I 1-33, The small equilibrium concentrations of EH + action studied in aqueous solutions in both direcmay be held against mechanism 1. For K , = tions w = 2 . 2 for the lactonization of y-hydroxylo7, (EH+) = 10-loM a t pH 3 where acid catalysis butyric acid while w = 6.1 and 8.5 (in HC1 and sets in. This concentration may not be too low for HC104, respectively) for the reverse reaction.2a a reactive species like E H f . Hydroxide ion catal- The difference of 4-6 units for hydrolysis and synysis begins for some reactions as low as pH 4 thesis of y-butyrolactone is greater than the 1-3 where (OH-) = 1O-l0JI. 9 n y objections on these units expected by both mechanisms I and 11. grounds however may be met by mechanism 11, Bunnett referred to the discrepancy of 4-6 units as the y-butyrolactone enigma" and ascribed the also symmetric. large difference to unusual hydration changes. Mechanism I1 The preceding analysis indicates that we should ki expect a w-value 1-3 units greater for hydrolysis R C O S + H20 HR RC(OH)(OH?+)S B anyway so that the enigma is reduced by up to k? 3 units to 1-3 units of w. Thus the magnitude of RC(OH)(OHz+)X RC(OH)?XH+; the effect necessary to ascribe to unusual hydration KD = [ R C ( O H ) ( O H ~ + ) X /[RC(OH)?XH ] +] changes is decreased considerably. Unfortunately only the above esterification reks action has been studied in predominantly aqueous RC(OH)?XH- t B RCOOH + S H + HB solutions. The reversible hydration of fumaric kp to malic acid yields w = 5 in both directions.1° Application of the steady-state assumption to both Values of w for hydration of crotonaldehydell tetrahedral intermediates that appear in K,, yields and crotonic acid12 are greater than 3.3 and greater for the initial rate of ester disappearance in the hydration than in the reverse direction. Hydrolysis of Amides.-In formulating a niechanism for the hydrolysis of amides, a symmetrical and for the initial rate of appearance of ester in a mechanism is unnecessary because the reaction is much less reversible than the hydrolysis of esters. solution with no ester originally Ordinary amides have w = 1.2 to 2.6, indicative of water acting as a nucleophile.*" Some amides yield higher values of w and are mentioned later. Mechanisms I and I1 yield equations of the same 4 general mechanism for amide hydrolysis is kinetically indistinguishable form. The rate con- Mechanism 111 stants of each mechanism have a different signifiRCOX + H+', cance however. Insertion of approximate numeriRC(OH)X+; K1 = (RCOX)ko/(RC(OH)X+) cal values into eq. 1-1, 1-2, 11-1 and 11-2 yields possible values for the various rate constants. The ki following discussion on esters is equally applicable RC(OH)X+ + HnO RC(OH)*XH+ to mechanisms I or 11: but in order to be specific kr the constants quoted will be those of mechanism I. k3 Equation 1-1 for ester hydrolysis implies w > 3.3 RCOOH + X H HB RC(OH)%XH++ ( B ) because water acts as a proton transfer agent in the (B) term in addition to a contribution of about In the last step the catalyst B may not always be The tetrahedral intermediate RC2 to w for H 2 0 acting as a nucleophile. Since necessary. neither (H20),(B) or (HB) appears in the denomi- (OH)*Xis much more basic than the corresponding nator of eq. 1-1 and 1-2, the inferred w-value is intermediate in ester hydrolysis, making mechaindependent of the relative magnitudes of k2 and nism I unnecessary for amide hydrolysis. Applicak ~ . In ester hydrolysis if k2 >> k 3 , the ka reaction tion of the steady state approximation to RCis rate determining, while if ka >> k2 the kl reaction (OH)?XH+ yields is rate limiting. In the acid-catalyzed hydrolysis (10) L. T. Rozelle and R A. Alberty, J . Phys. Chem., 61, 1637 of ethyl benzoateg ka,'k2 = 2 . 6 Thus the parti- (1957). (11) S. Winstein and H. J. Lucas, J , A m . Chem. Soc., 59, 1461 tioning of the tetrahedral intermediate is nearly equally divided between reactants and products, (1937); R. P. Bell, J . Preston and R . €3 Whitney, J Chrm. S o r . , 1166
+
2-
+
+
( 9 ) AI. I.. Bender, J . A m . Cheni SOL.,73, 1626 (1931); 11. I, Bender, K U Ginger a n d J. P. l i n i k , ibid , 80, 1044 (19.58).
(1962). (12) D. Pressman a n d H. J. I,ucas, .I. A m . Chrm. Sot, 61, 2271 I 1 o';e),
ACIDHYDROLYSIS OF CARBOXYLIC ACID ESTERS
Nov. 3 , 1962
In all amides that have been studied, the parti. tioning of the intermediate strongly favors the products13so that k3(B) > 100kz. Eq. 111-1becomes -d(RCOX)/dt
=
ki [RC(OH)X+](H20) (111-2)
which corresponds to w N 2. In this case it cannot be determined whether base is required in the third reaction. Since K , 'v 10 for ordinary amides, the range of acidities studied passes through the region where the amide is becoming protonated. The complete expression for [RC(OH)X+] = hoC~/(ho K J , where CAis the total concentration of amide, is required in eq. 111-2. I n conjunction with the declining activity of water, the maximum observed in the hydrolysis curves of amides can be accounted for in quantitative terms. The observed first-order rate constant for amide hydrolysis according to eq. 111-2 is
+
k' =
kia"i-120
ho/(ho
+ Ki)
(1)
At the maximum in a plot of k' versus HC, dk'/ dho = 0. Differentiating eq. 1 and setting the derivative equal to zero, we obtain K&h0 K I )= d log aR,o/dHo (2) A plot of the log ~ H versus ~ O Ho for HCl, HC104 and &So4 reveals that from 0.5to about 10.0 M acid the points lie on a single curve for all three acids.14 This curve is adequately represented by the parabolic function
+
- log U H ~ O= O.O34(
+ O.O18(-Ho)
-I- 0.01 (3)
Substitution of eq. 3 into eq. 2 yields K i / ~ ( h o Ki) = O.OSS(-Ho) 0.018 (4) a t the maximum in a plot of observed first-order rate constant v e w m acidity. For the particular case of acetamide in HC1, w = 2.472aand the maxioccurs a t about 3.05 M acid where14 - Ho = 1.06 and h o = 11.5, Substitution of these values into eq. 4 yields K1 = 3.3, a result nearly identical with the measured equilibrium value" of 3.0.
+
+
(13) M. L. Bender and R . D. Ginger, J . A m . Chem. Soc., 77, 348 (1955). C. A. Bunton, T . A. Lewis and D . R . Llewellyn, Chemistvy & Industry. 1154 (1954). (14) Values of ac;.dity function taken f r o m M ,A. Paul a n d F. A. Long, Chem. Reus., 57, 1 (19571, and of logarithm activity of water from ref. 15. T h e relationship of eq. 3 has been discussed another way; see P. A. H . W y a t t , Disc. F a v a d a y Soc., 24, 162 (1957). (15) Appendix of ref. 2a. (16) D. Rosenthal and I. T. Taylor, J . A m . Chem. SOC.,79, 2684 (1957); J. T. Edward, H . P. Hutchison and S. C. R. Rleacock, J . Chem. Sod., 2520 (1955). (17) N. F. Hall, J . A m . Chem. Soc., 52, 5115 (1930). T h e values estimated from ultraviolet absorption spectra (A. R. Goldfarb, A. Mele and N. Gutstein, ibid., 77, 6194 (1955)) are unacceptable. I n separate experiments on some of these amides a n d others in H?SO1 and HClOa we have confirmed t h a t a single p K , value will not fit the data. Measurements are performed a t 200 or 205 mp on steep slopes of absorption curves because the maxima occur further into the ultraviolet. Readings taken on steep slopes are subject to displacements as the character of the solvent is changed, This interpretation is supported b y the tendency of t h e ratio C B H ~ + / C B H I + of Goldfarh, el a l . , to approach unity as C B increases. Similar studies on protonation of benzamides reveal pronounced solvent effects.15 As the solvent blank changes, Iecording spectrophotometers with automatically varying slit width introduce an additional variable. By assuming t h e extinction of the protonated form t o be near zero a n d considering only readings taken in the most dilute acid solutions, we obtain results for acetamide consistent with t h e potentiometric results of Hall We feel no such explanation as Goldlarb, et al., offer is required. Until measurements can be performed xt the maxima, ultraviolet absorption
41 33
I t is now generally agreed that 0- rather than Kprotonation is predominant in amides. l 9 Some investigators,2b though admitting predominant 0protonation, have preferred to consider the Nprotonated species as the kinetically active species in hydrolysis and other reactions. They argue that the greater value of the equilibrium constant for N-protonation is more than offset by higher reactivity of E-over 0-protonated amide. Since there is no compelling reason to write the mechanism in terms of the less common protonated species, mechanism I11 has been written with the 0-protonated species kinetically reactive. Though a rate expression applicable to amides has been used to derive eq. 4, the result is a general one applicable to any rate expression where the rate varies linearly with concentration of protonated substrate and activity of water to some power. For this reason, eq. 4 also applies if the Nprotonated species were considered to be the kinetically reactive one. More complex derivatives are required to treat non-linear equations such as eq. I1I-1. The w-values of hydrolysis of some amides exhibit high values,2a w > 4.7. In three cases, picolinamide, nicotinamide and isonicotinamide, a plus charge already resides on the molecule so that further protonation is electrostatically inhibited. A higher K1 value is expected for these amides and no maximum is observed in their rate versus Ho curves. These amides may hydrolyze according to one of the mechanisms suggested for esters (the catalysts in the k3 steps of mechanisms I and I1 may not be required if k3 > kz),or in mechanism 111,kz might be greater than k3(B)yielding k' = kika(HsO)(B)/kz
(111-3)
When water is the base, w > 3.3 for this rate expression which involves the k3 rather than the k1 step as the rate-limiting one. The latter alternative has been suggested for the general base-catalyzed hydrolysis of N,N'-diphenylformamidinezo(w = 7.75) while mechanism I with kz > k3 has been proposed for the hydrolysis of piperazine-2,5-dione (w = 5.1). Mechanism I11 with k2 > k3 could be distinguished from mechanisms I or I1 with k3 > kz by oxygen isotope exchange studies. The complete mechanisms I, I1 and I11 with kz > k3 are not easily distinguishable experimentally. Acid hydrolysis of thioacetamide yields acetamide and HzS?'with w = 4.2 and 5.4 for HC1 and HC104, respectively.2a This reaction should proceed in a sequence analogous to mechanism I, which accounts for w > 3.3. Methyl benzimidate hydrolyzes in acid solution to give methyl benzoate and ammonia with w = 6.4. This reaction may also occur in a scheme like mechanism I. analysis of p K , values made on slopes of absorption curves in solutions of markedly channinp solvent composition must be suspect. (18) J. T . Edward and S. C. R . Meacock, J Chem. SOL.,2000 (1957). (19) G . Fraenkel, A . Lowenstein a n d S. Meiboom, J . Phys. Chem., 65, 700 (1961); A. R . Katritzky and R. A. Y . Jones, Chemistuy & Industry, 722 (1961); R. Stewart and L. J. Muenster, C a n . J . Chem., 39, 401 (1961); M. J. Janssen, Spechchim. Acta, 17, 475 (1961); D. Herbison-Evans and R. E . Richards, Trans. Faraday S O L ,58, 845 11962). ( 2 0 ) R . H. De Wolfe, J . A m . Chcm. Soc., 82, 1585 (1960). (21) E. A . Butler, D 0. Peters and E. H . Swift, Anal. Chem.. 30,
1379 (19%).
Vol. 84
R. BRUCEMARTIN
4134
Values of w from Mechanism.-A variant of on w of temperature and different acids are better eq. 111-3 provides an example of the principle understood. For substrates of a different kind enunciated in the introduction that the rate ex- than considered here, other values of p and espepression may indicate a greater w-value than would ically g and r may be required. It may not always be expected from simple inspection of the rate- be advisable to separate the effects of water acting determining step. If no base were required in the as a proton transfer agent into q and r contributions ka step, then spontaneous decomposition of the if changes in protonated and unprotonated subintermediate would by inspection of mechanism strate hydration values are not equal to r - q. I11 predict w < 0. A check with eq. 111-3 reveals, To be specific, however, we may choose 9 and q however, that without (B) the value of w is that for as about 4-2 and r as +4. These values yield water acting as a nucleophile or + l . 2 < w < +3.3. w values of $4 and +Z for ordinary ester and I n the Bunnett criterion of mechanism, water amide hydrolysis respectively. acting as a nucleophile shows up only when no It is necessary, in general, to determine w from proton transfer occurs to smother the former effect mechanism and not solely from the rate limiting by its greater contribution to w(or w*). It would step. Division of w into nucleophile, p , and proton seem that i t might be possible to break up the total transfer functions seems desirable. Further subinto a t least two parts representing water acting as division of the proton transfer functions into proton a nucleophile and as a proton transfer agent. This acceptor, q, and proton donor, 7 , contributions is division was implied in the discussion of ester primarily one of convenience in quickly ascertaining hydrolysis where the observed w was interpreted as the w-dependence of a given mechanism without the sum of nucleophile and proton transfer contri- having to derive the complete steady-state expresbutions. By virtue of this division a greater w- sion. If r contributions do not cancel and appear value is expected for ester hydrolysis than for in the final net w expression, a dependence on ha formation because no nucleophilic water contri- exists and special treatment is required. bution appears in the latter. Hydrolysis of o-Nitrophenyl Oxalate.-The hyIt may also prove useful to subdivide the con- drolysis of o-nitrophenyl oxalate passes through a tribution t o w of water acting as a proton transfer maximum in about 1.5 hl HCl,23and a Bunnett agent into two parts: that of water acting as a plot yields a curve of increasing slope as acidity proton acceptor and as a proton donor. Only one increases.15 This case may be accounted for by a water molecule is required to accept a proton, but modification of the full eq. 111-1 where there is a a hydrated proton is thought to be associated with change in the rate-limiting step as kz overtakes k3(B) four water molecules.22 This subdivision may be as the acid concentration increases. Let us assume performed without violation of the principle of that hydrolysis occurs when the anionic form (S-) microscopic reversibility in an acid-base-catalyzed of o-nitrophenyloxalate becomes protonated on the reaction if hydration differences between proton carbonyl ester oxygen (SH) and that protonation acceptor and donor are mirrored in acidic and of the ionized carboxyl group (S’H) yields a species basic forms of substrate with the result that the that competitively inhibits hydrolysis because same transition state occurs for the reactions in additional protonation a t the carbonyl ester oxygen either direction. Let the following symbols desig- (SH*+)requires a much higher acidity than for the nate the apparent power of the activity of water kinetically active (SH). Then we have the relafor the following functions of water: p , nucleophile; tionships q, proton acceptor; and r , proton donor. I n any S- H + Z S’H; K’ = (S-)ho/(S’H) mechanism an expected experimental w-value may S- H + Z SH; Ki = (S-)ho/(SH) be calculated by adding all p , q and r for forward reactions and subtracting for reverse reactions ki until the forward direction of the rate-limiting SH H?O If D step is just included. Any mechanism which kz upon this treatment yields a net w containing t ka must be given special attention because ha now apD +B products pears in the rate expression. One such special I n this mechanism, D may be considered a tetracase is treated in the last section. hedral intermediate which does not require a base Hydrolysis in mechanisms I or I1 with a rate- for formation because D- is quite basic. This situlimiting kl step yields w = p q, while if the k3 ation is analogous to that of amides. Application step is rate limiting, w = p p-r r =9 q of the steady state approximation to D yields again. Synthesis gives w = p for either the kl or kr step rate limiting. The difference between hydrolysis and synthesis is p units or that expected (S’H) and (SH) is of water acting as a nucleophile. Mechanism I11 where C, is the sum of (S-) for the hydrolysis of amides yields w = p for a rate- negligible. When the base B is water, the aplimiting kl step and w = p q if the k3 step is rate parent first-order rate constant may be written as determining in agreement with eq. 111-1 through 111-3. Logically the next step is to assign values to p , q and 7 . Even for a set of similar substrates ac- where p denotes water acting as a nucleophile curate assignments are not possible until the effects and p water acting as a proton acceptor. The
+ + +
+
+
+
-
+
+
+
(22) R. N. Bascombe and R. P. Bell, Disc. Faraday Soc., 24, 158
(1957).
(23) M. L. Bender and Y . L. Chow, J. A m . Chem. Soc., 80, 5380 (1958).
ACIDHYDROLYSIS OF CARBOXYLIC ACID ESTERS
Nov. 5, 1962
4135
are comparable and successive trials of k3/k2 could be attempted.) For the case where ho >> k 3 / k z , the apparent first-order rate constant from eq. IV-1 is given by
ordinate of a Bunnett plot is given by
Note that K'(cx0.6) and not the much larger K1 is k' = kIk3aaaoPKB/k~ho (IV-2) involved in the ordinate plot. Because y is a comwhere K B = (B)ho/(HB). Differentiation of eq. plex function of UH,O, a Bunnett plot will not yield a straight line in the region where kz and k3 are com- IV-2 yields parable. At low acidities where k3 U H * O ~> kt, the dHo -d -log k' (IV-3) last term on the right is log UH,# and a slope of about d log aHz0 -'+d= 1-3 is expected. At sufficiently high acidities the last term becomes log [k3u~,op+q/k2]and a slope The last term on the right may be estimated by greater than 4 is predicted. The two points a t high- substitution of eq. 3 into eq. IV-3 d log k' 1 est acidities23yield a Bunnett slope of 4. If p and q (IV-4) + 0.068 ( -Elo) + 0.018 d log aalo are taken as 2 and 3, respectively, a value of k d k z of about unity accommodates the data. More Equation IV-4 indicates the slope of a Bunnett plot exact fitting was not attempted due to lack of for a reaction described by a rate equation of the data a t sufficiently high acidities, Extension of form of eq. IV-1 when ha >> k3/k2. The slope is the studies to higher acidities and oxygen isotope not a straight line but is a function of Ho. For exchange experiments over a range of acid concen- Hovalues near zero the slope is markedly dependent trations would provide tests for the proposed on Ho, being about 60 when Ho 0. Even for interpretation. -Ho = 1.0, 2.5 AI HClOa, the slope is about 12 Though presented in terms of a specific case, p. Only when - HO> 4 does the magnitude of the rate equations of the same form as those of this last term in eq. IV-4 become similar to p . section may be of more general applicability. A The foregoing analysis reveals that when values reaction obeying a mechanism such as that offered of w greater than about +9 are obtained, mechfor the hydrolysis of o-nitrophenyl oxalate, where anisms yielding rate equations with a higher power w consists of different contributions from p and q of ho in the denominator than in the numerator depending upon the rate-limiting step, may yield should be considered. If the h o term is predomideviations from a Bunnett plot which are more nant in the denominator, a plot of log k' - HO informative than a reaction where a linear de- versus log U H ~ Oshould yield a straight line. In a pendence of pseudo first-order rate constant on similar fashion large negative w-values are expected water activity is observed. for reactions abetted by an unbalanced proton reAcid Inhibition of Hydrolysis.-There is a set of action in the forward direction in addition to equireactions, to be individually discussed at the end of librium protonation of substrate. this section, where a proton is liberated in one of Though complicated by salt effects, the rate of the steps of the mechanism. A simple scheme is hydrolysis of a c e t y l i m i d a ~ o l edecreases ~~ as the acid strength is increased, Bunnett plots do not Mechanism I V ki yield straight lines, and the initial values of w SH+ H ~ O B-' D +HB are as high as 19.6 for HC104. With PKa = 3.6, kz acetylimidazole is predominantly in the conjugate acid form in the moderately concentrated acid ka solutions considered here. In perchloric acid the D J_ products data are fairly well accommodated by postulating where S H + is a protonated substrate and D is an an acid-inhibited step (kv) in acid solutions analointermediate. In weakly acid solutions the k1 gous to that of mechanism IV. step may be rate limiting, while in solutions of sufficient acidity an equilibrium will prevail in the ki H+N N-C-CH3 + H10 + B HB+ first reaction with the result that the k3 step be+%/ I1 A2 comes rate limiting. Application of the steady 0 state assumption to the intermediate D yields for the initial rate of disappearance of protonated ks substrate N + CHZCOOH H+N\ ,N-C-CHB+ HN,
+
+
+
m
--= d(SH+)
dt
hkdSH+)(HzOXB) kdHB) k3
+
(Iv-l)
An important question arises as to the behavior of reactions with acid-inhibited steps when they take place in concentrated acid solutions. A reaction with a rate obeying eq. IV-1 will yield a straight line in a Bunnett plot only when k3/kZ >> ho. When ho >> kZ/kz, a straight line could be obtained if log k' - HOis plotted against log aHzo. Since this function is seldom plotted, i t is of interest t o investigate the effect of plotting only the logarithm of the first-order rate constant log k' versus log U H ~ Ofor the case where ho >> k3/kZ. (A more complicated situation prevails when ho and k d k 2
Equation IV-1 is also applicable to this mechanism and accounts for the data up t o about 1 AI in HCIOa, where the activity of water is still nearly unity, by setting k1 = 2.9 min.-l and k3/k = 1.2. In more dilute acid solutions ka/kz > (HB); eq. IV-1 predicts general base catalysis, which has been observed. 25 At higher concentrations of HC104 the observed rate is greater than that predicted by eq. IV-1. This difference may be due (24) S. Marburg and W. P. Jencks, J . A m . Chem. SOC.,84,232 (1962). (25) W. P. Jencks and J. Carriuolo, J . B i d . Chcm., 234, 1272 (1959).
4136
J. F. BUNNETT, E. BUNCELBND K. 1;. NAHABEDIAN
to participation of mechanism I11 already presented for amides or to salt effects. All other acids studied yield greater rates than equivalent concentrations of HC104. A plot of log k’ - HO versus log aH20 yields a curve which exhibits a negative w decreasing in magnitude as acidity increases. A straight line is not expected a t low acidities where, according to the above analysis, ha ‘v k 3 / k 2 . At higher acidities the method of analysis overcompensates, due to an additional mode of hydrolysis or salt effects as mentioned before. The effect of salts may be operative through the part they play in B and HB. The variation in effects of salts observed in acetylimidazole hydrolysis may be a more general phenomenon unappreciated because few studies exist of comparable scope on other compounds. Qualitatively the high w = 19.G value of acetylimidazole hydrolysis in perchloric acid is accounted for by mechanism IV. This mechanism predicts the occurrence of oxygen isotope exchange with acetylimidazole a t high acidities. *4reaction scheme similar to mechanism IV was used to account for acid inhibition in thiazoline26a,b
+ H 2 0 + B D + HB D +H B SH+ ~+ B
TH+
D
Z
S
(26) (a) R. B. Martin, S Lowey, E. L. Elson and J. T. Edsall, J . A m . Chem. soc., 81, 5089 (1969). R. B. Martin and A. Parcell, i b i d . , 83, [a) 4830, (b) 4835 (1961). A complete mechanism proposed for thiazoline hydrolysis and acetyl transfer in ref. 27 is unsatisfactory. I n designations similar t o those of ref. 27 we show in a foitbcoming publication
[CONTRIBUTION FROM
THE AXD THE
VOl. s4
and oxazoline2Gc compounds. In the case of some of the thiazoline compounds such as 2-methylthiazoline where the value of k 3 / k s = 0.1 is small, extraordinary powers of the activity of water would be required to account for the decrease in hydrolysis rate by the decrease in water activity. For 2-(l-acetamin0-2-methylpropyl)-thiazoline~~~ and 2-methyloxazoline a value of k3/k2 = 1.2 was obtained, the same as that for acetylimidazole. The identical value for these three compounds is probably coincidental but is nevertheless suspicious and indicates that some common factor, such as a decrease in the activity of water, might account for the results. In such borderline cases it is difficult to assign with certainty the mode of decomposition. By analogy with other thiazoline derivatives26b where k3,/kr = 0.1 to 0.3, the first compound probably decomposes according to mechanism IV. This conclusion is less certain for 2-methyloxazoline. More studies are required on both compounds with a greater variety of acids and salts. Least certain is the mechanism of hydrolysis of acetylimidazole. Even for this compound we favor mechanism I V with the reservations already discussed. Acknowledgments.-The author is indebted to Dr. I$-.P. Jencks for transmissicn of data on the hydrolysis of acetylimidazole in perchloric acid. Thanks are also due to Dr. R. J. P. IVilliams and the Division of General Medical Sciences for their forbearance while this work was performed on a fellowship granted for another pursuit. t h a t a mechanism of t h e following kind is a better representation of all t h e data. (27) R B Martin and R I Hedrick, J A m Chem S o c , 84, 106 (1962).
VENABLECHEMICAL LABORATORY, UNIVERSITYOF NORTHCAROLISA,CHAPELHILL, E.C., METCALF CHEMICAL LABORATORY, BROWNUSIVERSITY,PROVIDEXCE, R. I.]
The Mechanism of Acid-catalyzed Hydrolysis of Azoaryl Ethers1 BY J. F. BUN NETT,^ ERWINBUNCELAND K. V. KAHABEDIAN RECEIVED JULY 13, 1962 The phenolic products of hydrolysis of 4-(~-suIfophenylazo)-l-naphthyl methyl ether (IC) and 4-(p-sulfophenylazo)anisole (11) in oxygen-18-labeled water carry the oxygen-18 label in the phenolic hydroxy group. However, 4-(p-sulfopheny1azo)-phenol ( I V ) exchanges oxygen with the medium under the same conditions. Etherification of 4-phenylazo-1-naphtho1 ( I I I a ) and transetherification of 4-phenylazo-1-naphthyl methyl ether ( I a ) are accomplished by refluxing with ethanol under acid catalysis. The kinetics of hydrolysis of I1 in 1-6 M HCIOa solutions have been determined, and w and w * values computed. For several of these reactions a mechanism of nucleophilic displacement a t aromatic carbon is required, and for the rest it seems very likely. The protonated azo linkage is a strong activating group for aromatic nucleophilic substitution.
4-(Phenylazo)-l-naphthylmethyl ether (Ia) and it escaped general recognition. Recent ~ t u d i e s ~ ~ ’ ~ related ethers are exceptionally sensitive to acid- have dealt with the conditions, kinetics and mechcatalyzed hydrolytic cleavage, as represented in anism of such cleavages. eq. 1. Hydrolysis is quite rapid, for example, in Several features of this reaction suggest that i t 0.1 M hydrochloric acid a t 46”. Witt and S ~ h r n i d t , ~differs in kind from ordinary ether cleavage as who discovered this phenomenon in 1892, describe commonly effected by hydrobromic or hydroiodic i t as “gewiss bemerkenswerth.” Although several acid. The reaction occurs readily in dilute acid, other chemists took note of this phenomenon,4-8 it does not require highly nucleophilic anions such (1) This research was supported by t h e National Science Foundation (Grants G-2359 and G-6210). (2) Department of Chemistry, Brown University, Providence, R. I. (3) 0. N. Witt and C . Schmidt, Ber., 25, 1013 (1892). (4) K. H. Meyer, A. Irschick and H. Schlosser, ibid., 47, 1741 (1914). ( 5 ) W. Borsche, W. Muller and C. A. Bodenstein, A n n . , 472, 201 (1929).
(6) K. H. T. Pfister, J . A m . Chem. S o c , 64, 1621 (1932). (7) J. B. Muller, L. Blangey and H. E. Fierz-David, Helo. Chim. Acta, 35, 2579 (1952). (8) V. Ettel, J. Weichet and J. Sparil, Coll. Czech Chem. Comm., 16, 204 (1950). (9) J. F. Bunnett and G. B. Hoey, J . A m . Chem. Soc., 8 0 , 3142 (1968). (10) J. F. Bunnett and E. Buncel, i b i d . . 83, 1117 (19611,
