4182
COMMUNICATIONS TO THE EDITOR
Vol. 77
acetate (monoalcoholate: n1.p. 184-185' (dec.) fluoro to 10-fold iii the case of the bromo derivaafter melting and resolidification a t 126-139', tives. Dehydrogenation in the 6,7-position, on the other hand, effects in the two cases examined a two[ c ~ ] ~ +79" ~ D (CHC13), Ag;, 250 xnp (8,000), A?:' 2.85 p, 2.98 p, 5.80 p, 5.90 p, 5.98 p, G.24 p ; found: fold decrease in glucocorticoid xnd a 20-fold C, 54.48; H, 6.78; Br, 14.50; OCzHb,7.57; L.G. increase in salt-retaining activity. < 0.3)) 2-bromo-9a-fluoro-A4-pregnene-1lp,17a,21JOSEP FRIED triol-3,20-dione 21-acetate13 (m.p. 174-175' (dec.) ; KLAUSFLOREY EMILY F. SA50 [cx]'~D +138' (CHC13); 242 mp (12,200), THES Q U I B 5 INSTITUTE F O R LfEDICAL KESEAKCH JOSEF E. HERZ ' ' ;?A; 2.55-2.95 p, 5.78 p, 3.83 p , 5.94 p, G.15 p ; SEIV .4LBERT R. RESTIVO BKUSS\L.ICK, S . J . ALECKB O R M A N found: C, 55.70; H, 6.30; Br, 13.16; L.G. 1.0, I:KANK hf. S r s c > ~ r ~ Na lo), 6-dehydro-Sa-fluorohydrocortisoneaceRECEIVED J U N E 37, 1935 tateg.l4 (117) (m.p. 216-217, [ a ] 2 3 ~f123' (alc.), 4-135' (CHCI,), ,;A: 251 mp (23,000), ?'A:: 3.00 p (OH), 3.76 p, 5.81 p (acetylated side chain; 6.10 THE REACTION OF THIONYL CHLORIDE WITH p , (3.10 1.1, G.22 p (A4a6-3-ketone);found: C, 65.75; ALLYLIC ALCOHOLS1. H , 7.04; L.G. 5 , Na 20-30), the desired 111, and Sir: [ a I z 3 +73" ~ an isomer of 111 (m.p. 271-272', Several mechanisms are available for the reaction (alc.), xA;; 237 mp (15,200), A ~ 3.00 ~ p,~ 5.75' p ) of allylic alcohols with thionyl chloride.'b Without 5.92 1.1, 6.04 p, 6.18 p, 6.24 p ; found: C, 135.96; a solvent, mixtures of isomeric chlorides are always H, G.81; L.G. < 1). obtained. However, we have found that the SNi' Substitution of a hydrogen atom for a hydroxyl mechanism* may be made very dominant by the group a t C-21 results in a greater decrease of salt use of dilute ether solution, where the liberated retaining than of glucocorticoid activity.15 We hydrogen chloride is rendered quite inactive.s have therefore prepared the 21-desosy derivatives Under these conditions, crotyl alcohol yields 99% of TI1 and I V as follows. Saponification of 111 a-niethylallyl chloride and a-methylallyl alcohol and I V with potassium carbonate in aqueous yields 100% crotyl chloride. methanol yielded the respective dehydro-9aThe ether technique has now been found successfluorohydrocortisones ( A 1: m.p. 274-275' (dec.), ful even in some more reactive systems. With optically active trans-a,r-dimethylallyl [ a ] j 3+94' ~ (ale.), A& 238 mp (15,500); found: c, 66.GX; H, 7.16) and ( A 6 : m.p. 257-259') [ a ] " D alcohol (I) the more likely conformation of the + l O l " (ale.),,:A: 281 r n p (25,600); found: C, transition state of the SNi' process would lead to (X.30; H, 7.00), which were converted into the active trans-chloride (11) and the configuration of 21-mesylates in pyridine a t 0 ' ( A 1 : m.p. 220" the new asymmetric center would be opposite to that of the original one. The less likely conforma(dcc.), [ a ] 2 S ~+98' (ale.), x ~ C , , 235 mp (15,000); tion would give optically active cis-chloride. found: C, 5S.01; H, (3.36; S, 7.52) and (A6: n1.p. We have found that trans-alcohol (I) is converted to 237-238' (dec.), [ a I z 3+94' ~ (alc.),, : A: 281 mp trans-chloride (11) (100% trans-isomer) which has (27,5001; found: C, 58.19; H, 6.03; S, 7.54). the opposite configuration of the alcohols as illusThe latter were converted into 9a-fl~or0-A'~~-trated below. I n fact, the optical purity of the pregnadiene-1 1/3,17a-dio1-3,20-dione (m.p. 313- chloride is higher than that which results under 314' (dec.), [ a I z 3 D +47' (pyridine), APC,, 238 mp conditions favorable for direct displacement of (15,500); found: C, 69.47; H, 7.66; L.G. 4, chlorosulfinate ion by providing a soluble hydroNa < 0.1) and 9a-fl~oro-A~~~-pregnadiene-l1~,17achloride.3 diol-3,20-dione (m p. 294-296', +112" (dioxane), ,:A; 281 mp (26,000); found: C, G9.66; H, 7.48; L.G. 0.3, Na ) .
4183
COX~IUNICATIONS TO THE EDITOR
:lug. 3, 1935
TABLE I PRODUCTS OF REACTION OF ALLYLIC ALCOHOLS WITH THIONYL CHLORIDE Alcohol
Reaction conditions
CHKHClCH=CHz
SOCl2,no solvent
CHaCH=CHCHzOH
SOC12, no solvent
CHaCH=CHCHzOH CHaCHOHCH=CHz ( )transCHsCH=CHCHOHCHa CsHsCH=CHCHzOH CsHbCH=CHCH*OH
SOC12 in Et20 SOClz in Et20
-
SOCl, in EtzO 0.1 M R O H 0.1 MSOCl, in Et20 1 1.l ROH 1 M SOCL in Et20
+ +
show a-phenylallyl chloride is the product. This thermodynamically less stable secondary chloride is rearranged only very slowly in the reaction solution. Our present evidence is still insufficient to decide whether the SNi' mechanismZ involves a one-stage concerted process or ionization to an intimate, rigidly oriented carbonium chlorosulfinate ion pairI6 followed by internal returne of the chloride component of the chlorosulfinate anion to give rearranged chloride. It is very clear that the SNi' mechanisms does not involve a carbonium chloride ion pair of the type employed by Cram' in his preferred mechanism for the action of thionyl chloride on the 3-phenyl-2-butanols. A carbonium chloride ion pair in the a,y-dimethylallyl system would lead to a trans-chloride which is 100% racemic instead of the inverted chloride actually observed. Further, a carbonium chloride ion pair would not lead to the specific structural results obtained with the butenols and cinnamyl alcohol. The dominant role of the SNi' reaction is soinetimes difficult to preserve. I n the case of cinnamyl alcohol, even the use of 1 M concentrations of reactants changes the polarity of the medium and results in the product ion of a mixture of 60% cinnamyl chloride and 40% a-phenylallyl chloride from the reaction itself since a-phenylallyl chloride is stable under the conditions used. (5) E. Kosower, Ph.D. Thesis, U.C.L.A., 1952, page 97. (6) W. G. Young, S. Winstein and H. L. Goering, THISJOURNAL, 73, 1958 (1951). (7) D. J. Cram, ibid., 75, 332 (1953).
FREDERICK F. CASERIO GERALDE. DENNIS DEPARTMENT OF CHEMISTRY OF CALIFORNIA UNIVERSITY ROBERTH. DEWOLFE Los ANGELES,CALIFORNIA WILLIAMG. YOUNG RECEIVED APRIL28, 1955 7-NORBORNENYL
AND
7-NORBORNYL
CATIONS
Sir: We wish to record the synthesis of anti-7-norbornenol (I) and 7-norborneol (11), and a ratio of 10" in the solvolytic reactivities of the corresponding toluenesulfonates. unti-7-Norborneno1, m.p. 117-118°, was obtained: (i) as its acetate by reaction of ethylene with acetoxycyclopentadiene, generated in situ from acetoxydicyclopentadiene, a t looo, and (ii) by selective hydrolysis of the unsaturated dibromide (1) Dissertations (Harvard): (1950), C J. Norton (1955).
Product composition
33% CHaCHClCH=CHs 6770 CH&H=CHCHICl 71% CHoCHClCH=CH* 29 7 0 CH,CH=CCH CHzCl 99% CHsCHClCH=CHt 100% CHsCH=CHCHzCl 100% (-1 tramCHsCH=CHCHClCHs 100% CeHsCHClCH=CHz 60% CsHrCHClCH=CHz 40% CeHsCH=CHCHzCl
P Wilder, Jr. (l950), R. E. Vanelli
(111),one of the products of addition of bromine to bicycloheptadiene (IV), followed by zinc debromination of the resulting bromohydrin.
IV 7-Norborneol, m.p. 150-151°, was obtained by catalytic hydrogenation of anfi-7-norbornenol (I). The first order rate constants (kl)for acetolysis of the corresponding p-toluenesulfonates in acetic acid (0.1 M in potassium acetate, containing 1% Ac20), and other pertinent data, are I11
,OTs
m.p. 60.5-61.0"
v ki(205')
23.3 & 0.3 kcal./mole 5.7 & 2.0 e.u. 0.04 X sec.-l
AH*
AS* K1(25')
VI m.p. 54.7-55.7' 8.40 X sec.-l 35.7 & 0.6 kcal./mole -3.5 & 1.7 e.u. 6.36 X 10-'6 sec.-l
The striking situation brought to light by the new measurements is emphasized by the following reactivities a t 25' ~TOLUENESULFON ATE anli-7-Norbornenyl 104 103 exo-5-Norbornenyl2 Cyclohexyl* 1 endo-5-Norbornenylz 10-1 10-7 7-Norborny13
It is clear that the geometry of the norbornyl system is uniquely unfavorable for stabilization of a cationic center a t (2.7. We attribute the high reactivity of the unti-7norbornenyl derivatives to powerful anchimeric assistance to ionization at C.7, involving the 2,3 *-electron cloud (VI arrow). It will be noted that a homoallylic system4is present, which is geometrically unique in that a vacant orbital on C.7 can overlap the p orbital systems of the double bond (2) S. Winstein, H. M. Walborsky and K. Schreiber. THIS JOURNAL, 72, 5795 (1950); H. L. Schmid and K. Schreiber, unpublished work. (3) Qualitative mention of low reactivity for 7-norbornyl chloride and syn-7-norbornenyl chloride has been made by J. D. Roberts, P. 0. Johnson and R. A. Carbon, ibid., 76, 5fiR5 (1954). (4) hl. Simonctta and S. Winstein, ibrd , 76, 18 (19.51).
