Sept. 5, 1962
3295
HYDRIDE TRANSFERS INVOLVING CARBONIUM IONS
The individual pentenes were identified by comparison of retention times with those of authentic A.P.I. isomeric pentenes. Control Experiments.-A mixture of pentenes was analyzed, dissolved in ethanol, reseparated by the standard distillation procedure and reanalyzed. Within the experimental error ( &0.570) the mixtures gave identical analyses, showing that the distillation procedure separated the pentenes from the ethanol without fractionation of the pentenes themselves. A mixture of pentenes was heated in the rate bath with ethanolic sodium ethoxide for a period of one week. Subsequent analysis showed no change in composition. The uncorrected olefin compositions (base concentrations -0.2 N , ester concentrations -0.1 Ai) are listed in Table
[CONTRIBUTION FROM
THE
VI. The corrected olefin compositions are listed in Table IV.
TABLE VI UNCoBRECTED OLEFINCoMPoslT1oNs N %PENTYL ARENESULFONATES WITH 0.2 N ETHOXIDE I N ABSOLUTE ETHANOL 2-pentyl sulfonate
p -Aminobenzenep-Methoxybenzenep - ~ ~ l ~ ~
p-BromobenzeneP-Nitrobenzene-
F , ~ , , ~ ~ . ~ .F~ , , ~ ~ ~~~
0.385 .394 ~ ,398 ~ .408 ,425
0.194 197 .199 .199 .198 I
'.'
OF
SODIUM
. F ~~
.
,~ ~ ~~
0.421 .409 ,403 .393 .377
COLLEGE OF CHEMISTRY AND PHYSICS, THE PENNSYLVAXIA STATE UNIVERSITY, 'C'SIVERSITY P A R K , PENNA.]
Carbonium Ions. X.
Hydride Transfers Involving Arylcarbonium and Alkylcarbonium Ions BY N. DENO,G. SAINESAND M. SPANGLER' RECEIVED DECEMBER 20, 1961
A linear relationship between log k and A ~ K Rwas + found by Dauben and McDonough (ref. 3) for three series of hydride transfer reactions. This same relation has now been found for the hydride transfer reaction between xanthene and a series of four triarylmethyl cations. For a given APKR+value, the hydride transfer for xanthene is 103-104faster to diarylmethyl cations than t o triarylmethyl cations. For hydride transfers from aliphatic alcohols, esters and ethers to the diphenylmethyl cation, rates can be achieved in polyphosphoric acid that are a t least 104 faster than in aqueous sulfuric acids. This effect is interpreted in terms of acidity functions. The acid-catalyzed equilibration of aliphatic alcohols and ketones has been observed. The failure of t-butyl alcohol to abstract hydride from xanthene in 55% HtSOd is discussed.
Linear Free Energy Relations.-Dauben and to a t least 60% completion of the reaction and were McDonough3 demonstrated that for three reaction independent of the concentrations of reactants. series (triarylmethyl cations cycloheptatriene, Triphenylmethyl cation showed no variation in triarylmethyl cations trianisylmethane and tri- k between 55 and S70 H2S04 and 4-methoxyphenylmethyl cation triarylmethanes) , each triphenylmethyl cation showed no variation in k characterized by constant steric requirements a t between 45 and 50y0H2SOd.j the reaction center, a plot of log k against A ~ K R + TABLE I was linear. The symbol k is the rate constant and FOR THE HYDRIDE TRANSFER REACTION A ~ K Ris+the difference in ~ K Rbetween + the cation RATE COKSTAKTS AND ARYLALKY L CATIOXS BETWEEX XANTHENE reacting and the cation f ~ r m i n g . ~ For a given Devn. in a p K ~ +the , rate of hydride transfer from cyclolog k from Log k (k in l./mole correlating heptatriene to triarylmethyl cations was about Cation sec.) APKR"" eq. lo2 faster than for transfer between triarylmethyl Triphenylmethyl 1.39 5.79 0.14 systems. 4-Methoxytriphenylmethyl -0,907 2.56 0 The above kinetics were conducted in aceto,593 4.40 -0.12 nitrile solutions using perchlorate salts of the cat- 4-Methyltriphenylmethyl ions. \Ye have confirmed Dauben's relation using 4,4',4"-Trimethgltriphenylmethyl - ,807 2.72 0 sulfuric acid solutions, xanthene (I) as the hydride 4,4'-Dimethoxydiphenyldonor and a series of four triarylmethyl cations. methyl 3.29 4 . 8 7 (-0.60) The data fit the linear equation log k = 0.76. l,l-Bis-(4'-methoxyphenyl)( a p K ~ +-) 2.86, as shown in Table I. The rate ethyl 1.31 4.66 ( - 2 . 4 2 ) constants for the six cations studied were invariant
+
+ +
a ~ K R of + xanthyl cation minus PKR+ of the arylalkyl cation.
/
I (1) This research was s u p p x t e d in part b y a grant from the Petroleum Research Fund of the American Chemical Society and in part by a grant from the r a t i o n a l Science Foundation. Grateful acknowledgment is hereby made o f this support. ( 2 ) Research participant i n t h e National Science Foundation program for college chemistry teachers. (3) H. Dauben, J r , and L M . Mcnonough, P h . D . Thesis of L. M., University of U'ashington, 1960. (4) T h e method of evaluating p K x + is given in N. Deno, J. Jaruzelski and A. Schriesheim, J . A m . Chem. SOC.,77, 3044 (1965).
The linear relation between log k and APKR+ is equivalent to a linear relation between log k and u+ because A ~ K R and + CT+ are linearly related.6 Steric Effects.-The steric effect of replacing one of the aryl groups on a triarylmethyl cation by first methyl and then hydrogen can be estimated from the data in Table I. Since ~ K Rfor - dnzCH+ ( 5 ) Experimental details appeared in t h e P h . D . Thesis of G . Saines, t h e Pennsylvania State University, 1961. T h e rates were followed spectroscopically using t h e Xmnx of t h e triarylmethyl or diarylalkyl cation. (6) H. C . Brown and Y. Okamoto, J. A m . Chem. SOL.,79, 1913 (1957); N. Deno end W.L. Evans, i b i d . , 79, 5804 (1957).
~~
~~ ~ ~ ~ ~
AND M. SPANGLER N. DENO,G. SAINES
3296
(An = anisyl) and A4n2C+(CH8) are close in value and since d lng k / d A ~ K Rdoes + not vary more than 2- to 3-fold for the reaction series studied, a measure of the steric effect is the difference between the observed log k for the diarylalkyl cation and the value estimated by the correlating equation for the triarylmethyl series, log k = 0.76 ( A p k ' ~ +) 2.86. Replacing aryl by niethyl iricreases the rate (for constant A ~ K R +&fold ) and replacing aryl by hydrogen increases the rate 3600-fold. In the transition state of intermolecular hydride transfers, the C+-H--C+ system must be linear. Molecular models show that much deviation from linearity, particularly in the transfer of hydride between triarylniethvl cations, leads to interpenetration of atomic radii. For 1,5-transannular hydride models indicate that a linear arrangement is most probable. This is in contrast to 1,2-intramolecular hydride shifts in which the C-H-C system must form a triangle. Hydride Abstraction by Mineral Acid s.-A major difficulty arose in studying the hydride transfer reactions in which xantheiie acted as the hydride donor. The quantitative conversion of xanthene to xanthyl cation proceeds a t appreciable rates in sulfuric acid concentrations as low as 55%. The rate of oxidation increases with acid concentration so that in %yoHzS04 the oxidation of xanthene to xanthyl cation is instantaneous. In polyphosphoric acids from 6O-S3% P205,the rates were slow requiring hours for half completion of the oxidation. Similar oxidations to cations mere found for 4,4'dirnethoxydiphenylmethane and 4,4',4''-trimethosytriphenylmethane in 60-8G% H2S04. I n the polyphosphoric acid systems, the reaction may be a light-catalyzed air oxidation, since in 85% phosphoric acid degassing the system by repetitive freezing and evacuating combined with conducting the reaction in the dark led to marked reduction in the rate of oxidation. However, degassing and darkness were not effective in reducing the rate in 73% sulfuric acid. Relative Rates in Polyphosphoric and Sulfuric Acids.-Bartlett and McCollums reported that triphenylniethyl cation abstracted hydride from ethanol, 1-propanol, 2-methyl-1-propanol, %propanol, 2-butanol, 3-methyl-2-butanol, 3,3-dimethyl2-butano1, diethyl ether and diisopropyl ether. Although methanol and dioxane failed to react, presumably they would also have donated hydride to the cation if it were possible to suppress their protonation while retaining the triphenylmethyl system in the cation form. It is evident that to achieve hydride transfer with alcohols, esters and ethers, a system is needed which, relative to sulfuric acid, favors carbonium ion formation ROH 3. H + = R + $- H?O
more than protonation (7) A , C. Cope, G. A. Berchtold, P. E. Peterson and S. H. Sharman, J. A m . Chem. Soc., 82, 6366 (1960). ( 8 ) A . C. Cope, Abstracts of t h e Am. Chem. SOC.Meeting, March, 1001, p 19-0; V. Prelug, i b i d . , p. 20-0. (9) P. 1). Bartlett and J. D. McCollum, ibid., 78, 1141 (1956); for R m u r e Aetniled study v i t h 2-propano1, see S. G. Entelis, G. V. Epple aud S . bl. Chirkov, Doklady A k a d . N a u k , 136, 667 (1961).
B
Vol. 84
+ HC = BH+
,4n obvious solution is to decrease the water activity and for this reason polyphosphoric acid systems were investigated. Qualitative indicator studies, summarized in T&le 11, showed that carbonium ion formation, as measured by the HR acidity function, was favored relative to protolintion, 3s measured by the Ho function, v,lien compared with sulfuric acid systems.1° .kPPROXIMATE
yo PtOa in water
HR
ASD
( - €10)c
TABLE II Ha VALUES I N ACIDS 7$H2SO4 with same IIo value
POLYPHOSPHORIC
-HR
with '% H same ~SOI IIlr value
11 A8 12.5 74 14 80 4 15 83 5.5 70 15 83 7.5 81 16 87 85% H3POa. Commercial polyphosphoric acid. These values are in agreement with those given in ref. 10.
61.5" 65.8 70. 1 74.4 78.7 83.Ob
3.8 4 4
65 67 57 57
(I
Using the diphenylmethyl cation as the hydride acceptor, rates were measured in aqueous 83% H2S04 and 72% Pz05. These acid concentrations were chosen as being somewhere near to the acidity for optimum rate." The rate constant for 2 propanol, cyclopentanol, cyclohexanol, allyl alcohol, 1,2-dimethoxyethane, and dioxane were of the order of magnitude of 1 liter/mole sec. a t 2.5'. Ethyl ether and 3-oxa-lj5-pentanediol were a factor of 5-10 slower. 2-Propyl benzoate and ethanol were about 1/100 as fast. Borneol, methyl benzoate, ethyl acetate, butyrolactone and valerolactone were so slow that reaction was not detected. In 83% H2S04,every one of these compounds failed to show significant reaction rates so that these rate constants must be a t least smaller. All of the rates were followed spectroscopically using the 442 mp A,, of the diphenylmethyl cation. With triphenylmethyl cation, the effect was less dramatic because the greater stability of this cation allowed the use of acid concentrations close to those in which free alcohol exists in comparable concentration with the protonated form. The rate constants were much smaller, in keeping with the greater steric hindrance and greater stability of the triphenylmethyl cation. 2-Propanol, cyclopentanol, cyclohexanol and allyl alcohol were oxidized in 627* H2S04at rate constants of the order of magnitude of 10-3 liter/mole see. a t 25'. The rates in 68% P205were about twice as fast. The acid concentrations were chosen to give rapid rates, as with the diphenylmethyl cation.'l All of the other compounds tested with the diphenylmethyl cation were also tested with triphenylmethyl cation in 62y0 H2S04and 68% PzOj, but (10) T h i s accounts for t h e observation of D. E. Pearson ( J . A m . Chcm. Soc., 83, 1715, 1718 (1961)) t h a t t h e rates of Beckmann rearrangements are much faster in polyphosphoric acid t h a n in sulfuric acid solutions of t h e same Ho value. (11) T h e rate is proportional t o t h e product of t h e concentrations of t h e reactants, R + and t h e hydride donor (ref. 9). T h e acid concentration a t which this product maximizes can be computed using t h e Ho function t o calculate the concentration of hydride dunor (free alcohol, ester or ether) and tlir H R (Cu) function (ref. 4 ) t u calculate CE+
.
HYDRIDE TRANSFERS INVOLVING CARBONIUM IONS
Sept. 5, 1962
3297
no significant reaction was detected. The rates based on added xanthene. Isobutane was not were again followed spectroscopically using the formed when the xanthene was omitted. 434 rnp Amax of the triphenylrnethyl cation. 2-Methyl-3-buten-2-01 was the one aliphatic The direction of hydride transfers gives an order alcohol of many tested that clearly formed a xanthyl of stability of carbonium ions based on extent of cation from xanthene much faster than the direct the reaction RH = Rf H-. For example, oxidation. However, the reaction is complex CII:CH+OH and CH3CH+OCH&H3 are more since the absorption spectrum, although closely stable thzn triphenyltnethyl cation. resembling xanthyl cation, had, , ,A a t 385 mp inAcid-catalyzed Equilibrsltion of Aliphatic Ketones stead of 375 mfi, suggesting that an alkylated xanaqd Alcohols.-In refluxing 60% sulfuric acid, thy1 cation had been produced. The similar 22-butanone oxidized 2-propanol to acetone in a methyl-3-penten-2-01 and 2,4-dimethyl-3-penten-2yield of 5%. Under the same conditions, cyclo- 01 failed to abstract hydride faster than the direct hexanone oxidized 2-propanol to acetone in 2S% oxidation. yield. Both yields are calculated on the basis of It is significant that t-butyl alcohol fails to the 2-propanol added. Increzsing the acid con- abstract hydride from xanthene in 60% H~SOI. centration to 65y0or decreasing it to 55yoreduced On the basis of alkylation of acrylonitrile studies14 the vields of acetone. The acetone was contin- and by analogy to the protonation of 2-propan01,~ uously distilled off so that the yields do not repre- t-butyl alcohol is half protonated in 56% HzSOa sent concentrations present a t equilibrium. and it is the generation of protonated t-butyl alcoI t was anticipated that the reaction would be hol, (CH3)3C-OH2+, that quantitatively accounts most rapid in 60-G570 HzS04. From the pK for for the acid catalysis in nitrile a l k y l a t i ~ n . ' ~The acetone," the pK of 2-propanols and the Ho function, it was computed tha.t the product of the encumbered carbonium ion proposed for the dehyof t-butyl alcohollj is regarded by us as concentrations of the reactants, protonated ketone dration and free alcohol, would maximize in that acidity, conceptually equivalent. The abundance of (CH3)3C-OH2+ in 60% HpS04 and the failure to a t least a t 2.5'. abstract hydride from xanthene in this acid means t-Alkyl Cations.-An effort was made to detect that protonated t-butyl alcohol, although capable hydride transfer between xanthene and simple of alkylating nitriles, dehydrating to olefins and aliphatic alkyl cations. With t-butyl alcohol, no exchanging 0 1 6 , l 5 is~ ineffective in hydride transfer reaction was observed in 55y0HzS04. In more concentrated acids, although abstraction of hydride by reactions, presumably because it does not possess the t-butyl cation was overshadowed by the direct a sufficiently open sextet of electrons. Hydride transfer reactions of the t-butyl system oxidation by the sulfuric acid, a trace of hydride transfer was detected.13 Nitrogen was passed are found in concentrated sulfuric acid,17where the through a reaction mixture of xanthene and t- greatly lowered activity of water has shifted the butyl alcohol in 75y0H2SOa. Passage of the nitro- equilibrium from (CH3)3C-OHz+ to (CH3)3C+. gen stream through a liquid air trap and examining (14) 3 . Deno, T . Edwards and C. Perizzolo, I . A m . Chem. Soc., 79, the condensate by vapor phase chromatography 2108 (1957). R. H.Boyd, R. W. Taft, Jr., A. P. Wolf and D. R. Christman, revealed the formation of a O.OS~oyield of isobutane i b i(15) d . , 82, 4729 (1960).
+
(12) H . J. Campbell and J. T. Edward, Can. -7. Chem., 36, 2109
(lnoo), (13) These measurements were kindly made b y Dr. H. Skovronek.
(16) I. Dostrovsky and F. S. Klein, J . Chem. Sac., 791,4401 (1955). (17) J. W. Otvos, U. P. Stevenson, C. D. Tn'agner and 0. Beeck. J . A m . Chem. Soc., 73, 5741 (1951); J . Chem. Phrs., 17, 419 (1949).
