4924
LEONGOODMAN AKD B. R. BAKER
j0" and 3 m m . Recrystallization from ether-pentane gave 4.6 g. (55% yield) of t h e diketone as a yello\; solid with an irritating odor. T h e melting point was 37-38 . Anal. Calcd. for C 5 H 4 0 L :C , 62.5; H , 4.17; mol. wt., 96. Found: C , 62.3; H , 4.0; mol. mt., 96. I t s colorless di-O-Incthylo\ime melted at l l G 1 1 7 ' d t e r recrystallization from ethanol. A n d . Calcd. for C7HloS201: C . 54.54; H , 6.54; N, 18.17. Found: C , 54.60; H , 6.23; N, 18.35. (ii) Oxidation in Acetone.-To a cold stirred solution of the diol (20 g., 0.20 mole) in 200 nil. of acetone was added dropwise with stirring a solution of chromic anhydride (30 g., 0.3 mole) in a mixture of 60 ml. of concentrated sulfuric acid and 142 ml. of water. T h e reaction flask was surrounded by an ice-bath, and t h e addition regulated so t h a t the temperature in t h e flask did not rise above 0". After t h e addition was complete the reaction mixture was stirred an additional two hours at O", allowed t o warm t o room temperature, diluted with 400 ml. of water and extracted three times with 300 ml. of methylene chloride. T h e combined extracts were washed with 100 ml. of water, dried over anhydrous sodium sulfate, and t h e dione isolated as above. Yields are generally about 8 g . (42%), b u t despite t h e lower vield this procedure is more convenient t h a n the oxidation in acetic acid. Cyclopentane-1,d-dione .-To a vigorously stirred mixture of 5 g . of zinc dust in 25 ml. of glacial acetic acid was added dropwise cyclopentene-3,5-dione (0.50 g. ) in 25 ml. of acetic acid. T h e addition took 45 minutes, during which time the flask was immersed in an oil-bath a t 8,5". T h e reaction mixture was cooled, filtered and t h e acetic acid removed a t room temperature under vacuum. T h e residue vias extracted in a Soxhlet apparatus with chloroform, t h e chloroform removed and cyclopentane-l,3-dione sublirned at 120' (4 m m . ) ; yield 0.23 g. (4593) m . p . 149-150" (reportedI8 150-151 "). Catalytic Hydrogenation of Cyclopentene-3,5-dione.T h e dione (1.0 g . ) in 25 ml. of ethyl acetate was hydrogenated a t room temperature and atmospheric pressure using Xdams c a t a l y ~ t . ~ 7Absorption stopped after 2.7 molar equivalents of hydrogen had been used. T h e solution was filtered and t h e solvent removed. Titration of a n aliquot of the oily product showed t h a t not more than 2y0 of acidic material was present. T h e infrared spectrum showed a single carbonyl peak at 5.75 ,u and an intense hydroxyl band. Gas (27) An a p p a r a t u s which is especially convenient for this hydro. genation, a n d for others i n which a relatively large amount of hydrogen is absorbed (in this case G20 m l , ) , is deccribed b y E'. R. Story n n d C. H. D e P u y , in press.
81
chromatography showed the presence of two compounds in about equal amounts, one of which had t h e same retention time as did cyclopentanone. T h e 2,4-dinitrophenylhydrazone of t h e mixture was prepared and chromatographcd. Roughly equal amounts of the 2,4-DNP of cyclopentanonc (m.p. 142-143') and of cyclopenterione (m.p. 166-16T0) were isolated. Monotosylhydrazone of Cyclopentene-3,5-dione .-To 500 mg. (0.52 mole) of t h e dione in 10 ml. of hot methanol was added 915 mg. (0.049 mole) of tosylhydrazine. The mixture was allowed t o stand and then cooled to 0" and filtered. Recrystallization from 957, ethanol gave 1.15 g. (goyo yield) of the monotosylhydrazone, m . p . 1 8 5 1 8 6 " dec. Anal. Calcd. for Ci2H1203X2S: C, 54.54; €1, 4.58; S, 10.60. Found: C , 54.45; H , 4 . 6 8 ; N, 10.9. This tosylhydrazone was readily soluble in dilute XiaOH. It could be recovered in 90% yield by acidification after heating a t 100"with base. Although some coloring of the solution took place, no evidence for t h e presence of a d i u ) grouping could be found. Cyclopentadiene Adduct .-The adduct could be prepared in quantitative yield by mixing equimolar amounts of t h e dione and cyclopentadiene in four times their volume of banzene and allowing the mixture t o stand at room temperature. The adduct begins t o precipitate in a very short time. After a few hours t h e solution was filtered and t h e filtrate recrystallized from a mixture of chloroform-carbon tetrachloride and sublimed a t 120-130° (1, mm.). T h e adduct had a melting point of 169.5-170.5 , and appeared to be 100% enolic. Anal. Calcd. for CloHloO?: C, 74.1; H , 6.2. Foutid: C , 73.9; H , 6.05. Anthracene Adduct.-The dione (3.42 g., 0.036 mole) and anthracene (6.35 g., 0.036 mole) were refluxed in 40 nil. of benzene for four days. T h e precipitated adduct was filtered, dissolved in SaOH and again filtered, and precipitated by acidification with dilute HC1. T h e adduct weighed 8 . 7 g . (89% yield), m.p. 302-305" dec. Reaction of Cyclopentene-3,s-dione with Aqueous Base T o 500 mg. of t h e dione in 5 ml. of water was added 53 nil. of 0.1 N XaOH. After stirring a t room temperature for 20 minutes t h e bromnish-red solution was passed through an Amberlite I R 120 ion-exchange column t o remove sodium ions. T h e aqueous solution of t h e polymer vias then evaporated t o dryness leaving a t a n , highly enolic powder n hich was insoluble in organic solvents. 9 n d . Calcd. for C5H40?.Z/jH?O:C , 55.FiS; H, 4.97. Found: C, 55.44; H , 4.87. h I E S , IOWA
[ COXTRIBUTION FROM TIlE DEPARTMEYT O F BIOLOGICAL SCIENCES,
Potential Anticancer Agents.l
1701.
STANFORD RESEARCH IXSTITUTE]
XXII. The Conversion of Cyclopentene Oxide to Cyclopentene Sulfide
BY LEONGOODX.4N
AND
B. R.BAKER
RECEIVED MARCH 31, 1959 Reaction of cyclopentene oxide (6-oxabicyclo[3.1.0]hexane) ( I a ) with potassium thiocyanate led directly t o cyclopentene sulfide (6-thiabicyclo[3.1 Olhexane) (Va) in 20% yield. 00 reaction of t h e epoxide with ammonium thiocyanate, a n 11% yield of cyclopentene thiocyanohydrin ( V I ) was obtained, along with 37'0 of cyclopentene sulfide (Va). Treatment of trans-2-thiocyanatocyclopentylmethanesulfonate (1'111) with aqueous sodium hydroxide gave a 63y0yield of cyclopentene sulfide (Va).
In a previous paper of this series,2 a number of methods for the conversion of cyclopentene oxide to cyclopentene sulfide were reported which would proceed under conditions compatible with the (1) T h i s work was carried o u t under t h e auspices of t h e Cancer Chemotherapy National Service Center, Xational Cancer I n s t i t u t e , Contract h-o. SA-43-ph-1892. For t h e preceding paper of this series, cf. L. 0. Ross, L. Goodman a n d B. R . Baker, J . Org. Chent., 24, in press (1959) (2) L. Goodman, A. Benitez a n d B. R . Baker, THISJ O U R N A L . 8 0 , 1680 (1958).
chemistry of nucleosides. Additional methods have now been found and are the subject of this paper. A number of sulfur-containing reagents cause the direct conversion of epoxides to episulfides, alkali thiocyanates being the most frequently used. A mechanism for the thiocyanate reaction was advanced by Ettlinger4 and given support by (3) C. C . Culvenor, W . Davies a n d W. E . Savige, J . Chpm. Snr.. 4180 (19x2). (4) M. G. Ettlinger, THISJOURNAL,72, 4792 (1950).
CYCLOPENTENE SULFIDE FROM CYCLOPENTENE OXIDE
Sept. 20, 1959
aoHaoH
van Tamelens in his studies in the cyclohexane and cyclopentane systems and by Price and Kirk6 in their study of the stereochemistry of the reaction. For the case of an alicyclic oxide the mechanism may be written
I Ia, series, n 13, series, n
= =
1 2
3 S
(CH,),
V
I1
I11
NCS
SCN
vI
VI1
Aqueous potassium thiocyanate is much more basic than aqueous ammonium thiocyanate, and the conversion of I a to Va seems to require that the intermediates be in the anionic form, as suggested by the proposed m e ~ h a n i s m . ~An , ~ authentic sample of the thiocyanohydrin VI was prepared in 63% yield by the reaction of the epoxide Ia thiocyanic acid, as described by van Tamelen.5 The material prepared by this method showed strong infrared -SCN absorption a t 4.65 p which was accompanied by appreciable absorption a t 4.75 p ) probably attributable to the presence of trans-2-hydroxycyclopentyl isothiocyanate (VII) in the distillate. The extraneous 4.75 p band was present only in trace amounts in the thiocyanohydrin VI isolated fron the reaction of Ia with ammonium thiocyanate, and did not increase as a result of distillation; it is known that some thiocyano compounds rearrange to the more stable isothiocyanate isomers upon heating.8 bThen VI was heated with aqueous potassium carbonate, a mixture of epoxide I a and, predominantly, the episulfide Va resulted. The episulfide Va presumably resulted from the sequence I I a 4 I I I a + IVa + Va and the epoxide probably resulted from the reversion of IIa, as suggested by the mechanism. Under milder alkaline conditions, van Tamelen3 converted VI to unidentified, undistillable products; i t is surprising that he did not isolate cyclopentene oxide (Ia) from this transformation. The thiocyanohydrin VI was converted in good yield to its mesylate VI11 and its tosylate IX. The latter was a well-defined crystalline compound.
gQc3 I
-0CN8 (CHZ),
C -
4925
IV
van Tamelenj reported that cyclopentene oxide (Ia) was not converted to cyclopentene sulfide (Va) with potassium thiocyanate a t room temperature, conditions under which cyclohexene oxide (Ib) was converted to cyclohexene sulfide (Vb) in high yield. He also reported that cyclopentene oxide (Ia) was recovered in about 80% yield when it was heated with excess potassium thiocyanate a t 60" for 15 hours. van Tamelens rationalized these results by assuming that the intermediate IIIa represented a highly strained system of two transfused five-membered rings that would be expected to form with difficulty, if a t all. In view of the fact that allobiotin and epiallobiotin contain two trans-fused five-membered rings7 and the fact that thiocyanate ion should a t least convert cyclopentene oxide (Ib) to the thiocyanate IIb, the reaction was reinvestigated. I n these laboratories it was shown that reaction of I a with potassium thiocyanate a t about 100° gave a 207, yield of cyclopentene sulfide (Va). This was accompanied by a non-distillable, viscous residue which contained %yosulfur, showed no infrared absorption bands characteristic of -SH or -SCN groups, but did show distinct -OH absorption bands; this residue may have resulted from opening of the episulfide Va under the alkaline reaction conditions. When the reaction was carried out a t 60-75", approximating van Tamelen'sj conditions, the distillable product was largely the episulfide Va, accompanied by some of the epoxide Ia. Clearly, the conversion (Ia -+ Va) is much more difficult than that of Ib to Vb, but, if the above mechanism is accepted, the formation of intermediate IIIa is possible; the presence of the two heterocyclic atoms in IIIa must make the system more flexible than a carbocyclic system. When the reaction of cyclopentene oxide (Ia) and ammonium thiocyanate was carried out a t l o o o , only about 37, of episulfide Va was obtained. This was accompanied by about 11% of the thiocyanohydrin VI and a large, non-distillable residue whose infrared spectrum indicated i t to be a low molecular weight polymer arising from the reaction of cyclopentene oxide and water. (5) E. E. van Tamelen, THISIOURNAL, 78, 3444 (1951). ( 6 ) C. C . Price and P. F . Kirk, ibid., '76, 2396 (1953). (7) B . R . Baker, M. V. Querry, W. L. McEwen, S. Bernstein, S. R. Safir, L. Dorfman and Y.SubbaRow, J. Ovg. Chem., 12, 1% (1947); S. A . Harris, R . Mozingo, D. E . Wolf, A . N. Wilson and K . Folkers, THISJOURNAL, 67, 2102 (1945).
(IOR - a :: -Va
SCN VIII, R=SOzCHz IX, R =SOZC~H~CH~-P
When the mesylate was stirred a t room temperature with aqueous sodium hydroxide, cyclopentene sulfide (Va) was obtained in 63% yield. Only a low yield of Va could be obtained from the tosylate I X under alkaline conditions, but this could be attributed to its low solubility and the necessity of carrying out the reaction a t a temperature higher than the melting point of I X . Similar circumstances were responsible for the low yields of Va formed from a number of intermediates in the previous work.? Several attempts to prepare Va by the reaction of I a with thiourea3J gave only traces of Va. Most of the epoxide Ia was converted to undistillable material, probably as a result of hydrolysis of the intermediate in that reaction which would correspond to IIa. (8) Houben-Weyl, "Methoden der Organischen Chemie," Band IX, Georg Thieme Verlag, Stuttgart, 1955, P. 866. (9) F . G. Bordwell and H. M. Andersen, THISJOURNAL, 76, 49.59 (1963).
40%
LEONGOODMAN A N D B.K.BAKER
L-01. Sl
Reaction of VI with Aqueous Potassium Carbonate.--..\ Experimental lo Reaction of Cyclopentene Oxide (Ia) with Potassium Thio- mixture of 5.8 g. (40 mmoles) of thiocyanohydrin VI (fraction 2 , above), 16.56 g. (0.120 mole) of potassium carbonate cyanate.-A mixture [ i f 10.2 g. (0.103 mole) o f potassium thiocyanate, 3.90 g . (46.4 tiimclles) of cyclopentene ()side and 1 0 ml. of water was refluxed, with stirring, for 2 hours. i1a)Z aiid 3 0 nil. of water mas he'ited, with stirring, a t reflux The solution was extracted with three 25-nil. portioiis of peiifor 2 hours. Tlie resultant coltrless mixture coiitiiied two tane and one 25-id. portion of methylene chloride and the combined extracts were dried over niagnesium sulfate. ;Iflayers and was extracted with f : n r 30-ml. portions of penter filtratioii, the solvents were removed by distillation tane and one 30-ml. portion o f nietli>-letiecliloridc. T h e combined extracts were dried over iriagiiesium sulfate; the siil- through a short IYidnier column. T h e rcsitfue was cvnpnrat i d y distilled at 00 m n i . (bat11 temp. 48-00"j; yield 0.70 vents were distilled from the filtrate through a short Ii'idmer ~ \.apor-phase clironiatograpli)." indicated column and the residue was distilled in vacu(i tiircmgh a short g., n z 0 1.504S. the presence of 827, episulfide \'a and 18% epoxide Ia, in Vigreus column to yield 0.95 g. ( X O c / o ~of cyclopeiitene sulfair agreement with the proportions calciilatetl from the fide (Va), h.p. 68" (65 i i i m . ) , n2% 1.5240 (Iit.6 b.p. 69-70" refractive in In :tddition, vapor-phase clironiutog(65 mm.), nZ5D1.52221, whose infrared spectrum was identilled solvents showed t h a t almost 0.07 g. of cal with that of previous samples of Va.* The residue froni a mixture of Va and !a was present in the pentane-metliylthe distillation (2.8 g.) was a viscous, yellow liquid, eiie chloride distillate. 3.00 (OH), no SH near 3.0 or S C S at 4.65 p. trans-2-Thiocyanatocyclopentylp-Toluenesulfonate (1x1. A n a l . Found. C, 5
