Feb. 5, 1957
COMMUNICATIONS TO THE EDITOR
marized in the table. For the following reasons, we believe that the structure proposed by Sheline is untenable.
753
A NOVEL REARRANGEMENT OF N-BROMOSUCCINIMIDE
Sar. We have found that the reaction of N-broniosucPRINCIPAL INFRARED ARSORPTIOX R A S D S O F Fes(CO)12u cinimide (NBS) with allyl chloride, allyl bromide, Relativeb or P-methallyl chloride in chloroform solution with Frequency, cm. -1 intensity trace quantities of benzoyl peroxide yields /3-bromo2043 10 propionyl isocyanate BrCH2CH2COSCO (I). The 3020 8 product was isolated by distillation a t reduced 1997 -4 pressure to yield I, b.p. 68-69' (10 mm.), nZ7D -0.5 1858 1.4915, intense infrared absorption maxima, inter 1826 -0.5 alia, a t 2250, 1735, and 1400 cm.-'. A mass spec591 ... trometer was used to obtain the molecular weight 575 ... (calcd. : 177; found, 1'77); the cracking pattern a Solvents: CS2, CC14, CHCI,. * Estimated from a plot was consistent with the postulated structure. of (cm.-l X optical density) u s . cm.-' for the spectrum Reaction of I with water gave carbon dioxide in CSa in which the best resolution was obtained. and 0-bromopropionamide,' m.p. 115-116' (from The extremely low relative intensity of the ab- chloroform), undepressed by admixture with a sorption in the region lS20-18GO cm.-' strongly sug- sample prepared from P-bromopropionyl bromide gests that the bands are not fundamentals, and and ammonia. Reaction of I with methanol gave hence that there are no ketonic bridging carbon colorless needles (11), m.p. 137-1.38' (from methmonoxide groups of the type previously proposed anol). Anal. Calcd. for CbH803NBr: C, 28.60; for Fe2(CO)g,3and C O ~ ( C O ) BFurthermore, .~ if H, 3.84; N, 6.67; Br, 38.05. Found: C, 28.7; the spectra are taken in CSS, CC14 or CHC13, rather H , 3.8; N, 6.5; Br, 38.0. Reaction of I with than the toluene used previously,l two bands are aniline in methylene chloride yielded colorless clearly discernible ; the model favored by Sheline needles (111), m.p. 181-183' (from methanol). would produce only one fundamental in this region. Anal. Calcd. for C10H1lN2O2Br: C, 44.30; H . Although only two bands a t -2000 cm.-' are ob- 4.09; N, 10.33; Br, 29.47. Found; C, 44.7; H, served using rock salt optics (cf. also ref. l ) ,a third 4.3; N, 10.3; Br, 29.3. The infrared spectra of band a t -1997 cm.-' is resolved easily with a I1 and 111 were consistent with their formulation lithium fluoride prism. Throughout the range 450- as methyl P-bromopropionylcarbamate and N3000 cm.-' the only other significant absorptions phenyl-N'-P-bromopropionyl urea, respectively. To synthesize I, P-bromopropionic acid was were the two bands a t 594 and 575 cm.-', which seem best assigned as carbon-metal stretching fre- treated with phosphorus tribromide in benzene to quencies. It is possible to assign the very weak yield 597c of P-bromopropionyl bromide, b.p. bands a t 1858 and 1826 cm.-' as triple combinations 50-52' (4 mm.), n27D 1.5320. Anal. Calcd. for among the metal--carbon stretching frequencies, an C3H40Br2; C, 16.67; H , 1.87; Br, 74.03. Found: assignment more in keeping with their low inten- C, 16.4; H, 1.7; Br, 74.4. Treatment of the acid bromide with dry, powdered silver cyanate2 gave sity. Before any conclusion can be reached concerning 40% I, b.p. 68-69' (10 mm.), nZ9D1.4900, inthe structure of the free molecule, Fe3(C0)12,the frared spectrum superimposable on that of I isoband a t 1997 cni.-' must be assigned. There are lated from the NBS reaction. Anal. Calcd. for three main possibilities: (i) that it is a fundamen- CdH4N02Rr; C, 26.99; H , 2.26; X,7.87; Br, tal, though it is perhaps a little weak; (ii) that it is 44.90. Found: C, 26.9; H , 2.3; N, 7.8; Br, 44.9. Samples of I1 and I11 prepared from this sample of a combination involving one of the other C-0 stretching frequencies, but this would require the I had m.ps. of 137-138' and 183-184', respectively, presence of a fundamental frequency a t least as low undepressed by admixture with the previously deas -50 cm.-', which would seem unlikely although scribed samples. The infrared spectra of the not impossible; (iii) that it is the C13 analog of the samples of I1 and I11 from both sources were 2043 cm.-' band which would be expected in this identical. No satisfactory analysis of I isolated from the position, but our estimate of the relative intensities (-l/lOo expected) seems to rule this out. If the reaction of NBS has been obtained as yet; a 1997 cm.-' band is in fact a third fundamental, typical analysis is shown. Anal. Calcd. for C4then another structure (IV) considered by Sheline H4N02Br: C, 26.99; H , 2.26; N, 7.87; Br, 44.90. involving an equilateral triangle of three iron atoms Found: C, 26.8; H , 2.5; N , 7.2; Br, 49.0. Mass bonded directly to one another is again eliminated. spectra indicated the presence of 1-2Tc of a chlorobromopropene in the sample; X-ray absorption in(3) R . K. Sheline and K. S. Pitzer, THISJOWRXAL, 72, 1107 (1950). dicated the same sample contained 45 1% Br (4) J. W. Cable, R. S. Nyholm and R . K. Sheline. ibid., 76, 3375 and 2 f lyOC1. Presumably the chlorobromo(1084) propene was responsible for the discrepancy in the DEPARTMENT OF CHEMISTRY analytical data. OF TECHNOLOGY MASSACHUSETTS INSTITUTE The identity of the samples of I from the reaction CAMBRIDGE, MASS. F. A . COTTON DEPARTMENT OF CHEMISTRY of NBS with allyl chloride, allyl bromide, and
*
IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY (1) C. S. Hamilton and C. L. Simpson, THISJOURNAL, 71, LONDON, ENGLAND G. WILEINSON (1924). RECEIVEDOCTOBER15, 1956 ( 2 ) 0. C. Belliter, Brr , 36, 3213 (1903).
8168
754
COMMUNICATIONS TO THE EDITOR
methallyl chloride was established with the m.ps. and mixed m.ps. of samples of I1 prepared from each sample of I. The rearrangement has been found to occur only in chloroform solution as yet. The allylic halide is required for the rearrangement; in its absence YBS, chloroform, and benzoyl peroxide did not react (except to form a small amount of free bromine) when refluxed for periods four times as long as those required for the rearrangement to go to io% completion. Since the reaction was inhibited by picric acid or trinitrobenzene, it appears to be free radical in nature and, indeed, to be a free radical analog of the Hofmann hypohalite reaction of amide^.^ Work on this reaction is ~ o n t i n u i n g . ~
L'ol. 70
(17-ketone), 1508, 1503 (C=C stretching in aromatic ring), 1409 (unsubstituted CH2 a t C-16) and 1373 (C-18 methyl group) cm.-'. X prominent band a t 874 cm.-' in carbon disulfide was interpreted as the out of plane C-H deformation of isolated hydrogens in ring A. Comparable bands were noted in the spectruni of a dispersion in potassium bromide.
Anal. Calcd. for C19H24O3: C, 75.9i; H, 8.05. Found: C, 75.82; H
The compound formed a monoacetate, n1.p. 152-1535'; Anal. Calcd. for C~1H2604:C, 73.OG; H, 7.65. Found: C, 73.29; H , 7.93. With this information i t seemed that the coin(3) E . S. Wallis a n d J. F. Lane, "Organic Reactions," \'ol. 111, pound might be a methoxy derivative of estrone JI,hn Wiley a n d Sons, Inc., New York, h-. Y . . 1946, p. 267. and in view of the lack of polarity of the phenolic (4) T h i s work was supported in p a r t b y a University of California hydroxyl as evidenced by the partition coefficient Research G r a n t a n d by a grant from t h e Research Corporation. relative to that of estrone, a monomethyl ether of II. W. J. acknowledges a s u m m e r position a t T h e Dow Chemical Company, Midland, Michigan, during which p a r t of this work was com2- or 4-hydroxyestrone was considered probable. 1 )I et ed. The four possible monomethyl ether isomers of 2and 4-hydroxyestrone were synthesized by methods I~VISION O F PHYSICAL SCIEYCES I'NIVERSITY OF CALIFORNIA HARRY 11:. JOHNSON, J R . based upon the work of Mueller and Mills,4 and RIVERSIDE,CALIFORNIA DONALD E. BCBLITZ Horner and Stohr.6 The synthetic route to the metabolite involved the following reactions : 2RECEIVED DECEMBER 7 , 1956 nitroestrone, 2-aminoestrone diazonium salt, photodecomposition in methanol. The details of the synthetic work will be reported in the near future. The 2-METHOXYESTRONE, A METABOLITE OF new metabolite proved identical with 2-methoxyesESTRADIOL-l7@ I N THE HUMAN' trone (synthetic product, m.p. 184.5-188.5'; [aIz4D Sir: f178') as judged by identity of the infrared specAfter the administration of small or large doses trum6 in both carbon disulfide solution and potasof estradiol-17P-1G-C14 or estrone-16-CI4 to hu- sium bromide dispersion as well as the m.p. and mans, the phenolic fraction of urine contained rotation; there was no depression of the melting considerable radioactivity other than that asso- point on admixture of the natural and synthetic ciated with estrone, estradiol or any of the isomers samples. 2-Methoxyestrone failed to exhibit fluorescence6 of estriol. Counter-current distribution in the system 7091, aqueous methanol as the upper phase with sulfuric acid in a test commonly employed and carbon tetrachloride as the lower phase showed for estrogens and their metabolite^.^ Biologically6 a peak of radioactivity with a partition coefficient2 the compound is a very weak estrogen with activity less than 1!20,000 that of estradiol-17p as = 0.30 (estrone = 1.3). The same peak of radioactivity was observed after either hot acid or p- judged by the intravaginal assay of Emmens8 Quite apart from the interest attached to the glucuronidase3 hydrolysis of the urinary conjugates; the amount of material present was greater identification of another metabolite of the estroin the first and second days' urine than thereafter. genic hormone, the biochemical introduction of a The new metabolite was a ketone as evidenced by methoxyl group is a novel reaction in the steroid the essentially quantitative reaction with Girard field. The significance of this new type of transReagent T ; counter-current distribution studies formation is under further investigation. indicated that it was not identical with 3-hydroxy(4) G. C. Mueller a n d M. E. 32111s, personal communication; r / . A1*3~5('0)-estratriene-16-one. hlueller, N a t w e , 176, 127 (1955). T h e work of these authors related A pure sample of the new metabolite was iso- t o t h e proof of structure of 2-nitro a n d 4-nitroestrone. Independent lated from urine after the administration of 1 g. evidence for t h e correctness of these assignments has been obtainecl of estradi01-17/3-1G-C~~(specific activity = 191 in these laboratories from infrared a n d ultraviolet spectrometry. Werbin (Ped. PYOC., 15, 382 (1956)) h a s also reached similar concIIIcounts per minute per milligram (c.p.m../mg.)) over sions. a period of ten days. The product melted 187( 5 ) L. Horner and H . S t o h l , C ! w n . Bet'., 85, 993 (19623. ( 0 ) T h e a u t h o r s are indebted t o t h e several staff members uf t h i s 189.5'; [ a I z 8 D i-179' (ethanol); Xmax. 284.5288.5 mp ( E = 4000), Xmin. 254 mp (E = 420); Institute: Dr. William L. hfoney f o r t h e bioassays, Dr. C. D . W e i t o r t h e fluorescence analysis a n d Dr. G. Roberts for help with t h e i l l specific activity = 172 c.p.m./'mg. ; the infrared ffrared spectra. T h e technical asristance of Rosemarie Lehmarl. spectrum in carbon tetrachloride solution ex- Jerome Boxer and Albert Klutch is gratefully acknowledged. (7) I>.I,. Engel, Rec. Prop. i72 Hovmoize R e s . , 5 , 335 (19.50). hibited bands a t 3560 (hydroxyl group), 1743 (1) T h i s investigation was supported in p a r t by a g r a n t from t h e American Cancer Society a n d a research g r a n t (C-2271) f r o m t h e National Cancer I n s t i t u t e of t h e S a t i o n a l Institutes of Health, United
S t a t e s Public Health Service. ( 2 ) B. Williamson and I>, C. Craig, J . B i d Chrm., 168, G87 (1947). ( 3 ) Ketodase. obtained from t h e Warner Chilcott Laboratories, a division of Warner 1.ambert Pharmaceutical Cn., New York.
(8) C . W. Emmens, I r e d . Res Council, Spec. R e p t . Series 2 3 4 . H. lf,S t a t . Office (1939).
THESLOAN-KETTERIYG ISSTITUTEFOR CANCER RESEARCH SEW YORK21, NEWYORK
S. KRAYCHY T. F,GALLAGHER
RECEIVED OCTOBER 27, 1950
