JOURNAL O F THE AMERICAN CHEMICAL SOCIETY Registered in
U. S.Palenl
O&e.
@ Copyright. 1963, b y the
VOLUME 85, NUMBER 1
American Chemical Society
JANUARY
30, 1963
[ COh-TRIBUTIOX FROM THE DEPARTMENT OF CHEMISTRY, CORNELL UNIVERSITY,
ITHACA,
N. Y.]
A Comparison of the Bimolecular and Intramolecular Nucleophilic Catalysis of the Hydrolysis of Substituted Phenyl Acylates by the Dimethylamino Group B Y THOMAS
c. BRUICEAND STEPHEN J. BENKOVIC’ RECEIVED AUGUST3, 1962
The rate constants and activation parameters have been determined for the nucleophilic displacement of subm- and p-substituted stituted phenols from phenyl acylates (solv. HnO) in the systems I [trimethylamine phenyl acetates], I1 [m- and p-substituted phenyl y-(N,N-dimethylamino)-butyrates]and I11 [m- and p-sub. The values of AH* were found to be, within experimental stituted phenyl 6-( N,N-dimethylamino)-valerates] error, invariant for all the esters of system I, I1 and 111. Therefore, all electronic and steric factors-both being large-are reflected in AS*, providing linear free energy relationships (unusually) dependent only on AS*. The OH@-catalyzed hydrolysis of m- and p-substituted 0-acetylphenols was also found t o be characterized by a constant AH* value and a substituent controlled A S . The unimolecular processes of 11 and I11 involving five- and sis-membered transition states had TAS* higher by 3.6 A 0.6 and 4.7 z!= 1.1kcal. mole-’, respectively, than the bimolecular process of I. The Hammett p-values for I, I1 and I11 are very nearly identical ($2.2, +2.5, +2.& respectively). The constancy of AH* and p-for systems I, I1 and I11 as well as the uniform dependence of AAS* on the nature of the substituent group provides convincing evidence that the bimolecular and unimolecular processes take place by the same mechanism. The ratio of rate constants for the similarly substituted esters of I1 us. those of I11 was found to be 2.3. This value is but l/loo t h a t previously determined as the ratio for the formation of glutaric and succinic anhydride in the intramolecular carboxyl anion nucleophilic catalysis of the hydrolysis of monophenyl glutarates and succinates. This difference as well as the apparent change in mechanism previously noted for the conversion of carboxyl anion catalysis of substituted phenyl acetates to an intramolecular process are discussed.
+
initiated studies of intramolecular catalysis of the hyTo determine if and drolysis of esters and how a mechanism is altered on bringing reactant groups into close proximity it is desirable to have available as many measurable parameters for the reactions as possible. The determination of AF*, A€€+, T A S and p for a series of intermolecular reactions and their intramolecular counterparts has been difficult due to competing side reactions in one or another of the processes, instability of compounds, etc. The present study is concerned with a comparison of the three systems of Chart I. The tertiary amine was chosen as the nucleophile on the basis of its pK,‘-assuring readily measurable rates in both inter- and intramolecular processes-and its inability to partake in complicating general base-catalyzed attack a t the ester carbonyl group, as can secondary and primary amines.’5-17
Introduction Systems involving intramolecular participation of acidic and basic groups in ester and amide hydrolysis have drawn attention as partial models for hydrolytic enzymes.2 The intramolecular “models” and the enzymatic reactions share in common the bringing together of reactant species within a single molecule or complex, respectively (propinquity effect). The enhancement in efficiency obtained by the propinquity factor and considerations as to whether this might account in great part for the facility of enzymatic processes have been d i s c u ~ s e d . ~ ”The only means of ascertaining the kinetic significance of the propinquity effect is through the study of reactions occurring intramolecularly or w i a pre-equilibrium complex formation, In the conversion of a bimolecular reaction into a unimolecular process the possibility arises that the details of mechanism or the actual mechanism of the reaction might be altered. Thus, the bringing together of groups in an intramolecular or an enzymatic reaction may alter the course of the reaction by favoring reaction mechanisms which may not be observable in simple bimolecular reactions (change of mechanism) ; or the bringing together of groups may change the nature of a reaction by causing a step to become rate determining which in the bimolecular process would not be so (change of rate-limiting step). In order to ascertain the importance of these possibilities we have
Experimental Compounds. Trimethylamine hydrochloride (Eastman Kodak Co., white label) was recrystallized from methanol and dried over P20s.p-Nitro-, m-nitro-, -chloro- and p-methyl-0-acetyl phenols as well as 0-acetylpheno itself were from a former study18 and were repurified by crystallization or distillation prior to use. 7-(N,N-Dimethylamino)-butyric Acid Hydrochloride.-To 25.0 g. (0.24 mole) of y-aminobutyric acid was added 63 ml. of 90% formic acid (1.50 moles) and 48.5 ml. of 37% aqueous formalin solution (0.49 mole). The resultant solution was refluxed
f
(7) G. L. Schmir and T. C. Bruice, J . A m . Chem. Soc., 80, 1173 (1958). ( 8 ) T. C. Bruice and J. M. Sturtevant, ibid., 81, 2860 (1959). (9) T. C. Bruice, ibid., 81, 5444 (1959). (10) T. C. Bruice and U.K . Pandit, Proc. Natl. Acad. Sci., 46, 402 (1960). ( 1 1 ) U. K . Pandit and T. C.Bruice, J . A m . Chem. Soc., 82, 3386 (1960). (12) T. C. Bruice and U. K. Pandit, ibid., 82, 5858 (1960). (13) T. C. Bruice and Fritz-Hans Marquardt, ibid., 84, 365 (1962). (14) T. C. Bruice and T. H. Fife, i b i d . , 84, 1973 (1962). (15) J. F. Bunnett and G. T . Davis, ibid., 82, 665 (1960). (16) W. P. Jencks and J. Carriuolo, ibid., 82, 675 (1960). (17) T. C. Bruice and M. F. Mayahi, ibid., 82, 3067 (1960). (18) T. C. Bruice and G . L. Schmir, ibid., 79, 1663 (1957).
( 1 ) N I H Predoctoral Fellow (1961-present). Part of the work t o be submitted by S. J. Benkovic in partial fulfillment for the P h . D . degree, Cornell University. ( 2 ) T. C. Bruice, “Enzyme Models and Enzyme Structure,” Symposium X o 15, Biology Dept., Brookhaven Natl. Laboratories, 1962. (3) M. L. Bender, Chem. Reus., 60, 53 (1960). ( 4 ) D. E. Koshland, J . Cell. Compt. Pkysiol., 47, Suppl. 1, 217 (1956). ( 5 ) D. E. Koshland, J . Theor. Bioi., 8, 75 (1962). (6) R. Lumry. in “The Enzymes,” Vol. I, 2nd ed., P. D. Bayer, H. Lardy and K. Myrback, eds.. Academic Press, Inc., New York, N. Y., 1959, Ch. 4.
1
THOMAS C. BRUICEATD STEPHEN J. BENKOVIC
2
for 16 hr., cooled, and acidified with 28 m1. of concentrated hydrochloric acid. Evaporation t o dryness yielded a white crystalline solid which was recrystallized from acetonitrile t o give the pure amino acid hydrochloride in 70% yield, m.p. 148149' (lit.'@145-147'). 6-( N,N-Dimethylamino j-valeric Acid Hydrochloride.-To 10.0 g. (0.085 mole) of 6-aminovaleric acid was added 22 ml. of 90% formic acid (0.525 mole) and 25.4 ml. (0.255 mole) of 37% formalin solution. The amino acid hydrochloride was obtained in 857? yield as in the previous case after recrystallization from acetonitrile; m.p. 166-167". (lit .20 163-165'). The substituted phenyl esters of y-( N,N-dimethylamino)butyric and 6-( N,N-dimethylamino)-valericacids were prepared by the general method detailed below. T o a small test-tube containing 1.O g. (0.006 mole) of the y-( N,N-dimethylamino)-butyric acid hydrochloride or 1.0 g. (0.0055 mole) of the 6-(N,N-dimethylamino)-valeric acid hydrochloride was added 3.0 ml. (0.021 mole) of trifluoroacetic anhydride. The tube was tightly stoppered with a calcium chloride drying tube and warmed a t 35' for 20 minutes. The cooled homogeneous clear solution was then saturated with dry hydrogen chloride. T o this was added the desired phenol (in 1.8 mole excess to the acid hydrochloride), and with the drying tube in place the resulting mixture was heated in an oil-bath with occasional agitation a t 55' for 1.5 hr. The resulting clear solution was cooled and layered with ether (dried over sodium) causing either crystallization or oiling out of the ester. By grinding the oil under anhydrous ether there was obtained the crude ester hydrochloride in 50-80% of theory. The crude products were recrystallized from 3 : 1 (v./v.) chloroform-carbon tetrachloride solution by layering said solution with anhydrous ether. The recrystallized products were dried over P205. The ester hydrochlorides so obtained were found to be extremely hygroscopic and sensitive t o heat and light, necessitating immediate analysis. Chloride determinations were carried out by the Volhard procedure, nitrogen analysis by the Kjeldahl procedure, and phenoxy group analysis were carried out spectrophotometrically by comparing the t , absorbance of a solution of the ester in p H 9.5 borate buffer t o the absorbance of standard curves of the phenolate ions in pH 9.5 borate buffer. The analytical data are provided in Table I.
Vol. s3
For each experiment, equal volumes of a 1.0 M KC1 solution 5 X M in appropriate ester and the correct buffer were thermostated in the reservoir syringes, transferred to the thermostated stopped-flow syringes (kO.2'j and mixed in the stoppedflow cell. In the case of the m-and p-nitro esters the 1.0 M KCl solution was acidified t o ca. pH 2.0 prior t o solution of the ester t o minimize its solvolysis. The wave lengths employed to follow the formation of phenoxide ions were N
Substituent m ,u
p-CH3 280
H 275
p-C1 283
m-NO2 350
p-SO. 330
The p H of the run was taken as that of the spent solution ejected Five to twenty from the stopped-flow cell compartment a t t,. runs were made a t each pH and a t each temperature. Reported rate constants have been calculated from a t least five superimposable recordings of phenol release vs. time. The reactions were all found to follolv first-order kinetics with good accuracy and the method employed t o calculate the values of the rate constants was that of Guggenheirn.2' Bimolecular Reactions.-Solutions of trimethylamine -trimethylamine hydrochloride were prepared immediately before use by half-neutralization of solutions of the hydrochloride with standardized potassium hydroxide solution. All solutions were maintained a t a calculated ionic strength of 1.6 111 with KCl. The stock solutions so prepared were kept a t 0 ' in an ice-bath. Dilution of the stock solutions to obtain the correct amine concentration were made with 1.6 M p buffer (0") of the same pH as the amine solution. .Imine concentrations of 0.15 t o 1.5 M were employed. The various amine solutions were then drawn into 30-ml. Luer-lock syringes equipped with stopcocks, all air displaced and the syringes replaced in the ice-bath until initiation of the kinetic run. Prior to each run the amine solutions were injected into a stoppered cuvette and allowed to equilibrate iii a constant temperature cell block fitted t o the spectrophotometer. The reaction was initiated by adding 2 drops of the solution of ester from a calibrated dropper to the cuvette resulting in a final ,If. Formation ester concentration of approximately 3 X of phenolate ion was followed by recording the rate of change of absorption a t the wave lengths
TABLE I H p-c1 nt-NO? p-NO,. Suhstituent MELTISG POINTS AND ANALYTICALDATA FOR THE PHENYL 275 285 420 400 mu ESTERSOF y-( K,N-DIMETHYLAMINO)-BLJTYRATE HYDROCHLORIDE AND 6- Y,( ~-DIMETHYLAMINO)-VALERATE HYDROCHLORIDE The pseudo-first-order rate constants ( k o b r ) were determined as the slopes of plots of log (O.D. 33 - O.D.,) os. t. Cmpd. pK.' Determinations.-The pK,' of the dimethylamino group (Butyrate Xp., Calcd , ClO Found, % for the p-CH3, H and p-C1 substituted phenyl esters of y-( N,Nx-CsHa-OH, C1-, N x-CsHa-OH, C1-, S series) 'C. dimethylamino)-butyric and 6-( N,N-dimethyl)-valeric acids a t 20" were assumed t o be those of the pKappvalue determined 4 8 . 2 , 1 2 . 2 8 , 9 . 7 0 47.3,12.11,9.31 C6HhNO2-p 89-90 kinetically from the pH-rate profiles. In a similar manner the CcHSC1-p 134-135 4 6 . 2 , 1 2 . 7 5 , 5 . 0 4 46.1,12.E1,4.95 pK,' values for the phenyl y-(N,N-dimethylamino)-butyrateand CnHiH Si-88 38.6,14.75,5.74 38.1,14.55,5.86 p-cresol 6-( K,N-dimetliylamino )-valer,ate esters a t 10" and 34" C6HiCHJ-p 126-127 4 2 , 3 , 1 3 , 7 5 , 5 , 4 3 4 1 . 5 , 1 3 , 7 7 , 5 . 2 7 were assumed to be those of the pK,,, values determined a t these temperatures from pH-rate profiles. (Valerate series) The pKa' of trimethylamine a t each temperature was obtained C6HsNO?-p 97-98 46.0,11.71,9.25 46.5,11.82,8.97 by the method of half neutralization; pK,*'34' = 10.00, pK,'?o0 = 10.28, PKa'io0 = 10.50. C6HsCl-p 151-182 4 4 . 0 . 1 2 . 1 4 , 4 . 7 9 43.8, 1 1 . 9 1 , 4 . 9 2 Apparatus.-All pH measurements were made with a Radioni105-106 3 6 . 5 , 1 3 . 7 5 , 5 . 4 3 3 6 . i , 1 3 . 7 1 , 5,28 CcHsH eter model 22 pH meter using a combined Radiometer GK2021 CnHsCH3-fi 150-151 39.8, 1 3 . 0 5 , 5 . 1 5 3 9 . 2 , 1 2 . 7 5 , 4 . 9 1 electrode thermostated ( i 0 . 1 ' ) a t the same temperature as the accompanying kinetic reactions. The rates of liberation of sub(Butyrate stituted phenolate ions from the esters were followed with a series) C, H, x C , H, S Zeiss PMQII spectrophotometer. For the slower reactions the IZ-NO~" 127-128 4 9 . 6 5 , s . 9 3 , 9 . 7 0 49.07,6 0 3 , 9 . 8 8 spectrophotometer was fitted with a special hollow brass cuvette holder thermostated a t 30 i 0.01" by a Haake constant tempera(1-derate series) ture circulating bath. For the more facile reactions the spectro?~-N02" 115-116 5 1 . 5 5 , 6 . 3 2 , 9 . 2 5 5 1 . 0 4 , 6 . 6 0 , 9 . 0 5 photometer was equipped with a stopped-flow block designed a Midwest Micro Labs., Indianapolis, Ind. around the model employed by Sturtevant.22 I n our modified design the plunger syringes as well as the reservoir syringes were Kinetic Measurements. Intramolecular displacement of imbedded in a thermostated brass block. Also imbedded in the phenoxide ions from the phenyl esters of y-( N,K-dimethylamino) block were the mixing and observation cell (Plexiglas with quart7 butyric and 6-( N,S-dimethylamino)-valeric acids was followed windows) and the channels (Tygon). The temperature within by means of stopped-flow spectrophotometry. Buffer solutions the brass-block as well as that within the observation cell were were of a calculated ionic strength of 1.O M-though the reactions measured via thermistor probes. The change of absorbance with are insensitive to ionic strength. Buffers were prepared by comtime was recorded on a Brush model RD2321-00 inking oscillobining solutions A4with solutions B in the proper ratio: graph recorder. pH r a n g e
5-7 7-9.5 9.5-11
xi
B
0 . 3 3 M KH2P04 . 6 i Af KC1 ,40 M KHzB03 60 rzI KCl .20 M KiHPOi . 4 0 M KC1
0.33 M K2HP04 01 M K C l .40 M 1.0 MKCl 0.20 iM K ~ P O I 0.40 M KC1
(19) V. Prelog, Coli. Czechoslov Chem. Comm , 2, 712 (1930). (20) P. A Cruickshank a n d J. C. Sheehan, J . A m , Chem. Soc., 83, 2891 (1961).
Results Intramolecular Reactions.-In the solvolysis of systems I1 and 111, the rate of phenolate ion formation would be expected to be due to intramolecular nucleophilic displacement a t the ester bond by the unprotonated dimethylamino group ( k l ) and hydrolysis by (21) E. A . Guggenheim, Phil. Mag., 2 , 538 (1926). (22) T. Spencer a n d J. lf, S t u r t e v a n t , J . A i u C h e m . S n c . , 81, 1871 I1059).
Jan. 3, 1963
'
HYDROLYSIS OF PHENYL ACETATESB Y DINETXYIAIIYO
0.8
Il6
2f4 kob:'i
312 L
410
R
4.8
10"
PH.
Fig. 2.- Partial pH-rate profile for the intramolecular nucleophilic catalysis of p-chlorophenyl -,-( S,S-dimethvlamiiio)but! rate in water ( p = 1.0) a t 20' - - - 0 kcb;H
Y.
Fig. 1.-Plots of kobr is$. li,,t,, arr for the intramolecular nucleophilic catalysis of the hydrolysis of phenyl *,-iS,S-dimeth~.l-( S,S-dimeth>-laInino)amino)-butyrate (11-d),p-meth>-lphen>-l-, butyrate (11-e) and phenyl 6-( ~,S-dirnetli?-lami!io)-valerate( I I I d ) in water ( p = 1.0) a t 211".
lyate species ( k a ) d phenol/'& = kcCE,
- klCE
+
C
(1)
is the total concentration of ester ( L e , ,C E T = CE C&). At any p H the values of the observed first-order appearance of phenolate ion (Kobs) would then be given by 2 and 3 where
t
+ 6.00
IO'!
CET
rI
-4.001
-0.2
0
I
+0.2
I
I
+0.4 + O X
I
+OB
!
I
t1.0
-
+i.2
0. kobs
-
ko
=
ki
1 -7 K,
(ko5S
-
ilo)aH
( 3i
where K,' is the dissociation constant for the dimethylammonium group and aH the hydrogen ion activity, as determined by the glass-electrode. Plots of kobs as. kobs UH were found to be linear for both the phenyl and p-cresol esters of I1 and I11 (Fig. 1). Thus, 4o is insignificant for these esters. Since the p value for nucleophilic displacement on phenyl esters by amines is about twice that for nucleophilic displacement by OH8 and 100 times greater than displacement by H30e1' it is safe to assume KO may also be ignored for the p-C1, $-NO2 and m - K 0 2esters of I1 and 111. From the plot of kobs DS. kobs aH (Fig. l), we may determine Ktapp(slope = - 1/KTapp) and kl (intercept). For the p-C1 esters of I1 and 111 plots of kobs us. pH possessed the shape of theoretical dissociation curves from which the PK'app values of the dimethylamino group could be obtained (Fig. 2 ) . Replotting the data for the p-C1 esters of I1 and I11 in the form of hobs vs. K'app/(K'app UH) provided the kl values as the slopes. The values of kobs determined a t pH 8.0 (for the p NO2 esters the value of kobs a t pH 8.0 was extrapolated
+
Fig. 3.--Hmii::ett plot of the log of the true nucleopliilic catalysis constant L I S . cr for the bimolecular catalysis of substituted phenyl acylates by trimethylamine ( A ) , for the intramolecular catalysis of substituted phenyl y - ( X,X-dimeth>-lamino)-valerates (B) and for the intramolecular catalysis of substituted phenyl 6-( X,X-dimethylamino)-butyrates ( C ) in water a t 20".
from plots of log kobs as. PH when plotted in the Hammett manner (;.&, log hobs as. a)) provided h e a r plots 2 . 5 for both I1 and 111. IVith the assumpof p = tion that the p values for kl would also be 2.5 a plot of log k1 vs. was made using the determined values of kl for the P-CHs, phenyl and p-chlorophenyl esters of I1 and I11 (for the limited series p did equal 2.5) (Fig. 3). The values of kl for the m-NOz and +NO2 esters of I1 and I11 were then extrapolated from the plots. Having k1 for the nitro esters the PK',,, for these esters was then obtained from the expression 2 where kn = 0. The values of pKlappand k l so obtained a t 20' are recorded in Table 11. From the data of Table I1 one may calculate the ratio of kllI/kIII1 for the similarly substituted esters in the series I1 and I11 (Table 111). It may be noted that the average PK'app for the
+
+
+
4
THOMAS
c. BRUICEAND STEPHENJ. BENKOVIC
Vol. 85
constant (kJ and as intercept the values of koHK,/aH = koHCoHe. The values of kz determined in this manner (20') are recorded in Table V.
+5.401
I\
TABLE I1 APPARENT DISSOCIATION CONSTANTS(pKapp') A N D FIRSTORDER RATECONSTANTS (k,)FOR THE ISTRAMOLECULAR NUCLEOPHILIC ATTACKOF THE DIMETHYLAMINO GROUPOF I1 AND I11 AT THE ESTERBOND(20') Series
Substituent
~K'avp
kl, min.-1
I1
p-1'02 m -SO2
9 60 9.82 9.88 9.70 9.53
21,500 580 33.1 10.0 3.50
p-c1 H P-CHa I11
p-NO3 m-N02
p.c1 H P-CHt RATIOSOF
-2.601
Fig. 4.-Plots
I
ck1I-C
I
phenyl esters of I1 is 9.71
I
=t 0.12
* 0.08.
+
kaCE,CMpsN
+
kOHChTKw/UH
(4)
At any constant p H in the presence of excess amine the determined pseudo-first-order rate constant is given by 5. kobs
= kzCMstN
+ koaKv/aa +
(51
Therefore, the slope of a plot of k o b s os. the concenUH),where A T tration of free amine (ie., ATKa'/(Ka' = ( C M ~ CMerNH@)yields the true second-order rate
+
TABLE I11 RATECONSTANTS FOR THE RINGCLOSURE RBACTIONS IN THE ESTERS O F 11 AND 111 (20')
THE
Substituent
kp/k,*"
P-NOz m-NO8
2.15 2.27 1.67 2.57 2.73
P-CHI
while the average The slightly greater basicity of the esters of 111 as compared to those of I1 (0.28 pKa unit) is exactly that expected.2g I n order to determine the true activation parameters i t was essential to correct the rate constants determined a t the various temperatures for the sizable heat of ionization ( m i ) of the dimethylamino group. The value of AHi was determined in the following manner. The P K ' a p p values for the phenyl ester of I11 and the pcresol ester of I1 were determined from pH-rate profiles a t lo', 20' and 34'. The apparent AHi values were then determined from PKlapp V S . 1/T plots and were found to be 8.1 kcal. for I1 and 7.4 kcal. for 111. These values were then reasonably assumed to hold for all esters of I1 and 111, respectively. The pKlappa t any temperature (0' to 34') could then be calculated for each ester of I1 or 111. In turn kl a t each temperature was then calculated from the expression kl = kobs (K'app uH)/K'app. The values of k o b s employed to determine the true kl constants were determined a t 10'; 20°, 27' and 34' (Fig. 4). The determined activation parameters are included in Table IV. Bimolecular. -The rate of appearance of phenolate ion in the bimolecular reaction of trimethylamine with substituted 0-acetyl phenols follows the expression d phenol/dt =
255 19.9 3.90 1.28
p-C1 H
I/T x IO? of log (ki or k2),the true nucleophilic catalysis con. stants, for I, I1 and I11 US. 1/T.
pKlappfor the esters of I11 is 9.93
10,000
9.90 10.07 10.13 10.00 9.87
(23) E.J. Cohn and J. T. Edsall, "Proteins, Amino Acids and Peptides," ReinhoId Publishing Corp.. Rew York, N. Y.,1943, p. 122.
TABLE IV ACTIVATIONPARAMETERS FOR NUCLEOPHILIC DISPLACEMENT BY THE DIMETHYL AMINOGROUP IN I, II ASD 111 (25')' Substituent
p-NOn
I. kcal. mole-1 TAS*
AH+
XI, kcal. mole-' AH+
TA&
111, kcal. mole-' AH+
TAS+
11 9 -1.9 11.5 -2.6 -4.3 11.8 -4 4 11.5 15.9 -2.2 13.8 -4.1 H 12.5 -5.7 12.3 -6 4 P-CHI 13.7 -5.1 14.4 -5 5 0 The values of E, were determined from plots of log ki US. 1/ T , AH* = E, - R T , TAS+ = AF+ - AH+, and AF+ = RT2.303 log (KT/hk,) (Frost and Pearson, "Kinetics and blechanism," John Wiley and Sons, Inc., New York, S . l',, 1953, pp. 95-97) standard state used was I mole 1.-'
m-NOs p-Cl
12.3 12.1 12.5 12.9
-6.3 -8.0 -9.1 -9.4
-
TABLE V SECOND-ORDER RATE CONSTANTS FOR THE REACTION OF TRIMETHYLAMINE (pKat 10.27) WITH A SERIES OF SUBSTITUTED 0ACETYLPHENOLS (20') Eubstituent
kP, 1. mole-1 min. --I
p-hTos
4.29 0.342 .0308 .00795
m-NO,
p-c1 H
The value of k2 for p-cresol acetate could not be determined above that for koH. The values of k:, were determined for each ester a t lo', 20' and 34' and were corrected in each case for the AH;of trimethylamine (determined under our conditions to be 9.3 kcal.). From the various rate constants AH* and T A S (25') were determined and are included in Table-~IV. A plot of log ko vs. ET was linear with a slope ( p ) of -1-2.2 .~ (Fin. 3). I n t'he'kodection of data necessary to evaluate A F I , AS*, AH* and p for system I there was obtained the necessary data t o determine these parameters for OH@catalyzed hydrolysis of substituted 0-acetylphenols. Apparent kOH' constants (5) were obtained from the intercepts of the plots of k,,bs vs. C.U~,Nand related to
Jan. 5, 13C3 Ac rIVATION
pseudo-first-order rate constant equivalent to the true rate constant for the intramolecular reaction. Of course these concentrations are not obtainable, but the method serves to show the very large enhancements obtained and is equivalent to the comparison of AF* values. I n the second set Gaetjens and MorawetzZ4studied the intramolecular catalysis of the hydrolysis of substituted monophenyl esters of glutaric and succinic acids (Table VIII). They obtained a limited number of activation parameters for the intramolecular reaction and were able t o compare these to a smaller number of activation terms for the bimolecular catalysis of the hydrolysis of substituted phenyl acetates by acetate ion. The bimolecular reactions were studied in the range of 60' and the intramolecular reactions in the range of 25'. With this in mind, it may be noted that the intramolecular displacement reaction is much more sensitive to the effects of strongly electron-withdrawing p-substituents than the analogous bimolecular reactions-a p = $2.5 for the former compared to p> I - +1.1 for the latter. In the comparison of the
TABLE VI HYDROXIDE 10N-CATALYZED
PARAMETERS FOR THE
HYDROLYSIS OF SWSTITUTED 6ubstituent
0--4CETYLPHENOLS
AH+. kcal. mole-1
T A S f (25')
10.1 10.3 9.4 10 1
-5.3 -5.5 -7.0 -6.8
p-so, m-SOz
p-c1 H
TABLE VII RATIOO F
INTRAMOLECULAR :INTERMOLECULAR CATALYTIC R A T E
CONSTASTS FOR THE
HYDROLYSIS O F ff2- AND ACETYLPHENOLS (CHa)zh'-a
Nucleophile Substituent
p-c1
fi-SUBSTITUTED
0-
CtHnN?*
Af, 1260 1080 1700 5370
H
min.-1/1
mole-' m h - 1
24 23 32 9.4
m-NOt P-KOp a Intramolecular reaction, y-( N,N-dimethylamino)-butyrates. Intramolecular reaction, y-(4-imidazolyl)-butyrates.
ACTIVATION TERMS FOR
I'
TABLE VI11 HYDROLYSIS O F PHENYL
INTER- AND INTRAMOLECULAR CATALYSIS O F -hlonophenyI
Aeoe
P-OCHa P-CHI H
x
T A S ~
AH*
10' (600)
Kcal./mole
x
ki. sec.-1 104
16.6
.098 .37
-9.3
p-c1 fi-COOCHj P-SOz
rx~)
0.041 .048
0.102
1.5
15.7
4.8 53
-8.5
the true bimolecular constant ( K o H ) for hydroxide ion catalysis by the expression koa = kox' (aH/K,)
plutarates-
AH*
-Catalysiw-
kz, M. sec.-l Substituent
5
HYDROLYSIS OF PHENYL ACETATESBY DIMETHYLAMINO
(6)
The temperature dependence of K , was considered in our calculations. Plots of log koH vs. 1/T lead to the values of AH* and TAS+ which are summarized in Table VI. The p-value a t 30' was found to be 1.1 in agreement with previous studies."
+
Discussion Various attempts have been made to determine quantitatively just what type of acceleration might be obtained in converting an intermolecular nucleophilic catalysis to an intramolecular nucleophilic catalysis and to see if the mechanisms for both processes are similar. Until the present none have been completely satisfactory The earlier comparisons have been summarized by Bendere3 Three sets of data now appear to be worthy of consideration. All three involve a comparison of bimolecular and intramolecular nucleophilic catalysis of the hydrolysis of a series of m- and p-substituted phenyl esters. I n chronological order, the first involves the comparison of the bimolecular catalysis of the hydrolysis of the phenyl esters by imidazole to the intramolecular catalysis of the hydrolysis of phenyl esters of y(4-imidazolyl)-butyric acid.' I n this study activation parameters were not determined. A rough approximation of efficiency may be obtained, however, by dividing the first-order rate constant for the intramolecular reaction by the second-order rate constant for its bimolecular counterpart in Table VII. In so doing, a number is obtained in units of moles (i.e., time-'/l. moles-' time-' = M ) , and one may suppose that this would be the hypothetical concentration a t which the catalyst, in the bimolecular reaction, would have to be raised in order to obtain a
ACYLATES BY CARBOXYL
--
T AS*
ki, 8ec.-l
cKcal./mole(250)
19.3 19.5
-5.5 -5.2
20.3
-3.2
19.1
-1.8
x
''
ANION
hfonophenyl succinatesAH+ TAS+ -Kcal./mole-
104 (250)
6.2
(250)
19.0
-2.8
rates of hydrolysis of y- (4-imidazolyl)-butyrate esters with imidazole participation to the bimolecular hydrolysis of phenyl acetates by imidazole p changes from, +1.7 to +1.35 with the bimolecular hydrolysis being more sensitive. I n the present study the intramolecular catalysis for both the y-(N,N-dimethylamino)-butyrate and &(N,N-dimethylamino)-valeratesis more sensitive to electron withdrawing substituents than the corresponding bimolecular hydrolysis of phenyl acetates as catalyzed by trimethylamine, the p values changing from + 2 . 5 to +2.2. The small changes in p obtained on converting the bimolecular displacements involving imidazole and trimethylamine to their intramolecular counterparts most certainly do not indicate changes in mechanism. However, the large change in p accompanying the conversion of the bimolecular catalysis by acetate anion to its intramolecular counterpart definitely does suggest a change in mechanism In addition, the decided dependence of rate on AS* in the intramolecular reactions as compared to the bimolecular reaction as catalyzed by carboxyl ion (Table VIII) further substantiates a definite difference in mechanism for the two processes. Substituent effects on the ionization constants of phenol are related, in the main, to changes in AS but not to AH of ionization25,Z6 Substituent
P-NO* rn-hTO1 $5-c1 H P-CHs
AH
TAS (25')
Ref.
4.70 4.70 5.80 5.65 5.52
-5.00 -6.70 -7.00 - 8 00 -8.25
25 25 25 25,26 26
(24) E. Gaetjens and H. Morawetz, J. A m . Chem. SOL.,82, 5338 (1960). ( 2 5 ) L. P. Fernandez a n d L. G. Hepler, ibid., 81, 1783 (1959). (26) D.T.Y.Chen a n d K. J. Laidlex, Trans. Faraday SOL.,68, 4809 (163).
TIIOXAS C.BRUICE. ~ S D STEPHES J. BENKOVIC
6
The marked parallel dependence of the intramolecular catalytic coefficients upon AS* suggests that phenoxide ion is being formed in the critical transition state. In addition, the rates for intramolecular carboxyl group participation in the succinate monophenyl esters far exceed those for the glutarate series. This may be interpreted on the basis that the attack oi the carboxyl ion on the ester bond is also important in the critical transition state. On the basis of this evidence (i.e., p values, influence of ring size, AS* values), Llorawetz?4 postulates that the normal addition elimination mechanism expected for the bimolecular reaction 7 is supplanted in the intramolecular reaction by an S s 2 displacement not involving a tetrahedral intermediate
,----ailf== 1-11
p-N02 W2-N02 I3
RCOOR
1
2RCOR'
2, --?-
1111
RCOCCHj
(7)
ATAS*, 1-11
p-c1 H
CHI COO^
CCHa
$,OR
I' 0
0
+ 12'03
In S it is certain that anhydride is formed as an intermediate. Thus, Bruice and Panditll were able to identify maleic anhydride and 3,6-endoxo-A4-tetrahydrophthalic anhydride as products of the solvolysis of their monophenyl esters. I t now appears certain that a change of mechanism does occur in the conversion of the acetate anion-catalyzed hydrolysis of phenyl acetates to the intramolecular counterpart but that the change of mechanism is from general base in the bimolecular to nucleophilic catalysis in the intramolecular reactions. Thus Butler and Goldz7 have recently established that the reaction of acetate anion with 9nitrophenyl acetate does not, in the main, go by way of acetic anhydride as an intermediate. The latter reaction appears to proceed primarily via general base Catalysis. Most likely, then, the catalysis of the hydrolysis of phenyl acetate by acetate anion is strictly of the general base type. The latter conclusion arises from considerations of the relative sensitivity of 9nitro and phenyl acetate to general base-catalyzed a m m o n o l y ~ i s . ~This ~ change of mechanism (i.e., general base-catalyzed hydrolysis to nucleophilic catalysis) occurring on conversion of a bimolecular catalysis to its intramolecular counterpart is of profound interest and is receiving further study in this Laboratory. The intramolecular nucleophilic catalysis by carboxyl anion, as studied by Morawetz, shares in common with the bimolecular and intramolecular systems of this study (I, 11, I11 and OH@ catalysis) the fact that AH* is independent of the nature of the substituent group and that TAS* is dependent on the nature of the substituent group. The values of AH* for these reactions is dependent only on the nature of the nucleophile (i.e., R-N(CH3): vs. OHe) but is independent of whether the reaction is of an intramolecular (I1 and 111) or intermolecular (I) type. Except in the case of the p-chloro esters, the change in AH* accompanying the conversion of I + I1 or I + I11 amounts to less than a kcal./mole. (27) A. R. Butler and V. Gold, J. C h e m . SOL,1334 (1962).
1-111
0.8 0.3 -1.3 0 6
Comparing the values of TAS* for I to those for I1 and 111 shows that the formation of the five-membered ring of I1 is favored by ca. 4.7 kcal. mole-' (-16 e.u. a t 25') and the formation of the six-membered ring of 111 is favored by cu. 3.8 kcal. mole-' (-13 e.u. a t 25') over the bimolecular reaction of I.
p-KO2 in-S02
0 0
03
k P . l l . Illule -1-
0. 4 0.6 -3.4 0, 4
p-c1
(8) k,
17d.S3
l.cp.1
!r,"le-l-----1-111
-3.7 -6 9 -3.7
-3.7 -3.6 -5.u -3.0
- 4 . 7 Zk 1 . 1
- 3 . 8 zk 0 . 6
-1 4
The increase in efficiency of I1 and I11 with respect to I (Table VII) must reside in the gain in translational entropy brought about by bringing the reactant groups together, there being no evidence for any change in mechanism.2Y Clearly then, all steric and electronic influences on AF*, in the systems I, 11, 111, as well as in the case of OH--catalyzed hydrolysis of O-acetylphenols and in the solvolysis of the monophenyl esters of succinic and glutaric acids, are essentially apparent only in changes of AS+. The alterations in AS* on substitution are more apparent in I, 11, I11 and in the monophenyl esters of the dibasic acids than in the case of OH- catalysis. This is as expected on the basis that the p values for the former are at least twice that of the latter. There remains to be explained just why, in the systems discussed above, the electronic effect of subjtituent groups alters AS* and not AH*. As far as the authors are aware this is a rarity in kinetics,29though the dissociation constants of oxyacids in general are related primarily to alterations in AS. 30 Whether electronic effects on AF* are reflected in alterations of AH* or AS* may be dictated by the solvent. Thus, for the alkaline hydrolysis of substituted 0-acetylphenols the p value is always ca. 1.1 regardless of solvent (HzO, p = 1.1 (this study and ref. 23); 28.5% EtOH-H?O(v./v.), p = l.1517; 60% acetonethough the values of A F water (st.%), p = l.ljY2) are greatly solvent dependent. Of greatest interest is the fact that electronic eEects show up in AS* in aqueous solution of p = 1.5 (this study) but show up in AH* in GOYoacetone water3': -AH+
H m-CH3 P-CH, m-NOz
P-NOz
11.9 12.3 12.3 10.7 10.4
Kcai. mole-'---------. TAS* (23')
5.7 5.6 5.7 5.6 5.:
It has not been ascertained if the systems I, 11, I11 as well as the monophenyl esters of glutaric acid and succinic acid might behave similarly. A working hypothesis assumes a mechanism proceeding through a (28) M. L. Bender, E. J. Pollock and M. C. Xeveu, J . A m . Chem. Soc., 84, 595 (1962), have recently shown the reaction of trimethylamine with phenyl acetate does not exhibit a Da0 solvent isotope effect. (29) L. P. Hammett, "Physical Organic Chemistry," McGraw-Hill Book Co., Inc., New York, N. Y.,Chapt. VII, 1940. (30) L. Eberson and I . Wadso, Acfo C h e m . Scand., in press. (31) Extrapolated from the data of E. Tommila and C. N. Hinshelwmd. J . Chem. SOC.,1801 (1938).
HYDROLYSIS OF P H E N Y L ACETATES BY
Jan. 5 , 1963
DIMETHYLAMINO
Chart I I
24
II
\
I1 0
0
x
rc-0-
~
O
-
~
III
0 -
CH,-N-CH,
C
H
,
x
0 ~
O
-
~
CH,-N
0
%Cx
a
12
0
0‘ I,
4t 0
4
12
8
Pk,,,
16
20
.
Fig. 3.-Plot of A H 7 z’s. pKinsafor those nucleophiles whose catalysis of o-substituted phenyl acylates is characterized by a constant AH:.
tetrahedral intermediate (7) in which the free energy of the transition states associated with the formation of the bond between the nucleophile and the carbonyl carbon as well as that for the departure of the leaving group were comparable to that for the tetrahedral intermediate.32 Electronic effects on the formation of the bond between the nucleophile and the carbonyl group would be expected to be evident in AH* while the departure of the leaving phenoxide ion would be expected to be most evident in A S . For such a mechanism the solvent could determine if the formation of the bond between the nucleophile and ester or the departure of the phenoxide ion from the tetrahedral intermediate were of the greatest importance, and thus determine the effect of substituent groups on AH* and AS*. The importance of ring size on the magnitude of k, has been consideredgI1land for a mechanism of this type krate = k a ( k b / ( k b + k c ) ) where neither k b nor kc can be ignored; ring size would always be important regardless of whether the ka or k, term were of greatest significance ( 7 ) .33 It has previously been noted34 that the Bronsted C) adequately corequation (ie., log k2 = @Ka’ related the rate constants for displacement of p nitrophenol from 0-acetyl-p-nitrophenol when bases of the same type were employed ( i e . , 4(5)-substituted imidazoles) but that different plots were obtained if the type base was changed (imidazoles, anilines, pyridines, phenoxide ions, etc.). The present study suggests that a more meaningful plot might be that of 11.
C h a r t I1
+
AH*
=
aPKa 4- C
(11)
A plot of this type is presented in Fig. 5. It may be (32) Such a mechanism would be experimentally indistinguishable from a direct S N displacement ~ reaction. I t would be in agreement with, but not uniquely required t o explain, the known fact that in the alkaline hydrolysis of phenyl benzoates in Hz018 no 01% is back incorporated into the ester ( C . A. Bunton and D. N. Spatcher, J. Chem. Soc., 1079 (1956)), since in such a mechanism the values of kz and ka would be greater than the rate of proton exchange (M. Eigen, Disc. Faraday Soc., 17, 204 (1954)). Evidence ha5 recently been presented that the rates of proton exchange may be kinetically significant in the alkaline hydrolysis of +substituted methyl benzoates (M. L. Bender and R. T. Thomas, J. A m . Chem. Soc., 83,4189 (1961)). (33) Alternatively if kb became very much greater than kc, the importance of ring size might be determined by the relative stability of the carbonyl double bond in the product (H. C. Browu, J. Ovg. Chem., 22, 439 (1957)). (34) T. C. Bruice and R . Lapinski, ibid., 80, 2265 (1958).
noted that bases widely divergent in type fit one single line. U‘hether this is fortuitous or not awaits further investigation. It may be noted that the values of ANT for i m i d a ~ o l eand ~ ~ ,thiol ~ ~ anion36attack on O-acetylpnitrophenol exhibit negative deviations from the plot of Fig. 5 (10 and 8 kcal. mole-‘, respectively); however, it is not yet known whether these bases operate o b the same type mechanism (ie., constant AN* and varying AS* with substitution) in H20. The ratio of the rate constants for the formation of the five-membered ring to that for the formation of the six-membered ring ( k ,I1 lis. 111) were found to be 2.3 f 0.3; Table 111. This value is just one-hundredth the ratio obtained for the solvolysis of the glutarate (35) M. L. Bender and B. W. Turnquest, ibid., 79, 1652 (1967). (36) J. R . Whitaker, ibid., 84, I900 (1962).
8
D. S.BERNS,H. L. CRESPIAND J. J. KATZ
and succinate ester^.^^" A possible explanation for this difference could reside in the preferential conformation of the ground states for the dimethylamines and the carboxylic acids. Possibly owing to the solvation of the carboxyl anion the glutarate and succinate monoesters exist preferentially in an extended conformation, whereas the relatively non-polar amines exist in a coiled conformation due to formation of hydrophobic bonds (Chart 11). I n this connection i t should be noted that rate studies conducted in 50yodioxane-water and also in 8 M aqueous urea solution failed to alter this ratio. 37 (37) These experiments are probably not critical since the uncoiling of hydrocarbon chains due to clathrate formation requires at least a 7-membered chain (see L. F. Fieser and M. Fieser, “Advanced Organic Chemistry,’’ Reinhold Publishing Corp., New York, N. Y.,1961, p p . 131-133). Also, the water content of 507, dioxane-water would probably not be low enough t o bring about uncoiling.
[CONTRIBUTION FROM THE CHEMISTRY DIVISION,
Vol. 85
A second explanation would implicate steric hindrance imposed by the dimethyl substitution of the amino group on the rate of closure to form the five-membered ring (though the magnitude of the rate constants would not appear to suggest any steric hindrance; see Table VII). I t should be noted that the ratio of the rate constants for the formation of five- and six-membered rings in the cyclization of w-aminoalkyl bromides in water is 800.3s Our investigations in this area are continuing. Acknowledgments.-This work was supported by grants from the National Science Foundation and the National Institutes of Health. S. J. Benkovic particularly acknowledges support in the nature of a predoctoral fellowship from the National Institutes of Health. (38) G. Solomon, Helv. Chim. Acta, 16, 1361 (1933); 17, 851 (1934).
ARGONNE NATIONAL LABORATORY,
ARGONNE,
ILL.]
Isolation, Amino Acid Composition and Some Physico-chemical Properties of the Protein Deuterio-phycocyanin1 BY DONALD S. B E R ~ ’ sHENRY ,~ L. CRESPIAND
JOSEPH
J. KATZ
RECEIVED AUGUST15, 1962 The isolation and purification of a fully deuteriated protein, deuterio-phycocyanin, from blue-green algae grown autotrophically in 99.8y0 D20 is described. Sedimentation behavior in the ultracentrifuge shows the protein to be a system of reversibly interacting components. The amino acid compositions of ordinary and deuteriophycocyanin isolated from Plectonema calothricoides have been established and it appears that the amino acid compositions are identical within experimental error. Ordinary and deuterio-phycocyanin therefore probably differ only in isotopic composition. Thermal denaturation of ordinary and deuterio-phycocyanin, both dishas been studied by measuring the quenching of fluorescence. Deuterio-phycocyanin undergoes solved in H20, thermal denaturation a t a temperature 5’ lower than is the case for ordinary phycocyanin. Since the two proteins appear t o differ primarily in the isotopic composition of the non-polar side chains, differences in denaturation behavior are probably to be ascribed to differences in hydrophobic bonding.
Introduction The successful cultivation of fully deuteriated organisms on the large scale by Katz, Crespi and coworkers3 has made a great variety of fully deuteriated substances accessible for the first time. Proteins in which hydrogen has been entirely replaced by deuterium may be expected to make a useful contribution to problems of protein structure and function, and efforts have therefore been directed to the preparation of such substances. In the present communication the isolation, purification, amino acid composition and behavior on thermal denaturation of the fully deuteriated protein phycocyanin, extracted from the fully deuteriated bluegreen alga Plectonema calothricoides, are described ; a preliminary description of some aspects of this work has already a ~ p e a r e d . ~ Phycocyanin is a blue photosynthetic pigment widely distributed in blue-green algae. A member of the class of biliproteins, for which molecular weights of the order of 200,000 to 300,000 have been quoted, phycocyanin appears in fact to be a system of reversibly interacting components. The chemical properties of phycocyanin are reviewed by O’hEocha.6 For purposes of the present discussion the ordinary, hydrogen-containing phycocyanin extracted from algae grown in (1) Based on work performed under the auspices of the U. S. Atomic Energy Commission. Presented in part at the 141st National Meeting of the American Chemical Society in Washington, D. C., April, 1962. (2) Resident research associate, 1861-1962. (3) H. DaBoll, H . L. Crespi and J. J. Katz, Biofechnologr and Bioengineering, to be published. (4) D. S. Berns, H . L. Crespi and J. J. Katz, J . Am. Chem. Soc., 84, 486 (1962). (5) C. O’hEocha, in “Comparative Biochemistry of Photoreactive Systems,” M. B. Allen, ed., Academic Press, Inc., New York, N. Y.,1860, pp.
181-205.
HzO will be referred to as ordinary or protio-phycocyanin; the protein extracted from algae grown in 99.8% DaO will be designated deuterio-phycocyanin. The prefix deuterio- will also be used when the deuteriophycocyanin is dissolved in HzO and the exchangeable positions of the protein are occupied by hydrogen. The sharp distinction between the protein described here, in which the non-exchangeable positions in the molecule are occupied by deuterium, and ordinary proteins into which deuterium is introduced into exchangeable positions by treatment with Dz06 can be readily inferred from the discussion of Scheraga’ on hydrogen-deuterium exchanges in proteins. Experimental Isolation and Purification .-Deuterio-phycocyanin and protiophycocyanin were isolated from the blue-green alga Plectonema calothricoides by either of two procedures. In the first, approximately 25 g. of algae (wet weight) were frozen and thawed twice to rupture the cells. About 200 ml. of aqueous acetate buffer (pH 4.7, p = 0.1) was added and the solution allowed to stand, first for several hours a t room temperature, and then in the refrigerator a t 5’. After several days, the supernatant solution was quite blue and intensely fluorescent. The supernatant solution was removed by centrifugation, fresh buffer was added, and the extraction continued until most of the phycocyanin was removed from the algae. An olive-green appearance of the residue indicated that most of the blue pigment had been extracted. The aqueous extract was then centrifuged for 10 minutes a t 4’ a t 12,000 r.p.m. to remove debris. Phycocyanin was then precipitated from the clarified extract by adding ammonium sulfate to 50y0 saturation. The second procedure for the extraction of phycocyanin from algae omitted freezing and thawing. Instead, cell lysis was achieved by the action of lysozyme (Worthington 2 >
