《有机化学》课程教学资源（文献资料）Rearrangement of 1,2-Glycols, Halohydrins, and Azidohydrins

团购合买资源类别：文库，文档格式：PDF，文档页数：8，文件大小：1.12MB

VOLUME 41, NUMBER 14 0 Copyri ht 1976 by the American 8hemical Society JULY 9,1976 Molecular Rearrangements with Ethoxycarbonyl Group Migrations. 2. Rearrangement of 1,2-Glycols, Halohydrins, and Azidohydrins1*2 ,Jacques Kagan,' Dalmacio A. Agdeppa, Jr., David A. Mayers, S. P. Singh, Mary J. Walters, and Richard D. Wintermute Chemistry Department, University of Illinois, P.O. Box 4348, Chicago, Illinois 60680 Received September 29,1975 The ethoxycarbonyl group has been shown to undergo a facile 1,2 migration during the pinacol rearrangement of glyceric esters in fluorosulfonic acid. Pyruvic esters and 2-hydroxy-3-butenoic esters were also formed, and the prolonged treatment of some pyruvic esters in fluorosulfonic acid led to their isomerization into p-keto esters. The product of ethoxycarbonyl group migration was also found in the treatment of an azidohydrin with nitrosonium tetrafluoroborate, and that of halohydrins with silver salts. The rearrangement of tert- butyl 2-hydroxy-3-chloro- 3-phenylbutyrate (24) without fragmentation upon treatment with silver carbonate argued against a mechanism involving a carboxylium ion, and in favor of a process in which the alkoxycarbonyl moiety acts as an internal nucleophiie. The migration of a carboethoxyl group in preference to hydrogen, alkyl, or aryl groups in the well-known pinacol rearrangement has not been previously noted,3 but we were encouraged to look for it by its discovery in the acid-catalyzed isomerization of epoxides.ll4 Glyceric esters 1, obtained by mild hydrolysis of glycidic esters (I) or by hydroxylation of OH OH II - COOEt Ph-C-C-COOEt Ph -7 - 7 II R I he k ke R I 1 COOEt I OH I -CCKOOEt Ph-C-COR Ph-C-C-COOEt II I CH, R 2 3 4 I ph-(i Me Me Ph-C-C-COOEt 5 6 bR=H m,R=Me e,R=Et f,R = Ph cinnamic esters, were treated with sulfuric acid, alone or in acetic acid, trifluoroacetic acid, or boron trifluoride, with poor results. For example, treatment of lm with sulfuric acid in acetic acid yielded the allylic alcohol 4m with only minor amounts of 3-phenyl-2-butanone and of the desired rearrangement product 3m. Fluorosulfonic acid used as solvent was found to be a most convenient catalyst for the desired pinacol rearrangement, and the standard procedure in this work consisted in dissolving 1 in pure fluorosulfonic acid, both having been precooled to 0 "C, pouring the reaction mixture over ice after 3 min at this temperature, and extracting with carbon tetrachloride. The crude extract was free from starting material. The experimental results, presented in Table I, were very reproducible. They were generally similar to those obtained in the reaction of the related glycidic esters with boron trifluoride,ls4 and the same sequence of apparent migratory aptitudes was deduced. The allylic alcohols 3 were not detected, but independent treatment of 4m, 4e, and 4f under the reaction conditions confirmed their reactivity and their partial conversion into the lactones 5m and 5e and the indene 6, respectively. The synthesis of indenes from phenyl substituted allylic cations has been previously de~cribed.~ The intramolecular Friedel-Crafts reaction requires the allylic-benzylic cation to have the geometry shown in E-7. Whether or not the.other E-7 2-7 isomer was formed was not established directly. The observed cyclization to 6 from each diastereoisomer of If in the presence of boron trifluoride agreed with a facile interconversion of both cation^,^ but the absence of the lactone 5f indicated a slow rate of cyclization from 2-7. Note that the indene 6 had not been observed in the boron trifluoride catalyzed rear- 2355

2356 J. Org. Chem., Vol. 41, No. 14,1976 Kagan, Agdeppa, Mayers, Singh, Walters, and Wintermute Table I. Pinacol Rearrangement of Glyceric Esters with FSOIH at 0 "C for 3 min (Yields from NMR Analyses) % products Precursor 2 3 5 Others lh u 6.5 a U lm b 84 16 (threo + threo- lm b 84 16 erythro- lm b 84 16 lm-ds b 90 10 (threo + le (threo + 9 75 9 If (threo + 60 13 6 (12),8 (6),9 (9) erythro) erythro) erythro) erythro) Too complex for analysis. Trace. rangement of If, which yielded a single product, resulting from phenyl group migration, further indicating that 4f was not formed under the conditions of that reacti0n.l The dehydration of both isomers of 1m into an almost single rearranged product 3m (Table I) contrasted with the results of the boron trifluoride treatment of Im, from which the three rearranged products 2m, 3m, and 4m were 0btained.l However, the isomerization of lm into the allylic alcohol was the preferred process in weaker acids, such as trifluoroacetic or dilute fluorosulfonic acids, and the treatment of Im with pure fluorosulfonic acid under conditions identical with those used with lm gave a somewhat different product distribution, 52% of 3m, 48% of 5m, and a trace of 2m. The migration of the phenyl group from the 3 to the 2 position would also have given 3m from lm, but this process was ruled out by the study of lm-3-methyl-d~, which yielded 3m-d3 with no deuterium at the acetyl position, a result which agrees with the behavior of 1m.l The acid treatment of If yielded two unexpected products, ethyl 2,2-diphenylacetoacetate (8) and acetophenone (9), along with 2f, 3f, and 6 (Table I). This reaction contrasted with that of If with boron trifluoride, which yielded only 2f,2 and where the allylic alcohol 4f could not have been an intermediate, since its treatment with boron trifluoride yielded the indene 6 as the sole product. Shorter reaction times did not have a marked effect on the yield of 9, but reduced the formation of 8, and these studies also demonstrated the very fast rate of this pinacol rearrangement, which was complete in less than 15 s at -5 "C and where only 6% of starting material was left after 15 s at -50 "C. The formation of acetophenone described above must have taken place in the reaction of the starting material, since the reaction products did not yield it when submitted to the treatment with fluorosulfonic acid. The idea that an air oxidation reaction had taken place, formally the reverse of a pinacol synthesis, was dismissed by checking that bubbling air through a solution of If did not lead to any detectable fragmentation. The most reasonable mechanism for the fragmentation of glyceric esters into carbonyl products involves a protonation at the ester carbonyl, followed by a retro-aldol cleavage reaction. This type of acid-catalyzed /3 fission had not 07H RZ II + PhCOR, + R2-CH-COOEt I R3 Table 11. Isomerization of 2f Catalyzed by FSOBH at 0 "C -_ I_ Time, min 2f 3f 8 5 79 15 6 10 71 21 8 20 50 33 17 40 28 50 22 80 16 57 27 120 8 64 28 150 5 65 30 180 2 71 27 210 0 71 29 Table 111. Isomerization of 2m Catalyzed by FSO3H at 0 "C Time, min 2m 3m I1 3 95 3 2 90 48 32 20 180 19 46 35 300 5 55 40 450 0 56 44 been previously reported with esters, although it was known for carboxylic acids6 and ketone^.^ We also found that 2,3- diphenyl-3-hydroxybutyric acid yielded acetophenone and methyl phenylacetate upon attempted esterification with methanol catalyzed by sulfuric acid. However, a control experiment disclosed that fragmentation had taken place prior to exterification, since the ester formed by reaction with diazomethane did not undergo the fragmentation reaction when it was treated in the same acidic conditions. We previously reported on the acid-catalyzed formation of benzophenone from ethyl 3,3-diphenylpyruvate, which we tentatively ascribed to a retro-pinacol rearrangement to a glyceric ester, followed by the above fi fission.8 Later work on the rearrangement of chlorohydrins, which also led to fission? attracted our attention to the facile autoxidation of ethyl 3- phenylpyruvates, a reaction which best explains the formation of 9 in this case. In the effort to establish whether acetophenone had been formed by reaction of the starting material or of the reaction products, 2f was treated in the usual reaction conditions. To our surprise, it was converted into two other products, and acetophenone was not formed (Table 11). These two products were purified by preparative TLC and identified as 3f and 8, the @-keto esters isomeric with 2f, which were also observed in the direct reaction of If. Ph,-C - COCOOEt I Me 2f Ph Ph - FSOjH I I PhCO-C-COOEt + MeCO-C-COOEt I Ph I Me 3f 8 We do not know the exact mechanism of the a-keto to 6- keto ester rearrangement, for which there are at least two alternatives, differing by whether or not the ethoxycarbonyl group migrates,8 and we still hope to see this problem elucidated in the near future, by studying the rearrangement of properly labeled precursors. The generality of the fluorosulfonic acid catalyzed rearrangement of a-keto to @-keto esters has not been fully established, but we found that 2m [the main product formed in

Rearrangement of 1,2-Glycols, Halohydrins, and Azidohydrins J. Org. Chem., Vol. 41, No. 14,1976 2357 Table IV. Solvolysis of 17 Solvent Temp,OC Time,h Agsalt Ih 2h 3h 4h lh 17 Other None 132 22 None 50 50 Nonea 132 63 None 100 None 132 63 None 25 44 50% EtOH r.t. MeOH r.t. 4 Nitrate C6H6 r.t. 21 Carbonate 60 5 14 18 Acetone r.t. 21 Carbonate 70 10 14 5 MezSO rat. 21 Carbonate 17 3 HMPTA rat. 21 Carbonate c c0 b CCll r.t. 41 Carbonate 68 2 19 11 THF rat. 41 Carbonate 35 2 10 3 CH3CN r.t. 41 Carbonate 80 2 16 2 Hexanes r.t. 65 Carbonate 50 2 28 20 Ether r.t. 65 Carbonate 25 5 9 20 MezSO r.t. 96 Carbonate c co c DMF r.t. 96 Carbonate c cb C 88% HCOOH ' r.t. 16 None AcOH r.t. 16 None AcOH 118 16 None 50 Ether -10 1.5 Triflate 100 Ether 0 2 Tosylate C6H6 ret. 20 Tosylate C c 0 29 C6H6 r.t. 46 Mesylate 80 a In the absence of air. Detected by NMR. Mixture too complex for NMR analysis. the acid treat:ment of ethyl 2-phenyl-3,3-dimethylglycerate (lo)] did rearrange to yield the isomeric P-keto esters 3m and 11 (Table 111). OH OH II I MezC-C-COOEt -+ Me&-COCOOEt I Ph Ph 10 2m Me I Me I -+ MeCO-C-COOEt + PhCO-C-COOEt I I Ph 3m Me 11 Finally, in ,the course of this work glyceric esters having different subetitution patterns at the 3 position were also treated with fluorosulfonic acid, and the results are shown here: COOEt I I R1 OH OH R, II I II I R1 R2 PhC-CCOOEt - PhCCOCOOEt + PhCCOR, + PhCOR1 R, 12,R,=Ph;Rp==H 88% 0 12% 13, R, = H, R, = Me 0 low0 0 14,Rl=H;R,=Ph 95% 0 5% Ethoxycarbonyl Group Migration from an Azidohydrin Precursor. Following the report of easy generation of carbonium ions by the nitrosonium tetrafluoroborate decomposition of azides,1° the procedure was utilized with ethyl 2-hydroxy-3-azido-3-phenylbutyrate (E), which was readily obtained by treating Ih with hydrazoic acid. N, I "2 I Ph- C- CHOHCOOEt Ph- C- CHOHCOOEt I I 9 (Tr) 9 (31) 23 (100) 22 (100) 4 80 C bc 50 40 C C C C 100 100 AcO-4h (50) 100 0 C 14 14 (6) There was no reaction when 15 was treated with an excess of the salt at room temperature in acetonitrile, benzene, or dimethoxyethane for up to 60 h. The starting material disappeared completely upon refluxing for 6 h in benzene, and the product of ethoxycarbonyl migration was detected in 18% yield, along with a trace of 2h and 10% of the allylic alcohol 4h. Acid-catalyzed decarbonylation and condensation reactions were probably responsible for the low yield of identified products, a situation similar to that observed with Ihl and lh. The identification of 3h in this reaction mixture was sufficient proof that the ethoxycarbonyl grohp could migrate when the azohydrin was used in the generation of the initial cation, and no further work was devoted to this reaction. The synthesis of the amino alcohol 16 proved to be much more difficult than anticipated.ll The treatment of the epoxide Ih with ammonia or sodamide did not yield 16. Ring opening took place when Ih was treated with benzylmine, but the nitrogen was then doubly benzylic, and the hydrogenolysis cleaved the bond between the nitrogen and the adjacent tertiary carbon, rather than the desired one. Acid-catalyzed reactions, such as the condensation with a nitrile to form the required carbon-nitrogen bond, instead resulted in the dehydration into 3h. The reduction of 15 with sodium borohydride gave a low yield of 16, which was purified with difficulty. Upon diazotization of this product with isoamyl nitrite, a very complex mixture was generated, where 3h was absent as judged by NMR. Further work in this area will have to await the development of a convenient synthesis for 16. Ethoxycarbonyl Group Migrations from Halohydrins. Because of the ease with which the product of rearrangement with ethoxycarbonyl group migration could be detected by the NMR analysis of the aldehyde proton, much work was devoted to the chlorohydrin 17, derived from Ih by treatment with hydrogen chloride in ether, and which was a 1:l mixture of diastereoisomers. The solvolysis of 17 in 88% formic acid at room temperature for 16 h resulted in hydrolysis to the glyceric ester lh. No reaction was observed under the same conditions in acetic acid, but following reflux in this latter solvent for 16 h, the allylic alcohol 4h and its acetate were formed in equal amounts. Me 15 Me 16

2358 J. Org. Chem., Vol. 41, No. 14,1976 Kagan, Agdeppa, Mayers, Singh, Walters, and Wintermute Table V. Dehydrochlorination of 20 and 21 in the Presence of Silver Carbonate Compd Solvent Time, h Temp, "C COOCzHh migr Epoxide Starting material Others 20 C6H6 Ether DMF DMF CHsCN CHsCN Acetone 21 DMF C6H6 15.5 16 15 40 15.5 24 15 66 16 65 8 r.t. a r.t. 0 15 r.t. 82 r.t. a 68 r.t. 0 a Identified by NMR, but spectrum too complex for analysis. The reaction of 17 was then carried out in a variety of solvents. Nucleophilic displacement of the chloride, but no rearrangement, was the rule in hydroxylic solvents. In aprotic CI I I Ph-C-CHOHCOOEt Me 17 X OH II I II Ph- CH- C - COOEt Ph- C-C- COOEt Me Me Me 18, x = ci 20, X = Rr 22,X = OMe 23,X=OEt 19, x = C1 21, X = Br solvents, the dehydrochlorination took place in all possible manners, yielding the allylic alcohol and the epoxide as well as the products of molecular rearrangement, the a-keto ester 2h, and the aldehyde 3h. Qualitative variations in the product distribution were observed as a function of the solvent chosen, and they are recorded in Table IV. The nature of the silver salt used also had an effect on the product distribution, and no 3h was detected with silver tosylate, mesylate, or triflate. After this series of experiments was completed, silver carbonate was utilized with other halohydrins, in order to ensure that no strong Bronsted acidity would be developed during the course of the reaction, thereby ensuring that the migration of the ethoxycarbonyl group would not actually involve a protonated ester moiety. No reaction was observed at room temperature with the chlorohydrins 18 and 19, but the corresponding bromohydrins 20 and 21 did undergo dehydrobromination in these conditions, yielding the epoxide and the product of ethoxycarbonyl group migration in each case, as well as the allylic alcohol in the case of 21 (Table V).12 These observations confirmed the preference for the migration of the ethoxycarbonyl group over the methyl which was previously encountered in the rearrangement of glyceric and glycidic esters under certain conditions. Much more work remains to be done to correlate the effect of the temperature and the stereochemistry of starting materials with the product composition. However, the present results are in line with those described for the rearrangement of the corresponding glycidic esters. The allylic alcohol 4 was the major product at low temperature. In the absence of electrophilic assistance, the product of methyl migration 2h was formed next, and the formation of the product of ethoxycarbonyl migration 3h, which has the highest activation energy, required electrophilic assistance. 61 31 a a a a 100 a a 100 100 a 0 4m (100) 4m,a 5mn Conclusion The simplest explanation for the pinacol and pinacol-like rearrangements described in this work starts with an ionization at the 3 position. We have no information yet on the lifetime of the 3-carbonium ion generated, compared to the rate of the subsequent migration of one group from the 2 to the 3 position, but it is anticipated to depend on the reaction conditions. The concertedness reported in other pinacol rearrangement~,~~ and the recently disclosed concerted rearrangement of the epoxide Ih to 3h,14 suggest that a group migration from the 2 to the 3 position could well take place synchronously with the departure of the leaving group, resulting in inversion of configuration at the 3 position. Formally, a distinction may be established between a process in which the carbonyl carbon remains with eight electrons throughout (path A) and that in which the bonding electrons are first attracted to the adjacent electron-deficient site (written here as a full cation for convenience), yielding an enol-carboxylium ion pair, which further undergoes the acylation reaction required by the structure of the final product (path B). P "-YF= Me OY Path A +I PathB Ph-C C COOR' + IV Me R - Ph-C=C-OY + C-OR' II 0 II Me R The fact that all the attempted syntheses of carboxylium ions have resulted in their decomposition with loss of carbon dioxide15 gave a strong presumption against path B. Additional support for this view was derived from the study of tert-butyl 2-hydroxy-3-chloro-3-phenylbutyrate (241, which was treated with silver carbonate in benzene. If path B had been operative, decomposition to carbon dioxide and isobutene would have been expected. Instead, the rearrangement with ester group migration was observed, indicating that little or no positive charge was generated at the ester carbonyl, and that the carbon-carbon bonding electrons act as an internal nucleophile in the rearrangement. Each diastereoisomer of 24 was treated under the same conditions. Although they yielded the same products, 25,26, and 27, considerable stereoselectivity was displayed and one isomer gave predominantly the aldehyde resulting from ester group migration (ca. 50% of the products) with only 25% of epoxide, while the other gave only 33% of this aldehyde and

Rearrangement of 1,2-Glycols, Halohydrins, and Azidohydrins J. Org. Chem., Vol. 41, No. 14,1976 2359 C1 OH COOCMe, I1 I Ph-()-$-COOMe3 - Ph-c-CHO II Me H I Me 24 25 OH I + Ph-C-CH-COOMe3 + Ph-C-CHCOOMe3 II I CH* Me Ph- CH- CHO 26 27 I Me 28 56% of the diaotereomeric epoxide (27). The formation of acetophenone in these dehydrochlorination reactions is interpreted as resulting from the autoxidation of tert- butyl 2- keto-3-phenylbutyrate: giving a measure of the extent of hydrogen migration from the 2 to the 3 position. In contrast, the treatment of 27 with boron trifluoride in benzene resulted in complete cleavage to 2-phenylpropionaldehyde (28), coupled with the alkylation of the solvent, a decomposition reaction reminiscent of the thermal process.16 Although further work is obviously needed, the ability of the alkoxycarbonyl group to undergo 1,2 migrations in a variety of pinacol-like rearrangements is now firmly established. The ready availability of glycidic and glyceric esters, as well as chlorohydrins, suggests that the introduction of a carboxylic ester function adjacent to a carbonyl group via a pinacol-like rearrangement may occasionally compete with current methods based on nucleophilic reactions of enolate ions. From the practical viewpoint, a competition between two different substituents is usually the rule for the migration from the carbinol to the adjacent cationic position in these rearrangements. We previously demonstrated that a judicious choice of the reaction parameters often provided an effective control over the formation of the primary products.l Some selectivity over the subsequent isomerization of these products is also available, and may even lead to products not directly accessible from the starting material, such as 8 from If. Further work will be directed at gaining a more effective control over the selection of the group which migrates in these competitive rearrangement reactions. Experimental Section Rearrangement of Ethyl 3-Methyl-3-phenylglycerate (lh). Fluorosulfonic acid (1 ml) precooled to 0 OC was added dropwise with stirring to 472 mg of the diastereoisomer of lh melting at 93-94 OC.1' The reaction mixture was stirred for 3 min and poured onto an icewater-carbon tetrachloride mixture. After separating the organic layer and extracting the aqueous phase twice with 50 ml of Cc4, the com- bined organic layers were washed with two 50-ml portions of 5% NaHC03, dried, and concentrated. The NMR (CC14) was very complex, but showed the aldehyde signal of 3h at 9.90 ppm. Preparative TLC on silica gel, developed successively with EtOAc-petroleum ether (1:9) and EtOAc-CCL (1:9), led to many overlapping bands. The only product isolated in pure form was ethyl 2-formyl-2-phenylpropionate (3h, 20 mg), identical with an authentic sample. The same results were observed when either the lower melting or a mixture of both diastereoisomers of 111 was treated as above. Ethyl 2,3-Dixnethyl-3-phenylglycerate (lm). The hydration of 2.0 g of Im with 30% by weight of perchloric acid in 50% aqueous THF for 16 h at room temperature led, after workup, to 1.5 g of lm, as a viscous oil which was purified by chromatography over silica gel. A mixture of both diastereoisomers was obtained NMR (CDCl3) 7.0-7.6 (br, 5 H), 4.07 ((1, J = 7 Hz, 2 H), 3.55 (br, 2 H), 1.65 (s, 3 H), 1.40 (s, 3 H), and 1.13 ppm (tr, J = 7 Hz, 3 H) for one isomer, and 7.0-7.6 (m, 5 H), 4.03 (9, J := 7 Hz, 2 H), 3.55 (br, 2 H), 1.60 (s, 3 H), 1.43 (s,3 H), and 1.08 ppm (tr, J = 7 Hz, 3 H) for the other. The signal at 3.55 ppm disappeared when D20 was added. The fraction collected before lm was 4m, obtained in 10% yield, and identified by comparison of the NMR with the known material. The same procedure was used to hydrate 400 mg of Im-3-hfe-d3, yielding 400 mg (92%) of ethyl 2- methyl-3-trideuteriomethyl-3-phenylglycidate (lm-da), which was 72% deuterated from the NMR analysis. Rearrangement of lm. A 576-mg sample of lm was treated with 1.15 ml of FS03H at 0 "C for 3 min, poured into an ice-water-CC14 mixture, and worked up as usual. The crude reaction products were fractionated by preparative TLC using EtOAc in petroleum ether (once with 6% and twice with 4% v/v). Three bands were visible under uv light. Ethyl 3,3-dimethyl-3-phenylpyruvate (2m), an oil (3.7 mg, 0.7%), was isolated from the fastest moving band. Its NMR (CCLJ and mass spectra were superimposable onto those of the FSOsH-catalyzed rearrangement of ethyl 3,3-dimethyl-2-phenylglycerate (10). Ethyl 2-methyl-2-phenylacetoacetate (3m, 418 mg, 78%), an oil, was obtained from the second band. Its NMR (CCL) and mass spectra were superimposable onto those of the major product of the BFs-catalyzed rearrangement of ethyl 2,3-dimethyl-3-phenylglycidate (Im). The third band yielded 59.5 mg (14%) of 2-methyl-3-phenyl-4-hydroxy- 2-butenoic acid lactone (5m): mp 120-122 "C (lit. mp 121-122 OW); A, (cyclohexane) 211 and 260 nm; NMR (CDCl3) 7.45 (s,5 H), 5.04 (9, J = 2 Hz, 2 H), and 2.12 ppm (tr, J = 2 Hz, 3 H). The NMR analysis of the mixture before separation showed 84% of 3m and 16% of 5m, the signal at 1.18 ppm for 2m being barely noticeable. The same treatment was supplied to 300 mg of lm-ds, which yielded 2 mg of 2m-d3,209 mg of 3m-d3 in which the singlet at 1.67 ppm integrated for only 0.67 H, and 15 mg of 5m-d2, having a full methyl at 2.12 ppm. Rearrangement of threo- and erytbro-lm. The above reaction was repeated using 133 mg of either threo- or erythro- lm in 0.26 ml of FS03H. After workup, 107 and 90 mg of an oil were obtained, respectively, each analyzing by NMR for 84% 3m, 16% 5m, and a trace of 2m. Rearrangement of Ethyl 2,3-Dimethyl-3-phenylglycidate (Im). Treatment of 200 mg of Im with 0.43 ml of fluorosulfonic acid for 3 min at 0 OC and workup yielded 140 mg of an orange semisolid product which contained 52% 3m and 48% 5m plus a trace of 2m (NMR analysis). Reaction of lm in Dilute Fluorosulfonic Acid. Qualitative tests were performed using 200 mg of lm in (a) 0.4 ml of acid and 0.05 ml of water, (b) 0.3 ml of acid and 0.05 ml of water, and (c) 0.3 ml of acid and 0.1 ml of water. The major products were 3m (along with some 4m and 5m), 4m (along with some 3m), and 4m, respectively. In the last two experiments some starting material remained. Ethyl 2-Ethyl-3-methyL3-phenylglycerate (le). A solution of 482 mg (2.06 mmol) of E-Ie in 2 ml of tetrahydrofuran was added to 25 ml of perchloric acid in 50% aqueous tetrahydrofuran (30% by weight). The reaction mixture was heated in an oil bath at 45 OC for 1 h, diluted with 150 ml of water, and extracted with two 50-ml portions of ether. The combined ether extracts were washed with two 50-ml portions of 5% aqueous sodium bicarbonate, dried, and concentrated. Purification of the product mixture by preparative TLC, using 7.5% ethyl acetate in petroleum ether, yielded 385 mg (74%) of le, a thick oil, as a 1:l mixture of threo and erythro isomers: NMR (CC14) 7.0-7.6 (m, 10 H), 4.02 (q, J = 7 Hz, 2 H) and 4.13 (9, J = 7 Hz, 2 H), 3.33 (br s, exchanged with DzO, 4 H), 1.5-2.1 (m, 4 H), 1.58 (s, 3 H) and 1.5 (s, 3 H), 1.18 (tr, J = 7 Hz, 3 H), and 0.5-0.85 ppm (m, 6 H). Ethyl 2-ethyl-3-phenyl-2-hydroxy-3-butenoate (4e, 43 mg, 9%) was also obtained from the first band and was identified by NMR (CCL): 7.20(~,5H),5.45(d,J= lHz,lH),5.14(d,J=1Hz,1H),4.12(q, J = 7 Hz, 2 H), 3.28 (s, 1 H, exchanged with DzO), 1.90 (9, J = 7 Hz, 2 H), 1.12 (tr, J = 7 Hz, 3 H), and 0.90 ppm (tr, J = 7 Hz, 3 H). Rearrangement of Ethyl 2-Ethyl-3-methyl-3-phenylglycerate (le). This compound (385 mg, 1.53 mmol) was treated with 0.73 ml of FS03H at 0 OC for 3 min and worked up as described above. NMR analysis of the mixture showed 75% 3e, 9% 2e, and 16% 5e. These components were separated by preparative TLC developing the plates four times with 5% ethyl acetate in petroleum ether. Three bands were seen under uv light. The bands were scraped off and extracted with 10% ethyl acetate in chloroform. The lactone 5e was obtained in 14% yield (40 mg) from the slowest moving band as prisms, mp 79 OC when recrystallized from CC14- petroleum ether: NMR (Cc4) 7.50 (s,5 H), 5.02 (2 H, tr, J = 1 Hz), 2.22-2.75 (d, q, J = 1 and 7 Hz, 2 H), and 1.21 ppm (tr, J = 7 Hz, 3 H); mass spectrum mle 188 (M+), 187,159,143,129,128,115 (base peak), 91, 77, and 29. The second band yielded 203 mg (58%) of 3e identified by com- parison of the NMR and mass spectra with those of the known comp0und.l

2360 J. Qrg. Chem., Vol. 41, No. 14,1976 Kagan, Agdeppa, Mayers, Singh, Walters, and Wintermute The fastest moving band yielded 29 mg of an oil which was purified further by preparative TLC (the plate was developed once in 3% ethyl acetate in petroleum ether, twice in 2% ethyl acetate in petroleum ether, once in 2% chloroform in carbon tetrachloride, and once in carbon tetrachloride) to give 11 mg of products which were not identified and 18 mg (5%) of 2e, identified by NMR and mass spectral comparisons. Ethyl 2,3-Diphenyl-3-methylglycerate (lf). Perchloric acid (120 ml, 30% by weight in 50% aqueous tetrahydrofuran) was added to a solution of 3.5 g (0.0124 mol) of ethyl 2,3-diphenyl-3-methylglycidate (If) in 2.5 ml of tetrahydrofuran. The reaction mixture was stirred at 60-65 "C (oil bath) for 20 min, poured onto ice, diluted with 300 ml of water, and extracted with three 100-ml portions of ether. Washing and drying of the ether extracts and removal of the solvent yielded an oily residue. Crystallization from ether-petroleum ether yielded 1.73 g (48%) of a 1:l mixture of threo- and erythro- lf, white needles: mp 101-105 "C; NMR (Cc4) 7.0-7.9 (m, 20 H), 4.27 (9, J = 7 Hz, 2 H), 4.09 (q, J = 7 Hz, 2 H), 3.7 (s, exchanged withDzO,4 H), 1.67 (s, 3 H) and 1.42 (s, 3 H), 1.26 (tr, J = 7 Hz, 3 H) and 1.12 ppm (tr, J = 7 Hz, 3 H). Anal. Calcd for C18H2004: C, 71.97; H, 6.72. Found C, 71.91; H, 6.54. Rearrangement of If. A. Treatment of 500 mg of If with 0.8 ml of FS03H at 0 "C for 3 rnin and workup as above yielded a mixture containing (NMR analysis) 6096 2f, 13% 3f, 12% 6,6% 8, and 9% 9. This mixture, an oil, was treated by preparative TLC, developed four times in 1:3 benzene-carbon tetrachloride. Three almost overlapping bands seen under a uv light were scraped off and extracted with 1oOh EtOAc in CHC13. The first band yielded 49 mg of an oil which was a mixture of 2f (5%) and 6 (6%), identified by GLC-mass and NMR spectra. Pure 2f (215 mg) was obtained from the second band to yield a total of 51%. The third band yielded a mixture of 3f (38 mg, 8%), 8 (19 mg, 4%), and 9 (10 mg, 5%), after further TLC with 0.5 and 0.3% EtOAc in petroleum ether. The identification of 6 was based on the NMR (cc14) at4.21 (q,J = 7 Hz, 2 H), 3.77 (s,2H), 1.15 (tr,J = 7Hz, 3H), in addition to the aromatic protons between 7.00 and 7.50 ppm. The GLC-mass spectrum had signals at mle 264 (M+), 191 (base peak), and 189. The NMR (Cc4) of 3f was at 6.95-7.8 (m, 10 H), 4.05 (q, J = 7 Hz, 2 H), 1.78 (s, 3 H), and 0.98 ppm (tr, J = 7 Hz, 3 H), and its mass spectrum at mle 282 (M+), 239 (base peak), 194,166,77, and 43. The NMR (CCld) of 8 was at 7.00-7.80 (10 H), 4.18 (9, J = 7 Hz, 2 H), 2.05 (s, 3 H), and 1.17 ppm (tr, J = 7 Hz, 3 H), and its mass spectrum showed peaks at rnle 282 (M+), 239 (base peak), 194,166,77, and 43. The treatment of If (50 mg) with concentrated FS03H (0.08 ml) at -50 "C for 15 s, followed by the usual workup, yielded an oil which contained 9096 2f, 4% 9, and 6% unreacted If (NMR analysis). Various experiments were run at -5 "C with reaction times ranging from 4 to 45 s; 2% of starting material was left after 8 s, and none after 15 s. In this last case the product distribution was 91% 2f, 4% 3f, and 5% 9. Rearrangement of If with Dilute FS03H. A 75-mg sample of If was treated for 3 min at 0 "C with 0.15 ml of acid and 0.05 ml of water. The reaction mixture was poured onto an ice-water-CC14 mixture and extracted with CC4. After drying and concentrating 65 mg of oil was obtained, containing 97% of 2m and 3% of 9 (NMR analysis). Ethyl 3,3-Dimethyl-2-phenylglycerate (10). A mixture of 2.5 g (22.36 mmol) of ethyl 3,3-dimethyl-2-phenylglycidate and 150 ml of 30% aqueous perchloric acid was allowed to stand at room temperature for 2 weeks, and extracted with two 150-ml portions of ether. The combined ether extracts were washed with 5% aqueous bicarbonate, dried, and concentrated. The oily residue was purified by column chromatography to yield 0.963 g of 10 (35%), a viscous oil which crystallized on standing: NMR (MezSO-de) 7.2-7.9 (m, 5 H), 5.50 (s, 1 H), 4.42 (s, 1 H), 4.24 (q,J = 7 Hz, 2 H), 1.22 (tr, J = 7 Hz, 3 H), 1.17 (s, 3 H), and 1.10 ppm (s, 3 H). The signals at 5.50 and 4.42 ppm disappeared upon addition of DzO. Rearrangement of 10. A 440-mg sample was treated with 0.88 ml of FS03H for 3 min at 0 "C and worked up as above to yield 353 mg (87%) of ethyl 3,3-dimethyl-3-phenylpyruvate (2m): NMR (CCL) 7.18 (s, 5 H), 4.0 (q, J = 7 Hz, 2 H), 1.58 (s, 6 H), and 1.07 ppm (tr, J = 7 Hz, 3 H); mass spectrum m/e 220 (M+), 119 (base peak), 91, and 77. It gave a 2,4-DNP derivative in 63% yield after recrystallization: mp 153.5-154.5 "C; NMR (CDC13) 13.9 (br s, 1 H), 9.15 (d, J = 2 Hz, 1 HI, 8.52 and8.35 (dofd,J = 2,9 Hz, 1 H), 8.16 (d,J = 9Hz, 1 H), 7.28 (9, 5 H), 4.12 (q, J = 7 Hz, 2 H), 1.72 (s, 6 H), and 0.95 ppm (tr, J = 7 Hz). Anal. Calcd for C19HzoN406: C, 56.99; H, 5.04; N, 13.99. Found: C, 57.26; H, 5.01; N, 14.22. Ethyl 3,3-Diphenylglycerate (12). A solution of 2.000 g of ethyl 3,3-diphenylglycidate in 75 ml of 30% perchloric acid in 60% aqueous tetrahydrofuran was heated with stirring in an oil bath at 65 "C for 25 min, diluted with 300 ml of water, and extracted with two 100-ml portions of ether. The combined ether extracts were washed with 50-ml portions of 5% aqueous bicarbonate and water, dried, and concentrated to yield an oil which crystallized on standing. Recrystallization from aqueous ethanol yielded 900 mg (38%) of 12 as white crystals: mp 131-132 "C; NMR (CDC13) 7.05-7.78 (m, 10 H), 5.08 (s, 1 H), 3.98 (q, J = 7 Hz, 2 H), 3.35-3.40 (br, 2 H, disappeared in DzO), and 0.92 ppm (tr, J = 7 Hz, 3 H). Rearrangement of 12. A 200-mg sample of 12 was treated with 0.34 ml of FS03H at 0 OC for 3 min. After the usual workup the crude product was purified by preparative TLC, developing the plates twice with 5% ethyl acetate in petroleum ether. Two bands were seen under uv light and yielded 14 mg (12%) of benzophenone and 100 mg (53%) of ethyl 3,3-diphenylpyruvate: NMR (CDC13) 7.27 (s, 10 H), 5.91 (s, 1 H), 4.11 (q, J = 7 Hz, 2 H), and 1.11 ppm (tr, J = 7 Hz, 3 H) (an additional small singlet at 10.32 and a triplet at 0.7 ppm revealed the presence of the enol tautomer); mass spectrum rnle 268 (M+), 195, 167 (base peak), and 77. These spectra were superimposable onto those of the major product of the boron trifluoride rearrangement of ethyl 2,3-diphenylglycidate.l Ethyl 2-Methyl-3-phenylglycerate (13). A solution of 4 g of ethyl (E)-2-methylcinnamate and 4.65 g of m-chloroperoxybenzoic acid in 100 ml of chloroform was refluxed for 25 h and washed with 10% aqueous sodium sulfite and with 5% aqueous bicarbonate. The organic layer was dried and concentrated to yield 4.1 g (95%) of ethyl (E)- 2-methyl-3-phenylglycidate, an oil: NMR (cc14) 7.3 (s, 5 H), 4.3 (8, 1 H), 4.2 (9, J = 7 Hz, 2 H), 1.27 (tr, J = 7 Hz, 3 H), and 1.25 ppm (8, 3 H). A mixture of 2 g of this product and 100 ml of 3096 aqueous sulfuric acid was stirred for 24 hat room temperature, and extracted with two 100-ml portions of ether. The combined ether extracts were washed with 5% sodium bicarbonate, dried, and concentrated to yield a viscous oil. After purification by column chromatography, there was obtained 1.5 g (69%) of a mixture of threo- and erythro-13, an oil: NMR (MezSO-de) 7.1-7.62 (m, 5 H), 4.6-5.6 (broad, 2 H, disappeared upon the addition of deuterium oxide), 4.82 (s, 1 H), 4.78 (s, 1 H), 4.18 and 4.11 (eachaq,J= 7Hz,2H),l.24andl.l8(eachatr,J=7Hz,3H), 1.18 and 1.05 ppm (each as, 3 H). Rearrangement of 13. A 500-mg sample of 13 was treated with fluorosulfonic acid at 0 'C for 10 min. After workup the product (435 mg) had the NMR spectrum (cc14) of ethyl 2-phenylacetoacetate as a mixture of the ketonic (76%) and enolic (24%) forms: NMR (Cc4) 14.17 (br s, 1 H), 7.08-7.50 (m, 5 H), 4.13 (4, J = 7 Hz, 2 H), 1.80 (s, 3 H), and 1.13 ppm (tr, J = 7 Hz, 3 H) for the enol; 7.08-7.50 (m, 5 H), 4.61 (s, 1 H), 4.15 (q, J = 7 Hz, 2 H), 2.07 (s, 3 H), and 1.22 ppm (tr, J = 7 Hz, 3 H) for the keto tautomer. Preparative TLC yielded 280 mg of this product and 101 mg of material with broad NMR signals, which appeared to be polymeric in nature. Ethyl 2,3-DiphenyIglycerate (14). A solution of 2 g of ethyl (E)-2,3-diphenylglycidate in 75 ml of 30% by weight perchloric acid in 65% aqueous tetrahydrofuran was heated in an oil bath with stirring at 65-70 "C for 15 min. The reaction mixture was poured onto 50 g of ice, diluted with 200 ml of water, and extracted with three 50-ml portions of ether. Washing the combined ether extracts with 5% aqueous bicarbonate followed by drying and evaporation of the solvent yielded 1.4 g of a semisolid residue. Crystallization from carbon tetrachloride yielded 330 mg (15.5%) of 14 as white needles, mp 135-137 "C. After recrystallization from benzene, the mp was 139- 139.5 "C; NMR (CDC13) 7.17-7.77 (m, 10 H), 5.5 (br s,l H), 4.44 (9, J = 7 Hz, 2 H), 4.12 and 3.0 (each a broad s which disappeared in the presence of D20,2 H), and 1.37 ppm (tr, J = 7 Hz, 3 H). Only one of the two possible diastereoisomers was obtained, judging from the NMR spectrum. Anal. Calcd for CITH1804: C, 71.31; H, 6.34; Found: C, 71.35; H, 6.33. Rearrangement of 14. A 192-mg sample of 14 was treated with FS03H (0.32 ml) at 0 "C for 3 min and worked up as described above. NMR showed the crude product to consist of 60% keto and 40% enol tautomers of ethyl 3,3-diphenylpyruvate. However, purification by preparative TLC on silica gel (5% ethyl acetate in petroleum ether) yielded 7.5 mg (6%) of benzophenone in addition to 153 mg (85%) of ethyl 3,3-diphenylpyruvate: NMR (CDC13) 7.27 (s, 10 H), 5.91 (s, 1 H), 4.11 (q, J = 7 Hz, 2 H), and 1.1 ppm (tr, J = 7 Hz, 3 H) for the keto tautomer; 8.5 (s, enol H), 7.27 (s, 10 H), 3.98 (q, J = 7 Hz, 2 H), and 0.72 ppm (tr, J = 7 Hz, 3 H) for the enol form. The keto:enol ratio was 9O:lO. This NMR was superimposable onto that of the major product of the boron trifluoride rearrangement of ethyl 2,3-diphenylglycidate. The mass spectrum had peaks at m/e 268 (M+), 195,167 (base peak), and 77. Ethyl 2-Hydroxy-3-phenyl-3-azidobutyrate (15). A mixture of 35 g of sodium azide, 50 ml of water, and 100 ml of ether was treated with 30 ml of concentrated sulfuric acid, maintaining the temperature below 10 "C. The ether layer was dried over CaClz and distilled. In

Rearrangement of 1,2-Glycols, Halohydrins, and Azidohydrins 65 ml of this HN3 solution, 6 g of Ih and 60 mg of p-toluenesulfonic acid were allowed to stand at room temperature for 3 days. After concentration under vacuum, 50 ml of ether was added, and the solution was washed with 25 ml of water, 25 ml of saturated aqueous NaHC03, and 25 ml of aqueous NaCl and dried. The crude product, 15, was used in further reactions. A sample distilled bulb-to-bulb at 75 "C (0.07 Torr). This material was contaminated with a trace of 4h (seen by NMR). Ir showed the characteristic signal at 2150 cm-l. NMR (Cc4) 7.10-7.45 (m, 5 H), 4.26 and 4.20 (each a I, 1 H), 4.02 (9, J = 7 Hz, 2 H), 3.65 (1 H, disappeared in the presence of 1120), 1.80 and 1.75 (each as, 3 H), 1.08 and 0.90 pprn (each a tr, 3 H). The ratio of diastereomers was 1:3.9. Reaction of 15 with Nitrosonium Tetrafluoroborate. A solution of 665 mg of 15 and 1.25 g of NO+BF4- in 20 ml of benzene was refluxed for 6 h, concentrated under vacuum, treated with water, and extracted with ether.1° After drying and concentration, NMR analysis showed 18% 3h, 10% 4h, and a trace of 2h. No starting material remained. No reaction was observed at room temperature for up to 60 h when 15 was treated with NO+BF4- in benzene, acetonitrile, or dimethoxyethane. Ethyl 2-Hyclroxy-3-chloro-3-phenylbutyrate (17). A 6.39-g sample of Ih cooled at -190 "C was treated with 75 ml of HC1-saturated ether, allowed to warm up to room temperature with stirring, and kept for 18 h. After concentration under vacuum, 50 ml of ether were added and the solution was washed with 25 ml of saturated aqueous NaC1, '25 ml of saturated NaHC03, and 25 mi of saturated NaC1, dried over MgSO4, and concentrated to yield 6.50 g of 17, mp 53-59 "C, a mixture of diastereomers (1:l). Anal. Calcd for C1zH15C102: C, 59.42; H, 6.22; C1,14.60. Found: C, 59.44; H, 6.22; C1, 14.58. NMR (CC14) 7.2-7.6 (m, 5 H), 4.50 and 4.34 (each ad, J = 8 Hz, 1 H),4.12and3.96(eachaq,J= 7Hz,2H),3.15and2.99(eachad,J = 8 Hz, 1 H), and 1.15 and 0.95 ppm (each a tr, J = 7 Hz, 3 H). Reaction of ]Ethyl 2-Methyl-3-phenyl-2-hydroxy-3-butenoate (4m) with Fluorosulfonic Acid. An 82-mg sample of 4m was treated with 0.178 ml of'FS03H at 0 "C for 3 min and poured onto a mixture of ice, water, and 20 ml of CC4, and the aqueous layer was extracted with 20 ml of ether. After washing each of the organic extracts with 20 ml of 5% aqueous sodium bicarbonate, they were combined, dried, and Concentrated to yield 60 mg of a semisolid product which was analyzed by NMR to contain 70% 3m, 30% 5m, and a trace of 2m. Reaction of Ethyl 3-Phenyl-2-hydroxy-3-butenoate (4e) with Fluorosulfonic Acid. At 0 OC for 3 min, 90 mg (0.384 mmol) of 4e was treated with 0.18 ml of FS03H. The mixture was poured onto a mixture of ice, water, and 20 ml of Cc4. The aqueous phase was extracted again with 10 ml of chloroform and the organic layers were combined, washed with two 15-ml portions of 5% aqueous NaHC03, dried, and concentrated to yield 50 mg of a semisolid residue, which contained 45.5% 3e, 7.5% 2e, and 47% 5e (NMR). Reaction of Ethyl 2,3-Diphenyl-2-hydroxy-3-butenoate (4f) with Boron Trifluoride. Boron trifluoride was bubbled through a solution of 40 nig of 4f in 2 ml of carbon tetrachloride for 10 min at 28 "C. After quenching the reaction with aqueous sodium chloride, the mixture was extracted with 10 ml of carbon tetrachloride and with 10 ml of chloroform. The combined organic phases were washed with 5% aqueous bicarbonate, dried, and concentrated. The NMR spectrum of the product (90% yield) was superimposable onto that of 2- phenyl-3-carbethoxyindene (6). Reaction of Halohydrins with Silver Salts. A mixture of 200 mg of the halohydrin and ca. 1.5 equiv of silver salt in 20 ml of solvent was stirred in the dark. The spent silver salt was filtered over Celite and the filtrate was concentrated and analyzed by NMR (Tables IV and V) . Ethyl 2-Hydroxy-2-methyl-3-chloro-3-phenylpropionate (18). A solution of 500 mg of ethyl (E)-2-methyl-3-phenylglycidate in 20 ml of hydrogen chloride saturated ether was allowed to stand at room temperature for 1 h, washed thrice with water, dried, and concentrated to yield 565 mg of 18. The major isomer (ca. 90%) had NMR (CCL) at 7.10-7.60 (5 H), 5.00 (8, 1 H), 4.05 (q,J = 7 Hz, 2 H), 3.50 (br, 1 H), 1.51 (s, 3 H), and 1.16 ppm (tr, J = 7 Hz, 3 H). The minor isomer had its corresponding signals at 7.10-7.60,5.20,4.20,3.50, 1.15, and 1.16 ppm, respectively. Ethyl 2-Hydroxy-2-methyl-3-bromo-3-phenylpropionate (20). A solution of 200 mg of ethyl 2-methyl-3-phenylglycidate in 7 ml of hydrogen bromide saturated ether was allowed to stand at room temperature for 44 h. After dilution with 15 ml of ether, it was washed thrice with water, dried, and concentrated to yield 787 mg of an oil, a mixture of the two diastereoisomers of 20: NMR (cc14) 7.00-7.50 (m, 5 H), 5.05 (9,l H), 3.94 (9, J = 7 Hz, 2 H), 3.58 (br, 1 H), 1.53 (s, J. Org. Chem., Vol. 41, No. 14,1976 2361 1 H), and 1.07 (tr, J = 7 Hz, 3 H) for the major isomer (80%), and 7.00-7.50 (m, 5 H), 5.30 (s, 1 H), 4.07 (9, J = 7 Hz, 2 H), 3.58 (br, 1 H), 1.72 (s,3 H), and 1.24 ppm (tr, J = 7 Hz, 3 H) for the minor isomer (20%). Ethyl 2-Hydroxy-2-methyl-3-chloro-3-phenylbutyrate ( 19). A solution of 250 mg of Im in 10 ml cf HCl-saturated ether was allowed to stand for 26 h at room temperature, washed several times with water, dried, and concentrated to give a mixture of 4m (16%) and 19 (84%) as a mixture of diasteisomers: NMR (CC14) 7.15-7.65 (m, 5 H), 4.02 and 4.12 (each a q, J = 7 Hz, 2 H), 3.56 (br, 1 H), 2.02 (s,3 H), 1.38 (s,3 H) and 1.16 ppm (9, J = 7 Hz, 3 H). Ethyl 2-Hydroxy-2-methyl-3-bromo-3-phenylbutyrate (21). A solution of 250 mg of Im (mixture of E and Z) in 10 ml of HBr-saturated ether was refluxed for 2 h, washed with water, dried, and concentrated to yield a mixture of 21% 4m and 79% 21 (a mixture of diastereoisomers). Similar results were obtained at room temperature for 1 h. The NMR of 21 (CCb) had signals at 7.1-7.8 (m, 5 H), 4.09 and 3.95 (each a q, J = 7 Hz, 2 H), 3.78 (br, 1 H), 2.29 and 2.30 (each as, 3 H), 1.53 and 1.48 (each as, 3 H), and 1.35 and 1.10 ppm (each a tr, J = 7 Hz, 3 H). Acid Treatment of tert-Butyl 3-Methyl-3-phenylglycidate (27). A 200-mg sample of 2716 (mixture of diastereoisomers) in 20 ml of benzene was treated with BF3 for 30 s at room temperature. The solution was washed with aqueous NaHC03 and saturated NaCl and dried over MgS04. The NMR (CCl4) showed a mixture of 28 at 9.70 (d, J = 1.5 Hz, 1 H), 6.9-7.3 (5 H), 3.45 (4, d, J = 7 and 1.5 Hz, 1 H), and 1.35 ppm (d, J = 7 Hz, 3 H), and tert-butylbenzene at 6.9-7.3 (4 H) and 1.26 ppm (s, 12 H). This was proved by GLC against authentic samples. Using an internal standard of p-nitrotoluene, the two components were shown to have been formed in ca. 25% yield each. The remainder probably had polymerized, as in the reaction of the ethyl ester analogue 1h.l tert-Butyl 3-Chloro-3-phenyl-2-hydroxybutyrate (24). A 300-mg sample of 27 was treated at room temperature for 30 min with 10 ml of anhydrous ether saturated with HCl. After concentration under vacuum, addition of CCL, and concentration again, 345 mg of residue was obtained, which was a mixture of threo- and erythro- 24. Pure samples of both diastereoisomers were obtained after fractional crystallization in several solvents. The isomer of mp 79-80 OC had NMR (Cc4) at 7.67-7.10 (m, 5 H), 4.30 (d, J = 7 Hz, 1 H), 3.03 (d, J = 7 Hz, 1 H, exchanged in DzO), 2.00 (s, 3 H), and 1.30 ppm (s, 9 H), and the isomer of mp 36-38 OC had NMR (CC14) at 7.33-7.17 (m, 5 H), 4.30 (d, J = 7 Hz, 1 H), 3.25 (d, J = 7 Hz, 1 H, exchanged in DzO), 1.97 (s,3 H), and 1.15 ppm (s,9 H). Reaction of 24 with Silver Carbonate. A mixture of 130 mg of 24,207 mg of silver carbonate, and 10 ml of benzene was refluxed for 23 h, cooled, and filtered through Celite. The filtrate was concentrated under vacuum and the residue analyzed by NMR in cc14. From the isomer of mp 79-80 "C, the mixture contained ca. 8% of 26,11% of 9, 51% of 25,25% of E-27 and 5% of 2-27. From the isomer of mp 36-38 OC the mixture contained 4% of 26,7% of 9,33% of 25,56% of E-27, and a trace of Z-27. The NMR spectra of the individual compounds in CC4 follow. 26: 7.15-7.50 (m, 5 H), 5.40 (m, 1 H), 5.35 (m, 1 H), 4.80 (br, 1 H), 3.10 (br, OH), and 1.22 ppm (s, 9 H). 25; 9.78 (s, 1 H), 7.15-7.55(s,5H),1.57(s,3H),and1.45ppm(~,9H).E-27:~~7.20-7.55 (m, 5 H), 3.18 (s, 1 H), 1.72 (s, 3 H), and 1.50 ppm (s, 9 H). Z-27:19 7.20-7.50 (m, 5 H), 3.38 (s, 1 H), 1.68 (s, 3 H), and 1.05 ppm (s, 9 H). The phenylhydrazone of 25 had mp 151 "C; mass spectrum mle 324, 267,223 (base peak), 130,92, and 77; NMR (Me&O-&) 7.65 (s, 1 H), 7.10-7.43 (m, 9 H), 2.75 (s, 1 H, exchanged with DzO), 1.73 (s, 3 H), and 1.43 ppm (s, 9 H). Acknowledgments. We are grateful to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for partial support of this research. Registry No.-erythro- le, 59069-63-5; threo- le, 59069-64-6; erythro- If, 59069-65-7; threo- lf, 59069-66-8; lh, 38082-38-1; erythro- lm, 59069-67-9; threo- lm, 59069-68-0; erythro- lm-d3, 59069-69-1; threo- h-&, 59069-70-4; 2f, 38491-43-9; 2m, 54934-36-0; 2m 2,4-DNP, 59069-71-5; 3f, 38491-44-0; 4e, 59069-72-6; 4f, 59069- 73-7; 4m, 59069-74-8; 5e, 59069-75-9; 5m, 1575-48-0; 6,59069-76-0; 8,38491-45-1; 10,59069-77-1; 12, 5461-98-3; erythro- 13,40707-70-8; threo- 13, 59069-78-2; 14, 59069-79-3; erythro- 15, 59069-80-6; threo- 15, 59069-81-7; erythro- 17, 59069-82-8; threo- 17, 59069-83-9; erythro- 18,59069-84-0; threo- 18,59069-85-1; erythro- 19,59069-86-2; threo- 19, 59069-87-3; erythro- 20, 59069-88-4; threo-20, 59069-89-5; erythro-21, 59069-90-8; threo-21, 59069-91-9; 22, 59069-92-0; erythro-24, 59069-93-1; threo-24, 59069-94-2; 25, 59069-95-3; 25 phenylhydrazone, 59069-96-4; 26,59069-97-5; E-27,21309-21-7; 2- 27

2362 J. Org. Chem., Vol. 41, No. 14,1976 Jones and Turbini 21309-20-6; erythro-28,93-53-8; E-Ie, 19464-94-9; If, 59069-98-6; Ih, 77-83-8; E-Im, 59069-99-7; Z-Im, 59070-00-7; Im-da, 59070-01-8; ethyl 3,3-dimethyl-2-phenylglycidate, 59070-02-9; ethyl 3,3-diphenylglycidate 5449-40-1; ethyl 3,3-diphenylpyruvate 6362-64-7; ethyl (E)- 2-methylcinnamate, 7042-33-3; ethyl (E)-2 methyl-3-phenylglycidate, 7141-24-4; ethyl ,2-phenylaqetoacetate, 5413-05-8; ethyl 2-phenylacetoacetate enol, 59070-03-0; ethyl (E)-2,3-diphenylglycidate, 7042-27-5; ethyl 3,3-diphenylpyruvate enol, 59070-04-1. References and Notes (1) Part 1: J. Kagan, D. A. Agdeppa, Jr., S. P. Singh, D. A. Mayers, C. Boyajian, (2) Preliminary communication: J. Kagan and D. A. Agdeppa, Jr., Helv. Chim. (3) C. J. Collins, 0. Rev., Chem. SOC., 14, 357 (1960). (4) S. P. Singh and J. Kagan, J. Am. Chem. Soc., 91, 6198 (1969). (5) C. U. Pittman, Jr., and W. G. Miller, J. Am. Chem. SOC., 95, 2947 (6) D. Ivanoff, N. Marecoff, and B. Amidjine, Bull. SOC. Chim. Fr., 1214 (1963). (7) H. 0. House, "Modern Synthetic Reactions", 2d ed, W. A. Benjamin, New (8) J. Kagan and D. A. Agdeppa, Jr., Helv. Chim. Acta, 55, 2255 (1972). and C. Poorker, J. Am. Chem. SOC., in press. Acta, 55, 2252 (1972). (1973). York, N.Y., 1972, p 636. (9) D. A.Mayers and J. Kagan, J. Org. Chem., 39, 3147 (1974). (IO) M. P. Doyle and W. Wierenga, J. Am. Cbem. SOC., 94,3896,3901 (1972). (1 I) A similar observation was made by S. P. McManus, C. A. Larson, and R. A. Hearn, Synth. Commun., 3, 177 (1973). (12) The rearrangement of halohydrins in the presence of silver carbonate was recently described by M. Fetizon, M. Golfier, M. T. Montaufier, and J. Rens, Tetrahedron, 31, 987 (1975), and M. Fetizon, J. M. Louis, and M. Goifier, Tetrahedron Lett., 1931 (1973). The formation of allylic alcohols was not descrlbed by these authors. (13) For one example, see G. Berti, B. Macchia, F. Macchia, and L. Monti, J. Chem. SOC. C, 3371 (1971). (14) R. D. Bach, J. Domagala, and J. Wemple, J. Am. Chem. Soc., in press. We are grateful to Professor Wempie for informing us of this work prior to its publication, (15) P. Beak, R. J. Trancik, and D. A. Simpson, J. Am. Chem. SOC., 91, 5073 (1969); D. N. Keviii, W. A. Reis, and J. B. Kevili, Tetrahedron Lett., 957 (1972); W. E. Dupy, H. R. Hudson, and D. A. Karom, ibid., 3193 (1972); D. N. Kevili, "The Chemistry of Acyl Halides", S. Patai, Ed., Wiley-interscience, New York, N.Y., 1972, p 381. (16) E. P. Blanchard, Jr., and G. H. Buchi, J. Am. Chem. SOC., 85, 955 (1963). (17) H. H. House and J. W. Blaker, J. Am. Chem. SOC., 80, 6389 (1958). (18) P. S. Steyn, W. J. Conradie, C. F. Garbers, and M. J. de Vries, J. Chem. Soc., (19) V. R. Valente and J. L. Wolfhagen, J. Ofg. Chem., 31, 2509 (1966). 3075 (1965). Valence Photoisomerization of 1-Ethoxycarbonyl-lH-azepine and Its Thermal Reversion. Quantitative Aspects Including Energy Surface Relationships Guilford Jones, II,* and Laura J. Turbini Department of Chemistry, Boston University, Boston, Massachusetts 02215 Received March 29,1976 Photolysis of 1-ethoxycarbonyl-1H-azepine (3) at 325-385 nm gives quantitatively the valence isomer, 2-ethoxycarbonyl-2-azabicyclo[3.2.0] hepta-3,6-diene (4). The quantum yield in benzene is 0.013, virtually unchanged in other solvents including n-propyl bromide and with the addition of triplet quenchers. Sensitization with fluorenone, benzophenone, or valerophenone does not lead to valence isomerization. Azepine 3 acts as a quencher of the photoelimination of valerophenone (k, = 7.0 X lo9 M-l s-l) and the phosphorescence of biacetyl (k, = 5.3 X lo8 M-' 9-l). Laser photolysis at 457.9 nm affects isomerization 3 - 4 with a quantum efficiency of 0.013. On the basis of absorption, sensitization, and quenching data for 3, energies of low-lying excited states are estimated (Es, = 60 i 1 kcal/mol and = 55 i 1 kcal/mol). Pyrolysis of 4 gives 3 in a clean, exothermic, first-order reaction at 113-143 OC in diglyme-dlr (E, = 28.7 kcal/mol, A = 1012.3 s-l) or in hexadecane (k = 4.60 X s-l, 127.5 OC). A mechanism for 3 - 4 and related photochemical isomerizations is suggested, with emphasis on the proximity of ground and excited state potential surfaces at diradical or biradicaloid geometries. For cyclic triene systems, charge separation in diradical species appears to be important in facilitating photochemical ring closure and thermal back reaction. The photochemistry of 1,3,5-cycloheptatriene (la), its heterocyclic analogues, 1H-azepine (lb) and oxepin (IC), and their derivatives has received considerable attenti0n.l A reaction of general importance is valence photoisomerization to bicyclic dienes 2; in many cases the process is thermally reversible. We became interested in the photoisomerization of 1-ethoxycarbonyl-1H-azepine (3), since the reportedld behavior of this system showed promise for the storage and conversion of radiant energy (in principle, a portion of solar 0 "(Xh 1 2 a, X=CH, b, X=NR c, x=o C0,Et I qOzEt 0 I I\ hrNh 3 4 energy). In particular, photolysis of 3, which absorbs light in the visible, gives 4 cleanly and photochromatically (with bleaching). Photoisomer 4 is kinetically stable but reverts to 3 in a thermal reaction which is uncatalyzed and apparently exothermic.ld We wish to provide quantitative details concerning this isomerization of a sort not generally available for the cyclic trienes. Th'e data allow characterization of interconverting states and suggestions for pathways for photochemical and thermal reaction. Results The ultraviolet spectrum of 32 displays a weak transition at 330 nm. This absorption, which is blue shifted in polar solvent and probably n,n* in character,3 tails into the visible with onset at about 480 nm and gives 3 its orange color. Luminescence was not observed for 3, either in benzene at room temperature or in EPA glass at 77 K. Irradiation of 3 at 325-385 nm gave 4 without the appearance of side products (>97% yield by GLC and NMR). Quantum yields for isomerization as a function of solvent and the presence of additives are shown in Table I. Notably, the photolysis was not appreciably altered by (1) moderate concentrations of potential

点击进入文档下载页（PDF格式）

已到末页，全文结束

点击下载（PDF格式）

浏览记录