EDWARD J. KING
1OOG [COSTRIBUlI O S
S R O M 'I IIE
1701.
70
DEPARTMEST O F CHEMISTRY, BARNARD COLLEGE, COLUMBIA UNIVERSITY ]
The Thermodynamics of Ionization of Amino Acids. I. The Ionization Constants of y-Aminobutyric Acid BY EDWARD J. KING RECEIVED OCTOBER 2, 1953 The electrom3tive fsrces of cells with hydrogen and silver-silver chloride electrodes which contained buffer mixtures either of y-aminobutyric acid and hydrochloric acid or of the amino acid, its sodium salt, and sodium chloride were measured a t 5 " intervals from 10 t o 53". From these d a t a t h e thermodynamic ionization constants, K I for t h e ionization of the carboxyl group and KPfor t h e ionization of the ammonium group, were calculated. The changes in free energy, enthalpy, entropv a n d heat capacity associated with the ionization in the standard state can be obtained from the equations -log KI = (1209.07j T ) 3.7820 0.012605T and -log Kz = (2832.81/T) 0.5879 f 0.00187977'.
-
+
+
The ionization constants and related thermodynamic properties of twenty amino acids and one peptide are available from precise electromotive force measurements by the method of Harned and Owen.' The accumulation of more data of this nature is desirable in order to furnish a broad basis for the comprehensive examination of the thermodynamic properties associated with ionization. A comparison of the existing data for amino acids with that for substituted ammonium ions and fatty acids offers the possibility of new insight into the phenomenon of ionization. This will be dealt with in a subsequent communication. The thermodynamic ionization constants of three w-amino acids have been reported over a range of temperatures : g l y ~ i n e ,P-alanine4 ~,~ and e-aminocaproic acids5 Because of the wide divergence in behavior of the last two of these acids it is of interest to determine the thermodynamic properties of the ionization of ?-aminobutyric acid which follows /3-alanine in the series. All of these w-amino acids exist predominantly in the dipolar ion form in aqueous sdution. Their dielectric increments, for example, increase progressively with chain length, an indication of increasing separation of the -NH3- and -COOg r o u p 6 I n the case of T-aniinobutyric acid there are about 300,000 times as many dipolar ions, +NH3CH2CH2CH2COO-, as neutral molecules, NH2CH2CH2CHnCOOH, in aqueous ~ o l u t i o n .The ~ first ionization constant, K1, which corresponds to the reaction represented by
was derived from data on the cell Hz 1 'NH~CHZCHZCH~COO(mi), NH~CHZCHEH~COO (m - ) ,S a C l (m3) IAgCI-Ag
(11)
Measurements were made a t 5' intervals from 10 to 50' so that the related thermodynamic properties could be calculated. Experimental
Potassium phthalimide and y-chlorobutyronitrile from Distillation Products Industries were used a s starting materials for the synthesis of y-aminobutyric acid.8 T h e crude acid was purified by repeated reprecipitation from concentrated aqueous solution with redistilled absolute ethanol. The final product was dried first at 45-50' overnight9 and finally in a vacuum desiccator. It was found b y formol titration t o be 99.5 f 0.2y0 pure. It has been assumed in calculating the ionization constants t h a t the 0.5yoimpurity is inert. There was less than 0.2% water as determined by the Karl Fischer method."J Semi-quantitative tests" showed less than 0.004% chloride, iron, phosphate, heavy metals or ammonia.lz A n d . Calcd. for CJfsN02: C, 46.59; 11, 8.50; S , 13.58. Found: C , 46.60; FI, 8.73; N,13.35.'3 Standard hydrochloric acid was prepared by dilution of the constant boiling acid and its composition was checked by gravimetric analysis to within 0.06%. The sodium hydroxide solution was prepared b y diluting a clear 50% solution with carbon dioxide-free water. It gave no turbidity with barium chloride. T h e base was standardized against both the standard acid and potassium acid phthalate (sational Bureau of Standards Sample 54c) ; the two standartlizations agreed within 0.05%. T h e sodium chloride has hrcn described before. 3 The apparxtus and general techniques were substantially the same as in earlier ~ o r k . ~ , ' ~ T o, ' conserve ~ niaterials a single cell was filled with each buffer solution. Duplicate pairs of electrodes were used in each cell. T h e standard deviation between duplicate pairs was f 0 . 0 2 4 mv. for the whole set of measurements. Measurements over the tem+NH3CHzCHzCH?COOH 1x20 H 3 0+ + S H 3 C H ? C H ~ C H ? C O O - perature range were made as previously describedI4 except t h a t for cell I1 the measurements were made from 25 t o 50' has been obtained from measurements on the cell on the first day and at 2,5O anrl below on the day following. Agreement between successive readings a t 25', three or four H.2 1 +NHnCH?CH2CH?COO- ( t ~ l ) , in number, was used as a criterion of the performance of the cells. The standard deviations b e t w e n initial a n d final +SH~CH2CH2CHsCOOHCI(mz) IAgClb-Ag ( I ) readings a t 2.5' were f 0 . 0 4 8 mv. for cell I and 1 0 . 2 9 n i v . The second ionization constant, K 2 , for the reac- for cell 11. The poorer performnnce of cell I1 is nio.;t prob-
+
tion represented by
+
(8) C. C. Dewitt, Synlkescs, 17, 4 (1937). + 1 1 2 0 -3 (9) S. Dunn and L. B. Rockland, ".4dvanceu in Protein ChemisVol. edited by 51. L. Anson J. Edsall, Academic Press, try," H?O + S R ? C H ~ C H I C H ? C O O ~ Inc., New York, N. Y.,1947, p. 295. Org.
+SHsCH?CH?CHcCOO_
_
hf
+
( 1 ) H. S. IIarned and 13. B. Owen, "The Physical Chemistryof Electrolytic Solutions." '2nd ed.. Reinhold Puhl. Corp., X P WYurk, N. Y., 1950, Ch:ipter I,-,. (2) R . H Owen, THISJ O U R N A L , 56, 24 (1934). (3) E . J. King, i h i d . , 73, 15.5 (1951). (4) I f . l f a y and \V Felsinr, { h i d , , 73, 406 (1951). ( 5 ) E . K . B. Smith anrl P . R. Smith, J . B i d . C h n ~ 146, 157 (1912). ( 6 ) E . J. Cohn and J. T. Rds:rll, "Proteins, Amino Acids a n d Peptides." Reinhold Piihl Curp.. h'ew York, &*. \'., 1943, pp. I 4 G . 290. f7) Referenre 6 , p HCI.
111,
and
T.
(10) K . Fischer, Anueu. Chew., 48, 394 (1935). (11) 51. 1'. Stoddard and 11, S. D u n n , J . B i d . C h e m . , 142, 329 (1943. (12) T h e white precipitate which forms u p o n addition o f Kessler reagent t o a solution of the amino acid must be removed b y centrilugation and more reagent added in order to detect as little as 0.004% ammonia. (13) Analyses b y Schwarzkopf hficroanalytical Laboratory. Middle Village79, N. Y. T h e nitrogen wasdetermined b y the Dumas method. (14) E. J. Kiug and G. W. Kiug, THISJowRN.sL, 74, 1212 (l%j2). (I.?) E J . King. ibid., 75, 2204 ( 1 9 5 3 ) .
ably a consequence of the increased solubility of silver chloride in the alkaline buffer solutions. The major part of the decrease in electromotive force occurred after the cells had stood overnight. This has been corrected for by adjusting the inital readings at 25' and those which followed them on the second day t o correspond with those on the first day. The electromotive forces, corrected t o a hydrogen pressure of one atmosphere, can be represented as a function of temperature by thc equation Et = E26 ~ (t 25) b(t - 25)' (1) The parameters of this equation for electromotive forces in absolute volts and concentrations in moles per kg. of water are given in Table I . All wcighings have been corrected for thc buoyancy of air. The standard deviations between the observed electromotive forces and those calculated from equation 1 are 1 0 . 0 6 0 mv. for cell I and f 0 . 0 2 8 mv. for cell 11. TABLE I P A R A M E T E R S O F EQVATIOX 1
+
ml
= vi2
+
E25
Cell I 0.58516 ,57557 ,57204 ,56882 ,56439 .55990 ,55342 ,54854
0.00999 .01502 ,01756" ,02019 ,02458' .02997 .04020 .04970 mz/mi
rn1
a X 10'
b X 107
5.25 4.91 4.79 4.66 4.55 4.37 4.17 3.99
-9.2 -5.8 -6.5 -7.5 -6.5 -6.8 -8.0 -7.8
Ear
mdmi
a X
106
b X 106
Cell I1 0.02192 0,5062 0,5021 0.94555 ,02617 ,5062 .5021 ,94123 ,03025 ,5062 ,5021 .93779 ,03266 .5063 ,5021 ,93601 .04250 ,5063 .5021 ,92963 .04821 ,5063 .5021 ,92668 a Measurements were omitted a t 10'
2.00 0.40 -0.50 -1.10 -3.10 -4.30
-2.70 -2.66 -2.70 -2.78 -2.62 -2.75
Calculations and Results The thermodynamic ionization constants were obtained from the intercepts a t zero ionic strength ( p ) of plots of the functions pK', and pK', defined by the familiar relations' f m'II)mH/(mZ - m'H)] pK'2 = (3/2.3026RT)(Er1 - Eo) 41% [Wk(ml f m'OH)/(mZ - nZ'OH)]
pK'1 =
- log
[(ml
= (3/2.3026RT)(E1
tained from these calculations are given in Table 11. The standard deviations of the extrapolated values are =k0.00091 for ph', and *O.O01ll for pK,, the equivalent of *0.054 and *0.065 mv., respectively, in the electromotive forces. No values of the thermodynamic constants are recorded in the literature. Ley," on the basis of conductance measurements, reported apparent ionization constants of 5.9 X 10-j and 3.7 X lo-". These are in fair and 2.7s X agreement with the values 9.31 X lo-" a t 25' found in this investigation. TABLE I1 THENEGATIVE LOGARITHMS OF THE IONIZATIOU CONSTANTS OF -~AMINOBUTYRIC ACID Temp., OC.
9 Ki
10 15 20 25 30 35 40 45 50
4.0671 (0) 4.0459 (0) 4.0380 (6) 4.0312(-1) 4 0266 (-8) 4.0251 ( - 8 ) 4.0272(11) 4.0293 ( 8 ) 4.0320(-7)
lJKz 11.0260 (3) 10 8634 (2) 10.7087(-9) 10.5557 (2) 10.4091 (-7) 10.2694(3) 10.1336 ( 5 ) 10.0025(7) 9.8740(-8)
The variation of the ionization constants with temperature can be expressed by the following equations1* PKI = (1209.07/T) PKz (2804.84/T)
- 3.7820 + 0.012605T + 0.5879 + 0.0018797T
(F) (7)
The number in parentheses following each experimental value of pK in Table I1 is the difference in the fourth decimal place between that value and the one calculated from equation 6 or 7 . The changes in free energy (AFO), enthalpy ( A H o ) , entropy (A s o ) , and heat capacity (ACp0) associated with the ionization reactions in the standard state, as derived from equations 6 and 7 by thermodynamic methods,'* have the following values a t 25' : AFOs cal. mole-'
AH08 cal mole-'
5500 14401
405 12070
.
ASB, cal. mole-' deg.-'
ACPQ? cal. mole-' deg. - 1
-17.09 -7.82
-3-1 -5
(2)
First ionization Second ionization
(3)
The minimum probable errors in these quantities are, respectively, A2.1, *l6, A0.06 and A 2 based on random errors of A0.05 mv. in the electromotive forces and standard potentials and reasonable estimates of the weighing errors involved in making up the solutions. As a result of the lack of agreement above 25' between the standard potentials of Harned and Ehlerslg and those of the author,15 the values of the last three thermodynamic properties may be too low by Xi, 0.12 and 1.8 units, respectively. The assumption that the 0.5y0impurity is inert is another possible source of error. It is probable that the absolute errors in the thermodynamic properties are two to three times the miiiimum probable errors given above. These data show that in all respects y-aminobu-
The apparent hydrogen and hydroxyl ion concentrations were calculated from the electromotive force data by
- log m'H
1007
IONIZATION CONSTANTS OF ~-AMINOBUTYRIC ACID
Feb. 20, 1954
- E") f
+ A ~ ~ ' +, Zlog)mI2 (4) - Eo) + log ma - log K ,
[2~v',Z/(l - log m'on = (5/2.3026KT)(E11
(5)
The latest values of the physical constants were used.16 In the second term on the right-hand side of equation 4, which is the Debye-Hiickel approximation for the activity coefficient of hydrochloric acid in the buffer solution, the value of the parameter d was chosen as 3.85 A. to give extrapolations with low slopes. This choice of d does not affect the final value of the ionizaton c o n ~ t a n t . ' ~ 'The ~ negative logarithms of the ionization constants ob(Ifi) P. D.Rossini, F. Gucker, Jr.. €I. Johnston, L. P a u l i n g a n d G. Vinal, T H I S J O I I R N A I . , 74, 2690 (1952).
(17) H. Ley, Be?., 42, 354 (1909). (18) H. S. Harned a n d R. A. Rohinson, Trans. Faraday SOC.,36, 973 (1940). (19) H. S. Harned a n d R. W. Ehlers, TIIISJ O U R N A L 6 4 , 1350 (1932); 65, 652, 2179 (t!l8X)
IOOS
JOHN
E. DEVRIESAND E. ST. CL.UR GANTZ
tyric acid occupies an intermediate position between /3-alanine and e-aminocaproic acid. The changes in enthalpy and entropy probably reflect the decrease in flexibility of the hydrocarbon chain with removal of the carboxyl hydrogen in the first ionization and the increase in flexibility resulting from removal of the proton from the ammonium group in the second ionization. Further discus-
1'01. i ( i
sion of these properties in relation to those of the other amino acids will be dealt with in another communication. Grateful acknowledgment is made to Barnard College for the support of this work by a special grant and to Professor E. D. Stecher for helpful advice. S E W YORK,
s.Y .
[CONTRIBUTIOV FROM , 4 V A L Y T I C A L CHEMISTRY BRANCH,c.S.h T A V A L ORDNANCE TESTS r A r l O N ]
Spectrophotometric Studies of Dissociation Constants of Nitroguanidines, Triazoles and Tetrazoles BY JOHN E. DEVRIESA N D E. ST. CLAIRGANTZ RECEIVEDSEPTEMBER 8, 1953 The dissociation constants of several derivatives of nitroguanidine and related compounds have been determined by spectrophotometric means. The method of Stenstrom and Goldsmith has been applied to compounds exhibiting weakly acidic (monoprotic and diprotic) and basic dissociation. Each pK value was obtained graphically from a smooth curve plotted from several absorption observations over the useful pH range using buffered solutions. Sodium chloride was added t o maintain a constant ionic strength of one. Dissociation constants too low ( p K 10 t o 12) t o be determined by potentiometric titration were easily measured. Where potentiometrically determined values were available for comparison, good agreement was obtained.
Ultraviolet absorption spectra have been reported for a large number of nitro-substituted guanidines and tetrazoles. In some cases absorption differences were noted when acidic aersus basic solutions of these materials were measured. 1 - 3 No instances are reported where these differences were employed to arrive a t a measure of the dissociation constants of these compounds. The acidic properties of nitroguanidines and tetrazoles have been discussed and in cases where acid dissociation was extensive, potentiometric titration has been used to determine dissociation constant~.~-6In cases where there was only slight dissociation, potentiometric methods were not feasible. The need for information concerning relative acidities of these weaker acids and bases was evident, The application of a spectrophotometric procedure provided a convenient method for determining dissociation constants for several nitroguanidines. A variety of nitramines has been included in this study to illustrate the applicability of the method. Apparatus The absorption spectra were measured using a Beckman model DU spectrophotometer. A Beckman model G, pH meter equipped with general purpose glass electrode was used for all pH measurements. Sodium ion corrections for the glass electrode were applied where necessary.
Experimental The method of Stenstrom and Goldsmith? has been u>ed by several workers in determining di5sociatiorl constants ( 1 ) R. iY.Jones ant1 G , D. T h o r n , C a n . J . Re.yeui.ch, B27,828 (1949). (2) A . F. XIcKay, J. P. Picard and P. B. B r u n e t , Con. J . ChPm , 2 9 , 746 (1961). (3) 1'. J. S a b e t t a , D. Himrnelfarb a n d G. B. I.. S m i t h , THISJOURN A L , 6 7 , 2478 (1935). (4) R'. D. Kurnler and P. P. T. Sah, J . O i g . C h e m . , 18, 669 (19.53). ( 5 ) E. Lieber, S . H. Patinkin a n d H. H. T a o , THISJ O U R N I . . 1 3 , 1792 (1951). ( 6 ) J. S. Rfihina a n d R. X r , H e r b s t , J . Ovg. C h r m . , 1 5 , 1082 (14RO!. (7) U'. Stenstrom and N. Gold5mith. J . P h w C h r m . . 30, 1683 11926).
of weak acids. The method has been extended t o include two step ionization processes.*J The values for the ionization constants K l and K , .(re calculated as
where . A , = absorbance'o of the mixture of the two species, U I I ionized and singly ionized species .-I O = absorbance of the un-ionized molecule .1' = absxbance of the singly ionized ion :I? = absorbance of the mixture of singly aiid douhly ionized species A" = absorbance of the doubly ionized ion Absorbance values are used throughout the cquations rather than molar absorptivity as in the original tlerivation7.l' sirice cell length and molar concentration Lire constant for each series of measurements. The use of these equations is simplified by plotting the relationship between p H and absorbance. At the midpoint of the break in the curve, PH = pK since the log term equals zero. I n the present work, the absorption spectrum of each buffered solution was recorded in the ultraviolet range 200400 mp. IVave length peaks characteristic of the un-ionized molecule or ionized species mere selected for use. In general, one major absorption peak Tvas exhibited in the ultraviolet region by the compounds investigated and this peak most strongly illustrated the change in absorbance with change in pH of the medium. B y employing a large number of buffered solutions over the p H range (1-11),it was possible t o plot a smooth curve showing the relationship between pH and absorption. By careful measurement then the pK value could be easily obtained graphically. In the case of the compounds for which the second dissociation constant was investigated, the removal of the second proton resulted in little or no shift in the main absorption peak. 18) C. V. Banks and A. B. Carlson, Aitcl. C h i m A r t a , 7 , 291 11452) (9) h l . T . Rogers, T. 11.. Campbell and R. l l a a t m a n , THISJ V U R N A T . , 73, ,5122 ( 1 9 5 1 ) . (10) hhsorbance = d = log ( 1 ' I ) = o b c : molar ahsorptilit, = e = u.lf, I l l ) I.. J. l
