DIBASIC ACIDS AND AMPHOTERIC ELECTROLYTES.
'503
4. The tube A (12 mm.) projects up through the stopper and to the bottom of the bottle. 5 . The intake C should be 4 mm. or as small as possible and still give sufficient flow. Its height determines the height of the water-table. Note.-All calibrations are outside measurements. F. C. CLAPP. DIVISIONOF SOILS, MINNESOTA AGRICULTURAL EXPERIMENT STATION. SI. PAUL, MINNESOTA.
[CONTRIBUTION FROM
THE CHEMICAL LABORATORY OF THE UNNERSITY OF CALIFORNIA. ]
RELATIONS BETWEEN THE CONSTANTS OF DIBASIC ACIDS AND OF AMPHOTERIC ELECTROLYTES. BY ELLIOT QUINCY ADAMS. Received April 4, 1916.
This paper comprises some remarks on the ionization of polybasic acids, of their acid esters, of polyacid bases, and of amphoteric electrolytes.
1504
ELLIOT QUINCY ADAMS.
The fact that the two dissociation constants of a dibasic acid seldom m e r by a factor less than four or five appears in all experimental studies of such constants,' and Walker2 observed that the constants of acid esters of symmetrical dibasic acids are never3 more than half those of the acids from which they are derived. In the following discussion it will be shown that these results are to be expected. The term amphoteric is applied to two classes of compounds-hydroxyl compounds, like Zn(OH)z, which may lose either hydroxyl or hydrogen ion-and compounds like the amino acids which may gain or lose hydrogen ion; and in the latter case the practice of expressing the "base constant'' of such electrolytes in terms of hydroxide ion concentration has very greatly obscured the results. In this paper the constants for amino acids and amines will be expressed in terms of hydrogen ion concentration, e. g., for aniline5 K = (CeHsNH2)(H +)/(CsH6"3 '1. The first point to be investigated will be the relation between the two ionization constants of a dibasic acid. The ionization of such an acid is represented in Fig. I in which the two HA H ionizable hydrogens are distinguished by being &I \ K written, respectively, before and after the symHA' bol A of the negative radical. K1, Kz, Ks, AH' Kq are the respective constants for the. reac-
+
+
H+, HA- = H + A--, AH- = A-H+, namely, K1 = (H+)(AH-)/(HAH), etc. (It follows that K& KzK~.) Expressions for the first and second ionization constants as ordinarily defined, K' and KK-since (AH-) (HA-) is the total concentration of intermediate ion-may be derived as follows: Fig.
I.
+
From these general equations may be derived the simpler equations for certain special cases. Cf. E. E. Chandler, THISJOURNAL, 30, 707 (1908).
* J. Walker, J . Chem. Soc., 61, 715-7 (1892).
3 Adipic acid (Loc. cit.) is only an apparent exception, for the fall in the observed constant on dilution shows clearly, as Walker himself pointed out, the presence of some conducting impurity. As will be shown later. In the following equation (CsHsNHz) means the aggregate concentration of aniline and a l l of its hydrates, just as (H+) means the aggregate concentration of hydrogen ion and all of its hydrates, etc.
DIBASIC ACIDS AND AMPHOTERIC ELECTROLYTES.
1505
The simplest of these is that in which the acid is symmetrical and hhe, hydrogens independent of each other’s ionization. Then K1 = Kz = & = & = K and Equations I and 2 become The condition of symmetry is easily satisfied, that of adequate separation of the hydrogens not so easily, for in the “straight chain” of five or six carbon atoms ring formation is possible and hence the ends of the chain may be near one another even when no ring is formed. In the derivatives of triphenylmethane the rigidity of the benzene rings may be trusted to keep para-substituents in the several rings apart; and the best examples of Equations 3, 4 are found in such derivatives. Thus the two constants for phenolphthalein’ as determined colorimetrically by Rosenstein2 are K’ = 11.5 X I O - ~ O , K ” = 2.83 X IO-'^, whence K‘/K“ = 4.06, while the first two constants of crystal violet as determined spectrophotometrically by Adams and Rosenstein3 and expressed in terms of acid concentration (although crystal violet is a base) are 1.62 X IO-^ and 6.6 x IO-^, whence K‘/K” for the yellow salt, regarded as a dibasic acid, is 4.08. The extreme clostness to 4 is accidental, as the ratio, in the second case a t least, is uncertain to five per cent. or more. Chandler4 has shown that the ratio of the constants for dibasic organic acids approaches this value at high concentrations. For independently ionizable hydrogens in an unsymmetrical acid KB = K1, K4 = Kz and Equation 2 becomes
If the hydrogens are not sufficiently separated to ionize independently, we may generalize from the behavior of oxalic acid and its immediate homologs, whose first constants are markedly greater and whose second constants are distinctly lees than the ionization constants of the corresponding fatty acids; and state the principle that a negative group inUn-ionized phenolphthalein and its singly charged ion are colorless, and are therefore supposed to be entirely in the lactone form; the doubly charged ion is pink and consists of a mixture, in tautomeric equilibrium, of the ion of colorless phenolphthalein and that of an unsymmetrical quinoid phenol-acid, 0 : CsH4: C(C6H40H)CE.H~CO~H.Since the ratio of the two constants, determined experimentally, agrees with that for a symmetrical dibasic acid, the quinoid ion can constitute only a few per cent., a t most, of the tautomeric mixture. I,. Rosenstein, THISJOURNAL, 34, 1 1 1 7 (1912); cf. phenol, K = 1.1 X 1 0 - l ~ . E. Q. A d a m and L. Rosenstein, THISJOURNAL, 36, 1452 (1914).In crystal violet the quinoid basic group is not involved in the change from the violet to the green and yellow forms. The two remaining nitrogens are equivalent, hence the yellow ion is symmetrical with respect to the two hydrogens which it is able to lose. E. E. Chandler, THISJOURNAL, 30, 707 (1908).
1506
ELLIOT QUINCY ADAMS.
creases and a negative charge diminishes the ionizability of any hydrogen in the molecule, whence KB< K1, Kd 4. (7) K” That is to say, the first ionization constant will in general be more than four times’ the second, the limiting ratio of 4 being found in symmetrical acids whose hydrogens are widely separated (and, therefore, independently ionizable). For a tribasic acid it may be shown in similar fashion that the limiting ratio between successive constants is 3 ; indeed the calculation may be made for an acid with any number of hydrogens. The truth of the conclusion for dibasic acids may be seen in Table I. The data are taken, unless otherwise indicated from Scudder,2 and are, in general, a t 25’. The acid esters of dibasic acids will differ from the acids in two ways: by the loss of one hydrogen and by the effect on the other hydrogen of the substitution. If the second effect-which will be in general a decrease in the “polar character” of the molecule and, therefoie, of the ionization-be neglected, then, as Wegscheidera showed, the ionization constant for the acid ester of a symmetrical dibasic acid will be one-half the first constant of the acid, and the sum of the constants for the two isomeric acid esters of an unsymmetrical dibasic acid will be equal to the first constant of the acid. Table I1 shows in how far this expectation is borne out. T o amphoteric electrolytes Equations I, 6 and 7 derived above, apply directly if the constant of the basic part of the molecule be expressed in terms of hydrogen ion concentration. Much of the supposedly anomalous behavior of amphoteric electiolytes is due to the neglect of the possibility that the constants so expressed will be related in such a 1 Cf. E. E. Chandler, LOC. cit. 2 &udder, “The Electrical Conductivity and Ionizatian Constants of Organic Compounds,” Van Nostrand (1914). 3 R. Wegscheider, Monetsh. Chem., 16, 153 (1895).
DIBASIC ACIDS AND AMPHOTERIC ELECTROLYTES.
TABLEI.
K'.
Acid.
(C02H)z... . . . . . . . . . . . CH2(CO2H)z. . . . . . . . . . (CHz)z(COzH)z... . . . . . (CHz)3(COzH)z... . . . . . (CHz)d(COzH)z.. . . . . . . (CH*)j(COzH)a... . . . . . (CHz)a(COzH)z... . . . . . (CHz),(COzH)z... . . . . . (CHz)a(COzH)z... . . . . . H(CH2)7COzH... . . . . .
K".
x 2.0 x
4.9
0.10
1.6 6.8 4.7 3.6 3.4 3.0
X IO-^ X IO-^
x
2.7
X IO-^
x
10-6
1.4
X
10-2
x
10-3
x
1.7
X
IO-'
x
IO-' 10-6
6 3 . 1 X 10-6 'I 2 . 4 x 10-6 8 0 . 7 X 10-6 3 . 0 X IO-^ 2 . 5 x 10-6
X IO-^ X IO-^ X IO-^ 10-2
800 25
16.2 12 . g 13.1 13 .o 11.3 10.0
10-5
2.6 X
IO-*
x
ZOO0
10-6
X IO-^
62.2
X IO-^
x
K'/K". 10-6
2 . 9 x 10-6 2.8 x 10-0 l 2 . 6 X IO-^ 22.3 x 1082.4 x 1 0 - 6 2 . 6 X IO-^
10-6
X IO-^ 2 . 7 X 10-6 2 . 6 X IO-^
ci~-CzHz(COzH)z. ..... 1.3 trans-CzHz(COzH)r. . . . 1 . 0 o-C~HI(COZH)Z.. ...... 1 . 2 VZ-CP,H~(COZH)Z.. ..... 2.9 CaH14(COzH)z.. . . . . . . . 2 . 3 ci~-p-C~,Hio(COzH)z. ... 3.0 truns-p-C6H1o(CO~H)~ . 4.6 HzSO4.. . . . . . . . . . . . . 0 . 4 5 H2SO4.. . . . . . . . . . . . . I . 7 9 H2C03.. . . . . . . . . . . . . 3 . 0
I507
5.0 1.3
x x x
10-2
10-6 10-11
5oooo 45 390 I2
33 IO
18 26 3400 23-
way that the constant determined in acid solution is the acid constant and that in alkaline solution the basic constant. Aminoacetic acid, for example, has the constants KA = 1.8 x IO-~O, KB = 2.8 X 1 0 - l ~ which ~ become in the notation used in this article K"
=
1.8 X
IO-~O,
K' =
x
1.0
10-l~
=
3.7 X IO-^, which are to be
K B
1.0X 10-l~4.1X 10-l~ 2.4 X IO-^^, N-benzoyl aminoacetic acid, K = 2.2 X IO-^, acetic acid, K = 1.8 X IO-^, N-acetyl aminoacetic acid, K = 2.3 X IO-^, chloroacetic acid, K = 1.6 X IO-^, showing that aminoacetic acid is a t once an acid stronger than monochloroacetic acid and a base nearly as strong as methylamine. compared with the constants for methylamine K =
By inversion; by partition = 4.4 X IO-^. By inversion; by conductivity = 1 . 9 X IO-^; by partition = 3.7 X 1 0 - 6 . 3 By conductivity;.by partition = 3.3 X IO-^. By conductivity; by partition 2.0 X IO-^; by inversion 3.9 X IO-^, 5.5 X IO-'. 5 By partition and inversion; by conductivity 3.0 X 10-6. 6 By partition; by conductivity 3.9 X IO-^; by inversion 1 . 7 X IO-^, 2 . 2 X 10-0. By conductivity; by partition 2 . 7 X IO-^; by inversion 1.0X 1 0 - 6 . These values are given for d-, I-, and dl-camphoric acids. Chandler (LOG. cit.) gives for the second constant of d-camphoric acid (by conductivity) K" = 14 X 10-6, but he uses K' = 2.3 X IO-^ in his calculations (apparently by mistake). As camphoric acid is an unsymmetrical acid it was rather surprising to find for i t constants whose ratio is less than 2 . 9 Landolt-Bornstein Tabellen (1912 edition). a
I508
ELLIOT QUINCY ADAMS.
TABLE11. Symmetrical Acids. Acid. K'. CHz(COzH)z . . . . . . . . . . . . . . . . . . 1 . 6 X IO-^ (CH2)2(COzH)z. . . . . . . . . . . . . . . . 6.8 X IO-' (CH2)4(COzH)'. . . . . . . . . . . . . . . . 4.2 X IO-^ (CH2)6(COzH)z.. . . . . . . . . . . . . . . 3.0 X IO-' (CHz)s(COzH)z.. . . . . . . . . . . . . . . 2.8 X IO-' C ~ - C ~ H Z ( C O , H. .) .Z... . . . . . . . . I . 3 X IO-' tt~~ns-CzHz(COzH)z. . . . . . . . . . . . I .O X IO-* o-CsH~(CozH)z . . . . . . . . . . . . . . . . 1.2 X IO-3 d-(CHOHC02H)2... . . . . . . . . . . . 9 . 7 X 10-4 (CH3)2C(C02H)2.. . . . . . . . . . . . . 7.6 X IO-^ m-(CHCH3C02H)2... . . . . . . . . . 1.4 X IO-^ dl-(CHCH3C02H)2.. . . . . . . . . . . 2.1 X 1 0 - 4 ~,~-C~ZCBH~(COZH)~(I,~). . . . . . . 3 . 4 X IO-' Unsymmetrical Acids. K'. Ester. C ( C H S ) Z C H Z ( C ~ ~. H . .).Z. . . 8.0 X IO-^ Methyl 3-N02C6H3(CO~H)z(~,z). . . . . I . 3 X IO-^ Methyl 4-N02C6H*(C02H)~(~,2). . . . . 7 . 7 X IO-* Methyl ~ - N 0 2 C s H 3 ( C 0 2 H ) 2 (.~. ,.4. ) . I .9 X IO-^ Methyl ~ - B ~ C B H ~ ( C O ~ H .).Z. .( .I , ~6.2 ) . X IO-* Methyl 4-(OH)C6H3(C02H)2(~,z). ... I . 2 X IO-^ Methyl 2-(OH)C6H3(CO:lH)2(~,4). . . . 2 .7 X IO-' Methyl o-C6H4(C02H)(CHrCO;H).. . I .g X IO-( Methyl Ethyl (3,4)(CHs0)2CtH2(C0~H)~(~,z) I . 17 X IO-' Methyl Ethyl Propyl
Ester.
K. 0.46 X IO-^ 3.2 X 10-6 3.0 x IOSs 2.5 X 10-6 1.5 X 10-6 1.4 X IO-^
Ethyl Methyl Ethyl Ethyl Ethyl Ethyl Ethyl Ethyl Methyl Ethyl Methyl Methyl Methyl Methyl Ethyl
Acid.
0.11
1. 2.3 1.7
3.0 0.08 0.37 0.15 0.25 0.43 0.46 '0.17 0.15 0.15
x x
IO-' IO-'
X IO-^
x x
2.0
IO-'
5.0
IO-'
X IO-^ X IO-^ X 10-8 X IO-^
x
10-2
E. 2.6 X IO-^ 0.3 x IO-' 5.2 X IO-^
10-2
X IO-^
x
x
0.48 X IO-' 0.66 X IO-^ 0.55 x IO-' 4 . 6 X IO-^ 3.1 X IO-( 0.45 X 10-4 0.6 X IO-^ 1 . 5 x 10-2
IO-'
0.20
2.77
x x x x
10-2
IO-* 10-3
IO-'
0.76 X 1 0 - 4 0 . 7 1 X IO-^ 1.30 I .OI
0.93
x x x
IO-* IO-' IO-'
Since K' is several million times as great as K", and is unquestionably the measure of the ionization of the carboxyl, aminoacetic acid will be almost exclusively the "inner salt," N+H&HzC02-, with less than one part per million of true amino acid, NHZCH~CO~H. In the case of the aminobenzoic acids (see below) this fraction becomes about 1%. It should be pointed out that aminoacetic acid appears to be an exception to the general rule that a negative charge diminishes the ionization of any hydrogen in the molecule, for whereas acetic acid, C02HCH3, loses its hydrogen more readily than acid malonate ion, COzHCHzCOz-, methylammonium ion, N+H3CH3, loses hydrogen ion less readily than aminoacetic acid, H+N3CHzC02-. The aminobenzoic acids do not show this anomaly, for, as bases, they are stronger than aniline. The data are from Scudder, but the designations a and j3 have been interchanged in the case of the methyl and propyl esters to agree with Wegscheider, as Scudder quotes Wegscheider and assigns no reason for the changes.
DIBASIC ACIDS AND AMPHOTERIC ELECTROLYTES.
1509
Comparing the aminobenzoic acids with benzoic acid, K = 6.7 X IO-^ and aniline K = 1.0X 10-14/4.6 X I O - ~ O = 2 . 2 x IO-^, and with their respective N-acetyl derivatives : Acid.
K,
=
1.0 X lW14,'K'.
o-Aminobenzoic . . . . . . . . . . 1.37 X 1 0 - l ~ m-Aminobenzoic. . . . . . . . . I . z z X 1 0 - l ~ p-Aminobenzoic . . . . . . . . . . 2.33 x 10-12 Chloroacetic acid. . . . . . . . . ....... m-Toluidine.. . . . . . . . . . . . 5 . 8 X 10-l" p-Toluidine.. . . . . . . . . . . . . I .8 X 10-l"
K'. 7 . 3 X 10-3 0.82 X 1 0 - 8 4.3 x IO-* 1 . 6 X 10-8
KA = K".
K(acety1-).
1.1 X IO-^ 2 . 4 X IO-( I .6 X IO-^ 0.85 X IO-^ 1.2 x 10-6 0.51 x IO-^
......
x
......
1.7
......
0.56 X IO-^
10-6
......
...... ......
the meta-acid is thus slightly weaker than monochloroacetic acid, the ortho- and para-acids stronger, while as bases all three lie between mand p-toluidine, the ortho-acid being twice as strong a base as aniline. The strongly negative character of the NH2 group may also be seen in the antagonistic action of neighboring NH2 groups. KB a K,/K'.
K'.
1.8 X IO-^ 5 5 . X C Z H ~ N H5.6 Z X IO-4 I .8 X C8H7NHz 4.7 X IO-^ 2 . 1 X %o-C4HgNHz 3 . 1 X IO-( 3.2 X 2iso-C6H11NH2 5 .o X IO-^ 2 .o X 1"s
Kl
-x
N2H4 3. 1 0 - l ~ 8(CH2)2(NHz)z 0.85 IO-'' (CHz)a(NHz)z3 . 5 1 0 - l ~ (CHz)4(NHz)z 5.1 1 0 - l ~ (CH2)6(NH2)1 7.3 10-l~
K,/K".
K".
300. x X IO-( 11 .8 X X IO-^ 2.9 X X IO-4 a.o X X IO-( I .4 x 10-6
10-11
10-l~
IO-'' 10-l~ 10-l~
Summary. For polybasic acids the ionization constants K' and K" as ordinarily defined are related to the constants, K1, K2, for the separate removal of the respective hydrogens, thus : K' = K1 K2 (1) I.
+
whence
K' 4. (7) KK 2. Phenolphthalein and crystal violet give values for this ratio very near the limiting value. (At high concentrations of dibasic organic acids, E. J3. Chandler has shown that this same limit is approached.) 3. The acid esters of a symmetrical dibasic acid have constants equal to approximately one-half the first constant of the acid, while the sum of the constants of the isomeric esters of an unsymmetrical dibasic acid is approximately equal to the first constant of the acid. (This generalization is due to Wegscheider.)
->
Not given in Scudder.
* Normal amine not given.
-
For ethylenediamine K' = I .o X 1o-l4/6.o X IO+ 1.7 X IO-*; for the others it is not given. It is to be remembered that a larger constant means a stronger acid or a weaker base.
1510
F. B . DAINS, H. R. O'BRIEN AND C. I,.
JOHNSON.
4. The constants of aminoacetic acid and the aminobenzoic acids are explained by the strong negative character of the NH2 group. BERKELEY,CAI..
[CONTRIBUTION FROM "HE DEPARTMENT OF CHEMISTRY OF THE UNIVERSITY OF KANSAS.]
ON THE REACTIONS OF THE FORMAMIDINES. V. ON SOME PYRAZOLONE DERIVATIVES. BY F. B. DA~NS, H. R. O'BRIEN A Y D C. Received May 31, 1916.
L. JOHNSON.
This is a continuation of a previous investigation on the pyrazolones,' where it was shown that the methylene group in this class of derivatives reacted with the formamidines, the hydrogen being replaced by the grouping >CHNHR. The general equation can be formulated as follows, RC C H * + RN
/I
CH-NHR
RC-C
= CHNHR
I N - N(R).CO II
I
N - N(R).CO
+ RNHi
the reaction yielding a substituted 4-amino methylene pyrazolone and the 'free amine. The following paper includes further experimental evidence as to the general nature of this reaction and a partial study of the action of certain reagents upon the substituted pyrazolones.
4-m-Toluido-methylene-l-p-tolyl-3-methylJ-pyrazolone, CHsC
I/
N
C = CHNHC7Hr.
I - N(C?H?).CO
Molar quantities of the pyrazolone and di-m-tolyl formamidine were heated for twenty minutes in an oil bath at 150'. From the reaction product the new compound was isolated after treatment with dilute hydrochloric acid, which removed m-toluidine. This separated from alcohol in golden yellow crystals melting a t 122'. Calc. for C18H190N8:N, 13.78%.
Found, 13.85%.
Under like conditions the ppaaolone and di-o-tolyl formamidine gave ~-o-toluido-methylene-~-p-tolyl-~-methyl-~-pyrazolone, yellow needles from alcohol which melt a t I 76.5 '. Calc. for Cl&hONs: N, 13.78%.
Found: 13.86%, 13.69%.
The ease of reaction of different formamidines seems te vary greatly. While many require heating a t 120-150' in order to ensure condensation, others unite a t much lower temperature. Thus p-tolyl methyl pyrazolone and di-o-phenetidyl formamidine a t water-bath temperature gave the amine and ~-p-tolyl-~-methyl-~-o-ethoxyanilido-~-pyrazolone. It crystallizes from alcohol in canary-yellow needles with a melting point of 133'. The same reaction occurs slowly when the components are ground together in a mortar and allowed to stand. Dahs and Brown, THISJOURNAL, 31, I 153 (1909).
