1268
ARTHURA. FROST AND WARREN C. SCHWEMER
of the absorption bands arising from the NH2 stretching vibrations probably results from the fact that one of the hydrogen atoms of the amine group is bonded to the adjacent oxygen atom of the nitro group. As a result there is some uncertainty regarding the assignment of the 3300 absorption band to the asymmetric NH, stretching vibration. However, since all of the absorption bands in the vicinity of 3500 cm.-l which could be assigned to this vibration have the same polarization characteristics as those bands due to known planar vibrations, the NH2 group is assumed to be planar. This configuration would be expected for the NH2 group in this case because of bonding and resonance. o-Nitrobrom0benzene.-The information listed in Table I for o-nitrobromobenzene, obtained from the spectra in Fig. l D , illustrate the effect of a large ortho substituent upon the geometry of the nitro group. The absorption band a t 1582 cm.-' arising from the asymmetric NO2 stretching vibration is polarized in the same direction as the absorption bands due to non-planar vibrations and opposite to those from planar vibrations. Therefore, the plane of the nitro group is rotated well out of the plane of the benzene ring due to the steric influence of the adjacent bromine atom and is probably almost perpendicular to the ring. This steric effect is usually observed indirectly by measuring the change in properties (e.g., dipole moments, ultraviolet spectra, rates of reaction, etc.) resulting
[ c O \ l RIBUTION FROM THE CHEMICAL
VOl. 74
from the inhibition of resonance18or by more direct X-ray diffraction methods. l 5 o-Nitroch1orobenzene.-The observed spectra of o-nitrochlorobenzene are given in Fig. 1E and the polarization data are presented in Table I . The absorption band a t 1628 cm.-' corresponding to the asymmetric NO2 stretching vibration has polarization characteristics more similar to the absorption bands from the non-planar rather than the planar vibrations, Therefore, i t is assumed that the plane of the nitro group is not coplanar with the benzene ring. The polarization of the 1528 an.-' absorption band does not compare as closely to the i 7 6 cm.-' band as for the analogous absorption bands in o-nitrobromobenzene. This difference may be assumed to indicate a smaller angle between the benzene ring and the nitro group in onitrochlorobenzene than in o-nitrobromobenzene, which would be expected because of the difference in van der Waals radii between chlorine and bromine. However, i t is difficult to attempt such quantitative comparisons in view of the non-ideal nature of the preparations used. Acknowledgment.-The valuable assistance of Dr. F. Halverson is gratefully acknowledged. (18) G. W. Wheland, "The Theory of Resonance and its Application to Organic Chemistry." John Wiley and Sons, Inc., New York, N Y., 1944.
RECEIVED APRIL 16, 1951
STAMFORD, CONNECTICUT
LABORATORY OF ?;ORTH\VESTERN
UKIVERSITT]
The Kinetics of Competitive Consecutive Second-order Reactions : The Saponification of Ethyl Adipate and of Ethyl Succinate1 BY ARTHUR,4. FROSTAND WARRENC. SCHWEMER
+
ki
+
+
k2
The differential rate equations for the kinetics of the reactions of the type A B +C E and A C +D 4-E have been integrated for the special case of stoichiometrically equivalent amounts of the reactants A and B. Numerical tables and graphs have been prepared that enable (a) the determination of whether a reaction is of this kinetic type, and (bi the evaluation of the rate constants. Measurements of the rate of saponification of ethyl adipate and ethyl succinate have heen made and usrd as ail illustration of the application of the theory.
Competitive consecutive (series) second-order kinetic system once and for all in terns of general reactions, as usually exemplified by the saponifica- variables that would apply to any future experition of diesters were first accurately handled by mental results. Ingold2 who used a successive approximation Theoretical Treatment method to obtain the rate constants. Ritchie3 made use of a procedure of graphical differentiation The chemical equations of the system are while the work of Westheimer, Jones and Lad4 involved graphical integration. French5 has recently handled this kinetic problem in more general kz fashion but still requiring a graphical integration AfC-DfE for each kinetic run. The aim of the present work was to integrate this and the pertinent rate equations are (1) Abstracted in part from the Ph.D. thesis of Warren C. Schwemer a t Northwestern University, August, 1950. Presented a t the X I I t h International Congress of Pure and Applied Chemistry, New York, N . Y.,September, 1951. See also W. C. Schwemer and A. A. Frost, THIS JOURNAL, 78, 4541 (1951). (2) C. IC. Ingold, J. Chem. SOL.,2170 (1931). (3) M. Ritchie, ibid., 3112 (1931). (4) F. H. Westheimer, W. A. Jones and K -1.1.4, .I. C h i n . P i r y s . , 10, 478 (1942). ( 5 ) D. French, THIS JOURNAL, 71, 48Oti (19501.
dA/dt = -kJB - ktAC dB/dt = -klAB dC/dt = kiAB - krAC
(1) (2)
(3)
where A , B and C represent the molar concentrations of the corresponding chemical species. Let the initial concentrations of A and B be A0 and BO, respectively, and the initial concentrations of C
SAPONIFICATION OF ETHYL ADIPATEAND ETHYL SUCCINATE
March 5, 1952
1.0
and D both zero. From the principle of material balance '4 - 2B - C = do - 2Bo Imposing for convenience the initial condition that A0 = 2Bo (equivalent amounts), then C=A-ZB dA/dt = (2k2 - ki)AB
0.8
(4)
and equation (1) becomes after substitution
- kzAZ
(5)
0.6
a. 0.4
The mathematical treatment is somewhat simplified by the introduction of the following dimensionless variables a,/3 and T and parameter K . O(
= A/Ao
p
r = Boklt
K
0.2
B/Bo = ke/kl =
(6)
In terms of these, equations ( 5 ) and (2) become da/dt = (2K - 1)ap dB/dr = - Z a p
- 2a2
I269
0 0.I
(7) (8)
Dividing (7) by (8) yields (9)
I
IO
T , TIME PARAMETER. Fig. 1.-Plot of a, fraction of substance A, as a function of the time parameter T (T = Boklt), for various values of I / K= k2/kl.
which is a linear differential equation with the solution
where the constant of integration is determined by the initial condition that when a = 1, p = 1. If (10) is substituted for a in (8) there results
or
taking r = 0 a t = 1. The solution of the kinetic problem then involves the evaluation of the integral in (11) for various values of the limit p and the parameter K . The concentrations of the reactants and products can then be evaluated with the aid of equations (lo), (6) and (2) and similar material balance equations for D and E if desired. See the appendix for details of the integration and numerical evaluation. The results for a and 0 as functions of r are shown graphically in Figs. 1 and 2 and numerically in Table I for the most useful range of K . T is plotted on a logarithmic axis both to compress the time scale as well as to make possible the use of the Powell6 method for evaluation of the rate constants from experimental data. Because T is defined as Bokit log
T
=
log Bok1
+ log t
Therefore if experimental values for a! or /3 are plotted against log t a shift of the plot along the log t axis by an amount log Bok, should make the curve coincide with one of the infinite family of curves such as those in Fig. 1 or 2. From the shape of the curve K can be determined, at least approximately, and from this together with the shift log Bokt both k1 and kz can then be calculated. If the (6) R. Powell, private communication.
0.I
I
7 , TIME
IO
PARAMETER.
Fig. 2.-Plot of 8, fraction of substance B remaining, as a function of the time parameter T for various values of I / K=
k2kl.
experimental curve for a or 0 fits one of the theoretical family of curves it may be supposed that the kinetics are probably of the type being discussed. However, caution must be used since for certain special values of K the dependence of a or p on 7 degenerates into what would correspond to a reaction of simple order, e.g., simple second order for K = 0, ' / z or 00, Since K cannot be accurately determined by this curve-fitting procedure a better way was devised as follows. It will be noticed that the most distinctive effect of varying K is to change the slope of the curves of a or B as a function of log r in the neighborhood of a or /3 equal to 0.5. Now the inverse of this slope is approximately given by Ahg r / A a or taking in particular T = 70.6 a t a! = 0.6 and r = 70.4 at a = 0.4 the expression becomes -5 log (
[email protected]/T#.6) = -5 log (h.4/t0.6)
ARTHURA. FROST AND WARRENC. SCHWEMER
1270 T
*
1 :x
a P O 8
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.5 4.4 4.6 4.8 5.0 6.2 5.4 5.6 5.8
0.2500 .2525 ,2547 .2566 ,2583 .2599 ,2612 .2624 ,2636 .2646 ,2656 ,2664 .2672 ,2680 .2687 ,2693 .2699 .2705 .2711 ,2715 ,2720 .2724 ,2728 .2732 ,2736 .2740 .2743 .2747 ,2750 .2754 ,2755 ,2758 .2761 .2768 .2765 .2768 ,2770 ,2772 ,2774 ,2776 ,2778
ii.0
6.2 6.4 G.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0 8.2 8.4 8.6 8.8 9.0 9.2
n ,4 9 . ti 0 .8 10,o
AS A
TABLEI FUNCTION OF K
AND u
0.7
0.6
0.5
0.4
0.4286 .4354 .4414 .4468 .4518 .4564 ,4606 ,4643 ,4678 .4711 .4741 .4769 ,4795 ,4820 ,4843 .4865 ,4886 .4906 ,4924 ,4941 .4957 .4973 .4988 .5002 .5OlG .5028 .5041 .5053 .5064 .507,5 ,5085 .50!)5 .5105 .5114 .ti123 .5131 ,5139 .5147 ,5155
0.6667 ,6817 ,6953 .7079 .7195 .7305 ,7407 .7502 .IT591 .7676 .7756 .7832
1.000 1.030 1.058 1.084 1.109 1.133 1.156 1.178 1.199 1.220 1.239 1.258 1.276 1.294 1.311 1.327 1.344 1.359 1.374 1.389 1.404 1.418 1.432 1.445 1.458 1.471 1.484 1.496 1.508 1..520 1.631 1.543 1.554 1.565 1.576 1.586 1.597 1.607 1.617 1.627 1.637
1.500 1.557 1.613 1.667 1.719 1.770 1.820 1.869 1.917 1.964 2.011 2.056 2.102 2.147 2.191 2.235 2.278 2.321 2.364 2.406 2.449 2.491 2.532 2.574 2.615 2.657 2.698 2.739 2.780 2.821 2.862 2.903 2.944 2.984 3.026 3.066 3.108 3.148 3.189 3.230 3.270
.51R3
,5170
.'in03
,7972 .8037 .SO98 .8156 .8213 ,8267 ,8319 ,8368 ,8416 .8463
.8507 .8549 .8589 .8629 .8667 .8704 .8739 .Si73 .%07 .8839 ,8869 ,8899 ,8929 ,8937 ,8985 .!IO12 .9038 ,9004
since Bokl will cancel. It is apparent then that a ratio of the times for certain fractions of reactant 10.0
5.0
2.5
0 2
3
4
TIME
RATIO.
5
6
of I/K versus the time ratio for 60 and 40y0 reaction and for 50 and 20% reaction, in terms of substance A. Fig. 3.-Plot
VOl. 74
remaining will be a function only of K . Table I1 shows calculated values of various time ratios. Figure 3 is a graph of the data of Table I1 with certain extensions. In applying these results TABLEI1 TIMERATIOS AS A FUNCTION OF K tso/lzo is ratio of time for 60% A reacting to time for 20% reaction, or where a = 0.4 and 0.8, respectively. l/K
fW/hO
kI/fll
flQ/fU
fOO/tSO
f&O/fZO
fM/fI*
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0 8.2 8.4 S.6 8.8 9.0 9.2 9.4 9.6 9.8 10.0
6.000 6.167 6.332 6.495 6.655 6.812 6.966 7.121 7.272 7.422 7.571 7.718 7.864 8.010 8.154 8.297 8.439 8.581 8.723 8.863 9.003 9.143 9.283 9.421 9.559 9.698 9.836 9.974 10.112 10.250 10.388 10.526 10.664 10.802 10.940 11.078 11.216 11.354 11.492 11.632 11.772
3.500 3.577 3.653 3.729 3.804 3.878 3.952 4.025 4.097 4.169 4.241 4.312 4.383 4.453 4.523 4.593 4.663 4.732 4.801 4.871 4.940 5.009 5.078 5.147 5.216 5.285 5.354 5.422 5.491 5.560 5.629 5.698 5.767 5.836 5.905 5.975 6.045 6.115 6.185 6.255 6.325
2.250 2.285 2.320 2.354 2.389 2.423 2.457 2.491 2.525 2.558 2.592 2.626 2.659 2.693 2.726 2.760 2.793 2.827 2.860 2.894 2.927 2.961 2.994 3.027 3.061 3.094 3.128 3.162 3.196 3.229 3.263 3.297 3.332 3.366 3.400 3.434 3.469 3.503 3.538 3.573 3.607
1.500 1.513 1.525 1.538 1.550 1.562 1.574 1.586 1.599 1.611 1.623 1.635 1.647 1.660 1.672 1.684 1.696 1.709 1.721 1.733 1.745 1.757 1.769 1.782 1.794 1.806 1.819 1.831 1.844 1.856 1.869 -1.882 1.895 1.908 1.920 1.933 1.946 1.959 1.972 1.985 1.998
4.000 4.079 4.154 4.225 4.294 4.362 4.427 4.489 4.550 4.609 4.666 4.722 4.776 4.828 4.878 4.928 4.978 5.025 5.071 5.117 5.161 5.205 5.249 5.290 5.330 5.369 5.407 5.445 5.485 5.521 5.558 5.594 5.629 5.663 5.697 5.731 5.765 5.797 5.829 5.861 5.892
2.333 2.366 2.397 2.426 2.455 2.483 2.511 2.537 2.564 2.590 2.614 2.638 2.662 2.684 2.706 2.728 2.750 2.771 2.792 2.812 2.832 2.851 2.870 2.888 2.960 2.925 2.943 2.961 2.977 2.995 3.012 3.028 3.044 3.060 3.076 3.091 3.106 3.122 3.137 3.152 3.166
to experimental data the procedure is to obtain several independent values of K from various experimental time ratios and to inspect the K values for constancy using data from either a single kinetic run or several runs. After obtaining K reference is then made to Table I where for several values of 7 0 reaction T can be accurately found which then by comparison with experimental f and Bo values will yield values of kl. Application of the Theory to Experimental Results.-A search of the literature revealed no data with which a satisfactory comparison with the theoretical results described here could be made. Either the initial conditions did not correspond to those considered here, namely, equiv-
March 5 , 1952
SAPONIFICATION OF ETHYL ADIPATEAND ETHYLSUCCINATE
1271
alent concentrations, or the extent of reaction Table IV shows the results of calculation of 1 / ~ was not sufliciently great. Therefore measure- and K I and Kz. Q! is first plotted as a function of I ments of the rates of saponification of ethyl adipate on a large scale. From a smooth curve drawn and ethyl succinate by sodium hydroxide were through the experimental points the times for 20% made. up to 60% reaction are read. From the t ratios the The usual titration procedure for following corresponding 1/ds are found by interpolating in saponifications was used with certain modifications. Table I1 and then after taking the average the Samples of the reaction mixture were pipetted a t values of r and k1 are then determined using Table measured time intervals into a solution containing I. an excess of potassium acid phthalate which served TABLE IV to stop the reaction. This acid was used rather CALCULATIONS OF RATE CONSTANTS FOR RUNNO. 1 ON than hydrochloric or other strong acid because it ETHYL ADIPATE provided fewer hydrogen ions to catalyze either Percentages the forward or reverse reactions and also because % Rx. 1, sec. compared t ratio l/X 20 605 60/20 6.661 it provided a better end-point in the titrations. 2.808 30 1060 60/30 3.802 2.795 Carbon dioxide was carefully removed from the 40 1690 60/40 2.385 2.777 original solvents and kept from the system during 50 2595 60/50 1.553 2.850 reaction and titrations by a stream of nitrogen. 60 4030 , 50/20 4.289 2.785 The sodium hydroxide used for the reacting system 50/30 2.448 2.750 and for the titrations was prepared carbonate-free by the reaction of sodium with water. The esters Average l/r-2.79 % Rx. 7 kl1. mole -1 sec. -1 were obtained as the best grade of Eastman Kodak 20 0.2582 0.08570 Co. chemicals and purified by an adsorption column. 30 0.4516 .08555 Their purities were tested by their saponification 40 0.7189 .08542 equivalents which yielded molecular weights of 50 1.108 .08574 174.4 and 202.2 for the succinate and adipate, 60 1.716 .08550 respectively, as compared with the calculated Average k1 = 0.0856 kz = 0.0856/2.79 = 0.0307 1. molecular weights of 174.19 and 202.25. The solmole-' set.-' vent used was a one to three volumetric ratio of dioxane and water which was made 0.2000 N Table V summarizes the rate constants for the in potassium chloride, this latter being included so as to swamp out any effects due t o changes in six runs. It may be concluded that the values are ionic strength that would occur as the diacid ion good to within 0.2 to 0.3%. Table VI shows how was produced. Further details of technique and TABLE V apparatus may be found in reference 1. SUMMARY OF RATE CONSTANTS FOR THE SAPONIFICATION OF DIESTERSBY SODIUM HYDROXIDE
Results Three kinetic runs were made on each of the esters. Run No. 1 on ethyl adipate as shown in Table I11 is typical of the results.
Run No.
TABLE I11 RATE OF SAPONIFICATION OF ETHYL ADIPATEWITH SODIUM HYDROXIDE IN DIOXANE-WATER-POTASSIUM CHLORIDE SOLUTION Run No. 1: initial concentration of reactants, 0.00996 N; normality of sodium hydroxide for titrating, 0.01057 N . Normality of potassium acid phthalate, 0.01497 N; volume of sample, 9.98 ml.; volume of potassium acid phthalate solution, 9.99 ml.: temDerature. 25.00'. Time in seconds
NaOH in titration, ml.
0 165 370 510 628 757 965 1180 1357 1671 2062 2468 3077 3664 4460 527 1
... 5.36 5.99 6.39 6.68 6.97 7.36 7.77 8.03 8.49 8.93 9.33 9.84 10.18 10.60
10.94
(I
[OH-], N
0.00996 .00931 .00864 .00822 .00791 .00760 ,00719
.00676 .00648 .00599 .00553 .00510 .00456 .00420 .00376 .00340
= [OH-]/ [OH-la
1.0000 0.935 .868 ,825 .794 .763 .722 .679 .651 .601 .555 ,512 .458 .422 .378 .341
Initial concn., N
Ai, 1. mole-' sec-1
kg, 1. mole-] s e c - 1
1 2 3
Ethyl adipate 0.00996 0.0856 .00986 .0856 .00902 .0855 Average 0.0856
0.0307 .0304 ,0305 0.0305
1 2 3
Ethyl succinate 0.00988 0.190 .00903 .190 .00823 .191 Average 0.190
0.0354 .0352 ,0352 0.0353
TABLE VI COMPAR~SON OF RATECONSTANTS FOR THE SAPONIFICATION OF DIESTERS BY SODIUM HYDROXIDE Temp., Investigators
Solvent
OC.
Ethyl adipate Ingold Water 20.3 Westheimer, Jones and Lad (extrapolated) Water 20.0 This research Water, , 25.0 dioxane, KCl Ingold Ritchie This research
ki liters
kz mole-]. mk-1
3.52 0.704 5.97 1.23 5.14 1.83
Ethyl succinate Water 20.3 12.05 1.253 Water 25.0 12.3 1.93 Water, 25.0 11.4 2.12 dioxane, KCI
ARTHURA. FROST AND XARREN C. SCHWEMER
1272
these rate constants relate to those of other investigators. It must be realized that because of differencesin solvents and temperatures there is no direct comparison. Acknowledgment.-We are indebted to the Standard Oil Company of Indiana for a fellowship held by the junior author during a portion of this research. Appendix The integral of equation (11) may be evaluated for certain ranges of K as follows: Case I. 0 < K < '/2 Let K = m/n where m and n are both integers and n >2m. This situation is most common since for the saponification of symmetrical diesters k, will always be a t least 2kz purely on statistical grounds and will be greater than 2 k 2 due to repulsive effects between the hydroxide ion and the half-ester ion. By substitution
VOl. 74
tables of arctangent and natural logarithms were particularly useful.* Case I1 : l]2 < K < , K # 1 Let K = m !n as before, but let x = n
n 2m -)n-m - n
1
(
i3-'In; then
with the lower limit now
The integral over x is of the general form
which for p
For p
> q may be reduced as
< q the integral is7 1
xq
- 1
dx = - log 4
(X
-
1)
+
where
(17)
with q
...q-1
2r - s;
=
The integral over x is of the general form
1z
s = 0 or 1; q > 2 ;
- - log
v - 4
which for p
.y
> (I may be simplified to
For p < p the integral may be evaluated by the use of partial fractions and the result is7 dx=(-l)pzlog(x 4
(x2 - 2 x
cos -
2 Qv n
p
=
0,l
+ 1)
- . drctaii -
When m > n the exponent in the denominator becomes negative and the integral may be expressed as
+ 1) Special Cases For K = 0 the reaction is a simple second-order reaction with the solution
uith
(I =
2r
+ >;
3
= 0 2-
or 1; p
=
0, 1, . . . y - 1
i - 1 - 2 x cosd:' -a Y
+ 1)
For ti = i/z the reaction behaves as one of siiiiple second-order as far as A is concerned
Lv t 1 7
=
1
CY
For
K
=
The calculated values of Table I were obtained in this way for K = l/2, 4/9, 2/5, l / 3 , 2/7, l / 4 , ' / s , '/E., where l/6, l/7, l / g and ]/IO. An interpolation was then made for every 0.2 ._-___for 1/ti from 2.0 to 10.0. In these calculations (8) "Project (7) W Grdhner ani1 N. IIofrriter, "lnte~ralt.tfc1," 1 , SpringerVcrlag. Vienna, 1940.
- I
alld
012
=
p
1 7
=;Jl$dr x = 2
- ltlp
for Computation of ?rfathematicdl Tables," Federal Works Aeency, A S . Lowran, Technical Ilirector, 1841-19 12, Yew \-ark, h7 \'.
March 5, 1952
COMPOUND
FORMATION IN SOLUTIONS OF IODINE AND OLEFINS
or
1243
simple second order but here with the solution 1 7 = -
[Ei (2 - In 8)
- Er (2)l
e2
where Ei(x) is the Integral Logarithm function. For K = 03 the reaction again behaves as one of
r = 21( p1- 1 ) EVANSTON, ILL.
anda=P RECEIVED SEPTEMBER 19, 1951
[CONTRIBUTION FROM THE CHEMISTRY DEPARTMENT, BROOKHAVEN NATIONAL LABORATORY ]
Experiments on Compound Formation in Solutions at Low Temperatures. Olefins1
Iodine and
BY SIMON FREED AND KENNETH M. SANCIER Absorption spectra of solutions of iodine in olefins a t temperatures from 77 to 230°K. show the presence of ultraviolet bands which match those of iodine in aromatic hydrocarbons previously reported. These bands had been proved characteristic of 1: 1 molecular addition compounds between iodine and the hydrocarbons. Such addition compounds are formed by iodine with each of the following olefins examined either pure or dissolved in propane: propene, cis-butene-2, trans-butene2, butadiene-1,3, 2-methylbutadiene-l,3. Cyclopropane also combines with iodine in the same way in agreement with the double-bond character often attributed to it. The addition compounds form reversibly. However, they require activation energy. At higher temperatures another type of reaction sets in decolorizing the iodine solutions, probably the halogenation across the double bond. Especially good evidence that the addition compound is an intermediate in the hdlogenation of the double bond is furnished by 2-methylbutadiene-l,3 and iodine. The heats of formation of the addition compound vary from 200 to 500 cal./mole. The nature of the bond in these compounds is discussed.
It is to be expected that compounds may form and persist a t low temperatures which a t higher temperatures would not be present in detectable quantities. The lower temperatures may therefore make possible the observation of weakly bonded substances and permit the study of reaction intermediates. We report here phenomena of this kind in reactions between olefins and iodine. In these systems we also have examples of the reduction to be generally expected in the number of competing reactions when the temperature is reduced. Solutions of iodine in organic solvents have often been classified into those which give purple and those which give brown solutions, with the former exemplified for our purpose by iodine in hexane and the latter by its solutions in aromatic solvents such as benzene. Benesi and Hildebrand2 have shown that in addition to the absorption band in the visible region, the brown solutions possess an extremely intense absorption peak in the ultraviolet and that this peak is to be ascribed to a 1: 1complex between iodine and the aromatic hydrocarbon. They regarded such a complex as a Lewis salt between the generalized acid and base, iodine and aromatic hydrocarbon, respectively. Mulliken3 viewed the intense absorption peak in the ultraviolet as a forbidden transition between the energy levels of the aromatic conjugated doublebond system made allowed by the presence of iodine. Our experiments to discover whether olefins also enter into what we shall call addition compounds were carried out a t low temperatures to increase their concentration and to avoid the well-known reactivity of halogens for the “double bond” a t room temperature. (1) Presented a t the American Chemical Society Meeting, Chicago, Ill., September, 1950. Research carried out under the auspices of the U. S. Atomic Energy Commission. (2) H. A. Benesi and 5. H. Hildebrand, THISJOURNAL, 71, 2703 (1949). (3) R. S. Mulliken, ibid., 72, 600 (19.50). At the Chicago Meeting of the A.C.S., September, 1950, he presented a more general interpretation.
A double-beam Cary spectropbotometer was employed for measuring the spectra of the solutions in fused quartz tubes with light paths of 1 and of 8 mm. To eliminate theoptical effect of the fused quartz, refrigerant, solvent, etc., the tubes were also measured with all but iodine present. The iodine dried over phosphorus pentoxide was dropped into a tube previously filled with dried nitrogen. The tube was then evacuated while the iodine was kept at 193’K. By allowing the temperature to rise, the iodine was sublimed several times onto the surface of another part of the tube kept a t 193’K. and any released gases were pumped off. The final sublimate was obtained as a film. The propane4 was condensed upon the iodine a t 193’K. Some time was required to form the purple solution. The corresponding condensation of propene led a t once t o a brown solution a t the interface between the iodine and propene which quickly diffused into the body of the liquid. The solutions in propane and also in propene were obtained easily as supercooled fluids a t the temperature of liquid nitrogen. Solutions in mixtures of these hydrocarbons appear t o be stable fluids even below the freezing point of nitrogen. All our operations were carried out in dim light t o reduce any photochemical activity. Propene and Propane.-At 230°K. the color of iodine in propane is purple with an absorption peak a t 5200 A, while its solution in propene is amber with a peak a t 4800 A. At 77°K. both solutions have about the same brown color aad the absorption peak of the propene solution is a t 4400 A. We are uncertain of the position of the peak of the propane solution because a t our concentration there was evidence of colloidal iodine at this temperature. The area “under” the visible band of iodine-propane changes little with temperature but that of iodine-propene increased two to threefold when the temperature was reduced from 230 t o 77’K. (Fig. 1). In the ultraviolet region, the iodine-propane solution a t 230’K. exhibited no structure down to 2100 A. hut the iodine-propene solution showed a band a t 2750 A. about ten times as intense as its absorption in the visible. Both the intensity and wave length of the ultraviolet absorption match those of iodine in aromatic solvents as given by Benesi and Hildebrand and there seems no doubt that it arises from a corresponding iodine-propene addition compound. A t 230”K., our highest temperature, an irreversible reac(4) All the hydrocarbons except cyclopropane were research grade (Phillips Petroleum Co.) further fractionated several times by distillation. Cyclopropane was obtained from Matheson Co., several times distilled from its solution of iodine. ( 5 ) S. Freed and C. J. Hochanadel, J . Chem. Phys., I T , 664 (1949).
