1994
Association of Protons with Oxygen-Containin in Aqueous Solutions. IV. Esters C. F. Wells Department of Chemistry, University of Birmingham, Edgbaston, Birmingham 815 ZTT, England
(Received November 10, 1972)
The equilibrium constant K , for the penetration of (H20)4Haq+ by esters has been determined in dilute solution in aqueous HC1 using spectrophotometric measurements on p-nitroaniline. The influence of electron-releasing groups on K , is not as clear as for alcohols, ketones, and carboxylic acids due to steric hindrance involving the alkyl groups, but it is concluded that protonation uia hydrogen-bonded attachment occurs predominately on the carbonyl oxygen atom.
Previous investigations1 in this series have included measurements on the replacement of water molecules in the aquo proton (H20)4H,qf by alcohols, glycols, ketones, and carboxylic acids. Distribution of the proton affinity among the bonds in the protonated complex containing water and the oxygen-containing organic molecule, probably accompanied by rearrangement of the molecules within the complex, results in effective protonation of the organic molecule at an 0 atom. The concentration quotient K , = [ROHzas+]/[ROH][(H20)~H,q+] has been determined in aqueous HCI using p-nitroaniline (B). It has been shown that eq 1 is obeyed,1,2 where CO, C R , and c are [B] withotit added HC1 or ROH, with added HC1 and ROH, and with added HC1 only, respectively, and 1 ~ !and a are the concentrations of unprotonated water and the total added concentration of ROH.
-CCK CK -
KZF~WI COCR
COU:
1 CO - +--KiFiCL
I-___
KlFlCL
CR
(1)
KIF1 and K g z are the concentration quotients for reactions 2 and 3, respectively, with FZ = f ~ f ( R 0 H 2 + ) / (ROH2+ ) = ~ ( B H + ) ~ R oand H F1 = f ~ f p / f ( B H - ~ ) f ( H z O ROH(W20)3H,q+ and P = (H20)4Haqf).
B
B
+ (H,o),H,,"
.f
ROH(H20)3Haqfe BK,,'
=+
BH,,+ -I-water
+ ROH,,
(2) (3)
Experimentally, plots of C C R / ( C R - e ) us. cp./(co - CR) are linear with an intercept = cow/KXFla using KlFx/w determined in the absence of substrate, showing that the ratio F1 is independent of a. From the symmetrical nature of F1 it would be expected that F1 1, and the agreement of the intercepts with the calculated values suggest that any deviation of Fl from unity due to differential solvation effects must be small. In particular, the suggestion3 that solvation effects make variations in f~ predominate over variations in f p , f(BH)+, and f(H2O) is not in accord with these observations on the intercepts. Although f B decreases with increasing a,3 e.g., for methanol f n = 0.85 in 5% v/v and f e = 0.17 in 40% v/v, and for 2propanol f s = 0.83 in 5% v/v and f B = 0.054 in 40% v/v, any variation in F1 of a similar magnitude to that in f B would show clearly in changes in the intercept.2 It is also of interest in this connection that the linearity of the plots of (-d[Oz]/dt)--l us. [substratel-1 observed in the oxidation of secondary alcohols and a n ether by the photoexcited anthraquinone-2-sulfonate ion in acidic conditions re-
-
The Journalof Physical Chemistry, Voi. 77, No. 16, 1973
quires the symmetrical term f ~ o ~ f ~ / f ( R o H ~ + ) f ( Hto2 0 ) remain -1 with varying a. Values for K , are derived in two ways: (i) directly from the slopes of the plots of ccR/ ( C R - c) us. cR/(co - CR) or (ii) indirectly, by determining K2F2 from the ratio slope/intercept of the latter plots, from which [ROH(H20)3Haq+]is derived and then K,. K, from ii is independent of [He11 a t any particular value of a and the close agreement with K , from i supports the above conclusion that F1 is independent of a and -1.0; the relationship between these two methods for calculating K, has been critically examined.2.3 These values1 for K , also agree well with K , determined from conductivity, ionic transport, calorimetry, and the kinetics of acid-catalyzed reactions in dilute aqueous solutions of mineral acids.1,2,4,9 From the earliest consideration of these equilibria arising from observations of the photosensitized autoxidation of alcohols,6 it has been assumed that they occur between solvated protons containing the organic molecule and other solvated protons without ROH. In highly aqueous media, ROH from the bulk solvent replaces an H2O molecule in (H20)4Haq+followed probably by a rearrangement of the molecules and hydrogen bonds, and in media of low water content, e.g., H2SO4, ROH replaces either an H2O or a solvent molecule in a solvated proton quite different from (H20)4Haq+.1 Nevertheless, in both cases, a protonated, solvated ROH is involved, the difference in the equilibrium constants1 arising from the difference in the nature of the solvation. K , is independent of a when a 5 10% v/v and is usually determined at a = 2 or 5% v/v1 where fB 1.0 and is 20.85.3 Such values for K , have been determined over a range of temperatures,l,6 and ROH(H20)3Haq+ is oxidatively inert compared with ROHaq.l9* This method involving the spectrophotometric estimation of CR and c has now been applied to esters.
-
Experimental Section Materials. All the esters used were purified by fractional distillation. All other materials were as used and prepared previous1y.l Procedure. This was as used previously with co = 1.45 x M , except that, to avoid any acid-catalyzed hydrolysis of the esters, the HC1 solution was added to the aqueous mixture containing NaC1, p-nitroaniline, and ester just before sampling for the spectrophotometric
Proton-Ester Association in Aqueous Solutions
1995
measurement. Negligible hydrolysis occurred during the time required to measure the optical d e n ~ i t y . ~
Results and Discussion Figure 1 shows plots of C C R / ( C R - C ) US. C R / ( C O - C R ) for methyl acetate, ethyl acetate, n-propyl acetate, isopropyl acetate, and methyl n-propionate all in aqueous HC1 at 25.0” with ionic strength adjusted to p = 1.00 by the addition of sodium chloride. The concentrations of methyl acetate and ethyl acetate were 5% v/v, but for all the other esters the maximum concentration obtainable was 2% v/v. The plots for methyl and ethyl acetates, where the higher concentrations of 5% v/v could be used, show good straight lines with little scatter of the points and the intercepts agreeing well with those calculated from K1FlIw. The linearity of these plots and the agreement between the observed anld calculated intercepts is as good as found previously for the other substrates1 Although the plots for the other esters in Figure 1 are not as good, with a greater scatter of the points due to the necessity for using the lower concentration of 2% v/v, the agreement found for the other substrates between the calculated and the observed intercepts,l now extended here to methyl and ethyl acetates, allows the use of the calculated intercepts as an additional point with these esters. The deviation of the point with the lowest [HCl], i.e., the highest C,/(CO - CR), for n-propyl acetate probably arises from incomplete solubility of the ester a t low [HCl] in p = 1.00. Values for K, determined by method i are given in Table I together with values for K 3 2 determined by method ii. Values for K, determined via K2F2 as in (ii) are given in Table 11, and these are in good agreement with K, from method i, as found with all other substrates.l,2,6 Unfortunately, the solubility of other suitable esters is too low in these solutions of electrolytes for a complete set of observations over the range of [HCl] indicated in Table 11, but in two cases, methyl n-butyrate and methyl isobutyrate, CR is determinable a t the high end of the [HCl] range, and estimates of K, can therefore be made for these esters using these observations and the intercept calculated from K1F1/w with a = 0. These approximate values for K, are included in Table I. When the substrate is an alcohol ( = ROH), K, increases with the increasing electron-releasing inductive effect of R, but branching in R produces steric hindrance to the penetration of (H20)4Ha,+ by ROH and, similarly, for ketones RlRzCO and carboxylic acids R&OOH, K, increases with the increasing electron-releasing inductive effect of R1, R2, and R3 with a corresponding reduction for branching. Table I shows that for esters RqCOOR5 there is no similar straightforward explanation for the variation of K, with unbranched Rq and Rg. With R5 = CH3, K, decreases when R4 changes from CH3 to CH3CH2 and then increases when R4 becomes CH3CH2CH2, but, nevertheless, excluding R4 = CH3, K, increases with change in R4 in the order CH3CH2 < CH3CH2CH2 < CH&H2(CH3)CH. Likewise, with Rq = CH3, K, first increases when R,g changes from CH3 to CH3CH2 but then decreases again in changing Rg from CH3CH2 to CH3CH2CH2. However, contrary to the effect of branching in R for alcohols, in R1 and R2 for ketones, and in R3 for carboxylic acids, change of Rb from CH3CH2CH2 to (CH&CH with R4 remaining as CH3 increases K,, and a similar effect is also suggested by the increase in K, in going from R4 = CH3CH2CH2 to R4 =
3.0
I 2.5
X
1
I
1.5, I
1
0.5
1
L/’’
’
, “ /
0Y 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7, 0.8 0.9 cR/(cO - cR)
Figure
1. Plots of c c R / ( c R
-
c ) vs. cR/(cO
-
cR)
at
25’
and
= 1.00: 0,5% v / v methyl acetate, A = 1.0; 8 , 5 % v / v ethyl acetate, A = 1.0; A, 2% v / v n-propyl acetate, A = 0.10; X , 2% v / v isopropyl acetate, A = 0.10; 0 , 2% v / v methyl n-propionate, A = 0.20. p
TABLE I: Kc
Calculated by Method i for 25’ and j~ = 1.OO Hindrance relative to CH3COOCH3 at Ester
Po
KZFZ K,, M - ’
>C=O
acyl 0
25
0.64
...
...
20
0.78
Small
None
33
0.48
Large
Small
CH C/ / O \OCH(CH~L
27
0.58
Large
None
CHCH,C No
29
0.55
Small
Small
-0.67
Small
Large
-1.5
Small
Large
CHIC
‘OCH, CH,C
//O
‘OCH&H,
CH,C//O \OCH~CH,CH,
\OCH, CH,CH,CH,CNo \OCH,
CHs
‘CH-C
1
CH,CH2
Yo ‘OCH,
(CH~)(CH~CHZ)C with H R5 = CH3. The interpretation of the variation of K, for esters with the structure of R4 and R5 is complicated by the existence of two possible sites for the protonation. This ambiguity also exists in the interpretation of the mechanistic data on the acid-catalyzed hydrolysis of esters.8 Ingold and coworkersgJO prefer proThe Journalof Physical Chemistry, Vol. 77, No. 16. 7973
1996
C. F. W e l l s
k\o/H H \o/H
I
H
I + R,COOR,
4
Figure 2. Possible p e n e t r a t i o n of trigonal p y r a m i d a l (H20)4Haq+ b y an e s t e r R4COOR5 i n v o l v i n g p r o t o n a t i o n of b o t h t h e c a r b o n y l and acyl 0 a t o m s .
TABLE II: Values of K , Calculated by Method ii for 25" and p = 1.00
[HCI], M 0.10 0.16
0.20 0.28 0.40 0.60 0.80
1.00 Av
~%v/v methyl acetate
5%v/v ethyl acetate
0.67 0.67 0.66
0.84 0.82 0.78
...
...
0.64
0.79
...
.. 0.73 ... 0.79
0.61 .. , 0.65
.
~%v/v ~%v/v ~%v/v n-propyl isopropyl methyl nacetate acetate DroDionate
... ... 0.50 0.48 0.47 0.47 0.43 0.44 0.47
0.62 0.60 0.60
0.59 0.58 0.56
...
...
0.58
0.55
. ..
.
.
(a) A large hindrance at the carbonyl 0 atom reduces K , and a large hindrance at the acyl 0 atom has only a small effect on Kc, so that the effect on K , of hindrance a t the carbonyl 0 atom predominates over hindrance at the acyl 0 atom. (b) In general, an increasing electron-releasing effect in Rq increases K,, but an electron-releasing effect in R5 produces only a small increase in K,. The small change in K, arising from an increase in the electron-releasing effect in R5 may be a distant effect on the carbonyl 0 atom, and although it is not possible to say that Protonation occurs exclusively at that 0 atom, the conclusion from a and b is that protonation occurs predominately at the carbonyl 0 atom.
I
0.56 . ..
0.53
0.59
0.56
...
tonation on the acyl 0 atom, whereas Bender prefers protonation on the carbonyl 0 atom.ll Bender uses the argument developed by Stewart and Yates12 from a comparison of values for pK, for the corresponding conjugate acids of substituted acetophenones and substituted benzoic acids. There is a close correlation between pK, for the acetophenones and their carbonyl stretching frequencies vc-0 in the infrared, and although there is only a rough correlation between pKa and V C , O for the benzoic acids, pK, for the benzoic acids correlate linearly with pKa for the acetophenones. This latter evidence is convincing for carbonyl protonation in 44-96% MzS04 where, however, the structure of water is largely collapsed and the solvated proton ( H z O ) ~ H has ~ ~ n~ < ~+ 4.13 In these conditions, the apparent acid dissociation constant of protonated water is considerably less than that of the trigonal pyramidal complex (HzO)4Haq+ in aqueous solution,l,2 and Stewart and Yates' results are not necessarily comparable with those in aqueous solution. Indeed, on the picture of penetration of (Hz0)4Haq+ by the s u b s t r a t e , l ~both the carbonyl and the acyl 0 atoms could share in the proton affinity by hydrogen bonding, as in Figure 2, whereas this is less likely
The Journal of Physical Chemis:ry, Vol. 77, No. 16, 1973
with ( H z O ) & ~ ~ ~ and + (H30+)801vin concentrated sulfuric acid. It is instructive, therefore, to consider how the structures of R4 and R5 might inhibit the penetration into (H20)4Haq+, and to this end Catalin models were constructed of all the esters in Table I. Assuming free rotation round all single bonds insofar as this is allowed by neighboring groups, the hindrance to access at the carbonyl and acyl 0 atoms by R4 and R5 has been assessed visually relative to that in methyl acetate. To allow for the crudity of the assessment, a three-point scale is adopted, i.e., none, small, and large, and these are recorded in Table I. The following points emerge from consideration of these scales of hindrance in relation to the values for
References and Notes (1) C. F. Wells, Trans. Faraday SOC., 61, 2194 (1965); 62, 2815 (1966); 63, 147 (1967). (2) C. F. Wells, J. Chem. Soc., Faraday Trans. 7, 68, 993 (1972). (3) J. Sierra, E. Teixido, and P. A. H. Wyatt, J. Chem. SOC., Faraday Trans. 1, 68, 290 (1972). (4) U. L. Haldna and A. I. Talvik in "Chemistry of the Carbonyl Group," S. Pattai, Ed., interscience, New York, N. Y . , 1966, p 429; V. A . Palm and U. L. Haldna, Doki. Akad. Nauk SSR, 135, 667 (1960); U. L. Haldna,, L. Pioom, and A. Maroos, Zap. Tartu Gos. Univ., 127, 65 (1962); U. L. Haldna and R. K. Puss, Russ. J. Phys. Chem., 38, 1529 (1964); U. L. Haldna, Org. Reactiv. (USSR), 1, 184 (1964); 2, 381 (1965). (5) C. F. Wells, J. Phys. Chem., 77, 1997 (1973). (6) C. F. Wells, Discuss. Faraday SOC., 29, 219, 255 (1960); Trans. Faraday SOC.,57, 1703, 1719 (1961); C. F. Wells and G. Davies, ibid., 63, 2737 (1967); C. F. Wells, C. Barnes, and G. Davies, ibid., 64, 3069 (1968); C. F. Wells and C. Barnes, J. Chem. SOC. A. 1626 (1968); 430, 1405 (1971); Trans. Faraday Soc.. 66, 1154 (1970); 67, 3297 (1971); C. F. Wells and M. Husain, ibid., 66, 679, 2855 (1970); 67, 1086 (1971); R. Varadarajan and C. F. Wells, J. Chem. SOC., Faraday Trans. I , 69, 521 (1973); C. F Wells and A. F. M. Nazer, to be submitted for publication. (7) H. M. Dawson and W. Lowson, J. Chem. SOC.,2146 (1928). (8) H. 6. Watson, "Modern Theories of Organic Chemistry,'' 2nd ed, Oxford University Press, London, 1941, Chapter 9. (9) E. H. Ingold and C. K. Ingold, J. Chem. SOC.,'756 (1932); S. C. Datta, J. N. E. Day, and C. K. ingold, Trans. Faraday Soc., 37, 686 (1941). (10) C. K. Ingold, "Structure and Mechanism in Organic Chemistry," 2nd ed, Bell, London, 1969, Chapter 15. (11) M. L. Bender, Chem. Rev.. 60, 53 (1960). (12) R. Stewart and K. Yates, J. Amer. Chem. SOC., 80, 6355 (1958); Can. J. Chem., 37, 664 (1959); J. Amer. Chem. Soc.. 82, 4059 (1960). (13) P. A. H. Wyatt, Discuss. FaradaySoc., 24, 162 (1957).
