28UJ
COMMUNICATIONS TO THE EDITOR
interpreted if pyrrylmagnesium chloride is represented as the N-MgX derivative I or the ionic, but- riot necessarily highly dissociated, resonance hybrid 11. Although a clear choice between these two alternatives cannot be made a t this time, it should be pointed out t h a t the lack of complete identity of the n.m.r. spectra of pyrrylmagnesium chloride and the presumably ionicIh pyrrylsodiurn (see chemical shift data) need not argue against the ionic structure 11, since these differences might simply be due to variations in the solvation or degree of association of the two metals. Hopefully, studies which are presently underway in this laboratory to clarify this point will also shed some light on the marked differences in the chemical reactivity of the Grignard reagentsg and alkali metal saltsI6of pyrrole. Acknowledgment.-This research was supported by grants from the research committee of the Cniversity of California.
Vol. 85
TABLE I THECiS/tt'anS RATIOSI S 0 - d i - S T Y R E N E MIXTURES OBTAINED BY .%DDITIOT O F HEAVY LVATER TO GRIGSARD REAGESTS PREPARED' F R O M Cis- OR tTaYLS-P-BROMOSTYRENE I N Y A R I O C S S O L V E N T S
Solvent
From cis-&bromostyrene
From lvans-@-bromostyrene
Tetrahydrofuran 10: 1 1 : 1 Mixture of 1O:l T H F a n d EE Ethyl ether a Bromostyrene in solvent was added t o M g metal a t 26" and t h e mixture was stirred for 15 min.
1: l(J 1:s
1:l in 90 min.
ture, the highest and lowest field signals of the CYproton quartet' of styrene are separate from the CYproton signals of /3-d1-styrenes as seen in an example shown in Fig. 1. Lye can, therefore, easily estimate the amount of styrene contained in the mixture and make a correction for the P-proton signals. The cis trans ratios thus obtained for the P-dl-styrene mix11.5) 1. P a u l i n g . "The S a t u r e of the Chemical Bond," Cornell University Press, Ithaca. S x'' , 1960, p 93 tures from these corrected areas of the &proton signals (16) C F H o b b i , C K X l c ~ l i l l i n ,E P P a p a d o p u u l m and C . .4 V a n are shown in the Table. dera'erf. J At77 C'him Suc , 84, 4 3 ( 1 9 6 2 ) . The cis trans ratios are higher for the products from (17) I.i,ckheed Graduate Research Felluw in Chemistry. s u m m e r , 1963 cis-P-bromostyrene than those from the trans comDEPARTMEKT OF CHEMISTRY MANFRED G . KEISECKE pound. Honever, the ratios are dependent on the solUXIVERSITV OF CAIJFORNIA HARRY W. JOHXSON, J R . vent species used for preparing the reagents. If the RIVERSIDE, CALIFORNIA J O H N F. S E B A S T I A N ~ ~ Grignard reagent is prepared from trans-P-bromostyRECEIVED J C L Y 23, 1963 rene, the original trans configuration is much more completely preserved in the P-d,-styrene mixture with the Solvent Effect on cis-trans Isomerization in Grignard reagent prepared in tetrahydrofuran than with the Reagent Formation from p-Bromostyrenes reagent prepared in ethyl ether. The use of a 1 : 1 mixture of tetrahydrofuran and ethyl ether as solvent Sir : affords a product with a cis trans ratio intermediate \Ye have found t h a t a Grignard reagent prepared in between those for the single solvents, though the ratio the dark from cis- or trans-6-bromostyrene shown to be for the mixed solvent is closer to t h a t for tetrahydrofree from the counter isomer of the bromide by infrared furan. If a Grignard reagent is prepared from cisand n.m.r. spectra gives a mixture of cis- and trans,B-bromostyrene using tetrahydrofuran or the 1 : 1 mixp-dl-styrenes by addition of heavy water as can be seen ture as solvent, the P-dl-styrene mixture obtained is in Fig. 1 and in Table I . Though n.m.r. signals due to much more abundant in cis-p-dl-styrene than in the the &protons of p-dl-styrenes overlap with p-proton trans isomer. cis-/3-Rromostyrene does not form a signals' of styrene contained in the p-dl-styrene mixGrignard reagent in ethyl ether a t temperatures up to 4 the boiling point of the solvent. The yields of P-dlstyrene mixtures in the other cases were in a range of 30-5070 based on P-bromostyrenes, when the Grignard reagents were prepared a t 2%". I t can clearly be concluded from the following experiment that the isomerization and the solvent effect on it take place in the course of Grignard reagent forniation but not in addition of heavy water. LYhen the Grignard reagent prepared in ethyl ether from frans-Pbromostyrene is diluted with tetrahydrofuran, the cis '!vans ratio of 5-dl-styrenes obtained is equal to that from the same Grignard reagent diluted with ethyl ether. The cis frans ratio is also unchanged by dilution with ethyl ether of the Grignard reagent prepared in tetrahydrofuran from cis- or trans-P-bromostyrene The cis 'trans ratio is unchanged by the difference of time intervals up to 5 hr. a t 20' between the Grignard reagent formation, the dilution by solvent, and the addition of heavy water. This experimental result and the high degree of retention of original cis or configuration in tetrahydrofuran show that the Grignard reagent is stable in tetrahydrofuran. in which prompt isomerization of vinyllithium reagents is reported? to take place. If the temperature is allowed to rise to the boiling point of the solution in the preparation of the Grignard reagent, the yield of p-dl-styrene decreases. (D-tll-Sty-
-
--
Fig. I .--The n.ri1.r. spectrum (BO Mc.) of p-dl-styrene mixture obt:iined b y t h e addition of h r a v y water t o t h e Grignard reagent
prep:ired from f~nns-B-brornostyrenein tetrahydrofuran. n m . r . spectrum o f styrene is also given for comparison,
The
(1) As f o r the assignment of proton signal- r t f 5tyrene. see J h P w l e , W' G Schneider, a n d H J Bsrnstein, "High~resnlution Suclear Xlagnetic Resonance,' McGraw-Hill Book C o , I n c , Kew Y o r k , N. Y , 1959, p . 240 ( 2 ) D . Y . C u r t i n a n d W. J . Koehl, 3. A m C h e m Soc.,84, 1907 (1902)
Sept. 20, 1963
COMMUNICATIONS TO THE EDITOR
2861
renes are not practically obtainable from trans-@bromostyrene a t the elevated temperature, when tetrahydrofuran is used as solvent.) However the cis trans ratios of P-dl-styrene mixtures remain almost unchanged. Details will be reported in a later paper.
Q
BASICRESEARCH LABORATORIES TSLIXEO YOSHINO KAYOS COMPAUY YASUOMANABE KAMAKCRA, KASAGAWA, JAPAN RECEIVED MAY25, 1963 TOY0
1.7~ p.s.
Spin-Spin Coupling between Hydrogen and Steroid Angular Methyl Protons
Szr: Recently' it has been demonstrated t h a t in certain fluorosteroids, spin-spin coupling between fluorine and angular methyl protons can occur. To our knowledge, only a single e k m p l e of a split angular methyl resonance in a steroid nucleus not containing fluorine has been reported; vzz. the C-18 methyl resonance in the n.m.r. spectrum of the 9,19-cyclo-11-keto steroid I occurred as a three-proton doublet ( J = 0.7 C . P . S . ) . ~ I t was tentatively suggested t h a t the splitting may be due to long-range coupling. We have observed a similar
d)2P
H IIa, RI = H
