SIDNEY D.
4916
[CONTRIBUTION FROM
MolecuIar Compounds.
THE
ROSS AND
VOl. 77
MORTIMER M. LABES
RESEARCH LABORATORIES OF THE SPRACUE ELECTRIC Co.]
V. Solvent Effects on Complexing of 1,3,5Trinitrobenzene with Aniline and with Naphthalene BY
SIDNEY D.
ROSS .4ND
PITORTIMER M. LABES
RECEIVED DECEMBER 29, 1954 The equilibrium constants for complex formation have been measured for the 1,3,5-trinitrobenzene-naphthalene complex in chloroform-ethanol mixtures and for the 1,3,j-trinitrobenzene-anilinecomplex in chloroform-ethanol mixtures and in dioxane-water mixtures. The heats, free energies and entropies of formation of the complexes were determined in 75% chloroform-25% ethanol. The structures of the complexes are discussed.
The response of a reaction rate in solution to solvent changes is a useful criterion of mechanism' and can, in some cases, be treated quantitatively.2 I n discussing solvent effects on an equilibrium, the structure of the transition state may be eliminated from consideration, since an equilibrium constant measures the ratio of the rates of two reactions, both of which have the same transition state. For molecular complex formation, the effect ol the solvent will depend on whether the reactants or the complex are preferentially stabilized by solTVeiss4has predicted that charge-transfer complexing will be favored by increasing the polarity of the medium. This is based on the fact that the reactants are, in most cases, uncharged while the complex has a contributing structure with a charge ~ e p a r a t i o n . ~ This expectation has been fulfilled for the ?,Adinitrochlorobenzene-aniline complex6 but not for the N,N-dimethylaniline-sym-trinitrobenzenecotnp l e ~ . This ~ has led us to make a more detailed study of the effect of solvent changes on the equilibrium constant for complex formation, and we have measured the 1,3,5-trinitrobenzene-anilineequilibrium in chloroform-ethanol mixtures and in dioxane-water mixtures and the 1,3,5-trinitrobetizene-naphthalene equilibrium in chloroform-ethano1 mixtures. The heats of formation for the nitrohyclrocarbon-hydrocarbon complexes are generally low and usually less than 5 kcal.8 However, for complexing between 2,4-dinitrochlorobenzene and substituted anilines, the heats of formation are considerably higher, and in the case of the 2,4-dinitrochIorobenzene-o-methoxyaniline complex the heat of forma(1) (a) E. D. Hughes and C . K . Ingold, J . Chew . S o < , 244 (l9.%3J; I,. C. Bateman, M. G. Church, E. D . Hughes. C I;.Jngold and N. A Taher, ibid., 979 (1940); (b) I;. J. 1,aidler and €1. Egring, A n n .V, Y . A c o d . Sci., 99, 303 (1940); (c) E. Grunwald and S. \Tinstein, THISJ O W R X A L , 70,84(i (1948); S. \\-instein, li. i,
,
tion in ethanol is 18 kcaLQ Also, Xbrahams and have reported a direct observation of a relatively strong bond between an oxygen atom of a nitro group and a carbon atom of an adjacent benzene ring in crystalline p-nitroaniline, which forms a self complex. This suggests that the interaction of donor and acceptor may be more specific and more strongly localized in the amine complexes and that earlier views" of the structure of these complexes should be reconsidered. I t was our hope that, if there are structural differences for these two types of complexes, this would manifest itself in a different response to solvent changes. Experimental Solvents and Reagents.-J. T. Baker and Adamson Reagent Grade chloroform, \rhich contained 0.75% ethanol as stabilizer, was used without further purification. Absolute ethanol from the U. S. Industrial Chemicals Company was used without further purification. Titration with the Karl Fischer reagent shoired t h a t this alcohol contained 0.08% r a t e r . For measureincuts of the 1,3,5-trinitrobenzerleaniline complex in cliloroform-etli;inol mixtures, the ethanol was purified further b!; refluxing it over calcium hydride and distilling it from calcium hydride iininediately before use. After this treatmelit the alcohol contained 0.03% water. 1,+Dioxane was purified bl- the inethod of Eigenberger'*; b , p . 101-102" and stored over sodium. Eastman Kodak Company I'i7iite Label aniliiie n w distilled from calcium liydridc iiiimediatel>- before use, a n d a middle fraction of 11.p. 63" a t 8 mni. W:IY used. Naphthalene was crystallized several times from tiietlianol; colorless plates; m.p. 80-81". 1,3,5-Tritiitrobeiizelie, Eastmnii Kodak Company tA'hitc Label, was crystallized from chloroform; almost colorless needles, 1n.p. 121'. The Absorption Spectra Measurements.-A Beckinail model DU spectrophotometer was used throughout. Stoppered absorption cells were used, and the cell housing was maintained a t constant temperature by means of two Beckman thermospacers, through which water from a constant temperature bath i n i s circulated. Measurements were made a t a t least six different sets of concentrations of the donor and acceptor. In measuring the 1 ,:3,5-trinitrobeiizene-naphthalene complex iii chloroform -ethanol mixtures. the trinitrobenzene concentrations were varied from 5 X lo-..' - 9 X IO-' 31, and the naphthalene concentrations were u r i e d from 0.1-0.9 91. For the 1,:3,5-triiiitrobenzeneaniline complex iri chloroform-etilanol mixtures, the tri- 1 nitrobenzene concentrations were varied from 5 X X 10-2 .IT and the aniline concentrations were varied from 0.3-1.1 -11. I n the measurements in dioxane-water mixtures, the trinitrobenzene concentrations were varied from 1 X - 7 X 10-3 A 1 and the aniline concentrations (Y) S. D. Ross, 31, Bassin and I. l i u n t z , ibid., 16, 4170 (19541. B.R. Hamilton and D. L1. Hammick ( J . Chern. Soc., 1350 (1938!), on the other hand, have reported low values, comparable with those found with hydrocarbons, for t h e heats of formation for complexes of 1,3,9trinitrobenzene with a series of aromatic amines. (10) S. C . Abrahams and J. 31. Robertson, A d a Ci'ysl, 1, 2 5 2 (1948); S.C . .%brahams, T i m J O V R N A I , , 74, 2tiQ2 (1952). i l l ) See Cur eu.iinple, (;. li llciiiiett : i n t i 1; I T \Villis. .I C'iirin .Svr , ?,-,ti
~I!l'"J)
( 1 2 ) IC
l,:igeiil>erger,J . /,,
dk/
('iiiii;.,
( 21 1 3 0 , i . i I
l!l:j
I).
&pt. 20, 1955
COMPLEXINC OF 1,3,5-TRINITROBENZENE WITH ANILINE AND
NAPHTHALENE4917
were varied from 0.45-1.4 M. The equilibrium constants and extinction coefficients were calculated by methods previously described.*a The method of least squares was used in every case.
The 7.0% ethano1-93.0% chloroform mixture is an azeotrope boiling at 59.350.14 As can be seen from Fig. 1, the equilibrium constant in this solvent mixture deviates from the smooth curve obtained with Results the other mixtures. This deviation is real and was The equilibrium constants for formation of the 1,3,5-trinitrobenzene-naphthalenecomplex and the checked by preparing the azeotrope both by distillaextinction coefficients of the complex a t the wave tion and by mixing the requisite weights of the two lengths a t which the equilibrium constants were pure, anhydrous solvents. The results with the calculated for a series of chloroform-ethanol mix- azeotrope prepared by both methods are given in tures are given in Table I, and in Fig. 1 the equilib- Table I and Fig. 1. Table I1 lists data for the 1,3,5-trinitrobenzenerium constants are plotted against the dielectric constants of the solvent mixtures. It is of interest naphthalene complex in 7570 chloroform-25% ethathat in this series of determinations the extinction nol a t three temperatures. From an Arrhenius coefficients are reasonably constant in all the sol- plot of log K us. 1/T and using the method of least vent mixtures. This suggests that the structure of squares, we can estimate A I P for this complex, and the complex is unaltered by these solvent changes. using the expressions AFO = -RT In K and A P = AHo - T ASo, we obtain the thermodynamic values TABLEI given in this table. I n column 3 we have listed EQUILIBRIUM CONSTANTS AND EXTINCTION COEFFICIENTSthose values of K which would exactly fit the least squares straight line. FOR T H E 1,3,5-TF3NITROBENZENE-NAPHTHALENE COMPLEX IN
CHLOROFORM-ETHANOL MIXTURES AT 24.8 f 0.1"
% by wt. ethanol
0.75
Dielectric constant
Wave length, mp
Extinction coe5cient
K, I. mole-'
450 87 1.25 460 37 1.32 470 15 1.33 7.0" 6.3 450 92 1.13 460 43 1.08 470 18 1.02 7.0' 6.3 450 89 1.14 460 45 0.99 470 19 0.94 15.0 8.4 450 84 1.20 470 17 1.18 480 7.2 1.03 25.0 11.1 440 169 1.07 450 88 1.03 460 42 1.05 50.0 17.3 430 319 0.744 440 176 ,698 450 85 ,716 75.0 22.2 420 504 0.572 430 284 ,549 Azeotrope prepared by distillation. Azeotrope prepared by weighing.
TABLEI1 THE
4.8
4
I 0
I
5
I
I
I
15 20 Dielectric constant. Fig. 1.-The equilibrium constants for formation of the 1,3,5-trinitrobenzene-naphthalenecomplex 8s. the dielectric constants of the solvent in chloroform-ethanol mixtures at 24.8 f 0.1". 10
(13) R . M. Keefer and 2 . J. Andrews. THISJOURNAL, 74, 1891 (1952); S. D. Ross, M. Bassin, M. Finkelstein and W. A. Leach, ibid., 76, 69 (1954).
Temp., 'C.
COMPLEX 75% cHLOROPORM-25% ETHANOL K K
1,3,5-TRINITROBENZENE-"THALENE
(exptl.), (calcd.) 1. mole-1 I. mole"
AHO,
kcal.
AFO, cal.
IN
ASO, e.u.
24.8 1.05 1.07 -3.3 -28.3 -10.6 2.4 -10.9 30.4 1.00 0.96 -3.3 -10.5 39.6 0.81 0.83 -3.3 13.0 a This is the required value for K for the point to exactly fit the least squares line of log K ws 1/T.
The equilibrium constants and extinction coef ficients for the 1,3,5-trinitrobenzene-anilinecomplex in chloroform-ethanol mixtures at 24.8 f 0.1' are given in Table 111, and the equilibrium constants are plotted against the dielectric constants in Fig. 2. For this complex the extinction coefficients are definitely changed by solvent changes, with the maximum change being greater than 30%. The extinction coefficients decrease regularly as the equilibrium constants increase, and these changes are probably indicative of specific solute-solvent interaction in these systems. l6 The values of the dielectric constants in both Tables I and 111 are taken from measurements reported by Graffunden and Heymannl6 at 450 m. A plot of dielectric constant vs. wt. yo chloroform is a smooth curve over the entire range of solvent compositions. The equilibrium constant for formation of the 1,3,5-trinitrobenzeneaniline complex in reagent grade chloroform a t 22 f 2' has been reported previously by Landauer and McC0nne1l.l~They report an equilibrium constant of 5.1 f 0.7 mole fraction-'. It is not possible to translate exactly their concentration units into ours, since the aniline concentrations involved are as high as 1 M , but their value corresponds roughly to 0.41 0.06 1. mole-',
*
(14) "Azeotropic Data," American Chemical Society, Washington, D. C., 1962, p. 21. ( 1 5 ) For a general discussion of solvent effects in absorption spectrosCOPY see A. E. Gillam and E. S. Stern, "An Introduction to Electronic Absorption Spectroscopy in Organic Chemistry," Edward Arnold Ltd., London, 1954, p. 263. (16) W. Graffunden and E. Heymenn, 2. Physik, 72, 744 (1931). (17) J. Landauer and H. McConnell, THIS JOURNAL, 74, 1221 (1962).
4918
SIDNEY
D. Ross
AND
MORTIMER M. LABES
TABLE 111 EQUILIBRIUM COSSTANTSASD EXTIKCTIOS COEFFICIENTS F O R THE 1,3,5-TRINITROBENZEKE-ANILINE COMPLEX I N AT 24.8 f 0.1' CHLOROFORM-ETHANOL MIXTURES R b y wt.
Dielectric constant
ethanol
0.75
4.8
7.0
6.3
25.0
11.1
50.0
17.3
60.0
19.4
65.0
20.4
70.0
21.4
Wave length, mp
Extinction coefficient
464 75.0
22.2
85.0
23.5
100
470 460 470 480 490 456 460 464 470 440 450 460 470
24.6
K,
0.258 ,252 ,261 0.293 303 ,301 0,373 ,359 ,352 ,391 0.418 ,432 ,442 ,423 0.442 ,464 ,488 0,544 ,567
,565 0.382 ,389 ,348 0.355 ,359 . 3 10 ,333 0,405 ,370 ,368 ,410 0.389 389 ,403 ,405
which is higher than our value of 0.257 f 0.005 1. mole-l. l 8 The 1,3,5-trinitrobenzene-aniline complex in 7.57, chloroform-25yo ethanol was also studied a t three temperatures, and the resultant values of the equilibrium constants and the thermodynamic quantities are given in Table IV. TABLEI V THE 1,3,5-TRISITROBENZENE-ANILINE COMPLEX CHLOROFORM-25% ETHASOL Temp., 'C.
K
(exptl.),
1. mole-1
K
(calcd.),'' I . mole-'
4IP,
kcal.
AFo, cal.
590 -5.1 -5.1 734 878 -5.1 a This is the required value for K for the point fit the least squares line of log K 11s. 1/ T . 24.8 30.4 39.6
0.369 ,296 ,244
0.368 ,308 ,244
1
1. X mole-!
1231 1021 790 1110 1002 970 997 997 862 547 977 820 686 580 976 884 776 840 770 702 1108 1041 1031 1188 1013 973 759 1101 1131 1085 904 1300 1181 1015 875
460 470 480 460 464 470 450 460 470 490 460 470 480 490 460 464 470 460 464 470 460
o.6 0.5
VOl. 77
IN
i5Yo
5
10 15 20 25 Dielectric constant. Fig. 2.-The equilibrium constants for formation of the 1,3,5-trinitrobenzene-aniline complex us. the dielectric constants of the solvent a t 24.8 i 0.1': upper curve, chloroform-ethanol mixtures; lower curve, dioxane-water mixtures.
aniline complex in dioxane-water mixtures a t 24.8 f 0.1'. The solvent composition was varied so as to cover the same range of dielectric constants as was covered with the chloroform-ethanol mixtures. The values of the dielectric constants are taken from the measurements of Akerlof and Short,Igand the equilibrium constants are plotted against the dielectric constants in Fig. 2. The low equilibrium constants obtained in 100 and 90% by wt. dioxane are close to the limit of measurement with the particular absorption cells we used. In this case, too, the extinction coefficients vary but in a more random manner. TABLE 1' EQGILIBRICMCOSSTANTS A S D EXTISCTIOS COEFFICIESTS FOR THE 1 , 3 , 5 - T R I K I T R O B E N Z E N E - A " L I S E COMPLEX I N DIOXANE-~VATER MIXTURES A T 2.4.8 dz 0 . 1 % by wt. ethanol
100
YO
Dielectric constant
2 1
j
ii
80
10 7
i0
17 7
05
21 3
A.co,
to exactly
Table V lists the equilibrium constants and extinction coefficients for the 1,3,5-trinitrobenzene(18) We cannot account qatisfactorily for this discrepancy. I n part, it m a y be d u e t o differences in temperature (see Table I V j a n d differPnces in t h e solvent, since reagent gradp rlililroform \.:trip\ iii f l i p nmoiiot o r alrolml u(lrleil 2.: rt iil)ili7cr
460 384 410
460 464 470 460 464 470 460 464 470 456 460 464
470
e.u.
-19.1 -19.2 -19.1
Wave length, mp
(XJ
25 9
Extinction coefficient
2677 2909 2117 3427 ,3139 2802 1567 1454 1101 1260 1283 1279 1556 1584 1171 1303
-100
1355
464 470
1366 1262
K>
1. mole-'
0.079 ,068 083 0 . 060 062 063 0.155 ,158 150 0.2:34 .21:1 . 193 0 .23C ,218
225 23,: 0 2.3 ,212 ,242
Discussion The solvating ability?"of a solvent is not a precise concept, in spite of the fact that solvation is essen(19) G. Akerlof a n d 0. A. Short, THISJOURSAL, 68, 1241 (1'236). (20) For a more detailed discussion of solvation, see J H . Hildebrand, "Soluhili Cy o f N-on~Tilect rol yt e - , ' ' I? Finhold Pub1 C o r p . , New I H X , chap