrahydro furan solution containing boron trifluoride etherate as a catalyst. The milder reaction condition of the THF medium is presumably due to the relatively strong Lewis basicity of this solvent and the resulting decrease in the acidity of the complexed Lewis acid. Although addition of triethylamine has been suggested for the cleavage of sensitive epoxides, this proved to be of no advantage in this study.(equation 22), The stereochemical assignments at 0-9 in 26 andqu were based mainly on 1H NMR evidence. The 10.h Hz coupling constant between the 8 and 9-ahydrogens of 26 agreed with the assigned trans configuration. A similar coupling constant has been reported for the A-aromatic steroidal analogs. (J9‘385=9.5 HZ for 11-oxo-9a-estradiol-B-benzyl ether52. 24 as -E o + __u____, 25 Turner 76% #4 However one must now consider the abnormally large 8,9 hydrogen coupling (12.8 Hz) for the molecule assigned as the cis isomer 22. Clearly some kind of structural deformation is needed to explain this value. An inspection of molecular models suggest that a boat like conformation of ring B alleviating steric interaction between 1uac-methyl group and ring A, causes a near eclipsing of the hydrogens at C-8 and C-9. The relatively low chemical shift observed for 1hoc-methyl group (1.5 ppm) of 26 might result from the deshielding effects of 11-keto and 17-keto functions, assuming that ring C adopts a chair conformation 53. A study was then undertaken in order to determine the products derived under kinetic and equilibration conditions. Kinetic protonation of the enolate anion derived from 26 and 22 gave mixtures in which 22 (the cis epimer) predominated (75%) Protonation under equilibrating conditions (KOH/MeOH) yielded the trans isomer 26 and the enol ;g in the ratio 40:60. No observable amount Of‘EZ was found in the crude product mixture. However upon purification on silica gel, 29 underwent slow isomerization to g]. The enol proton of 29 gave a 1H NMR signal at 87.2 ppm and this disappeared on exchange with D20. The formation of the 9«:-isomer under thermodynamic control indicates that the 14oc-methyl group has a strong influence on this stereochemical outcome. In steroidal analogs bearing an 11-keto function, the 98. isomer is found to be more stable than the 25 9ec.isomer. (scheme 6). 1+5 1) LDA mu: monacozu SchemeG 46 Alkylation of the 11-keto derivative was next studied as a means of introducing a 9al-methyl group. Unfortunately all attempts to introduce a methyl group by enolate alkylation failed. The only product isolated was the enol ether 21. Similar O— alkylations were observed with ethyl bromo acetate and allyl bromide giving 3; and 33 respectivelyh(equation 23). Koed RI or RBr eq23 31; 3::(313 323 R = CH2C02C2H, 33’ R = CH,CHcH2 M7 Attempted cycloprOpanation 0f,21 with Zn/Cu couple and CHZIZEQ'resulted in the complete recovery of starting material. In situ generation of silyl ether of 26 and 22 followed by trapping with chloromethyl phenylsulfide was unsuccessfulEs. A possible route for the introduction of the alkyl group at C-9 would involve a Claisen rearrangement of 33 to the C-alkyl derivative 2?? If this reaction proceeds stereospecifically as shown, an efficient conversion of allyl group to methyl group can be acheived by ozonolysis followed by oxidation and decarboxylation(scheme 7). KOt-Bu BfCMzCl-I:CH2 2. no: 3‘.—co2 Schenwe'? 48 A general approach to cucurbitacins may be based on the application of the intramolecular Diels Alder reactiog? A retrosynthetic analysis starting from our C/D bicyclic intermediate suggests two similar strategies (A&B). A third strategy (C) is based on Stork's approach to the synthesis of 11-keto steroids. C As can be seen in strategies A and B, very similar inter- mediates are involved in these strategies. Of paramount importance in these sequences is the control of stereochemistry at the potential C-9 site. If the reductive alkylation step (equationéflr) proceeds with good stereoselectivity, the desired products may be readily obtained by adjusting the order of the alkylations. (ea. 24) 50 The bicyclic intermediates needed for these reactions may be prepared from the readily available diketone 1, The preparation of enone 36 from bicyclic dione l was reported previously by Jacob Tou58. However the sulfoxide and/or selenoxide elimination routes used gave only poor yield of the enone. The major problem encountered in these reactions was the difficulty in preparing the ac-ketosulfide‘3z or the eC-ketoselenide 28. Phs PhSe A careful examination of the reaction mixture derived from the sulfenylation of parent dione l’by treatment with LDA followed by diphenylsulfide revealed the presence of a substantial amount of starting material (50%). This observation agrees with reports by Trost for the sulfenylation of other carbonyl compoundng. Because the acidity of the initially formed oC-ketosulfides exceeds that of the starting material, unreacted enolates react faster with the product OC-ketosulfide 51 than with diphenyl disulfide. Normally, two equivalents of base may be used and excellent yields of products are realized. However, since dione l’forms a bisenolate on treatment with two equivalents of base, such an approach was considered unwise. More reactive reagents such as phenyl phenylthiosulfonate or phenyl selenylbromide gave no improvements in the yields of 22 or 2§.Therefore an alternate synthesis Of 36 was reqUired. N Kinetically controlled bromination of with 2-pyrrolidone ,1. hydrotribromide (PHT) was reported to give an excellent yield (84%) of 22611 By performing the bromination with bromine in glacial acetic acid at room temparature, the thermodynamically favored equatorial bromoketone £9 was isolated in over 90% as a white crystalline solid (equation 25) PETT Bmmme HOAc meta... 52 The dehydrobromination of 39 with Ca003 in dimethyl formamide61 proceeded smoothly to give 36 in 86% yield. Since enone 36 proved to be readily available, we next examined some of its reactions to determine whether potentially useful intermediates for cucurbitacin synthesis may be prepared in this manner. The addition of vinylmagnesium bromide to 26 provided only the 1,4 addition product 31. Reaction of 36 with one equivalent of allylmagnesium bromide gave 85-90% yield of the 1,2 adduct 55' The amount of allylmagnesium bromide in this case is critical because of the facile addition of allylic organometallic to the five membered carbonyl function of 26. A substantial amount of the his adduct E3 was in fact isolated from the reaction of 36 with excess allylmagnesium bromide. The facile addition of allylic Grignard. to the carbonyl functions can be rationalized 862. by the six membered transition state as shown in scheme 5.5 SchenxaB Oxidation of fig,with PCC gave the triene &3_as the major component. Only a small amount of the expected product E; was obtained with PCC(pwridinium chlorochromate). O The dehydration of 33 with PCC can be explained by the action of slightly acidic reagent upon a sensitive allylic tertiary alcohol 32. However, a facile oxidative transposition of £6 to 53 by PCC (equation 26) suggested that such an approach could well be applied to 33 after functionalizing the terminal double bond. PCC: 54 Thus on hydroboration with thexylborane followed by oxidation of the derived organoborane with alkaline hydrogen peroxide?3 32 gave the keto diol flg'in almost quantitative yield. The primary hydroxyl group of £8 was protected as the acetate with dimethylamino pyridine (DMAP) / acetic anhydride to give 32.64 PCC oxidation of £2 yielded the transposed enone 29 in 90% yield. (equation 27 ). C) 1 Thu ylbo re M g, 2 Ne OH P1202 eq. 27 When the diol 38 was treated with PCC, the expected spirolactone 51 was obtained as the only product. The formation of 51 can ~ ~ be envisioned as proceeding through the lactol intermediate. (equation 28 ). Both 29 and 51 should serve as potential intermediates towards the proposed intramolecular Diels Alder (IMDA) strategy for cucurbitacins and the synthesis of several 55 other tetracyclic triterpenes. 51 PH) Next we explored the chemistry of the homologous enone 52. ~ The enone 23 was prepared in over 80% yield from 26 by the alkylation of the potassium enolate (KOt-Bu) with methyl iodide. (equation29 ). Koed tBuOH-THF can An alternate route to the enone 23 through the alkylation of sulfoxide 53 followed by thermolysis was complicated by the F0 formation of considerable amount of the O-methylated derivative 54. (equationEX3). N 1-NIH3CHaI 2oToluene In contrast to the facile addition of allylmagnesium bromide to 22, 22 was found to yield several products by the attack of the organometallic on both carbonyl sites. Therefore a selective protectibn of the 8-keto group was required. During a related study, we discovered that the five membered ketone of 22 could be reduced with the NaBHu/CeCl3 reagentéfii in 81% yield, to give 22. The stereochemistry assigned for 22 is based on the 1H NMR signal of C-17 hydrogen (dd, J=8 Hz, 2 Hz) Addition of allylmagnesium bromide to 22 followed by oxidation with PCC gave 26; no further oxidation was observed with this reagent. However 22 underwent smooth oxidative transposition 57 JONESREAGENT e with Jones reagent66 to give the expected enone 23 in 91% yield. The intermediate 56 was also converted into 58, 59 ~ Na! and 29 by the procedures applied to the parent system. O| 58 so 59 58 We also explored the possible construction of functionalized analogs of diene E) which are expected to show enhanced reactivity towards electron deficient dienophiles. Although the addition of diene E’to various quinone dienophiles proceeds efficiently, an attempted addition of this diene to the novel dienophile 21 failed. A facile addition of diene E’or its = analog to 21 should ease the ringnA modification greatly, including the introduction of geminal dimethyl group which in the past has caused considerable problems to our synthesis of tetracyclic triterpenes. We hoped that this sluggishness could be overcome by introducing a siloxy substituent at the diene moiety of A. To this end we have prepared a triketone precursor 65 to the trisiloxy tetraene 66 as shown in schemeEL ah! N 59 ArSOgH 1' Epoxidat ion I 1.3eq LDA ——> 2.3eq TMSCI 2' Base 3. (o) Scheme9 60 In the course of this study, we observed an interesting kinetic selectivity in the acid catalyzed rearrangement of’23 The preparation of the thermodynamically more stable isomer of the trans diene 22 from the vinyl carbinol’égon treatment with either PTS or iodine has been reported . When the dehydration of’é'was carried out with PTS in refluxing toluene for 1 hour a mixture of trans dienes 22 and 22 were obtained in almost 1:1 ratio, as determined by high resolution NMR. However on prolonged treatment, this diene mixture underwent complete isomerization to the more stable isomer 22 (equation3q ) Since the formation of either 22 or 22 must proceed through the intermediate formation of diene .L-L’ we studied the rearrangement 0f.& under Bronsted acid condition. Treatment of fiiwith O.1 equivalent of p—toluenesulfonic acid in refluxing benzene for 1 hour gave the Z—isomer 22 in.85% yield. When the acid catalyzed isomerization of E’was monitored by 1H NMR, a 61 new set of signals appeared, reached a maximum after 20 minutes, at 75°C and then faded as the spectrum of 22 grew stronger. Careful treatment of 2 with p—toluenesulfonic acid in chloroform enabled us to isolate the intermediate 22 as a crystalline product to which the E-configuration was assigned on the basis of 1H NMR. A nuclear Overhauser effect for the 1.8 methyl doublet and the 6.5 multiplet was observed for 22, but no equivalent signal effect was found for 22?7Some of the important 1H NMR assignments for these compounds are shown in the accompanying formulas. The 13C data also confirms the assigned configuration for 22 and 22. A Afii6e61 has been observed for 8 the C-3 carbons (steroid numbering) of 22 and 22. - 355“") 0 85.65 (m) o H H esssH 8a40H (bud. «9.3) (bud.J-IO.I) H CH, CH 3 H 85.6 (m) a l.90 8|.80 85.35 (d. J-Za) (duh-7.0) (q. J-ZO) Since the corfigsponding isomerization of 6,6-dimethyl-1- vinylcyclohexene 229to its trans isomer‘22 did not exhibit the above phenomenon, the possibility that it reflected unexpected differences in conformer equilibria was considered. Scheme 10 illustrates a possible explanation for the rapid formation of the thermodynamically less stable diene 22 from 2: The barrier for conformational interconversion in simple dienes is low (6 K.cal/mole), and in the absence of steric hindrance effects, the s-trans conformer is more stable than the s-cis or s-skew conformers by approximately 2.1 kcal/mole?DAllylic carbocations on the other hand have conformational barriers in the 38-h3 kcal/mole range and such intermediates exhibit strong structural integrityz1lf protonation of E’were to generate intermediate cation X(t) preferentially we would expect to obtain diene 22, provided deprotonation of X(t) is faster than its isomerization t0 X(C)e 67 68 Scheme“) 64 It is known that the s-trans conformer of acyclic dienes generally has a larger molar absorptivity (El) than the corres- ponding s-cis conformer?2 From the examples cited in Table we see that the UV absorption of’i is similar to that of 1-vinyl cyclohexene, indicating a similar s-trans: s-cis equi- librium for these compounds (s- trans predominates). Diene 22 on the other hand , appears to assume mainly the s-cis conforb mation. Because of the low-barrier for conformational interconversion of the diene, such a preferential protonation-requires that the activation energy for h(t)-——§>X(t) be lower than that for 4(c) X(t)?3 Since this does not appear to be true for the protonation of 22, we have looked for a unique conformational factoe in X that might reflect in properties of‘&: Evidence for such a factor has been found in the UV absorption spectra of 4, 67, and related dienes. Absorption Maxims and Mo1ar Absorptivities of Some Conjggated Dienes Compound A Egg: (nm) 6 Ref. 3-methylenecyc10hexene 231 (hexane) 21.000 74 1-viny1cyc10hexene 231 10,600 ’5 1,2-b1smethylenecyclohexane 220 6,400 76 1,3-cyc1oheptad1ene 246 7,500 77 ,4, 235 11,900 This work 131 237 6,600 This work 65 The unexpected similarity of the molar absorptivity of 32*.» to that of 1-vinylcyclohexene rather than 22 may be attributed to structural distortion in 2' introduced byrthe trans-fused five-membered ring. Molecular models indicate that the six- membered ring in fl’is forced into a boat(or twist boat) con- formation. This permits the s-trans diene conformation to experience less steric crowding than it does in 22. We suggest that this factor helps to lower the activation energy of the b,(t)—->X(t) reaction relative to that for compound 22. The transformation of diene 22 to the non-conjugated enone 25 has been achieved previously. However the isolation of the sensitive epoxides from 22 has not been reported. When the epoxidation of 22 was monitored by 1H NMR two stereoisomeric epoxides were observed in almost 1:1 ratio. (equation 32) . 8 Recently, these two epoxides were separated in our laboratory7; by chromatography on silica gel. In contrast to the epoxidation of 22, 22 gave a single crystalline compound in quantitative yield (equation 55) on treatment with m-chloro perbenzoic acid. eq- 33 The non-conjugated enone 64 was found to be very sensitive towards exposure to air. For example 64 underwent smooth trans- formation to the triketone 65 (equation 34). Air Dichloromet hano ecu-34 The preparation of the triketone 65 was achieved in moderate yield(50%) from 64 according to the scheme The epoxidation of 64 with m-chloro perbenzoic acid under a variety of condition yielded, besides the expected epoxide 29, a substantial amount of the Baeyer Villiger product 25: Thus only about 50% of the required epoxide was obtained by this route.(equation 35) MCPBA 7(1 ‘+ We thought the formation of 2; could be totally suppressed by first reducing the ketone functionalities in 64 followed by epoxidation and reoxidation as shown in equation.q2§ .Unfortu— nately the reduction of 64 with sodium borohydrida/ ceric chloride was found to be capricious. 2-m CPBA 39cc Since the isolation of an epoxide from 6; was easy we studied some of its reactions as a means to construct a functionalized analog of 4, Schemell illustrates the transformation of 62 into the bis-silyl tetraene 22. The reaction sequences in scheme‘l1 ndght be very well applied to the epoxides derived from the thermodynamically more stable isomer of the trans diene 6; also. (3 C) LDA PCC —> THRHMPA ”a“! LDA THF HMPA TBDM SCI Skiuune 11 Treatment of 62 with 3 equivalent of lithium diisopropyl amide(LDA) in THF/HMPA isomerized it to the dienol Z; in 88% yield. Oxidation of 22 with pyridinium chlorochromate gave the dienone 23 in 87% yield. Upon treatment with excess LDA followed by trapping with tert-butyldimethylsilyl chloride 24 yielded the bissilyl ether 22 (60%). The intermediate Zé may be considered as a valuable synthon for the construction of various tetracyclic triterpenes. A thorough study of the Diels Alder reactions between this reactive diene and various dienophiles should give fruitful results. A preli- minary study involving 2? and p-benz0quinone has been partially successful. A rigorous structural elucidation for the adduct obtainec was not completed. An efficient addition of 22 to the novel dienophile él-should open new avenues for the total synthesis of tetracyclic triterpenes. 7O EXPERIMENTAL General Except as indicated, all reactions were conducted under dry nitrogen or argon, using solvents purified by distillation from suitable drying agents. Magnetic stirring devices were used for most small scale reactions; larger reactions were agitated by paddle stirrers. Organic extracts were always dried over anhydrous sodium sulfate or anhydrous magnesium sulfate before being concentrated or distilled under reduced pressure. The progress of most reactions was followed by thin layer chromatography (TLC) and/or gas liquid phase chromato- graphy (GLPC). Visualization of the thin layer chromatograms was effected by spraying 30% sulfuric acid with subsequent heating. Analysis by GLPC was conducted with A—90-P3 or 1200 Varian-Aerograph instruments. Flash chromatograhy was carried out on flash silica (37- 53 mesh) as suggested by Still et al. Melting points were determined on a Hoover-Thomas apparatus (capillary tube) and are uncorrected. Infrared spectra (IR) were recorded on a Perkin-Elmer 237B grating spectrophoto- meter. Proton magnetic resonance spectra (1H NMR) were taken in deuterochloroform or acetone -d6 solutions with either a Varian. T-60 or a Bruker 250 MHz spectrometer and are cali- 71 brated in most cases in parts per million ( 6) downfield from tetramethyl silane as an internal standard. In some cases the chloroform peak (7.26) was used as a standard for 1H NIH measurements. Ultraviolet spectra (UV) were recorded on a Unicam SP—800 spectrophotometer. Mass spectra (MS) were obtained with either a Hitachi RMU 6 mass spectro- meter or a Finnigan 4,000 GC/MS spectrometer. Carbon magnetic resonance spectra (13C NMR) were taken in deuterochloroform solution with a Bruker 250 spectrometer and are calibrated in parts per million (6) downfield from tetramethyl silane as an internal standard. Microanalysis were performed by Spang Microanalytical Laboratory, Ann Arbor, Michigan. 72 Preparation of cisoid diene 4 by copper sulfate dehydration of the vinyl carbinol 3. 500 mg of alcohol 3 was refluxed in 100 ml of benzene with 800 mg of c0pper sulfate. Water was removed azeotrOpically. After two days of refluxing, the c0pper sulfate was removed by filtration. The filtrate was washed with ether and the combined organic solvents were evaporated under reduced pressure. The product was purified by Kugel rohr distillation. 402 mg (94% yield) of cis diene was obtained as a semi solid. uv (EtOH) max235nm (loge 4.08); IR 1750. 1665, 1625 cm“; ‘H NMR 50.9 (s, 3H), 1.0 (s, 3H), 1.1-2.6 (m, 8H). 4.7 (d, 1H. J=11 Hz), 5.05 (d, 1 H, J=17 Hz). 5.4 (t, 1H, J=5 Hz), 6.0 (dd, 1H, J=11 and 17 Hz); MS m/e (rel. abund.) 190 (79), 175 (38), 133 (100), 119. Preparation of diketone 7. To a solution of lithium diisopropyl amide, prepared by reacting 2.5 ml of diisopropyl amine (17.8? mmol) in 50 ml of THF with 7.3 ml of a 2.42 M n-butyl lithium in hexane, (17.67 mmol),was added at -78’C, 2.85 gms of dione 1 (15.83 mmol) in 100 ml of THF. It was allowed to stir at -78°C for 30 min, warmed to room temparature and then 30 ml of HMPA was added. It was then stirred at this temparature for an additional 30 min, cooled to 07C followed by addition of 1.5 ml of methyl iodide. After 5 min the reaction mixture was quenched with water, extracted with ether, washed sequentially with sodium bicarbonate 75 and brine. The ether layer was dried over anhydrous magnesium sulfate and the solvent was evaporated to give a yellow crystalline material. This product was treated overnight with methanolic potassium hydroxide. The resulting solution was diluted with benzene, washed with water, brine and dried. Evaporation of solvents yielded a pale yellow solid weighing 2.66 g . It was first crystallised from ether and then from ethyl acetate-hexane to give 1.6 6 (54%) of white crystalline solid 7. M.Pt 118-12030; IR(CHC13) 1710, 1740 cm" ‘H NMR(CDC13) 0.9 (s. 5H), 1.1 (d, 5H. J=7 Hz), 1.2 (s,5H). 1.5-2.4 (compleX. 9H). MS (70 ev) m/e (rel intensity) 194 (56.4), 179 (50.06): 165 (19.65). 152 (45.88), 137 (29.85), 124 (79.09). 109 (85.04). 96 (70.72). 82 (100), 67 (70.22), 55 (48.99). Preparation of vinyl carbinol 8. To 290 mg of dione 7 (1.5 mmol) in 10 ml of THF was added 4 ml of 1.1 M vinyl magnesium bromide in THF. The reaction mixture was stirred under nitrogen for two days, quenched with saturated ammonium chloride and processed as usual to give 330 mg (98%) of alcohol 8 as a white crystalline compound. M.Pt 155- 137 c. IR (nujol); 3500. 1735 cm". ‘H NMR (cnc13) 0.95 (s, 3H). 0.9 (d, 5H, J=7 Hz). 1.3 (s, 5H). 1.4-2.5 (complex, 1039, 5.0. 6.2 (typical vinyl pattern, 3H). MS (70 eV) m/e (rel intensity) 222 (4.96), 207 (25.18), 204 (0.55), 126 (17.54). 111 (100). 97 (20.59), 84 (24.59). 74 Preparation of the Diels Alder adduct 10. To a refluxing solution of 6.5 gms of p-benzoquinone in 200 ml of benzene was added 4.0 gms of diene 4 in 150 ml of benzene under argon, over a period of one hour. The reaction mixture was refluxed overnight, cooled and washed with 100 ml of 10% sodium bisulfite solution. The organic layer was washed with brine and dried over anhydrous sodium sulfate. Removal of solvents yielded an oil which on trituration with ether gave 4.66 gms (77.66%) of slight yellow crystalline solid. The solid was found to be a mixture of two products. The major isomer ( 80%) displayed the following properties. M.Pt. 170— 173 C; IR 1740, 1690, 1600 cm"; 1H NMR 1.05 (s, 5H), 1.20 (s, 5H), 1.4:2.8 (m, 11H), 5.2 (broad s, 2H), 5.20 (q, 1H, J=2.9 Hz), 6.53 (dd, 1H, J=1.2 and 10.3 Hz), 6.62 (d, 1H, J=10.3 Hz); 130 NMR 219.2, 201.0, 198.9, 144.8, 141.0, 137.0, 115.6 ppm and eleven higher field signals; MS m/e (rel. abund.) 298 (14), 189 (27), 145 (25), 131 (27), 123 (100), 91 (65). The minor isomer displayed the following 1H NMR 0.94 (5H, s), 0.99 (3H,s), 1.5-5.4 (15H. multiplet), 5.45 (1H, quartet, J=3.36 Hz), 6.65-6.7 (1H, dd, J=1.52 Hz and 10.4 Hz). 6.75- 6482 (1H, d, J=10.4 Hz). Alcid treatment of Diels Alder adduct; preparation of 11. 500 mg of the Diels Alder adduct 10 was dissolved in 2 ml o;f glacial acetic acid (hot) and to this was added two drops 021‘ conc. hydrochloric acid. It was then allowed to cool to room temparature. The white crystalline solid formed was suction 75 filtered, washed with water and dried to give 447 mg (90%) of a mixture of 7,8 and 8,9 double bond isomers in a ratio 80:20 as determined by proton NMR. M.Pt. 220-240 C; IR 5200— 2400, 1725, 1600 cm"; The major isomer displayed the following 1H NMR resonances. (CD5COCD5) 1.12 (8. 5H), 1.5-2.7 (m, 8H). 3.03 (ddd, 1H, J=2.5, 7.6 and 22.3 Hz), 3.3 (m, 1H), 3.43 (dt, 1H, J=5.2 and 22.3 Hz), 3.85 (2H), 5.85 (dt, 1H, J=5.2 and 2.5 Hz), 6.5 (s, 2H); Ms m/e (rel. abund.) 298 (100), 285 (57), 265 (26). Preparation of bismethoxy derivative 15 To 100 mg of the 1,4-dihydroxy derivative 11 in 8 ml of THF at 0 C was added 75 mg of 99% sodium hydride under an atmosphere of argon. After stirring for half hour, 0.5 ml of methyl iodide was added and the reaction mixture was allowed to warm up to room temparature. After stirring overnight at this temparature, it was quenched with water and extracted with ether. The ether layer was washed with water and then with brine, dried over anhydrous sulfate. Removal of the solvent gave 96 mg (91% yield), of colorless solid which was found to be a mixture of 15 and its 8,9 isomer in the ratio 80:20 as determined by high resolution NMR. An analytical sample of 15 obtained by HPLC (silica, methylene chloride) displayed the following properties. M.Pt. 124-128’0; IR 1730, 1600, 1475. 1250, 1075 cm"; 1H NMR 1.05 (s, 5H), 1.26 (s, 5H), 76 104-300 (m, 8H), 3013 (m, 2H), 305 (dt, 1H, J=5.5 and 2206 HZ), 5.78 (s, 6H), 5.78 (dt, 1H, J=5.5 and 2.5 Hz), 6.7 (br s, 2H); MS m/e (rel. abund.) 526 (100), 511 (51), 295 (11). Preparation of the bishydroxy derivative 17. One gm of the Diels Alder adduct 10 was dissolved in 5 ml of hot glacial acetic acid and to this was added 5 drops of concentrated hydrochloric acid. The reaction mixture was allowed to remain at room temparature overnight. The colorless crystalline product formed was suction filtered, washed with water and dried to give 950 mg of a mixture of 17 and its 8,9 isomer in the ratio 80:20 as determined by NMR. (yield=95%) The major isomer displayed the following properties. M.Pt. 220- 225 c (d); 1H NMR 0.76 (s, 5H), 0.94 (s, 5H), 1.4-3.0 (m, 11H), 6.50 (d, 1H, J=8.5 Hz), 6.55 (d, 1H, J=8.5 Hz), 7.18 (dt, 1H, J=2.6 Hz); MS m/e (rel. abund.) 298 (100), 285 (26), 265 (24). Preparation of the bismethoxy derivative 16. 500 mg of sodium hydride (60% in mineral oil) was washed three times with pentane under argon. To this was added a solution of 1.27 gm of the bishydroxy derivative 17 in 10 ml of THF. After stirring at room temparature for 50 min. the reaction mixture was cooled in an ice bath and 0.5 ml of methyl iodide was added. It was allowed to react overnight and then quenched with water. The reaction mixture was extracted with ether, washed with water and then with brine. It was dried 77 over anhydrous sodium sulfate and the solvent evaporated, to give 1.287 gm(100%) of a crude solid which was crystallized from ether. 1.057 gm (81%) of 16 was obtained as a colorless crystalline solid. The mother liquor (0.25 gm) was found to be mainly the required bismethoxy derivative on high resolution analysis. 16 displayed the following properties. M.Pt. 160-162 C; ‘H NMR 0.86 (s, 5H), 1.06 (s, 5H), 1.5.3.1 (m, 11H), 3.85 (s, 6H), 6.72 (d, 1H, J=9 Hz), 6.80 (d, 1H, J=9 Hz), 7.17 (brd, 1H, 6 Hz); MS m/e (rel. abund.) 326 (100), 311 (22), 293 (7), 269 (10). Preparation of the B-aromatic 1,4-diketo derivative 18 900 mg of a mixture of the bishydroxy derivatives 11 and 12(ratio 80:20 respectively) was refluxed in 50 ml of xylene with 200 mg of Pd/C(10%) for 10 hours. The reaction mixture was cooled, the catalyst was removed by filtration and washed with ' ether. The combined solvents were evaporated under reduced pressure to give 900 mg of a crude product. Crystallization from ethanol gave 450 mg(50%) of pure 18 which displayed the following properties. M.Pt=158-141 C., IR(CHC13) 1755,1685 cm'1; 1HNMR 0.8(s, 5H), 1.1(s, 5H), 1.8-2.8(m, 6H), 5.0f. 00 I. d. .I H4 lehc ' 1 I .00 I I I I. . II I III 1 11 Figure 18. Mass spectrum of 15 110 . 11.1 15 p 141111‘ I,J 1‘1’. Id .. 1 1 1I 5.43 N 5.4.4.3444: 5 n mp mo ssnpooam maz : oI—..4 1.1.I11II111:II11-II]I—I111I I111 .‘IIJIJ 11--.5 . 1414. .IJ I P I III II 141:. m 4 4-. .- 1-31 .. F1 1 4 .m. ohsmas h 4% cook n:0. .I‘I‘ 1‘ . o mmsmruucem) ——.———- —-.—-- 5.” U-b ' 0v B‘Iv IRANSMH’TANCEHS) 2000 1800 1600 1400 1200 Figure 20. Infrared spectrum of 18 19.0 1000 112 Figure 21. 1H NMR spectrum of 18 1 I 1 q p 219 3.0-1 201 4 2:1J ‘ 1 1 r ’1'- '1'. ’1‘- 11" 1“ Figure 22. Mass spectrum of 18 113 100 M a . A A L 0.1 U A u... ..L L L l 11 1 v.1 .. L 4L .0 L 0.1 J A 1500 2500 “‘1‘... .-.. o. 3500 2.5 100 80 w w m s: 2E0~ I. —- + f- -'- —O I T I ‘ 1 1 - - - .. LlLllllll -. . 60 -7- “.....- . ... ...-4 -..— .-. . ..- . A I I . ‘0. i . . ; 1°- TRANSMITTANCUQM - .. . . . . . . a 1 1 . . n . . - ' .‘ 1 I0 --- ..- -.— . .- .. . - -o.- ,., o 0 - I u . . . 1 1 s a . _. - y—.— -o ..— 1 . l s . . ' . - 4000 3500 3000 2500 ' 3.U U.U a .v 100 TRANSMITTANCE Hal Figure 43. Infrared spectrum of 31 IJ.U ILU 14.0 10.0 RD 40 20 127 Pm mo asppommm mzz m. .¢¢ mpzmflm — N n 1 ._1 14.11JI4 I 1J1 1111111 41 I 41 J 1114.114111141. 114.1 14.11 1.114111 41114111 Ill—1,114.1- 4|, .-114. a 1 1.1 4 all TI} 2111031) I? %£—..—>.. ' C -- ._'. I ... u. | I .owmeZ<—:¢ h : ova-,1 13.x" . . _ f . P.- ”‘71, .