EUGENEH . CORDES AND WILLIAM P. JENCKS
832 [CONTRIBUTION No. 137
FROM THE
VOl. 84
GRADUATE DEPARTMENT OF BIOCHEMISTRY, BRANDEIS UNIVERSITY, WALTHAM 54 MASS.]
On the Mechanism of Schiff Base Formation and Hydrolysis1 BY E. H. CORDES AND W. P. JENCKS RECEIVED SEPTEMBER 5, 1961 The following lines of evidence suggest that N-p-chlorobenzylideneaniline formation, like that of oximes and semicarbazones, undergoes a transition in rate-determining step from rate-determining dehydration of the carbinolamine addition product a t neutral pH t o rate-determining amine attack under acidic conditions: (1) A pH-rate maximum, which is not due to general acid catalysis, occurs near pH 4. (2) The hydrolysis of N-p-chlorobenzylideneaniline shows a break in the pH-rate profile consistent with a transition in rate-determining step. (3) The influence of structure on reactivity in Schiff base formation is different on the two sides of the pH-rate maximum. (4) Schiff base formation is more susceptible to general acid catalysis on the acid than on the alkaline side of the pH-rate maximum. The rate of hydrolysis of N-9chlorobenzylideneaniline is markedly decreased in strong acid; the logarithm of the rate follows HOwith a slope of 1.24 in the range 7 t o 11 M sulfuric acid. The rate of hydrolysis in strong acid is decreased approximately twofold in deuterium oxide solution. These results suggest the involvement of several water molecules, functioning as both nucleophilic and proton transfer reagents, in the transition state. The pK,' of N-p-chlorobenzylideneanilinium cation has been estimated to be 2.8.
Introduction Evidence suggesting the involvement of azomethine intermediates in several enzymatic rea c t i o n ~ * has - ~ focused attention on t h e mechanism of Schiff base formation and hydrolysis in aqueous solution. Previous studies of the hydrolysis of Schiff bases in water-alcohol mixtures have established the susceptibility of these reactions to specific and general acid c a t a l y s i ~ . ~ -However, ~ the nature of the rate-determining step (eq. 1 and 2) remains unsettled. Evidence obtained by Willi suggests that in the hydrolysis reaction the addition of water to form the carbinolamine intermediate
behaves in a manner similar t o oxime and semicarbazone formation. The formation of Schiff bases in aqueous solution has not been the subject of kinetic investigation, since the low equilibrium constants for these reactions make such a study difficult.lOell Results presented in the preceding communication12demonstrate that Schiff base formation is the rate-determining step in aniline-catalyzed semicarbazone formation in aqueous solution. Thus, semicarbazide acts as a trap for Schiff base and carries the reaction to completion. This system, used as a tool for the study of Schiffbase formation, permits the direct kinetic study of this reaction in aqueous solution. Experimental
Kinetic measurements were carried out a t 25' as previously described.8,gJ2 The rate of Schiff base formation, NHR + measured as the rate of semicarbazone formation, was obtained by subtracting the rate of a control reaction which (the reverse of eq. 2 ) is rate determining a t neutral did not contain aniline from that of the complete reaction pH,6 while, on the other hand, Kastening, et a1.,6 mixture. Where necessary, the rate constants for the anihave concluded from polarographic measurements line-catalyzed reactions were corrected to 100yo reaction of that the decomposition of the carbinolamine (the the Schiff base with semicarbazide by dividing by a, which determined in separate experiments as previously deieverse of eq. 1) is rate-determining under these was scribed.12 This correction was usually small, involving a conditions. Studies on the related reactions, change of less than 25Y0 in the rate constants. Materials .-All reagents were purified and solutions preoxime and semicarbazone formation have demonstrated a transition from rate-determining dehy- pared as previously described.12 Deuterium oxide, 99.870, obtained from the Atomic Energy Commission through dration (water addition in the reverse reaction, was the courtesy of the Department of Chemistry, Harvard eq. 2) a t neutral p H t o rate-determining amine University. 8.24 M DzSOl was prepared by distilling SOa attack (carbinolamine decomposition in the re- from 30% fuming sulfuric acid into DnO cooled in an isoverse reaction, eq. 1) under acidic c o n d i t i o n ~ . ~propyl ~ ~ alcohol-Dry Ice-bath. DnS04 prepared in this manner was neutralized with NapC03 in deuterium oxide and an The studies reported below were carried out in an infrared spectrum revealed that this solution contained no attempt to determine whether Schiff base formation detectable quantity of H20. The rapid hydrolysis of protonated N-p-chlorobenzylidene(1) Presented in part at the 138th National Meeting of the American aniline makes the direct determination of the absorption Chemical Society, New York, N. Y.,Sept., 1960 (Abst. 46P). Supspectrum of the protonated Schiff base impossible by ported by grants from the National Cancer Institute of the National ordinary techniques. This spectrum was estimated as Institutes of Health (C-3975) and the National Science Foundation. iM N-p-chlorobenzylideneanifollows: 0.06 ml. of 2 X This investigation was carried out by E. C. during the tenure of a Preline was added to 3.0 ml. of methanol containing 1.15 M doctoral Fellowship from the National Heart Institute, United States hydrogen chloride and the absorption spectrum was imPublic Health Service. mediately recorded on a Cary recording spectrophotometer, (2) E. E. Snell and W, T. Jenkins, J. Cell. Comp. Physiol., 64, model 14. Under these conditions, methanol adds to the Suppl. 1, 161 (1959). protonated Schiff base fairly slowly (half-time for the (3) D. E. Metzler, M. Ikawa and E. E. Snell, J. A m . Chem. Soc., 76, addition is approximately 4 minutes). Extrapolation of 648 (1954). the absorption a t the maximum (337 mp) to zero time gave (4) G. A. Hamilton and F. H. Westheimer, ibid., 81, 6332 (1959). a molar extinction coefficient a t this wave length of 2.35 (5) B. Kastening, L. Holleck and G.A. Melkonian, Z . ElcRlrochcm., X lo4. The shape of the absorption spectrum of protonated H + or OH-
+)C=NR
+ €I20 (2)
60, 130 (1956). (6) A . V. Willi, Hclw. Chim. A d o , 39, 1193 (1956). (7) A. Willi and R . E. Robertson, Con. J . Chcm., 31, 361 (1953).
(8) W. P. Jencks, J . A m . Chem. SOL.,81, 475 (1959). (9) B . ZI.Anderson and W. P. Jencks, i b i d . , 81, 1773 (1960).
(10) P. Zuman, Coll. Czech. Chsm. Comm., 16, 839 (1981). (11) 0. Bloch-Chaude, C o m p f . rend., 139,804 (1954). (12) E. H. Cordes and W. P. Jencks, J . A m . C h e m . .So[, 84, 82(i
(1962).
March 5 , 1962
SCHIFF
BASEFORMATION AND HYDROLYSIS
--
P
'
0.50
. L
833
'
'
O
r
I
.. . . . . '
L
.
.
'
.
'
.
'
.
300 350 400 mp WAVE LENGTH, Fig. 1.-Ultraviolet spectra of 5.0 X 10-6 151 N-p-chloro,in 0.1 A l benzylideneaniline and its conjugate acid: tris-(hydroxymethy1)-aminomethanebuffer, pH 9.5; . . . . . . , in concentrated sulfuric acid; * . -, in methanolic HCl (1.15 11);- - - -, iii 1 M aqueous HCI. 250
-
- -
N-p-chlorobenzylideneaniline in water was obtained from M pan equilibrium mixture containing initially 1 X chlorobenzaldehyde and 0.3 M aniline hydrochloride in 1 M hydrochloric acid Absorption due to the free aldehyde and anilinium ion was determined independently and subtracted. The molar extinction coefficient a t the absorption maximum in water (335 mp) was assumed to be the same as in methanol. The spectra of this Schiff base in water, methanolic HC1, dilute aqueous HCl and concentrated sulfuric acid are shown in Fig. 1. The addition of methanol to protonated N-p-chlorobenzylideneaniline was studied as follows: 0.06 ml. of 2 X 10-8 M N- p-chlorobenzylideneaniline was added to methanol containing 1.15 M HCl, and a series of absorption spectra was taken at increasing time intervals. The absorption at 337 m p , due to the protonated Schiff base, decreased with time and was zero after 45 minutes. At this time no appreciable absorption a t 260 mp was seen, indicating that no appreciable amount of p-chlorobenzaldehyde was formed. Addition of a small amount of water caused the immediate appearance of an absorption band with a maximum near 260 m+ The product of this reaction was identified as p-chlorobenzaldehyde as follows: 420 mg. of N-p-chlorobenzylideneaniline was added to 10 ml. of methanol containing 0.75 M HCl and allowed to remain a t room temperature for 45 minutes. Dilution with water gave a precipitate of 202 mg. of p-chlorobenzaldehyde (m.p. 46-48') which was converted to the semicarbazone (m.p. 232-233"). It was not determined whether the product of the addition of methanol to the Schiff base is the acetal or the aminoacetal.
Results In Fig. 2 , the pseudo-first-order rate constants for the formation of N-p-chlorobenzylideneaniline a t 25' are plotted against PH. All points on the $H-rate profile have been extrapolated to zero buffer concentration. A rate maximum occurs near p H 4. On the alkaline side of the pH-rate maximum, the rate of Schiff base formation increases with increasing acidity, suggesting a predominantly acid-catalyzed reaction in this P H region. On the acid side of the pH-rate maximum, the rate decreases less rapidly than the concentration of free aniline and becomes independent of p H below PH 1,indicating the existence of a specific acid-catalyzed reaction as well as an uncatalyzed reaction of free aniline (eq. 3) rate = kl[RSH*][R'CHO] = k1[RNH2][R'CHO]
+ kz[RNHzl[R'CHO] [H+]
+ KPr[RPL'H8+][RCHO]
(3)
P H. Fig. 2.-The rate of formation of N-p-chlorobenzylideneaniline as a function of pH a t 25': p-chlorobenzaldehyde 3.3 X lo-' M , total aniline 0.004 M,total semicarbazide 0.0025 M , ionic strength 0.50. The buffers used are listed in Table I. All rates are extrapolated to zero buffer concentration.
5001 I
kobs
t
50
[Aniline]
0
pH 6.05 0 p + ao.0
IO
-
I
/
'9 /
a
5-
'
I -1.0 -0.6 -0.2
t O . 2 +0.6 + I . O
6 +
Fig. 3.-Logarithm of the second-order rate constants for the condensation of aniline with p-substituted benzaldehydes a t 25' plotted against u + ; pH 2.55, 0.04 M chloroacetate buffer; pH 6.10, 0.10 M phosphate buffer; ionic strength 0.50
The effect of structure on the rate of Schiff base formation from aniline and several p-substituted benzaldehydes was studied a t p H 2.55 and DH 6.10 (Fin. . - 31. At DH 2.55, on the acid side of the pH-rate maximum-, a plot of the logarithm of I
83.1
34
TABLEI CATALYTIC CONSTAXTS FOR GEXERAI, ACID CATALYSIS OF S-p-CHLOROBENZYLIDEXEANILISE FORMATION AT 250a Concn. range,
PK*
Acid 2 .c
I .c
i 0
-h C
Q
I
n X
-
pl
0
-1.c
-2
c
M
pH
kcat, k - 2
mh-1
2.45 0.015-0.0782 . 6 6 7 . 4 X lo4 2.86 .02 - .10 2 . 6 8 7 . 7 X l o 4 2.90 ,017- ,086 3.01 5 . 6 X lo4 3.75 ,035- .17 2.98 3 . 1 X l o 4 4.76 .05 - .30 2.50b 2 . 3 X l o 4 4.76 ,013- .10 5.10 2 . 5 X l o 3 Propionic 4.87 , 0 5 - .30 2 . 4 1 2 . 9 X l o 4 Ionic streiigtli maintained a t 0.50 by the addition of KCl. In the presence of 0.0:',I1 hrornoacetate buffer. Cyanoacetic Bromoacetic Chloroacetic Formic Acetic
TABLEI1 CATALYSIS O F hT-p-CHLOROBEXZYLIDENEAX1LI?iE FORMATIOX BY CYANOACETIC ACID< AniCyanok2, "E; line,'l
acetic
x.tl105
acid,b
M
ki , min:.'
M-! Ah!, niin.-'
min. -1
X
min.-'
X 10-3
n
10-1
0.00 0.0152 0.149 7.45 ,0152 ,513 0.364 4 , 9 0.74 6.65 0.00 ,0304 ,198 Fig. 4.-The hydrolysis of S-p-chlorobenzjlideneaniline as 7.45 ,0304 ,660 ,462 6 . 2 . 74 8.40 a function of pH at 25': 0, in H2O; phosphate (0.05 iW, 0.00 ,0451 .246 pH 5.6-7.6), tris-(hydroxymethy1)-aminomethane(0.05 M , 7.65 ,0451 ,770 ,524 6.85 .74 9.25 pH 8.4), carbonate (0.05.If, pH 9.9-10.6), reactions in 0.00 ,0775 .347 strongly acidic solution carried out in sulfuric acid; A, in 7.30 ,0775 ,955 ,608 8 . 3 5 .74 11.30 D2S04-D?O; 0,in 80% ethanol, in dilute HC1 or formate " As the free base. As the undissociated acid. Carbuffer (pH 3.9-5.8), rates extrapolated to zero buffer con- ried out a t pH 2.66 in the presence of 0.005 31 semicarbazide a t 25". centration. S-p-Chlorohenzylideneaniline 3.3 - 7 X .If; ionic strength 0.50 except in 80ycethanol (0.05) and con The hydrolysis of Tu'-P-chlorobenzylideneaniline ceritrated sulfuric acid. Rcactiutis followed at 320 or 310 was studied in the p H range 5.10 to 14. The pHIll!.&.
the second-order rate constants against u +-substituent constantsi3 gives a straight line with a p+-value of 0.39. On the alkaline side of the pH-rate maximum, on the other hand, the secondorder rate constants are independent of the substituent except for the #-OH and p-OCHa substituents, which show decreased rates relative to the unsubstituted compound. The rate of N-p-chlorobenzylideneaniline formation increases with increasing buffer strength a t constant pH. This catalysis is of the general acid, rather than the general base, type since acetic acid and propionic acid are effective catalysts a t pH 2.5, more than two pH units below the pKa of these acids. The catalytic constants for six aliphatic carboxylic acids, measured below fiH 4, are given in Table I . Representative data, illustrating the method of calculating the catalytic constants, are given in Table I1 for cyanoacetic acid catalysis. h s is seen from Table TI, several corrections must be macle to obtain the catalytic constants, which may, therefore, be in error by as much as *30%. The logarithms of catalytic constants for carboxylic acids and for the solvated proton (1.7 X lo6 min.-') fall near a Bronsted line of slope 0.25 when plotted against the appropriate PK, values. The catalytic constant for acetic acid obtained a t p H 5.10 is approximately tenfold lower than a t p H 2.50 (Table I). (13) H . C. Brown and Y. Okamoto. J . A m . Chem. Soc... SO.. 4878
(ii5sj.
rate profile (Fig. 4) shows base-catalyzed, acidcatalyzed and uncatalyzed (or solvent-catalyzed) reactions and is similar to the pH-rate profile obtained by Kastening, ef ~ 1 Willi . ~ had observed a break in the PH-rate profile in the acid-catalyzed region in the hydrolysis of N-benzylideneaniline in 50% methanoL6 I n water, this break occurs a t a PH a t which the rate of hydrolysis is too fast to measure. Such a break is observed, however, in the hydrolysis of N-p-chlorobenzylideneaniline in 80% ethanol (Fig. 4). Rates measured in 807G ethanol have been extrapolated to zero buffer concentration and p H values in SOYo ethanol were determined with a glass electrode and the Radiometer PHM-4b p H meter, standardized with 0.01 M HC1 in SOYo ethanol.'? In strong acid, the rate of hydrolysis of this Schiff base again becomes slow enough to measure. I n the range 7 to 11 ,$sulfuric .I acid, the logarithm of the rate of hydrolysis follows HOi5 with a slope of 1.24. In deuteriosuliuric acid, the rate is decreased by about twofold compared to the rate in sulfuric acid (Fig. 4). I n the range pH 5 to 7 A I sulfuric acid, the rate of hydrolysis of N-p-chlorobenzylideneaniline is too fast to measure. The rate of hydrolysis in this region has been calculated as follows: The pH-independent first-order rate constant for the formation of Schiff base is 0.19 min.-l in the presence of 0.004 M anilinium ion (Fig. 2), giving a value of K 1 (eq. 4) of 47.5 M-l inin.-'. The equi(14) V. Gold, "pH hleasurements," John \Tiley a n d Son, Inc., New York, S . Y., 1956. (13) hI. A . Paul and F. A Imng, Cheiiz. Revs.. 57, 1 (19>7).
IRXITION AND
>C=O
+ RNH3+
ki Schiff base,H+ K1
k-
(4)
1
libriuni constant for this reaction, K1, is 0.080 M-l (see below), and L 1is, therefore, 595 min.-l (dotted line in Fig. 4). The shape of the curve on the alkaline side of the rate maximum is determined by the pKa of the protonated Schiff base. In order to maintain the continuity of the over-all pH-rate profile, a break in the curve similar to that observed in SOY0 ethanol must occur, although the exact position of the break is not determined by the available data. The equilibrium constant, K1, defined as
HYDROLYSIS
835
a t 310 mp to the absorbance a t 340 mp (due mainly to the protonated Schiff base) was plotted against pH over the range pH 2.05 to 4.30. From the absorption ratios a t pH values a t which the Schiff base is completely protonated and completely unprotonated and the data in Fig. 1, a PK,' for the Schiff base cation of 2.75 was calculated.
Discussion
Three lines of evidence indicate that Schiff base formation undergoes a transition in ratedetermining step from rate-determining dehydration of the carbinolamine addition product a t neutral pH to rate-determining amine attack under acidic conditions. First, seinicarbazone and oxime (Schiff base.H-) formation, reactions related to Schiff base formaK ' - (aldehpde)(aniline.HT) tion, exhibit a pH-rate maximum similar to that for the addition of aniline to p-chlorobenzalde- in the formation of N-p-chlorobenzylideneaniline hyde under acidic conditions was measured a t (Fig. 2). The former reactions proceed by the twoseveral wave lengths by the method of Bloch- step reaction mechanism shown in eq. 1 and 2. Chaud6." The conditions were chosen so that Near neutral pH, the acid-catalyzed dehydration absorption due to the protonated Schiff base was of the carbinolamine addition product is rate much more important than that due to the un- determining. Such a rate-determining dehydraprotonated species. The extinction coefficients tion may account for a leveling off, but not for the protonated Schiff base were taken from a decrease in the rate with decreasing pH.8 the data in Fig. 1. The results are given in Table Xt low pH values the rate of dehydration 111. The value of O.OS0 iW1 for K 1 , together becomes very fast and, a t the same time, the rate of amine attack is retarded because of TABLE I11 conversion of the amine to its conjugate acid. EQUILIBRIUM COSSTASTFOR FORMATIOS OF K-~~-CHLOROConsequently, a change in rate-determining step BENZYLIDESEANILINIUM CATIOS IS ACIDIC SCLUTIOS AT occurs and the attack of the amine (eq. 3) becomes 2zoa rate determining, accounting for the observed deWave length, crease in the rate on the acid side of the pH-rate mii 0H R I ,iW-1 maximum. The opposing effects of general acid 340 2.05 0.08i catalysis and protonation of the attacking arninel6 340 2.20 ,087 cannot account for the pH-rate maximum in 340 2.40 .087 Schiff base formation since the rate constants (Fig. 340 2.55 .OB7 2 ) have been corrected for the effects of general acid 340 2.90 .087 catalysis, which are small under the experimental 360 -9. 05 ,081 conditions employed. In the reverse reaction, 360 2.20 ,087 Schiff base hydrolysis, a change in rate-determin360 2.40 ,079 ing step requires a break in the acid-catalyzed region ,076 360 2.55 of the pH-rate profile. Such a break has been 360 2.90 . OS7 observed by Willi in the hydrolysis of N-benzylid380 2.0% . 070 eneaniline in 50yo methanoL6 Willi's interpreta38U 2.35 ,072 tion of this break, although stated in somewhat 380 2.70 ,076 different terms, is equivalent to postulating a ,066 380 3.05 change in rate-determining step. In the hydroly380 3.50 .092 sis of N-p-chlorobenzylideneaniline in water, this Average 0.080 break is not seen in the region in which the rate is 2.0 X l W 3 JI p-cliloro,eiizaldcli~de, 0.05 JI aiiiline li>-drucliloride,0.10 JI broinoacetate buffer, ionic strength slow enough to measure. In SOY0 ethanol, however, this break occurs (Fig. 4) and is evidence for 0.50. a in rate-determining step. The pH prowith the dissociation constant for anilinium ion, filechange for hydrolysis water (dotted line), calculated K B ,and the equilibrium constant for the formation from the rate andinequilibrium constants for Schiff of the unprotonated Schiff base, Ks, allows the base formation and the PK,' of the protonated calculation of the dissociation constant for the Schiff base, exhibits a break similar to that found protonated Schiff base, K 4 , since in 80% ethanol. K, KdG/Ki Second, a comparison of the effects of structure Since'* Ks = 4.55 M-l, Kh is 1.57 X M-l and on reactivity on the two sides of the PH-rate maxithe PKa' of N-p-chlorobenzylideneanilinium cation mum provides further evidence for a change in rate-determining step in Schiff base formation is 2.80. The pKa' of N-p-chlorobenzylideneaniline cation (Fig. 3). -4t pH 2.55, in the region of rate-demay be determined independently from the change termining amine attack, the second-order rate in the shape of the Schiff base absorption curve constants for the addition of aniline to substituted with pH by the following graphical method: (16) L. P. H a m m e t t , "Physical Organic Chemistry," >ZcGraw-Ilill The ratio of absorbance due to total Schiff base Book Co., Inc., S e w York, N . Y., 10i0, 11. 333. (L
836
EUGENE H. CORDESAND WILLIAMP. JENCKS
VOl. SA
benzaldehydes follow u + with a slope of 0.39. Third, the differing susceptibility of S-pAt this pH the small p+-value may reflect an chlorobenzylideneaniline formation to general acid incomplete cancellation of the opposing effects catalysis on the two sides of the pH-rate maximum of polar substituents on protonation of the sub- is evidence for a change in rate-determining step. strate and attack of the amine. The correlation Although both steps are subject to general acid of the rates with u+, rather than u, is not surprising catalysis, the catalytic constant for acetic acid is since the carbonyl carbon atom will possess some approximately tenfold greater a t pH 2.50 than carbonium ion character iii the acid-catalyzed re- a t pH 5.10 (Table I). General acid catalysis action. The base strengths of p-substituted ben- of Schiff base formation in methanol has been prezylideneanilines, as measured by their complexing viously reported by Santerre, et u1.,l9 and ]Villi ability with p-nitrophenol in carbon tetrachloride and Robertson7have reported general acid catalysis solution'' and the protonation of aromatic alde- of Schiff base hydrolysis. In these studies it is hydes,'* both follow u+. At PH 6.1, in the region not clear which step of the reaction is being cataof rate-determining dehydration, on the other lyzed. hand, the second-order rate constants for all In moderately concentrated solutions of sulsubstituents except p-OH and P-OCHS are the furic acid, the rate of hydrolysis of 9-p-chlorosame. This behavior is analogous to that ob- benzylideneaniline is slow enough to measure served in semicarbazone formationQ and can be and decreases rapidly with increasing acid conaccounted for in terms of the opposing effects of centration (Fig. 4). In the region 7 to 11 ;I1 polar substituents on the equilibrium constant for sulfuric acid, the logarithm of the observed rate carbinolamine formation and on the rate of acid- constant for hydrolysis of this Schiff base is linearly catalyzed dehydration of the carbinolamine.18a The related to HO (slope 1.24), Ho log CH+ (slope non-linear Hammett plot reported here and those 1.34), and log aH2023.24 (slope 3.12). A number reported for similar reactions, including the con- of acid-catalyzed reactions, including the hydensation of n-butylamine with substituted piper- drolysis of a r n i d e ~ , ~t.he ~ jhydrolysis -~~ of the heteroonals, l9 the hydrolysis of N-ben~ylideneanilines,~cyclic amide N-acetylimidazole,3 0 t 3 1 and the h y substituted benzophenone oxime formation20 and drolysis of the Schiff base 4-(2,3-dimethylanilino)exhibit decreased rates in modsubstituted benzaldehyde semicarbazone forma- pent-3-en-2-0ne,~~ tion in 75% ethanol,21can be accounted for in two erately concentrated acid solutions in which the ways. First, as in the present case, the decreased substrates are completely protonated, and show rate constants for the @-OH and p-OCHs substit- similar behavior with respect to these parameters. uents may be attributed to the unequal relative Such behavior appears to be characteristic ol importance of electron donation by resonance in reactions involving several molecules of mkter the addition and dehydration steps of the reaction. in the transition state, functioning as both nucleoElectron donation by resonance, particularly from philic and proton transfer reagents. The followthe $-position, is more important in the stabili- ing lines of evidence suggest that water functions as zation of the benzaldehyde than in promoting a proton transfer agent, as well as a nucleophilic the dehydration of the carbinolamine intermed- reagent, in the hydrolysis of N-p-chlorobenzylideneiate,22thus accounting for the decreased over-all aniline: (1) General acid catalysis of this reaction rates with substituents which donate electrons demonstrates that proton transfer between substrongly by resonance. Second, a non-linear strate and catalyst is involved in the transition Hammett plot may result from a change in rate- state and suggests that the specific acid-catalyzed limiting step with changing substituent a t a con- reaction is, in fact, general acid catalysis by the stant pH as, for example, in the case of semicar- hydrated proton or its kinetic equivalent. ( 2 ) bazone formation from substituted benzaldehydes The twofold decrease in rate in D?SO( compared strongly suggests proton transfer a t p H 3.7L9 Regardless of the explanation for to that in the particular behavior found a t either of the pH in the rate-determining step. ( 3 ) The value of the values, the important point is that the behavior slope of the logarithm of the rate plotted against is different, supporting the thesis that different lTIois close to that for the rate retardation of protoii rate-determining steps are being studied a t the two transfer in aqueous solution,33and for the rate decrease in the diazotization of aniline in cotipH values. in Nlhich proton transfer centrated acid ( 1 7 ) I. Weinstein and E. AIcIninch, J . A m . Chein. Sor., 82, 6064
+
(1960).
(18) K . Yates and R . Stewart, Can. J . Chem., 37, 664 ( 1 9 Z O ) . (18a) NOTEA m m IN PROOF: E. F. P r a t t and M. J. Kamlet, J. Oq. Chcm., 26, 4029 (1961), have shown t h a t the acid-catalyzed reactions of benzaldehydes with aniline in benzene solution exhibit a P value of 1.54 for benzaldehyde substituents. T h e authors suggest t h a t this p
(23) W.F. Giauquc, E.IL' Hornung, J. E . Kunzler and T. I< R u b i n , ibid.. 82, 62 (1960).
(24) J . F. Bunnett. ibid.. 82, 499 (1960). (25) J . T . Edward and S.C . I< Meacock, J . Chem. SOL.. 2000, 2009
(1957). (26) L). Rosenthal and T . I. Taylor, J A m . C h e m Soc., 79, 2681 value is accounted for by a greater effect of substituents on t h e nucleo(1957). (2;) J . A . Leisten, J. Chem .SOL.,7'65 (19.59). philic reaction than o n pre-equilibrium protonation; it might also be accounted for by a greater effect on pre-equilibrium addition com(28) I < . H . I k W d f e , J A m Chem. SOL.,82, I S 5 (1960). pound formation than on dehydration I n either case it is evident (29) 11. Burkett, W. 11. Scliubrt-t, F. Schultz, I
