1852
Vol. 68
K. H. STOKES AND BARBARA J. LEVIEN
chloro-4-quinoline aldehyde was heated a t 90-95" for two hours. On cooling the aldimine crystallized. After two recrystallizations from benzene it weighed 2.9 g., and melted a t 125-125.5'. Anal. Calcd. for C12HllN20C1: C, 61.4; H , 1.73; S, 11.94. Found: C,61.36; H,4.96; N, 12.0. A solution of 2.66 g. of the aldimine in 100 ml. of absolute alcohol containing 0.05 g. of pre-reduced platinum oxide was shaken with hydrogen a t 40 lb./sq. in. until absorption ceased (forty minutes). The product crystallized on evaporation of part of the solvent, and melted at. 147-150" (dec.). ilnal. Ca1cl:t. for C 1 2 1 ~ l s ~ 2 0 C C,I W.9; : r5,'5.54; CI, 15.0. Found: C, 60.7; H, 5.23; C1, 15.2. 6-Chloro-a-( 4-diethylamino-l-methylbutylamino)-lepidine.-A misture of 4.1 g. of pure Koval diamine and 4.8 g. of 6-chloro-4-q,uinoline aldehyde was warmed a t 75-80' for forty minutes. It was then dissolved in 75 ml. of absolute ethanol and shaken with hydrogen in the presence of 0.1 g. ofjpre-reduced platinum oxide. Absorption was complete in ninety minutes. The alcohol was removed, the residue taken up in dilute hydrochloric acid, and the @Hadjusted t o 6, t o precipitate any unreacted aldehyde or its reduction product. The filtrate was made strongly basic, and the amine taken up in ether. Distillation of the extract gave 5.6 g. (67%) of a light yellow oil, b. p. 171-174'' (0.04 mm.), n% 1.5587-1.5593. The oil was dissolved in n-propanol and two equivalents of propanolic hydrogen chloride added. p n cooling the dihydrochlorideseparated, m. p. 149-151.5 . Anal. Calccl. for Cl9H30N3C13.H20: C, 53.7; H, 7.59; N, 9.89; C1,25.0. Found: (2,533; H, 7.6; N, 10.3; C1, 24.6. 7-Chloro-a-( p-hydroxyethy1amino)-lepidine, SN-8646.12 -A mixture of 5.54 g. of 7-chloro-4-quinoline aldehyde and 1.8 g. of ethanolamine was heated a t 80-90" for forty minutes, and then hydrogenated in 100 ml. of absolute ethanol in the presence of 0.1 g. of pre-reduced platinum oxide. The solvent was evaporated and the residue washed with a little cold acetone to remove adhering oil. The white crystalline product melted a t 123-125" and weighed 3.7 g. XI^^), I t was soluble in alcohol and hot dioxane, insoluble in acetone. (13) T h e numhers are those as.iigned hy the Survey of Antimalarial D r u p to identify the drugs in their records. T h e anti. malarial properties o f these substances will be tabulated io a forthc.,rnin:: moniicraith.
[CONTRIBUTION FROM
THE
Anal. Calcd. for ClzH18NtOCI: C, 60.9; H, 5.54; C1, 15.0. Found: C, 61.2; H, 5.4; C1, 14.9. The hydrochloride was prepared by evaporating a solution of the base in dilute hydrochloric acid. The hygroscopic hydrochloride melted a t 110 '. 7-Chloro-a-( 4-diethylamino-l-methylbutylamino)-lepidine, SN-11,197.-This was prepared as described for the 6-chloro isomer. The product was a light yellow oil, b. p. 179-184" (0.07 mm.) and was obtained in 67% yield. The dihydrochloride, prepared in propanol, melted a t 106-108". Anel. Calcd. for C ~ ~ H ~ ~ N ~ C IC,Y51.5; ~ H ~H, O 7.74; : N, 9.5; C1,24.0. Found: C,51.66; H, 7.89: N,9.1: C1, 24.3.
6-Ethylaminohexylamine.-6-Bromohexylphthalimide (67 g.) was added to a solution of 45 g. of ethylamine in 100 nil. of dry benzene, a t loo, and the reaction mixture was allowed to stand with occasional cooling, for twentyfour hours. The benzene solution was extracted with dilute hydrochloric acid, and the extract made strongly basic. The dark oil was hydrolyzed with hydrazine hydrate and hydrochloric acid.14 The amine, obtained in 10-15 g. yield, had b. p. 82-84' (5 mm.), T Z ~ O D 1.4512, dZ0a0.8431, h!fRDobs. 46.0, MRDcalcd. 46.2. Anal. Calcd. for CsHzoNz: N, 19.42. Found; N, 19.61. The diamine dihydrochloride melted a t 200-203 '. 7-Chlor0-~-( 6-ethylaminohexylamino)-lepidine, SN13,l54.-This was prepared from 7.2 g. of B-ethylaminohexylamine and 9.5 g. of 7-chloro-4-quinoline aldehyde. The product, a light yellow oil, was obtained in 7 g. (44%) yield; it had b. p. 190-195" (0.2 mm.), T Z ~ O D 1.5563. The dihydrochloride, prepared in propanol, melted a t 195-200'. Anal. Calcd. for C1SH28NzC13: C, 55.0; H, 7.16; C1, 27.1. Found: C, 54.7; H, 7.43; C1, 27.3.
Summary 1. Several lepidylamines derived from 6chlorolepidine and 7-chlorolepidine have been prepared. 2 . 7-Chlorolepidine, 6-chloro- and 7-chloro-4quinoline aldehyde and the corresponding carbinols have been synthesized. (11) Ilan5ke J Chem. SOL.,2348 (1926).
NOTRE DAME,IXDI.4NA
RECEIVED APRIL 5, 1946
CHEMISTRY DEPARTMENT, AVCKLAND UNIVERSITY COLLEGE]
Transference Numbers and Activity Coefficients in Zinc Iodide Solutions at 25' BY R. H . STOKES' AND BARBARA J. LEVIEN A comparison of the activity coefficients of the halides, perchlorates and nitrates of zinc with those of magnesium has shown2 that the halides of zinc exhibit increasing abnormality in passing from the iodide through the bromide to the chloride. A parallel behavior is found in the transference numbers, for while that of zinc perchlorate is normal up to the highest concentration measured (4 M ) , the cationic transference numbers of the chloride and bromide fall rapidly with increasing concentration, reaching negative
values above 2 11.1 in zinc chloride solutions3 and above 2.7 M in zinc bromide solutions.* The only work on concentrated solutions of the iodide appears to be that of H i t t ~ r fwho , ~ made only three measurements. The present paper reports activity coefficients and transference numbers of zinc iodide from 0.05 to 10 M , derived from isopiestic vapour pressure data and e. m. f. measurements on cells with transference, a t 2 5 O . Experimental.-The technique of isopiestic measurements on zinc iodide solutions has been
(1) Present address: Chemistry Department, University of Western Australia, Nedlands, Western Australia. (3) R. H. Stokes a n d B. J. Levien, TIIIS J O U R N A L , 68, 333 (1946).
( 3 ) A. C. Harris a n d H. N. Parton, Trans. Furaduy SOL.,36, 1139 (1940). (4) H. N . P a r t o n a n d J . W . Mitchell, ibid., 86, 758 (1939). (5) W. Hittorf. Poag. A n u . , 106, 513 (1859).
Sept., 1946
ACTIVITYCOEFFICIENTS IN ZINCIODIDE
1853
TABLEI11 Et = E. M. F. OF THE CELL:Zn-Hg ZnI
I 1
Zn12 Zn-Hg (2)PHASE-/0.3-& nl (2-PHASE). E = e. m . f. of the corresponding cell without liquid junction (calculated from the activity coefficients) n, = transference number of the zinc ion (all a t 25") m
0.05 .1
.3
.6 1.0 1.3 1.6 1.9 2.5 3.0 4.0 5.0 5.75 8.0 10.0
w n I Z
lylxaci
mmt
mNnci
mzni:
mNsci
1.167 1.350 3.113
2.208 2.548 4.960
1.253 2.234
2.369 3.847
1.335 2.554
2.519 4.263
mmZ
~ H ~ E O mznIz ~
2.204 2.880 3.814 4.183 4.595 4.829 5.593 5.714 6.392 6.498 7.963 8.186 9.994 10.54ij
~ H Z E O ~W~?BO~
2.737 3.331 3.825 4.190 4.950 5.156 5.848 5.966 6.542 6.652 8.454 8.761 10.960 '11.638
mznIz
3.682 4.079 4.033 4.354 5.192 5.364 5.956 6.070 6.848 6.965 9.464 9.934 11.892 12.638
E t , mv.
E , mv.
-41.81 -26.89 (0.00) 21.39 40.90 51.53 59.63 66.19 76.90 84.73 100.43 116.81 130.07 174.73 214.49
-65.11 -41.18 (0.00) 31.87 59.83 74.53 85.47 93.86 106.35 115.09 130.90 145.64 156.13 188.63 215.19
n+
6, mv.
0.382 ,363 ,332 ,317 .291 ,269 .234 .192 .115 .056 --.050 -.190 -.300 --.444 --.550
-0.06 .03
. 00 .08
-
-
.Ol .05 .02 .05 .09 .02 .03 .ll
. 00 . 00 .01
In the case of zinc iodide i t would be possible to measure directly the e. m. f.'s of the cell without transference, as was done by Bates using the zinc amalgam and silversilver iodide electrodes. Such e. m. f.'s would, however, beAsubject to large errors arising from the solubility of silver iodide in the electrolyte; Bates8 noted that this solubility was appreciable even in the region of 0.5 M . Hence we have preferred to calculate the e. m. f.'s from the activity coefficients of Table 11, using the relation E = 0.08871 log (mzyz/n~iyi)
The relation between Et and E was too complicated to fit to an equation, so the anion transference number, n- = dEt/dE, was obtained by interpolating Et to round values of E a t 10-mv. intervals, and then differentiating by the method The differential coefficients were of R ~ t l e d g e . ~ then plotted against E and read off a t the values of E corresponding to the round molalities in Table 111. These final values were checked by graphical integration of the relation E
- Et
=
iy
n+ d E
where n+, the transference number of the zinc ion, is equal to (1 - dEt/dE). The column titled 6 in Table I11 gives the difference between the observed values of Et and those found by the graphical integration, and shows the check to be quite satisfactory. Discussion.--The traiisfermce numbers of the zinc ion are plotted against 6in Fig. 1. The value of 0.410 for zero concentration was calculated from the accepted limiting mobilities, and is seen to be quite consistent with the results in Table 111. The gradual fall up to about 1 M is similar to that found with zinc perchlorate, and ( 8 ) K . G . Bates, THIS J O U R N A L ,60,2083 (1038). (9) G. Rutledge, Piiys. Rev., 40, 262 (1032).
R. H. STOKES AND BARBARA J. LEVIEN
1854
I
\
I
OC
4
'
r -0.2
1 I
-0.4
-0.6
II I
I-
n
,
1
2
3
t''l,l Fig. 1 -Transference numbers in zinc iodide solutioiis: 0,present work; 0 , Egan and Partington; El, Hittorf.
Hittorf's measurements were at an unspecified temperature, probably about 15". does not suggest any significant amount of abnormality in solutions below this concentration. Above 1 M , however, the transference number turns sharply downward, becoming negative a t 3.53 M. Tn zinc chloride, negative values are reached above 2 M , and in zinc bromide above 2.7 M ; consequently we may conclude that the tendency to form complex ions is greatest in the chloride, and least in the iodide. I t is noteworthy that this order is the reverse of that found among the cadmium halidzs: i t is however confirmed by the order of the activity coefficient curves, as was clearly shown in an earlier paper.2 The points of Hittorf6 a t 4.73 M (n = -0.167) and 1.277 ;If (n = 0.273) lie very close to our curve, but hi:; dilute solution a t 0.0278 M (n = 0.325) gives :I point well below it, and scarcely compatible with the value a t zero concentration. It seems possible that thp Hittnrf method does
Vol. 66
riot work satisfactorily in dilute solutions of zinc iodide, as Egan and Partington,'* using i t in solutions below 0.2 M , obtained resdts widely different from both ours and Hittorf's. They interpreted their results as showing the formation of complex ions even in such dilute solutions, but weakened their case considerably by quoting in support of it the extremely high values of the activity coefficients of zinc iodide found by them" in an e. m. f. investigation. I n reality complex ion formation must be associated with abnormally low activity coefficients, since it causes a reduction in the total number of ions present. For example, cadmium iodide, a well established case of complex ion formation, has an activity coefficient of only 0.068 at 0.2 &I, as compared with 0.478 for calcium chloride, a typical normal salt of the same valency type. Furthermore, it has been shown6 that the work from which they derived the high activity coefficients in question is quite incompatible with the isopiestic data for zinc iodide and with the e. m. f. measurements of Bates.8 These latter are in satisfactory agreement, and confirm the conclusions of the present paper in indicating normal behaviour below 1M. We wish to thank Dr. R. A. Robinson for his interest in this work.
Summary Isopiestic vapor pressure measurements have been made to determine the activity coefficients of zinc iodide in aqueous solutions up to 12 iM a t 25'. These have been combined with e. m. f . measurements of cells with transference to give the transference numbers of the zinc ion from 0.05 up to 10 M . The transference number is normal up to about 1 M, but thereafter falls rapidly, becoming negative above 3.5 M. Thus zinc iodide forms complex anions less readily than the bromide, which in turn does so less readily than the chloride. AUCKLAND UNIVERSITY COLLEGE, NEWZEALAND RECEIVED APRIL8, 1946 (10) D. M Egan and J. R. Parttngton, J . Chem. Soc., 191 (1945) (1 1) D M Eran and J R Partinpton. ibid , 157 (1943)
