A new cyclization leading to vicinal cyclopentanediones - American

Oct 3, 1983 - obtained which is under study. These results can be ex- plained by the consideration of the reactions pictured in eq 1-5. COCOOCH3...
0 downloads 0 Views 166KB Size
J. Org. Chem. 1984,49, 397-398 thietan-fused j3-lactams. Since the starting materials 1are easily obtained by acylation of thiobenzamides and the yields of these photoreactions are high, these reaction provides a useful synthetic method of some thietan-fused j3-lactams. Registry NO. la, 88295-93-6; lb, 88295-94-7; IC, 88295-95-8; Id, 88295-96-9; le, 88295-97-0; 2a, 88295-98-1; 2b, 88295-99-2; 2c, 88296-00-8; 2d, 88296-01-9; 2e, aa296-02-0; 3d, 7468-30-6; 4d (isomer l), 88296-03-1; 4d (isomer 2), 88315-26-8; CH2=C(CHdCOCl, 920-46-7; PhCSNHCHs, 5310-14-5; PhCSNHCH2CH3, 39203-76-4; PhCSNHCH(CH&, 5310-15-6; PhCSNHCHZPh, 14309-89-8; PhCSNHPh, 636-04-4. Masami Sakamoto, Yoshimori Omote Hiromu Aoyama* Department of Chemistry, The University of Tsukuba Sakura-mura, Ibaraki, 305 Japan Received October 3, 1983

397

of the known2 Michael addition product, 5, was obtained. From the mother liquor a small amount of 4-acetylspiro[cyclopentane-1,Y- [9HJfluorene]-2,3-dione, 6, wm obtained. The structure was established by IR, NMR, MS, and elemental analysis. This finding seemed important for two reasons: (1)a cyclization different from that which yields the known methyl 4-oxospiro[ cyclohex-2-ene- 1,9'-fluorene] -2carboxylate, 92 was at hand and (2) a new route to vicinal cyclopentanediones was dis~overed.~ When the reaction of 1 with 2 was carried out with 1 equiv of t-BuOK, a 57 % yield of 6 and a very small amount of 5 was obtained. Furthermore, when 5 was treated with 1equiv of t-BuOK, a complicated mixture of products was obtained which is under study. These results can be explained by the consideration of the reactions pictured in eq 1-5. yCOCOOCH3

t t-BuOXCDCOOCH3 H

(or B-)

e

A-

1

A New Cyclization Leading to Vicinal Cyclopentanediones

Summary: Reaction of methyl 9-fluoreneglyoxylate with methyl vinyl ketone, acrylonitrile, and phenyl vinyl sulfone in the presence of 1 equiv of potassium tert-butoxide produced 4-acetylspiro [cyclopentane- 1,9'- [9H] fluorene]2,3-dione, 4-cyanospiro[cyclopentane- 1,9'- [9H]fluorene]2,3-dione, and 4-[(phenylsulfonyl)oxy]spiro[cyclopentane- 1,9'- [9H] fluorene]-2,3-dione, respectively.

A - t 2 =

i=O°CH3 Cii&HCOCH3 E-

E- t l ( o r t - B u O H )

e

x~c00cH3

t A- (or t-Bu0-)

ti

COCOOCH3

1

(3)

CHzCHzCOCH3

5

Sir: The reaction of methyl vinyl ketone, 2, with ketones under Michael reaction conditions followed by cyclization represents a standard route to cyclohexenones. In the review article on the Michael reaction' only one reference was made of the reaction of 2 with a polyfunctional a-keto ester. The only example of a reaction of 2 with a simple

@F@

t t-BuOH

-

(4) 6

2, RCH2=CHR = COCH, 3, R = CN 4, R = SO,C,H,

(5) C-

/\

CH2 CQCOOCH3

I

CH2COCH3

5

W RO +

6,R = COCH, 7,

R = CN

C

H

3

Y 0

9

8, R = SO,C,H,

a-keto ester is that describing the condensation2of methyl 9-fluoreneglyoxylate, 1. In repeating this reaction, using catalytic amounts of potassium tert-butoxide, an 82% yield (1)The Michael Reaction: Bergmann, E. D.; Ginsburg, D.; Pappo, R. "Organic Reactions", R. E. Krieger: Huntington, NY,1959; Vol. 10;pp 179-555. (2) Plieninger, H.; Ege, G.; Ullah, M. I. Ber. 1963,96,1610.

0022-3263/84/1949-0397$01.50/0

The keto ester 1 is a mixture of keto and enol forms (NMR shows two CH3 groups at 6 4.05 and 3.75, corresponding to keto and enol forms). The amount of each varies with the time of contact with the CDC1, solvent before measurement and with the solvent used for recrystallization. In the condensation of 1 with 2 when a catalytic amount of t-BuOK is used the anion B- rapidly reacts with the acidic keto ester 1 to form 5, as shown in eq 3.4 However, when an equivalent of t-BuOK is used there is no 1 to react with B- and the latter cyclizes by displacing methoxide ion, as shown in eq 4. Then 6 is converted to anion C- as shown in eq 5. Interestingly,when 5 is treated with 1 equiv of t-BuOK under conditions similar to those used in the reaction of 1with 2 and 1equiv of t-BuOK, no more than a trace of 5 and 9 are p r o d ~ c e d . ~ The latter reaction is under study. (3) For references concerning synthesis of adjacent cyclopentanediones, see: Ho, T-L.; Liu, S-H. Synth. Commun. 1983,13,685. (4)A referee suggested that B- reacts with t-BuOH rather than 1. If this were the case, one would expect a good yield of 5 when 1 equiv of t-buOK is used, but this is not the case. (5)In the Plieninger reference* cyclization of 5 to 9 occurs in high yield when catalyzed by piperidine and acetic acid.

0 1984 American Chemical Society

398 J . Org. Chem., Vol. 49, No. 2, 1984

Preliminary studies of the condensation of acrylonitrile," 3, and phenyl vinyl with 1 afforded 7 (18%)and 8 (77%), but neither of these reactions was optimized. Compounds 6,7,and 8 are highly enolic. The generality of such reactions is indicated as any compound containing a double bond capable of undergoing a Michael reaction should be capable of yielding a cyclopentanedione under proper conditions when an a-keto ester is involved. Condensatiod of 1 with 2 using t-BuOK on a 0.02-mol scale yielded 6 (57%): mp 217-218 OC; MS: m l e 290.0988, calcd for Cl9HI4O3290.0939; NMR'O (Me4Si)6 2.56 (9,3, CH&, (6)Bruson, H. A. J . Am. Chem. SOC.1942,64, 2457. (7) Prof. Leo Paquette generously supplied phenyl vinyl sulfone. See: Carr, R. V. C.; Paquette, L. A. J . Am. Chem. SOC.1980, 102,853. Paquette, L. A.; Williams, R. V. Tetrahedron Lett. 1981 22, 4643. (8) See ref 1, page 536, for the use of a vinyl sulfone in the Michael Reaction. (9) Mr. C. R. Weisenberger is thanked for the MS data. (10) I thank Drs. C. Cottrell(500-MHz instrument),P. K. Sujeeth, and S. Veeraraghavan for the NMR spectra.

Communications

3.42 (9, 2, CH2),7.0-7.8 (m, 9, Ar H and enol H). Anal. Calcd C, 78.6, H, 4.8. Found (Galbraith): C, 78.5, H, 4.9. There was obtained similarly pure 7 (18%) [mp 202.5-204.0 OC; MS? m l e 273.0759, calcd for C18H11N02 273.0789; NMR (Me4Si) 6 3.24 (5, 2, CH2), 4.93 (m, less than 1, OH), 7.13-7.86 (m, more than 8, Ar H and enol H). Anal. Calcd C, 79.1; H, 4.1; N, 5.1. Found: C, 79.1; H, 4.2; N, 4.91 and pure 8 mp 189-190 "C, MS: m l e 388.0825, calcd for C23H1604S 388.0769; NMR (Me4&) 6 3.07 (s, 2, CHJ, 6.80-8.20 (m, 14, Ar H and enol H). Anal. Calcd C, 71.1; H, 4.2; S, 8.3. Found: C, 71.2; H, 4.2; S, 8.31.

Supplementary Material Available: Additional experimental data for compounds 5, 6, 7, and 9 (2 pages). Ordering information is given on any current masthead page. Melvin S. Newman Chemistry Department, T h e Ohio State University Columbus, Ohio 43210 Received August 30, 1983