4999
Sept. 20, 1960
COMMUNICATIONS T O T H E EDITOR STEREOELECTRONIC EFFECTS ON ORGANIC BASES. 1 1 1 4 , THE BASICITIES OF SOME SATURATED ETHERS IN AQUEOUS SULFURIC ACID
Sir: I t has been recognized for many years that organic oxygen compounds are bases and various means for estimating their relative basicities have been devised.‘ So far, however, analytical difficulties have prevented placing aliphatic ethers on the pH-Ho scale which would allow their direct comparison with stronger bases in aqueous solution. We report here the pKa values for the conjugate acids of several important simple ethers determined by a combination of the usual Hammett acidity function method2 with solvent extraction3 and gas chromatography. For distribution of the ether between a non-polar, inert, organic solvent and an aqueous acid solvent in which the ether is essentially unprotonated the distribution constant is K D = [ B ] o / [ B ] a q . The observed distribution ratio for the ether between the same organic solvent and an aqueous acid layer in which protonation occurs is D = [B]o/ ([Blaq [BH+]) where [ B ] a q and [BH+] are the concentrations of the free ether and its oxonium ion in the aqueous acid layer and [B]o is the concentration of ether in the organic layer. Combining these expressions with that for the ether protolysis equilibrium,2 Ka = ho[B],,/[BH+] leads to D = K D - Dho/K, (1) or in logarithmic terms
+
Ho = fiK,
+ log [ D / ( K D- 011
(2)
Because of small activity coefficient effects, KD is best estimated from equation 1 by a linear plot of D versus Dho rather than from the distribution constant for water and organic solvent. As is predicted by equation 2, a plot of log (D/(K= D)]versus EI0 is linear for all of the compounds listed below and the pKa is taken from the Ho where the logarithmic term becomes zero. In addition to this test of the data, a series of control experiments established that the organic and aqueous layers were completely immiscible, that the ether did not react with the acid, that within o.05H0 unit i t did not change the Elo of the acid layer2 (as measured by indicator), that K n is essentially constant over an eight-fold range of ether concentration in the organic layer and that the same pK;, may be obtained with different organic solvents even though the value of K D may vary markedly from one to another. Extraction was performed by equilibrating a 10ml. portion of 2% ether in cyclohexane or carbon tetrachloride with an equal volume of aqueous sulfuric acid solution of known Ho in a specially W.Gerrard a n d E. D . Macklen, Chem. Reus., 59, 1105 (1950). (2) X I . A. Paul and F. A . Long, i b i d . , 6 7 , 1 (1957). (3) ( a ) C. Golumbic a n d R I . Orchin, THISJ O U R N A L . 7 2 , 4145 (1950); (b) G. H. Morrison and H. Freiser, “Solvent Extraction in Anelytic:il Chemistry,” J o h n Wiley and Sons, New York, N.Y.,1957. (1)
-_
-__-
-
designed apparatus agitated by a Fisher Vibrostirrer. After separation and filtration drying of the organic layer, a fifty microliter sample was injected into a nonpolar g.1.c. column a t appropriate conditions on a Burrell K-2 Kromotog and eluted with helium. The amount of ether in the organic layer was found by comparison of the area of the solvent peak to that of the ether (relative to the original stock solution of ether) and the concentration in the aqueous layer was obtained by difference. Listed below are the pKa’s (from least squares treatment of equation 2) of several important ethers as estimated by this method a t room temperature, and data from two other sources showing the same relative order of basicity.
Ether
DKa
Dimethyl ether Diethyl ether Tetrahydrofuran Tetrahydropyran 1,4-Dioxane Anisole4
-3.83 j z 0.19 -3.59i. .10 -2.08 i. .18 - 2 . i 9 =I=. l 5 -2.92i. .12 - 6 . 5 4 & .02
OD hand shift for CHnO-p, cm.
AF238
for In-
ether complex’ kcal./ mole
...
...
966
1115
1.12 1.70 1.66 1.3
70‘
...
lli6
1156
It is clear from the table that the pK,’s of the ethers in aqueous acid follow the same general order of basicity as would be deduced from their effect on the OD stretching frequency of deuteriomethanol or the stabilities of their iodine complexes. We have found that these compounds and a number of other ethers to be reported later follow the same linear correlation of pKa versus OD shift that has been reported frequently for stronger base^.^,^ Furthermore, it is found that the pKa’sgiven above fall in exactly the region of basicity predicted from a plot of all known pKa‘s (of weak and strong bases) for which OD shifts have been reported. This lends further support for the general validity of this rough correlation originally suggested by Gordy and Stanford6 and substantiated recently by other work.8 The superior basicity of cyclic ethers has been amply demonstrated in systems using other solvents and other acid^^,^,^ and is powerfully confirmed by our results. Perhaps the most striking result in the above data and our other results on ether b a s i c i t i e ~is ~ . their ~ general failure to correlate with the basicity order of the corresponding amines in aqueous solution and (4) For comparison, n o t done b y t h i s method. See: ( a ) E. M. Arnett a n d C. Y . Wu, Chem. and I n d . 1488 (1959); (b) E. M . Arnett a n d C. Y. Wu, THISJ O U R N A L , 8 2 , in press (1960). ( 5 ) S. Searles, Jr., and M. Tamres, ibid., 73, 3704 (1951). (6) W. G o r d y a n d S. C. Stanford, J. Chem. P h y s . , 0, 204 (1941). (7) M. Brandon, If. Tamrea a n d S. Searles, Jr., T H I S J O U R N A L , 8 2 , 2129 (1960). (8) E. M. Arnett a n d C. Y. Wu, Paper 9 7 , Division of Organic Chemistry 136th Meeting, American Chemical Society, Atlantic City, h’. J , S e p t e m b e r , I!259. (9) H C . Brown a n d K . M. A d a m 3 , J . A m . C h r m . S o i . , 6 4 , 2557 (1942).
Vol. 82 therefore with Taft's U* values.10 This reflects an entirely different balance of inductive, steric and solvation forces in oxygen bases than in their nitrogen a n a l o g ~ e s . ~Very , ~ ~ likely solvation is the reason for the large disparity between our measurement of the pKa of dioxane in water and that estimated for the same compound (-4) by Lemaire and Lucas12using perchloric acid titration in glacial acetic acid. Acknowledgment. We are grateful for a very helpful discussion of solvent extraction with Professoi Quintus Fernando of this department. This work was made possible by a research grant from the Public Health Service, A-3G33 B.B.C. of the National Institutes of Health. (10) H. K. Hall, Jr., J. A m . Chem. Soc., 79, 5441 (1957). (11) P. D. B a r t l e t t a n d J. D. hIcCollum. ibid., 18, 1441 (1956). (12) H. Lemaire a n d H. J. Lucns, ibid., 73, 5198 (1951). CONTRIBUTION K O . 1070 DEPARTMENT OF CHEMISTRY UNIVERSITYOF PITTSBURGH
Xu
I
J.
j.
X
I1
IJI
li'
EDWARD M. A n r i E T T CHINGYONG\Vu
PITTSBURGH 13, PENNSYLVANIA RECEIVED JULY 20, 1960 xi11
A FREE RADICAL WAGNER-MEERWEIN REARRANGEMENT
among other products, some isocamphane (VII), which is structurally related to I1 in the WagnerSir: Meerwein sense. The failure of alkyl groups to migrate to adjacent Among the other products of the decomposition free radical centers has been demonstrated re- of VI are camphane (11) and 9-nienthene (VIII). peatedly.' I n most of the cases investigated,2 The formation of 1'111 suggests that a t high temalkyl group rearrangements do not compete ob- peratures the bornyl radical V suffers ring-opening servably with such processes as hydrogen abstrac- by @-elimination; cleavage of bond m would then tion, aryl group migratioc, dimerization, and dis- lead to an unsaturated monocyclic radical (XIV) proportionation. We now report a study of a which would give VI11 by hydrogen abstraction. system in which alkyl group migration is observed. Of the three remaining forinally possible P-elimiThe results demonstrate the occurrence of a formal nations of V, the one involving bond b is prohibited free radical analog of the IVagner-Meerwein car- by Bredt's rule; the one involving bond e is not bonium ion rearrangement. thus prohibited, and although we have not yet Decarbonylation of camphane-2.carboxaldehyde found the olefin that would result, i t presumably is (I) with t-butyl peroxide a t I 40-15Oo gives mainly present among the several still unidentified reaction the unrearranged product camphane (11). Small products, Cleavage a t bond o would lead to radical amounts of the disproportionation products bor- IX, which could give olefin X by hydrogen abnyleiie (111) and tricyclene (117) also are observed. straction, radical V by re-cyclization a t the original Under these conditions, the bornyl radical (V), position of attachment of the two-carbon chain, which is the proximate product of the decarbonyla- or radical X I by cyclization in the alternative sense. tion, does not rearrange. Generation of the radical Direct evidence that cyclization of I X to V and by decomposition of 2,2'-his-azocamphane V I X I occurs is provided by the decomposition of the a t higher temperatures (255-200') however, gives, azo compound XII. The latter is prepared by the route a-pinene oxide -+ (2,2,3-trimethyl-A3-cyclo(1) T h i s research was supported in p a r t by t h e United S t a t e s Air Force t h r o u g h t h e Air Force Office of Srientific Research of t h e Air penteny1)-acetaldehyde (ca~npholenaldehyde)~ Research a n d Development Command under contract S o . X P 18 campholenaldehyde azine, b. p. 164--1G6°(1.5 mni.), (600)-1544. Reproduction in whole or in p a r t is permitted for a n y n2'D 1.5028, Anal. Found: C , 79.G9; H, 10.75; purpose of t h e United S t a t e s Government. Also supported in p a r t N, 9.21) -+ the corresponding hydrazine + X I I , b y t h e Office of Ordnance Research under C o n t r a c t hTo. DA-04-495ORD-532. We are indebted to these agencies a n d t o t h e Alfred 3GO b.p. lt50-1320 (1.5 min.), ? Z ? ~ D1.485S, X,,, P. Sloan Foundation a n d the Richfield Oil Corporation for support. inp, log E 1.47, A n a l . Found: C, 79.37; H, 11.39; (2) (a) W. H. Urry a n d N. Sicolaides, THISJOURNAL,74, 5163 N, 9.20. Wolff-Iiishner reduction of canipho(1952); (b) F. H. Seubold, ibid., 76, 3732 (1954); (c) S. J. Cristol lenaidehyde gives 1,5,5-trimethyl- 4 - ethyl-1 - cycloa n d G. D. Rrindell, i b i d . , 76, 5699 (1954); (d) G. I h p o n t , R. D d o u und G. Clement, 6 u U . Sor. chim. Francc, 1002 (1951); (e) W. von E. pentene b.p. 68-68.5" (43 mm.), 7tZ5D 1.4429, Doering, hl. F a r b e r , M. Sprecher and K. R. Wiberg. TEISJOURNAL. Anat. Found: C, 86.6s; H, 13.21, which forms two 74, 3000 (1952); ( f ) J. A. Berson a n d W. M . Jones, i b i d . , 7 8 , 6045 nitrosochlorides, m .p. 95-96 O and 122.6- 123O . Ana 1. (1956); (g) J. W. Wilt a n d H. Phillip (Hogan), 1.Or#. Chem., 24, Found (for the 123' form): C, 5S.85; H, 8.85; 441 (1959); (h) J. D. Backhurst, J. C h o n . Soc., 3497 (1059); (i) N, 0.96. The thermal decomposition of XI1 gives S. Winstein a n d F. €1. Seuhold, THISJ O U R N A L , 69, 2916 (1947); (j) D. Y. Curtin a n d M. J. Hurwitz, i b i d . , 74, 5381 (1953); (k) W. H. qualitatively the same products that are obtained Urry a n d M. S. Kharasch, i b i d . . 66, 1435 (1944); (I) J. IVPinstock from VI, namely, VIII, 11, VII, 111, IV and X. a n d S. N. Lewis, ibid , 79, 6243 (195i); (m) C. G . Overberper and H. Gainer, i b i d . , 80, 4561 (195s). (3) E. Arbusov, Bcr., 68, 1434 (1935).
-
(x),
