Xug. 20, 1937
TRANSFERENCE NUMBERS
[CONTRIBUTION FROM
THE
O F POTASSICM AND
DEPARTMENT OF CHEMISTRY OF
THE
CHLORIDE
IN
FOKSI.\MIDE 4"!)
USIVERSITYOF K E N T U C K Y ]
Solvents Having High Dielectric Constants. VII. Transference Numbers of Potassium and Chloride Ions in Formamide at 2S01 BY LYLER.
DAWSOX A S D CARL
BERGER
RECEIVED . ~ P R I L 4, 1957 -1pproximate transference numbers of potassium and chloride ions in formamide a t 25' have been detertnincri 1)y the Hittorf method within t h e concentration range of 0.2 t o 0.6 N . Estrapolation of a plot of t h e Longsworth function yielded a limiting cation transference number of 0.406. From t h e present results and conductance d a t a reported previously the calculated limiting transference number of t h e solvated proton in a solution of hydrogen chloride in fortnarnide was found t o be 0.37 a t 25'.
T h e electrical conductances of solutions of potassium chloride in formamide have been reported in an earlier paper2 from this Laboratory. I t was shown t h a t the slope of the plot of A z's. dc corresponds closely to the theoretical slope below 0.01 N and t h a t potassium chloride is completely dissociated in formamide even a t concentrations considerably above 0.1 N . Transference data have been obtained by the moving boundary method in ethanol and in methano13 b u t numerous experimental difficulties prevented using this method for formamide solutions. The present investigation was designed to apply the Hittorf method in an attempt t o obtain information concerning the relative solvent-solute interaction in this medium and in other more familiar solvents for which reliable data are available in the literature. Experimental Apparatus.-Use was made of the familiar type of Hittorf cell4 which was constructed from large Pyrex glass tubing. T h e cell was held in a thermostated water-bath. Care was exercised t o avoid exposure of t h e solution t o the atmosphere and t o prevent its contamination in any other way. Silver and silver-silver chloride electrodes were constructed from No. 24 B. and S.gage silver wire. X device employing a 6F6 pentode tube on a power line with relatively small voltage fluctuation maintained a constant current. -1 Richards silver coulometer was used in the circuit. Specially constructed weight burets mere used for t h e transfer of larger amounts of solutions. Salt.-P(itassiuni chloride which had been purified especially for accurate PH work \vas purchased. T h e salt was dried a t 110' for ten hours and used without further treatiiient. Solvent.-Commercial grade formamide was distilled three times a t reduced pressure over calcium oxide. T h e middle portion retained from the third distillation was carried through several fractional freezing cycles using t h e procedure described by the present authors in a n earlier paper.5 Physical properties at 25' of the solvent used were as .follows: density = 1.129 g./ml.; viscosity = 0.0331 poise; dielectric constant = 109.5 at 10 mc.; conductivity = 1-5 X ohm-' cm.-l. T h e solvent was used as promptly as possible after final purification and was protected from light while in storage. Procedure.-Dissolution of the salt required several hours with frequent agitation of the solution. XI1 solutions were prepared out of contact with air and dry nitrogen was bubbled through a solution for one hour before it was transferred to the cell. Lubriseal was used on t h e stopcocks and the usual precautions were observed t o ensure tight seals and t h e absence of air bubbles in the cell. hfter t h e cell had been in the thermostat long enough to (1) T a k e n f r o m a Pl1.D. dissertation submitted by Carl Berger. ( 2 ) I,. R. Dawson. T . 11.Sewell a n d W.J . hlcCreary, T H I S J ~ u R N A L , 76, ti024 (1954). (3) J . Smisko a n d L . R. Dawson, 3 . P h y s . C h c n . , 69, 84 (liJZ3). (4) D . A . LIacInnes a n d AI, Dole. THISJ O U H N A I - . 53, 1357 (1931). ( 3 ) C. Berger and I,. R Dawson, Aizol. C h e n t . , 24, 944 (1952)
establish temperature equilibrium t h e circuit was closed aiid electrolysis allowed t o proceed for from 24 t o 30 hours at 1-3 ma. A4t higher concentrations the current becainc larger and the time was decreased. Changes in salt concentration were measured by potentiometric titration for chloride ion in t h e formamide scilutions. This titration procedure has been described previously , E
Results and Discussion T h e data presented in Table I represent averages of consistent duplicate or triplicate determinations. Owing to the instability of formarnide and the difficulty of maintaining i t anhydrous as well as the problems involved in potentiometric titrations in this solvent, there may be inherent errors in the results amounting to 2 or 3cj,. TABLE I PorAssIuM CHLORIDE IN FORMAMIDE Concn. x 10:
I (ma.)
20. 04 20.80
1. 0 1. 0 1.2 1.4 2.0
24.37 28.98 57.28 58.91
-3. -3
G. of A g ' deposited
0,1247 ,1214 .133i ,1612 ,1871 1817
1,
0.40 4(J
40 39 .38 38
A41thoughexperimental difficulties prevented extension of the study to dilute solutions, an attempt was made to obtain an approximate value for the limiting transference number by extrapolating a plot of the Longsworth function.b: t+O' = ( t + l ' + 1,/2pc'/'?)/(*i1 + BC'I!) I n this expression, ,I1 = .i,- ( a & fi)C'/2, /3 = S2.42'(DT)'/zqand Q = 8.204 X lo5 (D7')'/'.. D is the dielectric constant, 17 is the viscosity and C is the concentration in equivalents per liter. The entries in Table I correspond to the plot in Fig. 1 . A value of 0.406 for the limiting transference number of the potassium ion was obtained by extrapolation of the plot; therefore, the limiting transference number of the chloride ion is 0.594. Above 3 X l o p 2 N the plot deviates sharply from a straight line. Conductance data from another study in progress in this Laboratory, which support the general validity of these values, will be reported in a subsequent paper. Using data from the work of Dawson, Sewell and McCreary2 with a value of approximately 29.7 ohms-' cm.2 equiv.-' for the limiting equivalent conductance of potassium chloride a t 2.5" and our value for t~ the limiting equivalcnt ionic coil-
+
? ,
(li)
I.
(',
1,ongsworth. T H I SJ O I . R I \ A I . . 6 4 , 2 i 4 1 (19:3'2)
D. L. LEUSSING A N D R.
42iO
I
5,'
1
/ /
/
0 400 10
cx
30
20
102.
Fig. 1 -The concentration dependence of the Longsworth function for potassium chloride in formamide.
ductances become X"R, = 12.1, and X"C.~- = 17.0. Similarly from a value of 27.9 ohms-' cm.? equiv. -' for the limiting equivalent conductance of hydrogen chloride in formamide a t 23". and X"CI= 17.G) the calculated Xo value for the solvated hydrogen ion is 10.3 and its transference number is 0.37.
[ C O S T R I B E T I O S FROM THE
DEPARTMEST OF
C.HANSEN
Vol. i9
I n passing from water to^+ in KCI = 0.49) to formamide solutions the limiting transference number of the potassium ion in potassium chloride is decreased 1TYc. The relatively larger solutesolvent interaction with the potassium ion in the latter solvent probably is an ion-dipole effect resulting from the greater dipole moment of the formamide.' Evidence that the large dipole moment of formamide produces strong ion-dipole forces appears also in the decrease in transference numbers of the solvated proton from 0.52 in water to 0.37 in formamide. I n formamide dissociated protons are held b y the solvent molecules with sufficient force to prevent easy transfer from molecule to molecule as occurs in aqueous solutions. The limiting equivalent ionic conductanceviscosity products for the potassium ion a t ?so, in o h n r l cm.? equiv.-' poise, in methanol, ethanol, formamide and water are 0.2X3, 0.235, 0.400 and 0 . 6 6 . -1 complete explanation of the relative magnitudes of these values would involve not only direct ion-solvent interactions but also shortrange viscosity effects resulting from the influence of the ions in modifying the normal solvent structure. (7) W .1%'. Bates and 11. E. Hohbs, THISJ o U R s . 4 1 ~ ,73,2131 (1'331).
LEXINGTOS,KY..
CHEMISTRY, L-YIVERSITY O F IYISCOSSIS]
The Copper (11)-Pyridine Complexes and their Reaction with Hydroxide Ions BY D. L. LEUSSING AND R.C. HANSEN' RECEIVED FEBRUARY 21, 1957
+
T h e formation constants have been evaluated a t a n ionic strength of 1.0 and 25.0' for the reactions: C u + + py & Cup!--', C u r + t 2py Cupy2+', . . , C u f + 5p)Cup>-:-'. The values found are 3.92 X IO2, 2.14 X lo4, 8.5 X 106, ,'3.,5 X IO6 and 1.06 X lo7,respectively. Mixtures of these complexes react with hydroxide t o forni a dimer: p y r C u The abI p y & p!.,C~i(OHMhpy~++is 5.1 X (OH)2Cupyr-+. The constant for the reaction 2 C u + - f 'OH5orption sFectra of each of the complexes were determined and the spectral behavior of the solutions was found to be in RCcordarice with the interpretation based upon equilibrium concentration measuretnents.
+
+
Recent publications have described the formation of mixed hydroxy complexes of copper(I1) with ethylenediaminetetraacetate, '' ethylenediamine, 1 ,1O-phenanthroline,l a,a'-dippridy16 and T h e hydroxide was shown to occupy a fifth or sixth coordination position on the copper in the case with his-ethylenediamine and ethylenediaminetetraacetste. However, with l,10-phenanthroline and cy,a'dipy-ridyl the coordination of copper is lower (excluding water) as only one molecule of the bidentate ligand per copper was observed in the nixed complexes. The existence of the hydroxy (1) T h e authors wish t o express their gratitude for partial support of this work b y t h e IVisconsin Alumni Research Foundat,on a n d t h e d u P n n t Grant-in-Aid. Preliminary experiments \\-ere carried o u t hy li', B Ohnesorge a t the XIassachusetts Institute of Technology s u p pnrtetl liy R research grant of t h e Atomic Energy Cummission. 1') \I C Bennett and h7.0 Schmidt, 7'rniis. F n r o r i n y S u i , 61, 1412 [ 1!1 i3 ) . (3) H B. Jimassen, R E . Reeves a n d I-. Segal THISJounr;.41, 77, 2748 fl9.55) f4) R T Pflaum a n d W. Ti' Brandt, i b i r i , 7 6 , ii2l.i (1934). (i) T IVagner-Jauregg, €3 E Hackle?, J r , 'r, A 1,ie.. 0 0 . Owen. and I< Proper, ibid.,77, (12'2 (1'3.13). Q ( I ; ) J . Bjerrum, C . J . Ballhausen a n d C li. Jorgensen. A ~ / ChCm. S c a i r d . , 8 , 1'273 (1934).
ammonia complexfiwas inferred from spectral cvidence and its composition was assumed to be Cu(NH:I):jOHf. Only in the case of ethylenediaminetetraacetate was the constant for the formation of the mixed complex evaluated. This was found to have a pk' of 11.3 for the reaction Cu\=
+ O H - I-CuYOH'
The reactions of the copper-pyridine complexes were investig2ted in order to throw more light on the nature of the reactions of the copper ammines with hydroxide. Pyridine was chosen because it is a monodentate ligand and its activity in aqueous solutions can be determined very easily b y nieasuring the distribution into toluene. The conventional pH method of Rjerrum' cannot be used i n the present case to determine the nature and stability of the mixed complexes because the pyridinium ion does not exist in appreciable quantities a t the PH values optimum for the reaction with hydroxide. In Rjerrum's method the concentration of ( 7 ) J Bjerrum, "XIetal .4mmine Formation in Aqueous Solution," P. Haase and Son, Copenhagen, Denmark, 1'341
