INTERACTION OF AMINO ACIDS AND SALTS II. SODIUM CHLORIDE AND THALLOUS CHLORIDE* BY NORMAN R. JOSEPH (From the Department of Physical Chemistry, Harvard Medical School, Boston) (Received for publication, July 5, 1935) The interaction of neutral salts and a-amino acids in water depends both on the nature of the salt and on the length of the hydrocarbon chain of the amino acid (9, 10). The activity coeffi- cients of amino acids in media of low dielectric constant depend largely on electrostatic forces and only to a small extent on the specific properties of ions and zwitter ions (2, 7). In order to describe further the interaction in aqueous systems, the electro- metric method of studying salt activity in cells without liquid junction has been applied to systems containing amino acids in the presenceo f thallous chloride and sodium chloride, respectively. Thallous Chloride-Thallium amalgam was prepared by the electrolysis of a saturated solution of thallous sulfate, with a mercury cathode. The thallous chloride and amino acids were C.P.p roducts, further purified by recrystallization from water. The experimental procedure was identical with that employed in the investigations with zinc amalgam electrodes (6). The observed E.M.F. values were constant within 0.2 or 0.3 millivolt, and duplicate determinations were reproducible within approxi- mately the same limits. There was no discernible drift in the readings. The results at two temperatures are recorded in Table I. Amino acids decreaset he activity coefhcient of thallous chloride. Four amino acids have been studied at l”, the concentration of * These studies are derived from a thesis submitted to Harvard Uni- versity in partial fulfilment of the requirements for the degree of Doctor of Philosophy. They were presented before the American Society of Bio- logical Chemists in 1934 (J. Biol. Chem., 106, xliii (1934)). 489 This is an Open Access article under the CC BY license. 490 Amino Acids and Salts. II salt being 0.005 M. The value of -log y3/~ao at 0.4 M amino acid decreases from approximately 0.06 for glycine to 0.03, 0.02, and 0.01 for amino acids containing respectively two, three, and four CHll groups. The two salts, thallous chloride and zinc chloride, can also be compared in their interaction with these amino acids. At any TABLE I E.M.F. of Cell AglAgCZ/TZCZ(,,IHgTZ,1TZCZ(,,,, Amino Aci&,,,~ lAgCZ]Ag (Approximatety 1 Per Cent Amalgam) - Molality of Molzdity of elwtrolyte. ns amino acid, ml E.M.F. --Log 7,/r: -- oolt Glycine at 25” 0.005 0.10 0.0013 0.011 0.005 0.20 0.0025 0.021 0.005 0.40 0.0048 0.041 0.005 1.00 0.0128 0.108 0.015 0.10 0 .OOlO 0.009 0.015 0.20 0.0020 0.017 0.015 0.50 0.0055 0.047 0.015 1.00 0.0108 0.091 0.015 2.00 0.0211 0.179 Glycine at 1” 0.005 0.20 0.0033 0.030 0.005 0.40 0.0065 0.060 0.005 0.60 0.0099 0.091 0.005 1.00 0.0161 0.148 Alanine at 1” 0.005 0.20 0.0016 0.015 0.005 0.40 0.0034 0.031 0.005 1.00 0.0073 0.067 or-Aminobutyric 0.005 0.40 0.0025 0.023 acid at 1” 0.005 0.60 0.0036 0.033 dLValine at 1” 0.005 0.20 0.0005 0.005 0.005 0.40 0.0012 0.011 - concentration, mz, -log ys/yaO in dilute solutions of zinc chloride has a value approximately 3 times as great as the corresponding value for thallous chloride. The difference can be largely attributed to the difference in valence type of the two salts, for in both cases the function (6 log y&p), where P is the ionic strength, is of the same order of magnitude. Somewhat similar results have been obtained by Failey (5), who N. R. Joseph determined the solubility of thallous chloride in the presence of various amino acids. He has reported the value of -log y3/y30 at 0.2 M glycine to be 0.032 at 25”, compared with our values, 0.021 for 0.005 M TlCl, and 0.017 for 0.015 M TlCl. Failey’s solubility determinations have been essentially confirmed by Straup-Cope and Cohn (15), but our value at lo, 0.030, is in better agreement with Failey’s value than are our results at 25”. Sodium Chloride-The study by means of amalgam electrodes of solutions containing amino acids and sodium chloride involved the consideration of a source of experimental error not encountered with amalgams of less reactive metals. When sodium amalgam is brought into contact with a solution of glycine, there is an evolution of hydrogen, which is quite conspicuous at concentra- tions of glycine higher than about 0.2 M. A similar effect is noted with all other amino acids that have been studied. Sir Humphrey Davy, the discoverer of sodium amalgam, observed that, ‘in the presence of solutions of ammonium salts, sodium is displaced and an ammonium amalgam formed (4). Methyl ammonium salts (16) and trimethyl ammonium salts (11) have also a tendency to decomposes odium amalgam. That amino acids also decompose sodium amalgam suggests that the ionized amino groups of zwitter ions are responsible for the effect.’ Electromotive force measurements in sodium chloride solutions containing glycine have been made at 1.4” and at 25’. At the lower temperature evolution of gas proceeds relatively slowly, and essentially reversible potentials are obtained even in dilute salt solutions, the errors becoming appreciable only at concentrations less than about 0.05 M. Even at the higher temperature the 1 Ions of heavy metals,s uch as zinc, thallium, or silver, in the presence of ammonium chloride prevent the formation of ammonium amalgam. In the presence of glycine they prevent the evolution of hydrogen. The cat- ions that form unreactive amalgams have a greater tendency than ammo- nium ions or ionized amino groups to displace sodium from its amalgam. Ions of the alkali and alkaline earth metals do not, have this effect, and the corresponding amalgams are unstable in the presence of amino acids or ammonium ions. The rate of decomposition of sodium amalgam is considerably diminished by the previous introduction of ammonium to the amalgam. The possi- bility of improving the electrometric methods for cations of alkali metals in this manner will be further investigated. 492 Amino Acids and Salts. II E.M.F. values at salt concentrations higher than about 1 M approach the reversible values estimated from independent experimental results, while at concentrations lower than this irreversible poten- tials are obtained. The experimental error in a mixture of 0.01 M NaCl and 0.4 M glycine, which is of the order of 10 millivolts at TABLE II E.M.F. at .S?J6" Of Cell UP: P- TC kmately 0.1 Per Cent Amalgam) t? Mlecotlraoliltyyt e, ofm a B. mMinoola litay cid. ofm s F.X.F. +hx r,/r,o -- - oozt Glycine 1.00 0.50 0.0007 0.006 2.00 0.50 0.0005 0.004 2.00 1.00 0.0013 0.011 3.00 0.50 0.0005 0.004 3.00 1.00 0.0011 0.009 4.00 0.50 0.0005 0.005 4.00 1.00 0.0010 0.009 Alanine 1.00 0.50 0.0007 0.006 1.00 1.00 0.0011 0.009 2.00 1.00 0.0012 0.010 4.00 0.25 0.0004 0.003 4.00 0.50 0.0006 0.005 4.00 1.00 0.0011 0.009 Lu-Aminobutyric acid 1.00 0.50 0.0012 0.010 1.00 1.00 0.0023 0.020 4.00 0.50 0 .OOll 0.009 4.00 1.00 0.0021 0.018 dl-Valine 1.00 0.10 0.0005 0.004 1.00 0.20 0.0012 0.010 2.00 0.20 0.0013 0.011 dl-Leucine 0.50 0.0694 0.0006 0.005 1.00 0.0585 0.0004 0.003 2.02 0.0460 0.0004 0.003 - - 25”, becomes about 1 millivolt at 1.4’. The approach to reversible values at low temperatures can probably be attributed to a decrease in the rate of decomposition of sodium amalgam. The approach to reversible values at high salt concentrations can be explained on the assumption that any increase in sodium ion concentration N. R. Joseph 493 at the amalgam electrode is small in comparison with the molality of sodium ions in concentrated solutions.2 The results at 25” with concentrated solutions of sodium chloride are given in Table II. They indicate a salting-out effect, maximal for leucine and valine, and minimal for glycine and alanine. This result agrees with those of Pfeiffer and Wiirgler (lo), who have found that the aqueous solubility of leucine is diminished in the presence of NaCl, and also with those of Cohn, McMeekin, and Weare (3), who have found a corresponding effect in solutions of cY-aminobutyric acid. In estimating the activity coefhcients of amino acids from their effects on the activity coefficients of salts, we have previously employed a relation derived by Bjerrum (1). 6-- -l-o--g= yY-2 6 log Y3 (11 6m3 6m2 By means of this relation, salting-out effects can be quantita- tively described and compared with those determined by the solubility method. The results plotted in Fig. 1 demonstrate that log y3/73O at high salt concentrations increases linearly with mz and that the proportionality factor is independent of m3. According to Equation 1 these results indicate that as a first approximation, and within the range of the concentrations studied, log y2/y20 increases linearly with m3 and that the proportionality factor is independent of mz. The slopes of the straight lines in Fig. 1 yield the values of the salting-out coefficient (6 log y2/6m3) estimated from E.M.F. measure- ments and permit their comparison with values derived from solu- bility determinations. Table III is a comparison of these values. The agreement of the two methods is quite satisfactory. While the results at high temperature and salt concentration indicate that the salting-out effect predominates, the E.M.F. deter- 2 No significant improvement in the irreversible readings was obtained by working with amalgams more dilute than 0.1 per cent. Any advantage gained by diminishing in this way the rate of decomposition is offset by the introduction of another error. The latter apparently results from the fact that any change in the composition of the electrode itself due to the displacement of sodium is more serious in dilute than in concentrated amalgams. 5” 2 at Cl a N er d ms Wat ntrate O$ nce O co O 0 d per I, upon ci s AlIthO Writ Acid Valine Leucine 0 A 0 B A 8 0 Slope from Slope from MC Meekin Weiffer and Cohn and Widler Moles Amino A ence of amino acid u Alanine A A A 1. Infl Clvcine 0 0 0 FIG. Salt Concentration 1.0 2.0 40 N. R. Joseph 495 minations in dilute sodium chloride-glycine systems at 1.4“ indicate that the activity coefficients are decreased, an effect corresponding TABLE III Values of (6 Log T2/8m3) at High Salt Concentration Crdeulated from System N&l plus E.M.F. Solubility measurements measurements Glycine ............................ 0.02 0.00 (9) Alanine ............................ 0.02 a-Aminobutyric acid ............... 0.035 0.04 (3) Valine ............................. 0.09 Leucine ............................ 0.12 0.09 (9) The figures in parentheses represent bibliographic references. TABLE IV E.M.F. at 1.4” f 0.2” AgIAgClINaCZc,f,IHgNa,INaCl(,,), Glyciwm,)IAgCIIAg (Approtimately 0.1 Per Cent Amalgam) - - 1 :.M.F. (calculated -Log Y,/YIO eleMctorolalylittey . of nr amMinool slity acid, of nz2 pofrinotm mfreeeazsiunrge - I E.M.F. (observed) E.(Mob.sFe. rvemd easureb-y ments ( 14) ) IWXltS) - volt volt 0.01 0.10 0.0013 0.0012 0.011 0.20 0.0025 0.0023 0.021 0.30 0.0036 0.0031 0.029 0.40 0.0047 0.0040 0.037 0.02 0.10 0.0012 0.0010 0.009 0.20 0.0023 0.0022 0.020 0.40 0.0044 0.0039 0.036 0.05 0.10 0.0011 0.0009 0.008 0.20 0.0021 0 .oom 0.018 0.40 0.0039 0.0036 0.033 0.10 0.10 0.0009 0.0009 0.008 0.20 0.0018 0.0017 0.016 0.40 0.0034 0.0032 0.029 to an increase of solubility. These results are thus in conformity with those obtained in dilute solutions of zinc and thallous chloride. At 1.4” the initial readings were somewhat less reproducible 496 Amino Acids and Salts. II than those in thallous chloride solutions. There was a tendency of the E.M.F. to decrease by several tenths of a millivolt when the flow of amalgam was interrupted. Each value in Table IV repre- sents the mean of two to four reproducible initial readings. These observations are compared in Table IV with E.M.F. values calculated from the very accurate freezing point determinations of Moles of Glycine per l,OOO$mWs ater FIG. 2. Influence of glycine upon dilute NaCl at 1.4” Scatchard and Prentiss (14). Their results for the activity coeffi- cient of sodium chloride can be expressed in the relation --log y3/y30 = l/2.303 (0.32633 - 0.55888m3~ + 0.57278m3 - 0.30496m& m2 - (0.07445 - 0.10539m3+)w2 (2) The calculated E.M.F. values of the third column of Table IV are obtained from this relation by means of the expression, E = -0.1089 log y3/y3”, the factor 0.1089 representing the value of the ratio (vRT/2.303NF) at 1.4”. N. R. Joseph 497 As Fig. 2 illustrates, the results of the two methods are in good agreement, particularly at the higher salt concentrations. The decrease in the activity coefficient of sodium chloride in the presence of glycine becomes less pronounced as the salt concentra- tion is increased, illustrating again the tendency of the salting-out effect to become predominant at higher salt concentrations. TABLE V Values of (- 6 Log y~/Bp) in Dilute Solution - M”Yty I I Leuoine Glycine e1eotdyte -___ . . _ N&l ................ 0.01 -0.09* 0.2413 0.228 TICIT ................ 0.005 0.06 0.11 0.16 0.30 CaCl*. ............... 0.01 0.04* 0.3011 0.14 0.14 0.15 0.25 0.32 znc1*t.. . . . . . . . . . . 0.01 I I 0.14 0.14 0.16 0.26 0.33 - - Figures in bold-faced type refer to a temperature of 0” f 0.2”; other figures to a temperature of 25”. * Amino acid solubility, Pfeiffer and Wiirgler (10). t ~3f.F. measurements. $ Freezing point measurements, Scatchard and Prentiss (14). $ Failey’s (5) value from TIC1 solubility is 0.32 for saturated TlCl at 25”; this agrees with unpublished results of Straup-Cope and Cohn (15). 11T he value 0.30 is estimated from freezing points of Pfeiffer and Angern (9) by comparison with freezing point effects calculated for zinc chloride (6). It is significant only to about f0.05, and characterizes not only CaCL, but also many other salts of the same valence type. DISCUSSION The influence of the various salts upon the different amino acids studied is most conveniently compared in terms of the coefficient (6 log y.&p). This coefficient, although it is in most casesa pproxi- mately independent of the amino acid concentration, is generally a function of salt concentration in dilute salt solution, particularly in the case of glycine systems. According to the results obtained with zinc chloride systems (6), it is more nearly independent of salt concentration the longer the hydrocarbon chain of the amino acid. The behavior of valine in this respect approaches that of 498 Amino Acids and Salts. II leucine, in that log y2/yXo is approximately a linear function of salt concentration. Considering these facts, it is advantageous to compare values at the lowest salt concentration experimentally studied. These are recorded in Table V. These results indicate that in water the sign and magnitude of the salt effect depend not only on the nature of the amino acid but also on the nature of the salt. The value 0.33, describing the effect in the zinc chloride-glycine system, as has been pointed out in Paper I, is in agreement with the results of measurements of amino acid solubilities as a function of ionic strength in media of low dielectric constant. In such media, this value as a first approxi- mation is independent not only of the number of carbon atoms of the amino acid, but also of the nature of the salt. Deviations from the value 0.33 can from this point of view be assumed to measure forces other than the electrostatic attraction of ions and zwitter ions. The data indicate that these deviations in aqueous systems are of the order of magnitude of the electrostatic attractive forces and that they depend on the nature of each dissolved com- ponent. The order of magnitude of these salting-out effects is about the same as in mixtures of salts and various non-electrolytes (8, 12, 13). As a measure of these forces, we may define the coefficient (Y by the relation Q = 0.33 + (6 log Y/&L), assuming that the salting-out effect is 0 in the zinc chloride-glycine system. It thus becomes evident that in the systems described in Table V (Y varies for a given salt by as much as 0.33 and for a given amino acid by as much as 0.23. These differences are of about the same order of magnitude as variations of the salt effects in mixtures of salts and acetic acid or its chloro derivatives (12). It may be concluded that in aqueous systems, because of the low order of magnitude of the electrostatic attractive forces between ions and zwitter ions, it is necessary to take account of salting-out forces. Unlike the attractive forces, which evidently depend only on the valence type of the ions and the dipole moment of the zwitter ions, other forces vary with the nature of the salt as well as with the number of carbon atoms of the amino acid. Unlike the salting-out effects in simple salt solutions, those between ions and zwitter ions are significant even at low salt concentra- tions, because both the electrostatic forces and the salting-out