media, and it is not clear from the preparative procedures what solution conditions are most favorable for forming IC14-. Solutiolis prepared by dissolving IC1, in 0.3 aiid I . l i HC1 are quite stable, and the A,,, values coincide n-ith thGJSe of IClc-. -Assuming that the absorptim is due to IClc- alone leads to the cow clusion that half of the IC& is converted to IC12The spectra and titer of these solutions remained unch:ingetl over a period of 27 days. There was no evidence that C1, could be lost from the soluticins either by aging or heating. On the other hand, when is dissolved in :3 .II HC1, a definite odor of chlorine can be detected. The spectrum of a freshly prepared solution is not itlent.ica1 to that of IClz- but has a contribution from another substance (Cl, and 'or Cla-) in the region around X15 nip, ;Is the solution ages, the spectrum approaches that of IC12-, the extent of absorption indicating that all of the ICl;;is cmverted to IC12-. The iodine content of these solutions was determined by reducing to iodide and then performing ;I potentioinetric titration with AgNOa. This value was then compared with the oxidizing power tieterinined by iodometric titration. These experinimts showed that in 0.3 JI HC1 the iodine is present in 3 form or forms equivalent in oxidizing power t:) ICl;i, whereas in :: .\I HC1 loss of chlminc reduces the oxidizing power to almost that equivalent to IC1. All these facts are consistent with the following reaction occuring in 0.3 JI HCl L'ICI.
-
:3I-I,0 +I O i r
+ ICI2- + 1C1- + iiH--
In :I -11and higher HC1 concentrations. on the other hand, the ICl2 decomposes according to IC13
+ C1
.
+IC12
+ Cl,
This interpretation is also consistent with the behavior of 1 0 3 - when added to HC1 solutions. If the concentration of HCl is 3 JI or greater. the 224 mp peak of ICls- appears within the order of minutes, along with the odor of Cl?. If the HC1 concentration is 0.3 -11, there is no evidence of reaction.
[ C o s ~ ~ i u r -S~o~. 1045 o s FROM
THE
-4 study of the potentiometric titr:ition"6 of
with R I in hydrochloric acid revealed that the e n d - p i n t appeared prematurely when the HC1 cancentration was greater than 2 aiid that the exact position of the end-point depended c)ii the time taken for titration. This effect can iiow be seen to result f r m i the forniatioii and loss ( i f chlorine gas. Fialkov and Kagan"' report the use of solutions of ICl:] for volumetric analysis. Their solutions were prepared either by dissolving solid ICl:] in 11.2-0.4 -11 HC1 or by adding KI and RIO:, in a mole ratio of 1 : 2 to the acid solution. On the basis of the evidence presented so far it would appear that such solutions do not contain ICL of I C k , but instead, equal amounts of IC1?- and IO3-. -1s a further test of this view solutions in 0.3 -11 HC1 were prepared by adding KIOa and K I in 4 : 2 and 1 : 2 mole ratios, keeping the amount 91" IC1 the same in both. If the only reaction that occurs is 103-
IO,,-
+ 21-
- OCI-
+ iiH'
+31C12.- + 3H,O
excess 103- ions ill remain in the 4 : 2 solution, and the two solutions ought to have identical spectra in regions where IO:{does not absorb. On mixing the solutions a rather intense yellow brown color developed immediately. probably due to the formation of Ia-. U-ithin a few minutes i t was replaced by the expected light lemon-yellav color characteristic of IC!?-. Spectral measurements in the 270-40;) m p range showed the two solutions to be identical except fnr the small range 270-29;) inp, where the 4 : 2 solution had a slightly higher absorbance. This cauld be ascribed to absorbance by the excess IOj- ions. Thus under the conditions employed in these experiments no evidence was found for the existence of ICla or IC14-. Of course, other conditions, such as the presence of excess Cl,, may prove favorablr to their existence. (:tC) H . T. S. Britton, R.E. Cuckaday and J. K . Vureman, J . Cl!eiii Soc., 3 8 i 7 (1952). (37) Y . A Fialkov anti 1'. E . Kacan, L I ! . K l l l i l , . ,?ll!!i. , 13, .ii ( 19.i2) ,
STERLISC CHEMISTRY LABORATORY OF YALEUSIYERSITY,SEIY HAVES,COSXECTICUT
Electrolytic Conductance of Salts of Several Cyanocarbon Acids BY JOHN E. LIXD,JR.,'
AND
R.-\E'MOND h l . Fuoss
RECEIVED SOVEMBER 12, 1960 T h e conductance a t 25" i l l the tetraethy1amr~ir)tiiums:ilts of S,S,L',3,3-~entac?-anoprc,pene and of 1,1,~,4,3,;j-liesacyano-3xzapcntadiene have been measured in acetonitrile and in eth? lene dichloride; t h e penta-salt vas also measured in water. 'The electrolh-tic behavior is typical for t h a t of s d t s c:f strong acids. Association is negligible in water and in acetonitril~ ID = 3i.01 I ; in ethylene tlichloride (D = 10.35,, associa?iciti constants are 2600 and 2000 for the penta- and hem-salts, respectively. These values, a5 well as t h e ciintact distances derived from them, show that the negative charge is widely dist r i h t e d i l l bcith anions.
A number uf cyaiiocarbon acids? have been prepared recently, as well as a variety of their salts. Briefly described, these acids are essentially
hydrocarbons in which all hydrogens except one or two have been replaced by nitrile groups; the resulting carbanion is so highly stabilized by
April 20, 1961
CONDUCTANCE
O F SALTS OF
CYANOCARBON
A\CIDS
1829
expect their salts to show association to ion pairs to an extent determined solely by the size of the ions and the dielectric constant of the medium. Dr. V. A. Engelhardt of E. I. duPontdeNemours and Company very kindly gave us samples of the tetraethylammonium salts of I,1,2,3,3-pentacyanopropene (anion I) and of 1,1,2,4,5,5-hexacyano-3azapentadiene (anion 11). (For convenience, we shall refer to them as the “penta-salt” and the
cs
YC \
I
cs /
,c=c-c~\c s
NC
I
sc c s cs c s \ c=c-s=&--cg I / /
\
cs
SC
I1
“hexa-salt.” Only one structure is shown for I and 11; in I , the negative charge is distributed over the five peripheral positions and the two terminal carbons; in 11, over the six peripheral positions, the two terminal carbons and the central nitrogen.) The conductance of the penta-salt has been measured a t 25’ in water, acetonitrile and ethylene dichloride ; the hexa-salt in acetonitrile and ethylene dichloride. (The latter was insufficiently soluble in water to permit measurement. For the penta-salt, saturation in water was about 0.0021 N ; this low solubility required measurement to quite low concentrations, with a correspondingly large solvent correction reaching about 2y0 maximum.) Methods of measurement have been described in other papers from this L a b ~ r a t o r y . ~The cells used had constants4 equal to 0.125052 and 0.038757. The data are summarized in Table I. As will be brought out in
Fig. 1.-Conductance of EtrN.CsiCS)S (top curve), Etaru’.C~s(CN)s(middle curve) and Et&.BPhn (buttom curve) in acetonitrile a t 25”.
The phoreograms in acetonitrile are shown in Fig. 1, where for comparison the curve for a typical unassociated salt (tetraethylammonium tetraphenylboride6) is shown. I t will be noted that the points for all three salts lie above the corresponding limiting tangents and (except for differences in .iodue to different sizes of the ions) the three curves show about the same curvature. Incidentally, these curves serve to illustrate an interesting point: straight lines could be drawn through the observed points, which would give TABLE I CONDUCTANCE O F TETRAETHYLAMMONIUM 1,1,2,3,3-PEXTA-values of A” about 0.5 unit too low. This beC Y A N O P R O P E S I D E A N D 1,1,2,4,5,5-HEXACYANO-3-AZAPENTA- havior is due to the opposition of the linear and transcendental terms in the conductance equaDIENIDE tionl 104 c i 10‘ c A HzO-Cyj
19.575 16,264 13.149 9.948 7,1715
CH3CN-Cya
64.10 64.36 64.62 64.94 65.24
CHZCN-Cyj
13.2688 10.8i95 8.7456 5,9148 2.9191
157.00 158.10 159.21 160.97 163.46
14.8539 12,1250 9.2526 6.2409 3.2496
143.82 144.89 146.21 147.91 150.17
CZHIC~Z-CYO
4.8578 3.8542 2.8541 1.9791 0.9858
42.10 44.05 46.57 49.55 54.71
C~H4C12-Cy.j
5,7818 4.5729 3.4857 2.2968 1.2648
40.61 42.78 45.32 49.20 54.48
the subsequent discussion, normal behavior for ionophores5 was found in all the systems studied. (3) H . Eisenberg a n d R . M.Fuoss, J. A m . Chem. S O C . ,2914 ~ ~ ,(1953). (4) J. E . Lind, Jr., J. J. Zwolenik a n d R . hl. Fuoss, ibid., 81, 15.57 (1939).
( 5 ) R. M .Fuoss, J . Chem. Educ., 32, 527 (1956).
A
A o - SC”*
+ J C + EC log c
(1)
which is valid for unassociated electrolytes. They nearly cancel in our working range of concentration; a t a concentration given by8 cg
=
exp ( - J / E ’ )
(2 )
the curve will cross the limiting tangent (c” is shown by the arrows in Fig. 1) and then approach it from below in the limit of very low concentrations. Considerable association to ion pairs is shown by both salts in ethylene dichloride, where D = 10.33. The values of the association constants are given in Table 11; they were obtained by IBL1 650 analysisg of the data, using the appropriate equation for associated electrolytes.1° The value F = 2.0 for the Einstein viscosity coefficient was used in the calculations. If the association constant is given by1‘ the equation K A = 2.523 X 10-3 63 antilog (243.4/&0) (3) (6) D. S. Berns and R . M. Fuoss J. A m . Chem. S a . , 8 2 , 5585 (1960). (7) R . M. Fuoss and F. Accascina, “Electrolytic Conductance,” Interscience Publishing Co., New York, N . Y., 1959; chapter 15. (8) Ref. 7, equation 1.5.61. (9) R. L. K a y , J. A m . Chem. Soc., 8 2 , 2099 (1960). (10) Ref. 7, equation 17.7, with inclusion of Jz t e r m , 11.104. (11) Ref. 7, equation 1F.19.
11.H. FORD-SMITH AND N.
1830
it solution for d , using the observed K A values, gives 6.64 and G 99 for the penta- and hexasalts, respectively. (The values of K A = 8 in acetonitrile in Table I1 were calculated from the above &values, using equation 3 ) As would be expected from the relative sizes of the two ions and the weaker field of the hexa-anion due to a greater spatial distribution of its single charge, the hexa-salt is noticeably less associated than the penta-sal t. TABLE I1
DERIVED CONSTANTS Sdt
Solvent
-4 Q
dAo
K.4
Hi0 67.19 z!c 0 . 0 1 0 N e C N 169.57 3z . O 1 (8) C2HaC12 70.81 =!= .17 2570 (i bleCN 156.35 =t ,131 (8) D C?H1C12 66.94 =t .13 2000 5
3 5
dIio
5.49 ,. 5.61 .. 5 . 5 0 6.64 6.10 .. 6.32 6.99
dJ
5.2 5.62 8.7 6.19 9.0
The hydrodynamic radiiI2 calculated from the limiting conductances of the penta-salt in three quite dissimilar solvents (water, acetonitrile and ethylene dichloride) agree within a few percent., showing that no special CN-CN interaction ( e g . , “selective solvation”) occurs in acetonitrile between solvent and anion. The hexa-salt was too low in solubility to be studied in water; the U A values in acetonitrile and in ethylene dichloride, however, are also in good agreement with each other. (12) R . h I . I‘unss, Proc. Sail A r a d Sci
[ CONTRIBUTIOK
, L‘ S , 45, 807 (1969).
FROM THE CHEMISTRY DEPARTMENT,
%TIN
Vol. 83
The values of U A from A, and of U J from J agree well for both salts in the solvents of higher dielectric constant, but U J > u-4 in ethylene dichloride. This discrepancy has been noted beforeI3 in solvents of low dielectric constant; its origin is suspected to lie in several of the approximations made in deriving the conductance equation14 which, while valid for large values of D ,become less reliable as D decreases below about 20. The values of the contact distance, calculated by any one of the three methods, are large enough, however, to show t h a t the charges on the polycyanoions cannot possibly be localized to any one site in the ion. The salts are colored (penta, yellow; hexa, orange) and also show strong absorptions in the near ultraviolet (molecular extinction coefficients E = 2 . 2 X lo5 for the penta-salt in ethylene dichloride a t X = 417.5 mp; E = 4.4 X lob for the hexa-salt in the same solvent a t X = 469.0 mp. These wave lengths correspond to the long wave length members of a double peak in the absorption spectrum). With the thought that ion pair formation might distort the anionic electron distribution, some exploratory measurements were made. The molar extinction coefficient of the penta-salt was, however, found to be unchanged over the concentration range 6.4 X to 6.4 X where the fraction of paired ions varies from about 25 to lY0. (13) E. Hirsch and R. hl. Fuoss, J . A m . Chem. S O L , 82, 1018 (1960). (14) R. M . Fuoss and L. Onsager, J . Phys. Chern., 61, BG8 (19Tr7); R. l f . Fuoss, unpublished calculations.
BROOKHAVEN KATIOSAL LABORATORY, UPTOS,
S E W YORK]
The KiEetics of the Reactions of Substituted 1,lo-Phenanthroline, 2,2’-Dipyridine and 2,2 ’,2 ”-Tripyridine Complexes of Iron(II1) with Iron(I1) Ions1 BY bl. H. FORD-SMITH AND N. SUTIN RECEIVED SEPTEMBER 29, 1960 The b inetics of the oxidation of iron(I1) ions by a number of substituted 1,lO-phenanthroline, 2,2’-dipyridine and 2,2‘,2“ trip) ridine complexes of iron(II1) have been investigated in both sulfuric and perchloric acid solutions using a rapid-mixing arid flow technique. The logarithms of the second order rate constants were found to be linearly related t o the averaqe basicity per nitrogen atom of the ligands for oxidations by the unsubstituted and the &nitro-, 5-chloro-, 5-phenyl- and 5meth) 1-phenanthroline complexes of iron(II1). The free energies of activation for the oxidation of ferrous ions by the unsubstituted, the 5-substituted, the 5,6-dimethyl- and the 3,4,7,8-tetramethyl-phenanthrolinecomplexes of iron(II1) are linearly related t o the standard free energy changes of the reactions The absence of specific steric effects suggests t h a t the electron-transfer betrveen the iron(II1) complex and the ferrous ion takes place in a n activated complex in which the ferrous ion has penetrated the space between the phenanthroline groups.
Introduction The kinetics of the oxidation of iron(I1) ions by tris-(1,lO-phenanthro1ine)-iron(II1) ions have recently been investigated.* I n a comparison of this reaction and several other iron(I1) ion oxidations, it was found t h a t the free energies of activation could be linearly related to the standard free energy changes of the reactions. We have extended this work to include the oxidation of ferrous ions by a number of substituted 1 , l O phenanthroline, 2,2’-dipyridine and 2,2’,2!’-tri(1) Research performed under t h e auspices of t h e U. S. Atomic Energy Commission. (2) K ,Sutin and B. hZ. Gordon, J . A m . Ckem. Soc., 83, 70 (19131).
pyridine iron(II1) complexes in both sulfuric and perchloric acid solutions. With various groups substituted in the phenanthroline nucleus, it is possible to obtain iron(I1) complexes with formal oxidation potentials ranging from 0.81 to 1.25 volts.3 Substitution in the phenanthroline nucleus alters the basicity of the nitrogen atom and hence its electron-donating properties. Since an increase in the electron-donating ability of the nitrogen atom will tend to stabilize the iron(II1) complex relative to the iron(I1) complex, the formal bxidation potentials of ‘the iroi(I1) complexes (3) 0.F. Smith and W. M. Banick, Talonta, 2, 348 (1959).