4688 (pK'a 7.70); its infrared spectrum lacked a band in the 3.8 to 4.0 p region characteristic for )KH+. From the pyridine mother liquor there was obtained after dilution with water and alkalinization a small amount of a chloroform-extractable substance which melted a t 310-312" after recrystallization from methanol. The analysis (calcd. for C:gH3&20&3: C, 66.38; H, 6.92; N, 5.34; S, 6.11; found: C, 65.91; H, 6.51; K, 5.35; S , 5.59) and the infrared data (bands a t 8.56, 8.95, 9.71 and 9.94 ,u characteristic for the tosylate ion, no absorption in 3.8-4.0 region) suggest the quaternary tosylate corresponding to the above chloride. It follows from these results that the hydroxymethylene group in IV and, hence, contrary to our previous postulate, the 16-carbomethoxy and 18hydroxy functions in reserpine are trans to the Cl6 and Cz0 hydrogens (;.e., the formation of the quaternary tosylate 11 from 111 occurred with retention of the configuration a t CIS). Taking into account our previous deduction regarding the H absolute configuration of this carbon atom,' the two quaternary salts obtained from reserpinol then have to be formulated as Va and Vb, respecca,ooc/y L'a, X = C1 I1 OCHI Vb, X = OTS OCHs tively, with the ring junction hydrogen atoms CYoriented. Considering further that Ca in reserpine The formation of the quaternary tosylate I1 from is readily epimerizable, and that its carbomethoxy methyl reserpate 18-tosylate (111) on refluxing in group can be saponified with alkali without change collidine furnished conclusive proof that rings D and of configuration2a and hence must be equatorial, E in I are cis-linked. On the assumption that the we suggest that the stereochemistry of reserpine is displacement of the 18-tosyloxy group involves a best expressed by VI (Cz-Ca-axial to ring D) and concerted mechanism leading to inversion a t C18, that of its 3-epimer isoreserpine3 by VI1 (C?-C., we advanced the postulate that the 18-hydroxyl equatorial to ring D). function and hence also the 16-carbomethoxy OR group2 in I and I11 were cis-oriented in respect to the hydrogen atoms a t the ring junction carbon atoms C16 and (220. Application of Hudson's rotation rule to reserpine lactone2a then led to the assignment of the @-configurationto the functional group a t Cu and hence also to the other substituents mentioned above. I n order to test the validity of the above assumption, we have recently essayed the preparation of the 0-tosylate of the primary alcohol reserpinol (IV),*bsince i t was clear that the formation in this case of a quaternary salt would be incompatible with the postulated cis-relationship of the 16- and 18-substituents and the ring junction hydrogens. VI \I1 Indeed, elimination of the tosyloxy group occurred already on treatment of IV with ptoluenesulfonyl (3) H. B. MacPhillamy, L. Dorfman, C F. Huebner, E. Schlittler chloride in pyridine a t room temperature. The and A. F.St. AndrC, ibid.,'7'7, 1071 (1955). crystalline product which deposited in 50% yield PATRICK A . DIASSI from the reaction mixture (m.p. 352-353", [a(Iz3D THESQUIBB INSTITUTE FOR FRASKL. WEISENBORN M. DYLION RESEARCH CHRISTIANE f73.6" (c, 0.713 in water) after recrystallization MEDICAL N. J. OSKARWINTERSTEINER from methanol) proved to be a quaternary chloride YEWBRUNSWICK, RECEIVED JULY 13, 1955 C22HzsNzOzCl (calcd. C, 67.94; H, 7.32; K, 7.20; C1, 9.12; found: C, 67.92; H, 7 . 5 3 ; 5,7.17; C1, 8.90) by the following criteria: a free base could not A COMPOUND WHOSE MOLECULE IS SUPERPOSbe extracted from alkaline solution by organic ABLE ON ITS MIRROR IMAGE BUT CONTAINS NO solvents; i t could not be titrated with perchloric PLANE OR CENTER OF SYMMETRY acid in acetic acid solution, nor with base in aqueous Sir: solution as could be reserpinol hydrochloride The presence in any molecule of a plane or (2) (a) L. Dorfman, A . Furlenmeier, C. F. Huehner, R. Lucas, center of symmetry is a sufficient condition for H. B. &lacPhillamy, J. M. Mueller, E. Schlittler, R. Schwyzer and optical inactivity,' and so far as we are aware, one A. F. St. Andre, Helu. Chim. Acta, 3'7, 59 (1954); (b) C . F. Huebner,
5,
'Y
H.B. MacPhillamy, A. F. St. Andre and E. Schlittler, THISJOURNAL, 7'7, 472 (1955).
(1) We are here concerned only with optical activity in the fluid state.
Sept. 3 , 193.5
COMMUNICATIONS TO
or both of these molecular symmetry elements is possessed by every inactive, non-racemic substance hitherto reported. However, i t has long been recognized2 that any molecule lacking a plane or center, but possessing a fourfold alternating axis of symmetry, would still be superposable on its mirror image; and it has been predicted2 that any compound consisting of such molecules would be experimentally inactive and non-racemic. We wish now to report the synthesis of what appears to be the first actual substance possessing this rare combination of symmetry elements, namely, the meso "trans/trans" diastereomer3 of 3,4,3',4' - tetramethylspiro - (1,l') - bipyrrolidinium (I). p-toluene~ulfonate~
THE
EDITOR
4GSD
tartaric acid, the "levo-levo" tartrate was obtained and after twelve recrystallizations showed [ c Y ] ? ~ D -36.2", m.p. 187--189O, Found: N, 5.68. The liquid dextro pyrrolidine regenerated from the tartrate yielded a 9-toluenesulfonate of m.p. 148-150.5", [ c z ] ~ ~ D 17.5" (water); the levo tosylate melted a t 150-153', [cY]'~D -20.15". Reaction of levo-3,4-dimethylpyrrolidine with the above levo-ditosyldiol in hot dioxane finally gave the desired meso trans /trans bipyrrolidinium tosylate (I or V) which after recrystallization from tetrahydrofuran melted a t 160-162", optical rotation zero within experimental errors; Found : C, 63.93; H, 8.92; N, 4.09; S, 8.94. The identical product can be obtained by employing the dextrorotatory forms of both reactants. However, reaction of the dextro pyrrolidine with levo-ditosyldiol gave a different and active trans, trans diastereomer IV, p-toluenesulfonate m.p. 150-152.5', [CYl3'D +19.8S0 (water). Since in this comparable reaction optical activity was not destroyed, it appears improbable that the observed inactivity of V is due merely to racemization. Using appropriate intermediates, we have also prepared the m-cisltrans diastereomer 111, ptoluenesulfonate m.p. 181-182') Found: C, 64.51 ; H, 8.97; N, 3.57, and its dextrorotatory form, 12.23' (water). Likem.p. 177.5-179') [cY]"D wise we have made the DLlcislcis diastereomer9 11, p-toluenesulfonate m.p. 185.5-187", Found : C, 64.27; H, 8.88; N, 3.57. The limited evidence reported here is consistent gith the prediction that a fourfold alternating axis of molecular symmetry would be a sufficient condition for optical inactivity.
+
H3cp@/-' CHI ys
N
/ \Hacr
\CHI (1)
Four diastereomers (three active, one meso) would be predicted for this structure. All four have now been prepared. The configurations can be depicted conveniently by the "swastika" projection formulas5 11-V. Only diastereomer V possesses the desired symmetry characteristics.
-
-1
-
-+A ' , Active cislcis
Active cisltrans
I1
111
Active transltrans IY
m,eso tmnsltrans Y
+
DEPARTMENT OF CHEMISTRY
G. E. MCCASLAND'O
OF TORONTO USIVERSITY meso-2,3-Dimethylbutane-1,4-diolea was con- TORONTO, CANADA STEPHEN PROSKOW verted to its di-p-toluenesulfonate, m.p. 108RECEIVED JULY28, 1955 109.2", Found: C, 56.06; H, 6.21; S, 14.56. (8) Polarimetric observations with sodium D light a t about 29" The epimeric DL-diol gave a ditosylate melting a t were made by three observers using two different solvents and con107-108.3" (mixed m.p. with meso epimer de- centrations. pressed), Found: C, 56.43; H, 6.35; s, 14.71. (9) C. R. Noller and C. E. Pannell. THISJ O U R N A L , 7 7 , 1862 (19551, The dextrorotatory diol (m.p. 44-43', [ c Y ] ~ ~ Dhave recently prepared t h e &/cis diastereomer of t h e homologous f5.38" (ether)) gave a levorotatory ditosylate, m.p. hipiperidinium compound. (10) On leave a t Ohio S t a t e University, 1955-1956. 90.3-91 O , [CY]"D - 7.67" (benzene) ; the levo-diol7 gave a dextro ditosylate, 90.5-91.3O, [ c Y ] ~ f~7D. 7 5 " . The preparation of meso and of ~ ~ - 3 , 4 - d i m e t h y l pyrrolidine has previously6b been described. The DIRECT ACTION OF COBALT GAMMA RADIATION ON NITRATE ION IN AQUEOUS SOLUTION DL pyrrolidine treated with dextro-tartaric acid Sir: gave the crude "dextro-dextro" tartrate, m.p. In the decomposition of water by ionizing radia+32.2" (water). Using Ieoo187-189", [CY]''D tion, the 100 ev. yields of the intermediate H, OH, (2) For a n excellent discussion see G. W. Wheland, "Advanced HZ and HL02 may be denoted by GH, GOH,GH? Organic Chemistry." 2nd Ed., John Wiley and Sons, Inc., New York, 1949, pp. 147-151. F o r earlier mentiun see Freudenherg, "Stereo. and G H ? ~ ?respectively. , In the reduction of chemie," F. Deuticke, Leipzig, 1933, p. 601. Ce+-++ in 0.8N sulfuric acid by ionizing radiation, (3) T h e designation "lmns/lrans" signifies t h a t hoth rings indethe 100 ev. yield of Ce+++ has been postulatedl pendently have trans configurations f o r t h e pairs of methyl groups. to be equal to 2 G ~ , o , GH - GOHaccording to a (4) Strictly speaking it is t h e quaternary ammonium cation which has t h e symmetry elements described. mechanism in which H atom reduces Ce++'+ and ( 5 ) T o interpret t h e formulas, consider coaxis of the t w o rings t o OH radical oxidizes Ce+++. be perpendicular t o paper. T h e intersecting lines represent t h e planes Cobalt60 gamma radiation induces2 a reaction of t h e t w o rings. T h e short arms on each line show whether t h e between Ce++++ and TI+ which increases the 100 substituents are cis or trans. ev. yield of Ce+++ from 2.52 to 7.85 in 0.8N (6) G. E. McCasland and S . Proskow, (a) THIS JOURNAL, 76, 3486 (1954); (b) i b i d . , 7 6 , 6087 (1954). sulfuric acid. The initial 100 ev. yield of C e + + + (7) Iewo-2,3-Dimethylsuccinic acid was converted t o its dextro is independent of the concentration of TI+ (10-6 dimethyl ester ( A n n . , 688, 1 (1939)), which on reduction gave t h e
+
levo-diol, m.p. 44.5-45.5', larly prepared.
[a]% -5.42".
The dextro-diol was sirni.
(1) A. 0. Allen, Radiation Research, 1, 87 (1954). (2) T.J. Sworski, forthcoming publication.
