4005
HYDRATION OF ALDEHYDES
The Hydration of Propionaldehyde, Isobutyraldehyde, and Pivalaldehyde. Thermodynamic Parameters, Buffer Catalysis, and Transition State Characterization1 by Y. Pocker2and D. G. Dickersona Department of Chemistry, University of Washington,Seattle, Washington 98106
(Received February 18, 1969)
Values of the thermodynamic parameters AGO, A H ” , and ASo for the reversible hydrations of propionaldehyde isobutyraldehyde, and pivalaldehyde were determined using a spectrophotometric method. Short extrapolations of absorbancy to time aero allow the accurate determination of the molar extinction coefficient for the n-u* absorption of the unhydrated aldehydes in water solvent. Kinetic data on the hydronium ion, acetic acid, acetate ion, dihydrogen phosphate ion, monohydrogen phosphate ion, diethylmalonate dianion, hydroxide ion, and water catalysis are presented for all three aldehydes. The data are shown to be consistent with a concerted mechanism of hydration-dehydration.
Introduction Although the zinc metalloenzyme erythrocyte carbonic anhydrase was once thought to catalyze only one reaction-the reversible hydration of carbon dioxide-it has been shown in these laboratories that the enzyme very efficiently catalyzes the addition of the elements of water to a great variety of carbonyl sy~tems.~-gWe have recently shown that the reversible hydrations of aliphatic aldehydes are useful reactions for delineating the steric requirements of this enzyme.’O These reactions proceed a t an appreciable rate in the absence of enzyme and are furthermore susceptible to general acid-base catalysis. Consequently, as a prerequisite to the study of enzymatic catalysis, it was necessary to characterize thermodynamically and kinetically the nonenzymatic hydrations of propionaldehyde, isobutyraldehyde, and pivalaldehyde. The first part of the present paper reports the thermodynamic parameters associated with the reversible hydrations of these aldehydes and compares them with the results of previous studies on related compounds. 11-16 The second part of the present paper examines the catalytic coefficients at 0.0” for water, hydronium ion, hydroxide ‘ion, diethylmalonate dianion, acetic acid, acetate ion, monohydrogen phosphate, and dihydrogen phosphate ions for the hydration of each aldehyde and for the dehydration of the corresponding hydrate. Earlier studies from our laboratory have kinetically characterized the catalytic coefficients a t 0.0” for these buffer components with respect to the hydration of a~etaldehyde’~ and 2- and 4-pyridine aldeh y d e ~ ’ *both , ~ ~ in HzO and DzO. For proton transfer reactions, the Bronsted catalysis law has often been interpreted on the basis of free energy curves.2o The interpretation is apparently
valid when curves cross on the reasonably straight part of the energy walls.21~22However, as Bronsted himself suggest’ed,28one could not expect the law to (1) This work was supported by U.8. Public Health Service grants from the National Institutes of Health. (2) Author to whom correspondence should be addressed. (3) (a) National Science Foundation Fellow, Summer 1966; Chevron Research Company Fellow, 1966-1967. (b) Taken in part from the Ph.D. Thesis of D. G. Dickerson, University of Washington, 1967. (4) (a) Y . Pocker and J. E. Meany, “Abstracts of the Sixth International Congress of Biochemistry,” Vol. IV, 132, New York, N. Y., 1964, p 327; (b) Y. Pocker and J. E. Meany, J . Amer. Chem. Soc., 87, 1809 (1965). (5) Y. Pocker, J. E. Meany, D. G. Dickerson, and J. T. Stone, Science, 150,382 (1965). (6) Y. Pocker and J. E. Meany, Biochemistry, 4, 2535 (1965). (7) Y. Pocker and J. E. Meany, {bid., 6,239 (1967). (8) Y. Pocker and J. T. Stone, ibid., 6, 668 (1967). (9) Y. Pocker andD. R. Storm, ibid., 7, 1202 (1968). (10) Y. Pocker and D. G. Dickerson, ibid., 7,1995 (1969). (11) E. Lombardi and P. B. Sogo, J . Chem.Phys., 32,635 (1960). (12) (a) R. P. Bell, Advan. Phys. Org. Chem., 4, 1 (1966), and references quoted therein; (b) R. P. Bell and P. G. Evans, Proc. Boy. Soc., A291,297 (1966). (13) L. C. Gruen and P. T. McTigue, J. Chem. floc., 5217 (1963). (14) (a) J. Hine, J. G. Houston, and J. H. Jensen, J . Org. Chem., 30, 1184 (1965); (b) J. Hine and J. G. Houston, ibid., 30, 1328 (1965). (15) G. E. Leinhard and W. P. Jencks, J . Amer. Ghem. Soc., 88, 3982 (1966). (16) (a) P. Greenaaid, 2. Lua, and D. Samuel, J. Amer. Chem. Soc., 89, 749 (1967); (b) P. Greenaaid, Z. Rappoport, and D. Samuel, Trans. F w a d a y Soe., 63,2131 (1967). (17) Y. Pocker and J. E. Meany, J . Phys. Chem., 71,3113 (1967). (18) Y. Pocker and J. E. Meany, ibid., 72,655 (1968). (19) Y. Pocker and J. E. Meany, %bid.,73, 1867 (1969). (20) See, for example, R. P. Bell, “The Proton in Chemistry,” Corne11 University Press, Ithaca, N. y . , 1959, p 168. (21) J. E. Leffler and E. Grunwald, “Rates and Equilibria of Organic Reactions,” John Wiley and Sons., Inc., New York, N. Y . , 1963, p 157. (22) R. A. Marcus, J.Phys. Chem., 72,891 (1968). (23) J. N. Bronsted and K. J. Pedersen, 2. Phys. Chem., 108, 185 (1924).
volume 78, Number 11 November 1960
4006
Y. POCKER AND D. G. DICKERSON
hold over wide values of the acid dissociation constant
K . Paradoxically, it is the proton transfer reactions which obey the Bronsted catalysis law over too wide a range of pK that are now in need of explanation. 12b~17,24 The observations reported in the present study are discussed in terms of possible mechanistic pathways for both general acid and general base catalysis.
c
8
aI a
Y
CI,
Experimental Section Materials. Propionaldehyde, isobutyraldehyde, and pivalaldehyde (J. T. Baker) were factionally distilled under nitrogen. An electrically heated 20-cm column of Heli-Pak was used for this purpose. I n this manner, the small amounts of carboxylic acids usually present in these aldehydes were removed. After purification, the acid content was shown to be less than 0.05 mol % ' as determined titrimetrically. The acid content remained below 0.20% for several weeks provided full vials of the aldehydes were stored in the dark at -5' in nitrogen-flushed containers. Sodium salts of the mono- and dianion of diethylmalonic acid, prepared by previously described procedures, were employed as buffers in the pH range 6.38.2. Phosphate and acetate buffers were prepared from commercially available analytical grade reagents and were titrated with standard hydrochloric acid or sodium hydroxide in order to verify concentrations. Reaction rates were determined with a Beckman Model DU ultraviolet spectrophotometer. A specially constructed cell compartment containing a motor-driven stirring device and calibrated thermometer was filled with a methanol-water mixture. The contents of this compartment were maintained at 0.0" by circulating a coolant (ethylene glycol-water) at a precisely regulated temperature through a coil in the compartment. The coolant was supplied by a Forma-Temp Jr. Model 2095-1 refrigeration unit which contained, in addition to the thermostating mechanism, a pump to circulate the coolant. The low temperature used in these studies made the frequent change of desiccant in the phototube housing imperative. All pH determinations were made with a Beckman Model H2 glass-electrode pH meter. The pH of reaction solutions was determined directly without dilution. Standardization was accomplished with Beckman standard buffer solutions. Method. Aqueous solutions of all reactants except aldehyde were pipetted into the silica cells which were then placed in the cell compartment and allowed to equilibrate thermally. The differences in absorbancy between the various reactant cells and the reference cell were recorded. Aldehyde was withdrawn from a stoppered glass serum vial with a calibrated Hamilton microliter syringe and injected into the cell to initiate the run. It was necessary to shake the contents to ensure that all material was in solution. The rate of diminution of absorbance (propionaldehyde; Xmax 277.5 The Journal of Physical Chemistry
0
J
Figure 1. Semilog plot of kinetic data from typical run; [propionaldehyde] = 0.0433 M . Kinetics followed a t 277.5 mp; 0.0".
mp; isobutyraldehyde; A, 284 mp; pivalaldehyde; Xmax 285.5 mp) was taken as a measure of the rate of approach to equilibrium. The absolute value of the slope of a plot of In (At - A , ) os. time (with At representing the absorbance a t any time t during the course of the reaction and A , representing the absorbance at equilibrium) is numerically equal to the pseudo-firstorder rate constant, kobsd. Because the rate of approach to equilibrium is determined, kobsd = k f IC,, where ICr is the rate constant for hydration and k, that for dehydration. These kinetic plots were linear for a minimum of three half-lives, as shown in Figure 1. The rates were such that half-lives ranged from 0.5 to 25 min. Data for the faster reactions were recorded on tape for subsequent playback and analysis. The equilibrium extinction coefficients at the wavelength of maximal absorbance, E,, were determined by two methods. The first of these involved weighing various amounts of aldehyde into buffers and determining equilibrium absorbance. The second technique involved calculating the molar extinction coefficients at equilibrium from runs in which the aldehyde was introduced via calibrated Hamilton microliter syringes. The two techniques furnished virtually the same value for em. The spectra of the aldehydes prior to hydration and following equilibration a t 0.0" are depicted in Figure 2.
+
Results Short extrapolations of the kinetic plots to t = 0 furnished an intercept, A0 A,, and, with knowledge of the equilibrium absorbance A,, it was possible to
-
(24) M. Eigan, Discussions Faraaag SOC.,39, 7 (1966).
HYDRATION OF ALDEHYDES
4007
Table I : Thermodynamic Parameters for Aldehyde Hydration
a
Aldehyde
eo (Xmax)5'b
K,,b,C
CHaCHsCHO (CHs)&HCHO (CH3)sCCHO
18.7 f 0.2, n = 110 (277.5) 22.7 f 0.2, n = 88 (284) 1 8 . 7 f 0.4, n = 25 (285.5)
1.98 f 0.02, n = 149 1.58 i 0.017, n = 110 0.47 f 0.007, n = 23
M - l cm-l (mp).
'
refer to 99% confidence limits based o n n observations.
A H O ~
0.0".
-5.4 f 0.2 -5.8 f 0.2 -4.4 f0.2 kcal mol-'.
APe
-18.3 f 0 . 6 -20.4 f 0 . 7 -17.6 f 0 . 8
cal mol-' deg-1.
dehyde are very sensitive t o temperature.13 In order to permit such a comparison, then, it was necessary to vary the temperature of equilibrated solutions of the aldehydes and observe the corresponding changes in absorbancy. A few kinetic runs at 25" established that EO is, as expected, temperature invariant over the range 0-25". Utilization of the kinetic method t o determine EO a t even higher temperatures was not considered to be sufficiently accurate due to the greatly enhanced rate of reaction at these higher temperatures. Thus, these studies furnished values of K,, as a function of temperature. as presented in Table 11. The values
0.%
0.6 0 C
0
-f20 v)
0.4
0.3
1
0.2
mP Figure 2. Ultraviolet spectra in aqueous solution a t 0.0'. Upper three curves prior to hydration: - - -, 0.0547 M propionaldehyde; 0.0449 M isobutyraldehyde; -, 0.0544 M pivalaldehyde. Lower three curves: same solution following equilibration.
.
calculate the molar extinction coefficients, EO, of the unhydrated aldehydes in aqueous solution and equilibrium constants, Keq, as defined by eq 1.
- [Hydrate] fH H - -- (Ao- A m ) f K,, = (aHydr)sq (1) ~
, e d& ( [Aldehyde1 fA A m fA The ratio of activity coefficients of hydrate and aldehyde, (fH/fA), was considered to be unity except for certain special experiments in which very large concentrations of salt were specifically utilized to study this ratio. Presented in Table I are the experimentally observed values of EO and K,, as determined a t 0.0" for each of the aldehydes studied. It is interesting to compare values of K,, obtained in this work with the results of previous workers, but most of these older data were obtained either at an unspecified temperature or a temperature other than 0.0". Gruen and McTigue pointed out that the values of K,, for the hydration of propionaldehyde and isobutyral-
IO3 T-'deg-'
Figure 3. Logarithms of equilibrium constants as functions of reciprocal absolut'e temperature: 0 , propionaldehyde; A, isobutyraldehyde; W, pivalaldehyde.
of AH" and AS" obtained from Figure 3 are presented in Table I. The values of Keq reported by earlier workers are presented in Table 111 together with values of the present study a t the same temperature as determined from Figure 3. The agreement is generally poor with respect to those workers who utilized spectral methods to measure the equilibrated solutions and various indirect methods to estimate the spectral characteristics of the same solutions prior to hydration. None of these workers utilized a kinetic technique which permits virtually direct observation of the aldehyde solutions prior to hydration, t = 0. However, agreement is excellent with those studies in which K,, was determined by examining the integrated nmr spectra of aqueous aldehyde solutions. This latter technique, which provides information on both comVolume 78, Number 11 November 1969
4008
Y. POCKER AND D. G. DICKERSON
Table I1 : Hydration Equilibria a t Various Temperatures
a
T ,Q K
Kega
273.0 276.7 278.0 280.1 282.5 282.7 285. 4 289.0
T,OK
Keqa
T,OK
1.98 1.78 1.70 1.54 1.46 1.44 1.29 1.18
Propionaldehyde 289.2 1.14 293.0 1.04 294.2 1.02 297.5 0.87 298.0 0.89 301.5 0.788 303.8 0.788 306.0 0.70
307.0 310.0 313.3 315.5 319.0 319.3 322.8
0.67 0.61 0.58 0.55 0.51 0.468 0.428
273.0 276.7 280.1 282.7 284.3 285.4
1.58 1.43 1.22 1.11 1 .ooE 0.99
Isobutyraldehyde 289.5 0.88 292.6 0.76 297.9 0.63 298.0 0.64 301 .O 0.58 305.9 0.50
310.2 311.0 314.8 319.3 321 .O 322.6
0.45 0.42 0.37 0.34 0.33 0.32
273.0 276.7 280.1 282.7
0.47 0.45 0.39 0.348
Pivalaldehyde 285.4 289.0 292.6 295.5
297.5 301.0 303.8 305.9
0.235 0.22 0.21 0 208 #
Kes as defined in eq 1.
ponents of the equilibrium, is expected to be more reliable than the other methods previously employed. As first pointed out by Gruen and McTigue, changes in water activity induced by high salt concentrations will disrupt the equilibrium ratio of [Hydrate]/ [Aldehyde].13 This feature was examined in the present study also since the results obtained by Gruen and McTigue in the absence of salts were not in agreement with those obtained in this work and by Hine.14 The presence of large quantities of salt lowers the equilibrium ratio of the hydrate to aldehyde, and it is concluded from (1) that the ratio.fH/f* is therefore increasing. Gruen and McTigue pointed out a method for correlating changes in fH/fA with changing salt con~entrati0n.l~This method, developed by Robinson and Stokesz5and subsequently modified by Gluekauf,26 takes the form of eq 2 log
0.33 0.31 0.28 0.26
Keqa
(.f~/f~)=
+
+
(h’ 1){[0.018ml(rl hi - 2)/2.3(1 0 . 0 1 8 ~ ~ 1 ~ 1 )log ] a,]
+
-
(2)
Equation 2 relates the ratio of activity coefficients of the hydrate and aldehyde to the salt molality, ml, in terms of the difference in hydration number of the hydrate and aldehyde, h’, the ratio of the partial molar volume of the electrolyte to that of water, rl, hydration number of the electrolyte, hl, and water activity, a,. This equation for 1:1 electrolytes ( v = 2), although disregarding electronic factors, is expected to furnish reasonably accurate estimates of the parameter h’. The ratio [Hydrate]/ [Aldehyde] was determined in the presence of varying quantities of KaCl; these ratios, The Journal of Physical Chemistry
in conjunction with the values obtained in very dilute M ) , provided values of f ~ / f ~ , buffers (5-10 X which as evidenced by Figure 4, are most consistent 1) = 2 for all aldehydes. This would indiwith (h’ cate the equilibrium to be best expressed by
+
where n is either zero or an integral value. In view of the different ultraviolet spectra obtained in water and nonaqueous solvents, it is likely that n is not zero. Also, n is not likely to be greater than 1, so the aldehyde is most probably hydrogen bonded to one water molecule and the hydrate to two. The rate of hydration of each of the aldehydes was studied as a function of pH in each of three buffers: acetate, diethylmalonate, and phosphate. The reactions are general acid-general base catalyzed; hence, kobsd may be expressed by the sum of terms kbsd
=
ko
+ k~,o+[H30]+f ~oH-[OH-]-k ~ H A [ H A-b] ~ A - [ A - ] (4)
The values of the catalytic coefficients for hydronium ions, JCH~O+, hydroxide ions, OH-, the protonated constituent of the buffer, k H A , and for its conjugate base, IC*-, as well as the spontaneois term ko, were deter(25) R. H.Robinson and R. A. Stokes, Trans. Faraday Soc., 53, 301 (1957). (26) E. Gluekauf, “The Structure of Electrolytic Solutions,” W. J. Hamer, Ed., John Wiley and Sons, Inc., New York, N. Y.,1959,P 97; cf. also, E. Gluekauf, Trans. Faraday SOC.,51, 1235 (1955).
HYDRATION OF ALDEHYDES
4009
Table I11 : Comparison of Thermodynamic Parameters for Aldehyde Hydration SO,
M-1 om-'
CHsCHzCHO (CHs)2CHCHO
18.3,' 17.7' 22.3,b 17.7"
(CHa)sCCHO
18.4,b 18.5'
" As defined by eq 1.
AH",koal/mol-1
Keq"
25': 25': 35": 25': 14.5': 4':
' Present work.
0.87,' 0.69" 0.66,' 0 . 6 l I d0.44' 0.43,b 0.47' 0.24,' 0.24,80.1,'0.24' 0.32,b 0.24O 0.42,b 0.33'
' Reference 13.
Reference 14.
-5.4,b -6.5' -5.8,b -7.3,'
-6.5d
-18.3,' -20.4,'
-22.5' -26," -26.gd
-17.6,b -18'
-4.4,b -4.4'
' Reference 16a.
Ahso, csl mol-' deg-1
'Reference 15.
'Reference 16b.
Table IV : Comparison of Catalytic Coefficients for Aldehyde Hydrations a t 0.0"" CHsCHO kHzOb k H a 0 +' kOH-' kA_"d
kHO-$,' kOA-2 kH*PO,-' kHPOla-'
0.0011/55.5 130 7.91 X 10' 0.035 0.0805 0.0154 0.189 0.42
CHsCHzCHO
(CHs)zCHCHO
0.000931/55.5 121 2.35 X 10' 0.040 0.0751 0.0146 0.146 0.312
0.000515/55.5 97.5 1.77 X 10' 0.0245 0.049 0.098 0.123 0.202
a Coefficients utilized to calculate rate constants, koalod, for comparison with kobsd as per Table in M-' sec-l. 'M-1 sec-1. Catalytic coefficient for diethylmalonate dianion.
0.3
X-Log
(Iw
Figure 4. Solvation number of gem diols relative to unhydrated aldehydes in aqueous solution: 0 , propionaldehyde; A, isobutyraldehyde; H, pivalaldehyde.
mined in the following manner. Plotting the rates obtained in 0.0100 M acetate buffer as a function of hydrogen ion concentration furnished a linear relationship, the slope of which was considered a reasonable
(CHs)sCCHO
0.000125/55.5 31.4 6.30 X 10' 0.0016 0.0157 0.00256 0.032 0.045
v. ' kaZo = k0/55.5; ko in sec-';
~ H I O
first approximation to ~ H ~ o + . The data obtained in 0.0100 M diethylamalonic acid was plotted as a function of hydroxide ion concentration to furnish a relationship which was linear over a substantial pH range. The slope of the linear portion was taken as a reasonable approximation to T OH- and the extrapolated intercept was taken as the first approximation to ICo. With these values as a starting point, all coefficients were evaluated by a series of successive approximations continued until further extension of the procedure did not alter the parameters. The values so obtained, summarized in Table IV, reproduce the observed rate constants of the 88 runs used to deduce the constants to within the limits of experimental uncertainty (Table V). This comparison refers to 0.01-0.1 M buffers. Previous workers have concluded that, with reference to aldehyde hydration reactions, catalytic coefficients for various bases show poor correlation with their relative strengths as bases in terms of the Bronsted relationship. A recent investigation of glucose mutarotation in our laboratory, however, shows excellent accord, excepting diethylmalonate dianion which was 0.8 log unit below a line correlating 11 other bases.27 Furthermore, as indicated by Figure 4, the present study furnished a linear relationship over 17 pK units for all three aldehydes examined, and the slopes of these three plots were virtually identical (Table V). Again, it was noted that the dianion of diethylmalonic (27) Y . Pocker, J. W. Long, D. Dahlberg, and T. C . Lacalli, unpublished observations. Volume 7S, Number 11 November 1960
4010
Y. POCKER AND D. G. DICKERSON
Table V : Comparison of Calculated and Observed Rate Constants" Propionaldehyde-
7 -
PH 0.0100 M
Diethylmalonic aoid
0.0100 Y Aoetio
aoid
0.100 M
Aoetic acid
Isobutyraldehyde--
c
loakoalod
loa kobsd
5.38 6.32 6.62 6.90 7.17 7.20 7.45 7.56 7.81 8.00 8.13 8.20
2.23 1.57 1.59 1.67 1.77 1.78 1.92 2.00 2.19 2 -40 2.58 2.71
2.26 1.57 1.61 1.66 1.77 1.79 1.94 2.05 2.18 2.40 2.56 2.70
6.12 6.36 6.63 7.14 7.19 7.21 7.46 7.61 7.79 8.05 8.23
1.01 1.oo 0.98 1.08 1.10 1.11 1.20 1.28 1.39 1.62 1.86
4.28 4.38 4.44 4.49 4.61 4.68 4.84 4.90 4.99 5.22 5.52 5.81 5.92
11.70 10.20 9.27 8.65 7.07 6.42 4.70 4.29 3.77 3.19 2.44 2.09 1.94
12.28 10.27 9.20 8.43 6.83 6.17 4.85 4.37 3.92 3.04 2.26 2.01 1.93
4.14 4.22 4.40 4.56 4.73 5.04 5.31 5.50
4.17 4.59 5.00 5.34
24.9 14.7 9.25 6.61
24.7 14.6 9.14 6.64
6.01 2.75 2.77 6.60 3.08 Phosphate 7.16 3.51 7.61 3.72 7.81 Calculated and observed rate constants
0.00500 M
2.77 2.76 3.12 3.47 3.75 are reported in
PH
lor kobad
10' koalod
1.02 0.98 0.99 1.04 1.10 1.09 1.21 1.22 1.41 1.61 1.85
6.12 6.21 6.46 6.80 6.97 7.39 7.77 7.89 8.10 8.30
0.478 0.464 0.441 0.437 0.444 0.488 0 584 0.645 0.765 0.964
0.457 0.480 0.435 0.486 0.480 0.502 0.566 0.647 0.738 0.976
13.45 11.58 8.10 6.01 4.48 2.78 1.98 1.65
13.57 11.50 8.23 5.96 4.28 2.84 2.17 1.76
4.20 4.30 4.50 4.72 4.92 5.24 5.61
7.18 5.88 3.97 2.64 1.88 1.19 0.78
7.20 5.85 4.00 2.59 1.90 1.21 0.75
4.20 4.32 4.72 5.19
18.8 16.0 9.47 5.50
18.3 16.3 9.53 5.74
4.17 4.33 4.73 5.19 5.50
11.6 9.13 5.25 2.91 2.08
11.7 9.2 5.42 3.05 2.16
5.96 6.63 7.11 7.60 7.71 sec-1.
2.03 2.01 2.15 2.40 2.49
2.04 2.14 2.10 2.36 2.51
6.01 7.12 7.58 7.71
0.99 1.01 1.11 1.16
1.01 1.01 1.14 1.16
acid was 0.8-1.4 log units below these lines defined by HzO,OH-, CH&O2-, and HP02-.28
Discussion I n the past, many workers have attempted to measure the equilibrium composition of aqueous solutions of the presently investigated series of aliphatic aldehydes. The most commonly employed technique has been the measurement of the intensity of the n-?r* carbonyl absorption band around 280 mp which diminishes on hydration. These studies have always suffered from the fact that accurate values were only available for equilibrated solutions; the main uncertainty in these measurements lies in the value to be ascribed to eo, the maximum extinction coefficient of the unhydrated carbonyl compound. It has been the custom to equate €0 in HzO with that in a nonhydroxylic solvent such as hexane, but this is not strictly valid since the intensities of n-a* transitions and the shape of the bands vary The Journal of Physical Chemistry
PiValaldehyde-loa kobsd
PH
10' kcslcd
I
somewhat with solvent. More recently, nmr measurements have permitted estimation of the concentration of both components, but a t higher aldehyde concentrations which in turn led to a decrease in water a ~ t i v i t y . ~ ~ * ~ ~ The technique described in this paper has permitted virtually direct observation of the intensity of the n-r* transitions of aldehyde solutions prior to hydration, as well as following equilibration, and, therefore, provides very accurate values of K,, over a sufficiently wide (28) For all three hydrations, the HPOa2- anion is ca. nine times more active as a catalyst than imidazole. Similarly, the catalytic constants for -OH and HzO are ca. eight times as great as those predicted from 8. line drawn through monofunctional basic catalysts (pyridine, imidazole, CeCbO-, SOaZ-, COaZ-, and HzBOa-).aT It is attractive t o suggest that HaO, CHaCOa-, HP042- and OH- operate as bifunctional catalysts when hydrogen bonded to one or more water molecules.~~ (29) Interpretation of Kea values from nmr data on concentrated solutions of aldehydes may be vitiated by the presence of dimers such as (RCHOH)ZO.~~ (30) M. L. Ahrens and H. Strehlow, Discussions Faraday SOC.,39,
112 (1966).
HYDRATION OF ALDEHYDES
4011
temperature range to permit accurate evaluation of A H o and AS". As indicated in Table 11, the hydrations are all exothermic but are accompanied by an unfavorable entropy change. Any mechanism proposed to explain the hydrationdehydration reactions must be consistent with the observed general acid-general base catalysis. It would seem reasonable to impose the further restriction that the rate-determining step involve inter alia a change in hybridization of carbon rather than just a proton transfer to or from oxygen, a step which is now commonly considered to be controlled only by diffusion rates.24 Mechanisms i and ii below for general acid catalysis and iii and iv for general base catalysis are consistent with these restrictions. BH
+
+
fast
+ >C=O
>C=OH
+ B-
+ B- + HzO
>C=OH
I
6-
at
1
H HzO
1
B-
*
*
+ HzO
+ OH- + >C=O I
HO. .C-O..
I
CHiCHiCHO (CHa)2CHCHO (CHa)3CCHO
6*
6-
BH
a+
Table IV: Values of the Parameters Gb and p of the
Aldehyde
6+
BH
where K , = [Hy-1 [H+]/[Hy ] or ~ C B - [B-] = k*[BH]/[H+]; loi4kg- K B R t = k*. Evaluation of ii* for the water-catalyzed dehydration of propionaldehyde furnishes IC* N 3 X 10l2 M-' sec-'. Eigen refers to the fastest of diffusion processes as being characterized by coefficients of 1011-1012.2 4 Hence, it would appear that although mechanism iii would be reasonable for most bases, it wouId not be in good accord with water catalysis; yet, water fits on the same Bronsted line as the other bases (Figure 5 ) . Similar analysis of mechanism i, performed with a value of -Pofor the pK, of RCHOH+, furnishes unrealistic coefficients for catalysis by a species such as H2P04-. See TableVI.
Bronsted Relationship kb/q = Gb(p/K,,)p
1 + I H2O* *C-O* *Ha * B H20-C-OH + BI I fast I + I H20-C-OH + BHO-C-OH + BH (ii) I 1 I + HzO + > G O B . - - H . . . O . . .C=O I I H I *
[B-][Hy] = k*[Hy-][BH] =
k*K a [HY 1 [BE I/ IH* 1
+ H B + >C=O a+
B-
kg-
I
+
B * . . H . * * O - . - C ~ O H Z B HHO-C-OH
I
hydrate anion, [Hy-I, protonated base, [HB], and a rate coefficient, IC*, for the actual dehydration step
fast
BH
=
6+
.Ha. . B
+ HO-C-OI
+ OH1
HO-C-OH
I
a
10s ab"
8"
7.9 6.3 2.5
0.45
0.45 0.46
Deduced from Figure 5 using pK, values at 0.0'.
Consequently, both mechanism i for general acid catalysis and its general base counterpart, mechanism iii (the two differ by one proton), would appear not to be suitable; the pair of mechanisms represented by (ii) and (iv) would be the better choices. However, with some catalysts the observed rates would require velocity constants higher than diff usion-controlled values for these mechanisms as well. Eigen has suggested that a more reasonable physical picture is obtained if two or more water molecules are included, when it becomes possible to replace two steps in mechanisms i-iv by a concerted process involving cyclic hydrogen bonded transition states (v) and (vi).lZb A mechanism proceeding through (v) (shown for general acid catalysis) incorporates the acid or base in a ring, ~
+ B-
7
(iv)
Mechanism iii requires that the product of the observed catalytic coefficient for a general base, k g - its concentration, [B-1, and the concentration of hydrate, [Hy] equals the product of the concentrations of the 'Volume 73, Number 11 November 1969
1
~
~
4012
Y. POCKER AND D. G. DICKERSON
whereas in (vi) (shown for general base catalysis) the ring is composed (formally) only of water molecules and aldehyde. Similarly, water molecules could be incorporated as bridges in (ii) and (iv). The advantage of utilizing water bridges, in either a cyclic or acyclic mechanism, is that acids and bases would be expected to show good correlation as observed with the Bronsted relation-regardless of the structural type of the acid or base. Transition states similar to (v) or (vi) may well possess one advantage not immediately apparent in (ii) or (iv). The salt effects, as analyzed in the preceding section, could be presented as in (vii).
4
PROPION ALDEHYDE
0 DEM'
-4 ISOBUTYRALDEHYDE
so
h
Y
Y
0
_11
-4
R
\C=O---H /
H
2
\ 0 /
+2H20 -2
-6
H
-4
H
/
R
'\\
0
\/ C /\ H 0 /
'\\\
H
H
/ 0 \
H (vii)
0
/\
H
H
Inasmuch as the type of solvation depicted in (vii) stabilizes the aldehyde and hydrate, it would be attractive to invoke similar stabilization of the activated complex leading to this product. The similarity between product vii and transition state vi is immediately apparent. It is interesting to compare the relative values of catalytic constants for hydration and dehydration within the series of aldehydes which has been examined. I n general, the values of the catalytic coefficients, relative to acetaldehyde as 100, are approximately 85 (propionaldehyde), 60 (isobutyraldehyde), and 20 (pivalaldehyde); the relative values of the coefficients for dehydration are then approximately 43.5 (acetaldehyde hydrate), 43 (propionaldehyde hydrate) , 38 (isobutyraldehyde hydrate) , and 21 (pivalaldehyde hydrate). The inductive effect of the three methyl groups would tend to neutralize the dipolar character of the carbonyl group of pivalaldehyde and it would be expected to hydrate slower than, say, acetaldehyde. The transition state is probably more polar than the hydrate, so
The Journal of Physical Chemistry
0
4
8
12
16
log (p/K,q) Figure 5. Bronsted plots of log kb/q us. log (p/K,q) for propionaldehyde, isobutyraldehyde, and pivalaldehyde. (Values of p K , taken a t 0.0').
this same inductive effect would tend to enhance the rate of dehydration of pivalaldehyde relative to aldehydes with fewer methyl substituents. The transition state is also more sterically crowded than the aldehyde so that large aldehydes would be expected to hydrate more slowly. In terms of the addition model discussed by Karabatsos,al this steric effect would be most pronounced in going from isobutyraldehyde to pivalaldehyde, because no amount of rotation of the latter would relieve the unfavorable eclipsing by a methyl substituent of the incoming nucleophile. Indeed, the greatest drop in relative hydration coefficients is observed a t this point in the series. It is difficult to predict the exact effect of steric factors in the dehydration, since the precise geometry of the transition state is unknown. Additional methyl substituents could result in either slight steric acceleration or slight steric inhibition, depending on the values of the bond angles relative to the distance between the leaving group and the rest of the molecule. At any rate, steric effects should not retard dehydration nearly so much as hydration. These suggestions are all in accord with the observed trends. (31) G . J. Karabatsos and N. Hsi, J . Amer. Chem. Soc., 87, 2864 (1965).