AMIXE-AMMONIUM SALTINTERACTIONS
Nov. 20, 1964
on p is observed for the pQ and pk’ values. In the case of the pk13’ value (Fig. 2b) this dependence is similar to t h a t observed previously for proton ionization from the -NH3+ group of alanine.” Gorin and Clary2 measured the concentration change with 1.1 of the ionized sulfhydryl proton of CYS. They observed t h a t the concentration of this species changed very little as a function of p except at low p-values, and concluded t h a t R was likewise invariant with 1.1. As seen in Fig. 2a, our results show that R does change appreciably with p. However, the results of Gorin and Clary can be understood by reference to the data in Fig. 2b. Gorin and Clary followed the pk12’ ionization and i t is seen that over the 1.1 range used in their study 0.03 to p a . 5 = 2.5) there is little change in pk12’ except a t low p-values. On this point our results and theirs agree. However, R depends upon the relative magnitudes of pk12’and pk13’and not upon that of pkI2’ alone. It is seen that pk13’ has a considerably larger variation with 1.1 than has pk12’. Thus, R is seen to have a considerable 1.1 variation. Furthermore, the data in Fig. 2c show both pk132’ and pk123’ to have much greater 1.1 dependencies than the corresponding dissociations in the lower pH region. The A S values reported in Table I1 for proton dissociation from ?ILIA and S M C are of the expected magnitudes considering the ionic charges involved. The relative effects on pQ of the - - A S and AH changes between p = 0 and p = 1 are shown in Table 111.
-
(11) A. C . Batchelder and C. L . A . Schmidt, J . P h y s . C h e m . , 44, 880 (1940).
[ CONTRIBUTIOS OF
THE
DEPARTMEST OF
4783
TABLE I11 THEEFFECTOF CHANGING p ON TAS, A H , b(--TAS)lp = 0
Reaction HzCYS- (I) = (11,111) H+ (11, 111) = cys-2 (IV) H+ HMAA- = MAA-2 H+ HSMC = SMCH+
+
+
+ -
+ +
I,
a,
= 1
AXD
pQa
b ( A H ) i w = 0 2 3RTb(pQ)’p = 0
TI, = 1
,I
a,
= 1
0
-0
44
-0
44
-1 8
+o
72
-1
I
2
0
-1
2
-1
-0.66
+0.22
-0 44
a Data in kcal./mole and taken from Tables I and I1 and Fig. 1 ; temperature = 25’.
The change in TAS is seen to be the determining factor in the observed decrease of pQ with increasing p in all cases except the first. The data in Table I11 also show the large decrease of the pk1Z3’ and pk132’ values compared to t h a t of the pk12’ and pk,,‘ values with increasing p (Fig. 2c) t o be the result of the much larger TAS change with changing p upon proton ionization from the 11, I11 species. The method described here for microconstant determinations is more accurate than other reported methods when R is at“or near unity and t h e heats of proton dissociation of the groups involved vary considerably. Acknowledgments.-The authors gratefully acknowledge the valuable assistance in this study of M r . Lee Hansen, M r . Blaine Blad, and M r . Nolan Edmunds.
CHEMISTRY, UNIVERSITY O F SOUTH CAROLINA, COLUMBIA, SOUTH CAROLINA]
Ion-Solvent Interaction. The Interactions of Some Amines with Tertiary Ammonium Salts in Low Dielectric Solvents BY EARLK. RALPH,111,’A N D W. R . GILKERSON RECEIVED MARCH 23, 1964 The conductances of solutions of tertiary ammonium salts have been studied a t 25” in o-dichlorobenzene and chlorobenzene as a function of the concentration of added amines, such as pyridine and tri-n-butylamine. The results are consistent with the formation of a 1 : 1 cation-amine complex. The following order of increasing complex stability with tri-n-butylammonium ion in o-dichlorobenzene is observed : 4-cyanopyridine < acetonitrile < tri-n-butylamine < pyridine < 4-methylpyridine. The effect of solvent on the cation-amine complex equilibrium can be accounted for by postulating that the amine displaces one tightly bound solvent molecule in forming the complex.
Specific interaction between an electrophile and an anion or between a nucleophile and a cation, in solution, may be regarded2 as a displacement of a solvating solvent molecule from the ion by the incoming electrophile or nucleophile. It would be of interesttostudy such interactions as a function of solvent, structure of anion and cation, and basicity (or acidity) and structure of the molecule displacing the solvent. It is hoped t h a t some of the factors of importance in ion-solvent interaction may be elucidated in this manner. Fuoss and co-workers2 are actively engaged in studying anion interactions with solvents and added electrophiles. We report here the results of a study of the conduct(1) Guy F . Lipscomb Fellow, 1962-1963. (2) A . D’Aprano and R . M . Fuoss, J . P h y s . C h e m . , 6 7 , 1722 (1963). (3) For recent leading references, see M . A. Coplan and R . M. Fuoss, i b i d . , 68, 1181 (1964).
ances of several amine salt-nucleophile systems in the solvents o-dichlorobenzene (ODCB) and chlorobenzene (CB) at 25’. It was hoped t h a t the results could be interpreted in terms of cation-nucleophile interaction. Incidental to the foregoing study, the conductances of tri-n-butylammonium bromide and iodide were studied in ethylene chloride (EC). Increasing conductivity with age of solutions of amines in this solvent, indicative of amine-solvent reaction, precluded studies of salt-amine systems in this solvent. Experimental Materials.-ODCB (Allied Chemical Co., Solvay Process Division) was passed through alumina (Alcoa grade F-20) on a 30 X 2 cm. column and distilled on a 35 cm. column packed with glass helices. The middle fraction, b.p. 48’ ( 5 m m . ) , was stored over alumina which had been fired to approximately 800” for 1 hr.
EARLK . RALPH,111, A N D W. R. GILKERSON
4784
Vol. S(j
Just prior to use the solvent was passed through a 30 X 2 cm. Results column t h e bottom half of which was packed with fired alumina T h e equivalent conductances, A, of the various and the top half with molecular sieve (Linde Air Products, type salts in the three solvents with and without addends 5A). T h e specific conductance ranged between 3 and 8 X IO-" mhos/cm. CB (Dow Chemical Co.) was treated in exactly the are listed in Table I. In each case, the specific consame manner as ODCB, but had b.p. 45" (33 m m . ) and specific ductance of the solvent, one-hundredth or less of the conductance 7 X lo-" mhos/cm. Ethylene chloride ( E C ) total, was subtracted from the observed specific con(Dow Chemical Co.) was treated in the same fashion, b.p. 83.7", ductance to yield the specific conductance presumed specific conductance 4 X lo-" mhos/cm. due to the salt. Also included in Table I is the conPyridine ( P y ) (Merck reagent grade) was distilled from potassium hydroxide and then from barium oxide, b . p . 115-116". ductance of a 1.804 X M solution of TBXHPi in Solutions (0.01 M ) of P y in ODCB had specific conductances in ODCB as a functic.n of the concentration of added the range 1-10 X 10-lo mhos/cm. 4-Cyanopyridine ( I - C N P y ) MeCN. The treatment of the conductance data to (Columbia Organic Chemicals Co. ) was recrystallized from yield derived constants is described below for each benzene and dried under vacuum, m.p. 78.0-79.0" (lit.4 m . p . 78.5-80.0O). 4-Methylpyridine (4-MePy) (Columbia Organic salt. Chemicals Co.) was distilled under reduced pressure, b.p. 56' TBAHPi in ODCB and EC.-The method of Shed(32 mm.). I t was necessary to use freshly distilled material since lovsky15 was used, it turns brown on standing. A 0.01 M solution in CB had a specific conductance of 5 X lo-" mhos/cm. Tri-n-butylamine ( T B A ) (Matheson Coleman and Bell Division, T h e Matheson Co., practical grade) was refluxed under reduced pressure over where S is the Shedlovsky functionlj and y*z is calcupotassium hydroxide for 2 hr. a t 3 m m . and then distilled. T h e middle fraction was refluxed over barium oxide a t 3 mm. for 1 hr. lated from the Debye-Huckel theory. Both of these and then distilled. T h e middle fraction was collected, b.p. 5 i o (3 quantities depend upon the value of the limiting equivamm.). Methanol (MeOH) (Matheson Coleman and Bell Divilent conductance, .lo,but since neither are different sion, The Matheson Co., reagent grade) was refluxed over barium from unity by more than 3Tc in any of these systems oxide for 2 hr. and distilled in a Vigreux column. The middle cut the results are not particularly sensitive to the value was taken, b.p. 64.8'. Acetonitrile ( M e C N ) (Matheson Coleman and Bell Division, T h e Matheson Co., spectral grade) was of .io,ised in calculating them. A plot of I/(-iS) u s . distilled through a 10 cm. column packed with glass helices, b.p. ACSy,2 should yield a straight line. Above 2 X 10V4 81". M , curvature in the plots of eq. 1 were observed in the Picric acid ( H P i ) , recrystallized from ethanol, was supplied by pure solvents, presumably due to triple-ion formation. R . E. Stamm. Tri-n-butylammonium picrate ( T B A H P i j was prepared by the method of Witschonke and K r a ~ s . T~h e salt The plots were linear below this concentration. In was recrystallized three times from 95Yc ethanol, m.p. 106.T" ODCB the intercept was about 0.5, corresponding to (lit.5 m.p. 106.5"). If the salt is subjected to heating under to -io of 2, and in E C the intercept was 0.05, correspondvacuum its melting point decreases, so freshly recrystallized ing to a -ioof 20. These values were not used in calsamples were dried under vacuum a t room temperature just prior culations of either the S function or the activity cot o use. Tri-n-butylammonium bromide ( T B A H B r ) was prepared from TBA and 487, hydrobromic acid in diethyl ether. I t efficient. The reason forRthe low values of A. will be was recrystallized seven times from ether, m.p. 74-75' (lit.6m . p . discussed later, but more realistic values were esti7.5'). Tri-n-butylammonium iodide ( T B A H I ) was prepared mated by assuming the ratio of ,iofor the salt to that from 4iTo hydriodic acid and the amine in 957; ethanol. Reof tetra-n-butylamnionium picrate ( B u s Pi) in either crystallization from ethyl acetate ten times gave a salt, m.p. 101.0-101.5" (lit.' m . p . 101.2-101.7"). Pyridinium picrate solvent was equal to the same ratio in nitrobenzene, (PyHPi)-pyridine solutions were prepared by the addition of where more unequivocal values have been obtained. picric acid t o a n excess of pyridine in ODCB. This method was The following values of -iowere used in these estitnanecessary due to the slow rate of solution of the salt in this soltions: Bu4NPi in PhNO2l627.9, in EC" 5 i . 4 , and in vent. ODCBg 36.8; TBAHPi in PhN025 28.9. The values Solutions.-All salts were weighed in a nitrogen-filled drybox. The solutions were made up by weight and the molarities calso estimated appear in Table I1 in column 3 in culated using the density of t h e solvent. Solutions for conductparentheses. From the slopes of the plots of ey. 1, ance measurements were made by adding a concentrated stock values of the product Ao'K may be obtained where K solution from a weight buret directly t o the conductance cell is the apparent ion-pair dissociation constant for the containing a weighed portion of solvent. salt. In the pure solvents, the quantity R o , the ion Measurements.-All conductance measurements were made a t 25.00' in a n oil-filled thermostat. The temperature was conpair dissociation constant for the salt in the pure solstant within f0.005'. T h e two conductance cells used were of vent, was obtained by dividing the product Ao2& by the Kraus type L ~ " ? bright platinum electrodes. T h e cell conthe estimated value of .io2. The value of Ao2Koof stants, 0.01800 and 0.03063 cm. -', were determined by compari7.18 X lod5 in E C compares favorably with the value son with a similar cell with platinized electrodes and a cell conc stant of 0.6297 c m . ? a s determined using a 0.01 Demal potas1.36 X 1 0 F obtained by Mead, Fuoss, and KrausIR sium chloride solution.* using somewhat higher salt concentrations than those T h e conductance bridge used has been described p r e v i o ~ s l y . ~ employed in this work. When addends were present, Densities were determined using a Lipkin pycnometer, calibrated the ratio, R,of the slope in pure solvent, l/(AOzK~), with water and mercury. to t h a t in a given addend solution, 1/(Ao2K),was calPhysical constants .-These are given for the three solvents in the order ODCB, E C , and CB: viscosity (cp.), 1.272,'" 0.785," culated. R is plotted us. addend concentration in O.i52l2; density (g./ml.), 1.3007, 1 .2455,11 1.1011; dielectric Fig. 1 and Fig. 2. The Shedlovsky plots were linear constant, 10.06,l3 10.23," 5.621.14 ~-
J . A m . C h e m . S O ~ 67, . , 79 (1945) (5) C . K . Witschonke a n d C. A . K r a u s , ibid., 69, 2472 (1947). (6) J. A. Geddes a n d C . A . K r a u s , T r a n s . F a r a d a y S o c . , 82, 585 (1936). (7) H.S. Young a n d C . A. K r a u s , J . A m . C h e m . S o c . , 7 3 , 4733 (1951). (8) G . Jones and B. C . Bradshaw, i b i d . , 66, 1780 (1933). (9) H.I,, Curry a n d W . R . Gilkerson, chid., 79, 4021 (1957). ( I O ) F . Accascina, E. L . Swarts, P . L. Mercier, a n d C . A . K r a u s , Proc. Y a t L . A c a d . Sct. G. S., 39, 917 (1953). (11) D . I-. Fowler a n d C . A . K r a u s , J . A m . C h e m . Soc., 62,2237 (1940). ( 4 ) D. G 1,eisand B. C. C u r r a n ,
below 2 X 1W4.1/1 in the pure solvents and were linear
(12) K . I.. M c I n t o s h , D. J. AMead,and K . M . Fuoss, i b i d . , 62,BO6 (1940). (13) P.H.Flaherty a n d K . H. S t e r n , i b i d . , 80, 1034 (1958). (14) A. M a r y o t t , U. S.National Bureau of S t a n d a r d s , Circular No. 514. (15) T.Shedlovsky, J . F r a n k l i n I n s l . , '226, 739 (1938). (16) E. G. Taylor and C . A , K r a u s , J . A m . Chein. Soc.. 69, 1731 (1947). (17) D.J , M e a d , R . M. Fuoss, and C . A . Kraus, T r a n s . F a v a i i a y Soc., 3'2, 594 (1936) (18) D . J. M e a d , R . M.Fuoss, and C. A . K r a u s , J . A m . C h e m . Soc., 6 1 , 3257 (1939).
AMINE-AMMONIUM SALTINTERACTIONS
Nov. 20, 1964
4785
40
30 R.
1000
R. 20
R.
500
10
0
0
2
4
6
0
Fig. 1.--R os. addend concentration; 0, T B A H P i with TBA in O D C B ; 6,TBAHPi with MeCN in O D C B ; 0 , TBAHPi with 4 - C S P y in O D C B ; Q, TBAHPi with MeOH in E C ; 0, TBAHBr with P y in ODCB ( t h e abscissa in this case is 1000[addend]).
a t higher concentrations in the presence of addends. I n ODCB-4-MePy solvent mixtures, the Shedlovsky plots were quite linear but gave a small negative intercept (about 0.5 reciprocal A-units). This is attributed to the presence of a small amount of impurity in the base which could not be removed even by using freshly distilled base. It was assumed t h a t the slope obtained from the four points corresponding to the highest salt concentrations in each case was more nearly representative of the correct value. Included in Fig. 1 are results calculated from data obtained in the titration of a 1.804 X M TBAHPi solution of ODCB with MeCN, Table I. A t this concentration, the salt is dissociated to such a small extent t h a t eq. 1 reduces to A2Cykz Ao2K. The ratios R in this case were obtained by dividing A2Cyk2 a t a particular acetonitrile concentration by t h a t value in the absence of acetonitrile. T h e slopes of the straight lines in Fig. 1 and 2 drawn through the points are recorded in Table I1 as K L . MeOH was the only addend whose effect on the conductivity of TBAHPi solutions in E C could be determined. T h e nitrogen bases apparently reacted with the solvent as noted in the introduction. As can be seen in Fig. 2, there is considerable departure of this system from a straight line in an R vs. [MeOH] plot. T h e value of K L recorded in Table I1 was obtained from a straight line drawn approximately half-way between the two points. TBAHPi in CB.-Triple-ion formation was evident at the concentrations of the salt employed. Values of Ao2Kin this solvent were obtained by the procedure of Fuoss and Kraus,lg taking into account the increased conductivity of the solutions due to triple ions. Here the conductance equation is
ACo.5g(C)=
f (K0.5A3/k)(l- A/&)C
1
3
5
lOO[Addend]
100[Addend].
(2)
Fig. 2.--R v s . addend concentration; left-hand ordinate: 0 , TBAHPi with 4-MePy in O D C B ; 6 ,TBAHPi with 4-MePy in C B ; 0, T B A H P i with P y in O D C B ; 3,TBAHPi with P y in C B ; right-hand ordinate: 0,P y H P i with Py in ODCB.
where g(C) is a known1gfunction of Ao, taking into account ion atmosphere effects; g ( C ) is very close to unity for these solutions. The constant k is the tripleion dissociation constant and is assumed to be the same for both possible ion triples. In practice this is .justified by the linearity of a plot of 1iCO.j vs. C (eq. 2 ) . The value of A0 was estimated as described in the case of ODCB and E C as solvents, utilizing in addition .lo (Bu4NPi in CB) = 36.8.20 K Oin Table I1 was thus calculated from the intercept of a plot of eq. 2 in the pure solvent. Values of the ratio R were obtained from the intercepts in the presence of added pyridine and 4-methylpyridine and are plotted v s . addend concentration in Fig. 2 . The slopes of these latter plots appear in Table I1 as K L . TBAHBr and TBAH1.-These salts were studied in ODCB in an effort to test the effect of changing the anion in the salt on the interaction with added bases. Conductance data for both salts were sensitive to the solvent conductivity, the iodide more so than the bromide. Data for both in E C are reported here, b u t only the bromide data are given for the solvent ODCB. These salts showed evidence of triple-ion formation in the to l o F 3 A4 concentration range, b u t plots according t o eq. 2 showed decreasing slope with increasing salt concentration. A rough plot of A2C vs. C resulted in a fairly straight line. This suggested to us that perhaps here was another example of unsymmetrical triple-ion formation similar to the ( F H F ) - ion. We suppose the one formed here is the cation type, (TBAH)ZBr+. The conductance function in such a case has been derived by WoosterZ1 (eq. 3) (Ay*/m)'C/(l - A/Ao) = Ao2K[1f hC(1 - A/Ao)/k+] (8) (19) R. M. Fuoss and C. A. Kraus, J . A m . Chrm. Sor., 6 6 , 2387 (1933). (20) J. B. Ezell a n d W , R . Gilkerson, J . Phys. Chem., 68, 1581 (1964). (21) C. B. Wooster, J . A m . Chem. Soc., 69, 377 (1937).
EARLK. RALPH,111, A N D W. R. GILKERSON
4786
Vol. 86
TABLE I EFFECTOF ADDENDS ON EQUIVALENT COXDUCTANCES OF TRISUBSTITUTED AMMONIUM SALTS AT 25' 104c
0,2084 0.4457 0,7636 1.119 1.491 1.870 2,343 2.790 3.266
1OA
+ 0.00890 M Py 1.347 3.026 4 995 7.140 9,400
+
0.02902 0 . 7553 1 ti91 3.331 5.381 7.755
104c
TBAHPi in ODCB 0 2666 1.369 0 5407 0.9473 0,9493 0.7303 1.344 0.6083 1.753 0.5307 2.166 0.4757 2.559 0.4285 2.974 0,3945 3.397 0.3669 2.105 1.418 1.114 0.9385 0.8240
M Py 4.929 3.282 2.343 1.881 1.550
+
+
0.01951 0.4878 0.8773 1.812 2 619 3.956
+
0.03559 1,220 2.311 5.012 7.529 9.821
+ 0.00654
1OA
1.217 0.8643 0.6572 0.5571 0.4904 0 4438 0 4100 0,3824 0.3600
M Py 5.162 3.673 2.559 2,132 1.741 M Py 4,223 3.064 2.088 1 712 1.503 4-MePy 6 449 4,249 3.308 2.683 2.281 2,047 1 ,878 1,741 1.620
0.00374 M 4-MePy 0,2444 5.520 0.4726 3.857 0.8754 2.776 1.211 2 341 1.616 2.015 2.011 1.801 2.413 1.639 2,816 1.515 3,258 1.407
0,2698 0.5742 0.9154 1.359 1.852 2.282 2,699 3.132 3.612
+
+ 0.00751 M 4-CXPy
0.01219 M 4-MePy 0.2368 9.810 0.5073 6.404 0.7947 5.002 2.041 3.014 2.490 2.715 2.822 2.544 3,347 2.330
+ 0.01113 M 4-CNPy 0.2270 0,5007 0.8123 1 194 1 637 2.143 2 596 0.01787 iM 0 3315
+
0,5823 0.9827 1 302 1.779 2 264 2.724 3 206
1.611 1.101 0.8720 0.7255 0.6254 0.5510 0.5027 4-CXPy 1.375 1.044
0.8317 0.7058 0.6068 0.5409 0 4951 0 4584 0.01145 M TBA 1,147 1.493 1 026 2 369 4.022 0.7814 5.614 0.6619 7.417 0.5771
+
0.3105 0.6351 1.139 1.575 2 121 2,619 3.090 3,585 4.032
1.305 0.9251 0.6983 0.5978 0.5193 0.4703 0.4354 0.4066 0.3857
+ 0.02465 M 4-CNPy 0 3060 0 6695 1 010 1 401 1 885 2 443 2 952 3 545 4 064
+
1 525 1 043 0 8550 0 7310 0 6341 0 5607 0 5135 0 4709 0 4422
0 01980 M TBA 0 8151 2 089 1 700 1 446 1 172 2 596 3 513 1 009 0 8998 4 423 0 8104 5 488
104c
0 1 2 3 4
8148 651 632 545 711
104c
A
+ 0 03218 M TBA 2 1 1 1 0
10'C
329 643 307 129 9825
1OSA
C (Salt) = 1.804 C, MeCN 0 0000 0 00259 0 00586 0 01063 0 01619 0 02096 0 02666
A
M
X
0 0 0 0 1 1 1
10'C
5452 6746 8109 9806 151 285 433 10314
TBAHPi in CB 1 2 3 4 6
065 397 595 722 099 7 205
1 0 0 0 0 0
104 8982 8428 8246 8096 8149
+ 0 00894 M 4-MePy 0 1 1 2 3 4 4 5 6
6203 302 836 502 297 094 924 474 198
0 1 2 3 4 5 5
0 01872 6800 479 270 250 087 084 960
+
5 610 3 3 2 2 2 2 2 2
958 406 999 704 513 370 303 228
8.331 1.449 2.773 4.501 6.406
+ 0 01497 M 4-MePy
104c
7 935
0 0 1 2 3 4 5 5 6
4485 9537 772 643 565 280 011 819 598
0 1 2 2 3 4 5
f 0 03781M Py 6745 6 874 303 5 159 162 4 139 911 3 679 701 3 372 3 166 469 144 3 019
M Py 4 701 3 324 2 784 2 443 2 266 2 128 2 045
0.8245 0 9944 0.8644 0.8247 0 8195
5 4 3 3 2 2 2 2
483 130 497 114 935 795 682 595
104c
A
A
TBAHPi in E C
0 1628 0 2971 0 6405
2 091 1 559 1 073
0 9740 1 742 2 508
0 8749 0 6588 0 5525
+ 0.02743 M MeOH
+ 0 06909 M MeOH
0 2218 0 5192 0 7836
0 0 0 1 1 1
1 083 1 554 2 289
2 131 1 586 1 299 1 112 0 9319 0 7725
l04C
lO2A
2331 5430 8391 197 633 947
104c
3 2 1 1 1 1
482 326 889 591 357 260
102A
TBAHBr in ODCB 0 6222 1 630 3 408 5 341 8 097 0 4338 0 6417 1 097 2 275 3 791
3 3 3 3 3 3 3 3 3 3
587 378 377 401 487 688 422 221 434 471
+
0 0 1 2 3 4 5 6 7
0 00167 3528 8542 752 898 780 765 643 660 399
M Py 3 560 3 208 3 084 3 048 3 016 3 017 3 046 3 035 3 042
AMINE-AMMONIUM SALTINTERACTIONS
Nov. 20, 1964
478 7
TABLE I (:Continued) 10'C
10'C
102A
+ 0.00268 M Py
0.5603 1.343 2,247 3.105 4.072 4.857 5.746 6.449 7.263
+ 0.00375 M P y
3.377 3.045 2.971 2,943 2.928 2.922 2.921 2.920 2.923
+
0.00411 0.5507 1,259 2.172 3.130 4.119 5.260 6.154 7.000 10'C
105c
1OzA
0.6271 1.022 1.640 2.374 3.127 3,959 4.821 5.775 6.579
3.526 3.279 3.130 3.058 3.022 3.003 2.993 2.988 2.985
+
M Py
0.006505 0,9003 1.798 2,730 3.785 4.656 5.683 6.604 7.380
M Py 3.726 3,260 3.105 3 049 3.020 3.006 2.999 2.999
A
0.3104 0.2354 0.2008 0.1904 0.1824
5.658 6.874 8.000 9.275
0.1765 0.1735 0,1715 0,1701
TABLE I1 DERIVED CONSTANTS AT 25" Solvent
ODCB
EC CB
AP
2 (38 1)
20 (59 4) (43 0)
ODCB EC
(45 2) (70 5)
EC
(69 5)
ODCB
(46 8)
10'QKo
Addend
TBAHPi 2 86 TBA PY 4-MePy 4-CNPy MeCN 203 MeOH 0 0048 Py 4-MePy TBAHBr 0 022 Py 8 07 TBAHI 157 PyHPi 0 35* Py
KL
560 f 40 1370 f 70 3240 i 210 40 i 7 212 f 6 41 820 f 50 1830 f 80 1450 f 90
19,000
z!z
500
Value preceding ( ) obtained by extrapolation in pure solvent As determined in Value in ( ) estimated as described in text the presence of 0.01313 M HPi a
+
0.1731 0 3324 0,5939 0.7683 0.8392 1.154 1.499 1.675 3.307 4.961 6.197
6.355 4.654 3.512 3.073 2.971 2.546 2.249 2.109 1.523 1.259 1.137
104c
1OA
+
where h = ~ X O / A O- 1 (1 - X O / A O ) ~ 4/ ( ~k + / C ) . The function m Z 1corrects A for departures due to ionatmosphere effects, Xo is the limiting equivalent conductance of the salt corresponding to the lone triple ion and counter-ion, and k+ is the triple-ion dissociation constant. Equation 3 was fitted to the conductance data assuming Xo/Ao = 0.667 The Wooster plot, eq. 3, for the bromide in ODCB was quite steep, and gave an intercept so close to zero as to be of doubtful value in obtaining Ao2Ko. The intercepts were first
0,1477 0 2237 0 3032 0.3897 0.8447 1.366 2.522 3.644
104c
PyHPi in ODCB
0.03608 M P y 1,175 6.874 2.978 4.332 4.346 3.580 6.305 2.944 7,862 2.620 9.013 2,440 0.02325 M Py 0.8339 6.444 1 638 4 623 2 656 3 546 4 060 2 919 5 791 2 418
+
TBAHBr in E C 0.5277 1.222 2.359 3.168 4.269
A
TBAHI in E C
3.652 3.311 3,201 3.145 3.122 3.109 3.106 3.104
:04c
A
10'C
A
0.7647 0.6430 0,5740 0.5265 0.4205 0.3813 0.3526 0.3453
10a
+
0.04038 M Py 1.673 6.127 3,816 4.080 6.057 3.229 8.396 2.730
+ 0.01313 M HPi 12,43 31.32
0.08496 0.06539
+
0 1 2 4 5 6
0.02418 M Py 8451 6 527 725 4 591 715 3 657 042 2 978 213 2 656 677 2 282
obtained in the presence of added pyridine, plotted v s . pyridine concentration and extrapolated to zero pyridine to give the value of Ao2Ko. The values of A0 in nitrobenzene used to estimate 'ioin the solvents ODCB and E C were calculated from A. for the picrate in PhNOz and the bromide,22 iodide,z3 and picratez3 ion conductances in P h N 0 2 . The values of K O calculated from the intercepts in the two solvents and the estimated value of hoare given in Table 11. The ratios, R,for TBAHBr in ODCB in the presence of added pyridine are plotted vs. pyridine concentration in Fig. 1. In EC, k+ was found to be 4.0 X for the iodide and 7.5 X for the bromide. In ODCB, for the bromide. k+ was found to be 1.7 X PyHPi in 0DCB.-The conductance data in the presence of excess HPi was treated by the method of Fuoss and Kraus,l9 using eq. 2, to yield the intercept, Ao2Ko. The estimated value of A. appearing in Table I1 was calculated using the value 34.7 as determined in PhN0z.j KOwas calculated using this estimate of A. and appears in Table 11. In the presence of excess pyridine, triple-ion formation was no longer apparent. The values of AozK,from which the ratios, R,were calculated, were obtained from the slopes of Shedlovsky plots (eq. 1). R is plotted vs, pyridine concentration in Fig. 2. The conductances as a function of added TBA are shown in Fig. 3 for PyHPi in ODCB-Py solution and for 4-methylpyridinium picrate in ODCB-4-methylpyridine solution. J26 s o h The addition of 0.023 M Py to a 1.5 X ( 2 2 ) H. Sadek and R . M. Fuoss, J . A m . Chem. Soc., 81, 4507 (19.59) (23) E. Hirsch a n d R. M . Fuoss, ibid., 82, 1018 (1960).
EARLK . RALPH,111, AND W. R. GILKERSON
4788
Vol. 86
the slope is unaffected by acid-base dissociation. Kraus and U'itschonke, observing similar effects in PhNOz as solvent, attributed their results to the interaction of the added amine with the acidic hydrogen atom on the cation, resulting in increased ion-pair dissociation. We believe their interpretation to be correct. No such effects were obtained upon addition of either Py or MeCN to solutions of Bu4NPi in ODCB. Such interaction as occurs with TBAHPi can be viewed as proceeding according to the following equilibria, in addition to eq. 4
2.5
1OA.
2.0
E A H f + X - K O= [AH*] [X-]y,2/[AHX] (6) 4HX + B AHBX KS = [ A H B X ] / [ A H X ] [ B ] (7) AHBX AHB- + X- K B [AHB+][X-]y+2/[AHBX] AHX
1.5
=
I
(8)
-42
0
1.n
0.5
[Added T B A J / [Salt 1.
Fig. 3.-Open circles, left-hand ordinate, A for 5.652 X 10-4 A4 PyHPi in 0.02326 M Py in ODCB rs. mole ratio of added TBA to salt. Upper curve, closed circles, right-hand ordinate, a similar plot of A for 3.715 X M 4-MePyHPi in 0.01658 M 4-MePy in ODCB 21s. mole ratio of added TBA t o salt.
tion of Bu4NPi in ODCB increased the conductance by only 2%. The addition of 0.033 :If acetonitrile to a 1.87 X M solution of the same salt in the same solvent gave a 6% increase in the conductivity. Discussion T h e values of A 0 appearing in Table I1 for TBAHPi in ODCB and EC, obtained by extrapolation, are seen to be quite low compared with what are believed t o be the more nearly correct values, those in parentheses. Elliott and FuossZ4 and Kraus and cow o r k e r ~ ~have , ' ~ shown this effect to be due to acidbase dissociation in the case of amine salts (eq. 4) AHX
X
+ HX
Kd
=
[A] [ H X ] / [ A H X ]
(4)
where A represents the amine, H X the acid, and A H X the ion pair. Activity coefficients of all uncharged species are assumed to be unity. According to Elliott and F u o s ~ the , ~ ~conductance equation (eq. 1) is replaced by l/(AS)
=
(1
+ Kdo.'/KoO'')//ho+ -iCSy*'/KoAo'
where for generality B may not be the same as the amine A . Kolthoff and co-workersZ6have interpreted conductometric titration curves in acetonitrile in terms of similar equilibria involving complex formation between anions and suitable weak acids. Complex formation between base and protonated base has been postulatedz7to explain the behavior of the potentials of cells containing amines and their corresponding ammonium salts in acetonitrile. I n eq. 6, 7, and 8, AHBX represents any complex between base and the ion pair, ,4HX. AHB+ represents a complex between cation .4H+ and the base B. In the following derivation of the conductance equation, acid-base dissociation will be assumed to be repressed, as i t would be for B the same as A. I n the presence of a large (compared t o AHBX or AHB f) and constant concentration of B, the equivalent conductance, corrected for ion-atmosphere effects, of a salt solution of-concentration C is given by AS = C Y O ~ ~ O CYBAB, where a0 = [ilH+]/C, CYB = [AHB+]/C, and AB is the limiting equivalent conductance of the salt AHBX. Utilizing the fact t h a t [AHX] Y [AHBX] = C(1 - CYB)and the relation K2,we find that
+
[AHX]
=
C(l -
CYO
-
+ Kz[B])
CYB)/(~
Using the relations KO and K3, as well as Kz, we find CYO
=
iiS/[Ao(l
+ ABKL[BI/Ao)]
w h e r e K ~= K2K3/Ko. Also
(5)
It is seen from eq. 5 t h a t the effect of acid-base dissociation is to change the intercept of a Shedlovsky plot, leaving the slope unchanged (as long as the ion concentration is low enough t h a t the Shedlovsky function and activity coefficients are not too different from unity). Accordingly, we calculate that TBAHPi in ODCB has a Kd of 9 X lo-*, while in EC i t is 7 X 10-6. Pearson and Vogelsong,?j using a spectrophotometric method, found Kd to be 3.41 X 10-5 in E C at 25' for the salt of the same amine with the weaker acid, 2,4-dinitrophenol. The increased conductance of solutions of TBAHPi in ODCB in the presence of added TBA is due to some factor other than repression of acid-base dissociation. The ratio R appearing in Fig. 1 and 2 in each case is calculated from the slope of a plot of the modified Shedlovsky equation, eq. 5 . As pointed out above, (24) M . A . Elliott a n d R . M . Fuoss, J . A m . Chem. Soc., 6 1 , 294 (1939). (25) R. G. Pearson a n d D. C. Vogelsong, ibid., 80, 1038 (1958).
CY0
f
CYB =
[x-]/c=
hS(1 f KL[BI)/ [&(I -k ABKLP]/Ad
+
1
Inserting these relations into KO = CCY~(CYO CXB)Y*'/ [AHX],and rearranging, we obtain 1/(AS) = (1
T h e ratio 9, by
I?
+
+
K L [ B ] ) / & ( ~ ABKL[BI/&) f hSCy*'(1 Kz[Bl)(1 KL[BI)/ [KoAo2(1f ABKL[B]/&)'] (9)
+
+
R can be seen to be given, according to eq.
= (1
ABKL[B]/Ao)'/[(~ f Kz[BI)(1
+ K L[BI)I
(10)
The intercepts of the Shedlovsky plots cannot be determined with sufficient precision (they are too close (26) F o r leading references, see I. M . Kolthoff and M. K . Chantooni, ibid., 8 6 , 2195 (1963). (27) J . F. Coetzee, G. R. P a d m a n a b h a n , a n d G. P. C u n n i n g h a m , T d ~ ~ l a , 11, 9 3 (1964).
AMINE-AMMONIUM SALTINTERACTIONS
Nov. 20, 1964
to zero) in the solutions we are dealing with to allow evaluation of the quantity A B / A o . We shall assume the latter ratio to be unity here. This is reasonable considering that the value of the ratio Ao(Bu4NPi)/ &(TBAHPi) is 0.965 in P h N 0 2 . 5 Then eq. 10 reduces to
R e (1
+ K L I B I ) / ( ~+ Kz[BI)
(11)
The curve in Fig. 1 for TBAHPi in ODCB-TBA mixtures was calculated using KL = 560 (Table 11) and KS = 25. Due to the approximation that h B / & = 1.00, we believe these to be the minimum values for KI, and K 2 . Curvature in R vs. [ B ] plots was not observed for any of the other salt-addend systems studied here. It is noteworthy t h a t TBA was the strongest base used in the present work. Similar curvature can be found in the work of Kraus and Witschonke in the case of the relatively strong bases E t a N and piperidine added to trisubstituted hydroxyammonium salts in P h N 0 2 . j The (TBA)2HPi complex can be represented by structure 1 wheresthe plane of the benzene ring in
4-
O2N
0- N 02
+ Bu, N/ H.*. Bu/ : ', Bu Bu
v
