POLARIZATION OF CHARGE-TRANSFER BANDS
July 20, 1964
A factors can be rationalized in terms of a loose complex in which enhanced entropy of activation arises from low frequency stretching and bending modes involving the I atom. Heat of Formation and Bond Strength The observed value of El, given by (j),may be used to derive the bond dissociation energy, D(i-C3H7-H), if E z is known. Flowers and BensonGbhave reported a value of E2 - E3 = 0.8 f 1.0 kcal. for the Me radical, Hartley and BensonIO give E2 - E3 = 0.9 kcal. for the C2H5 radical, and O'Neal and B e n ~ o n ~give ~ , ~E 72 - E3 = 1.5 kcal. for the CH3C0 radical. In the absence of any experimental finding on the isopropyl radical, we may assume it to be between those for E t and CH3C0, with a chosen value, Ez - E3 = 1.1 kcal. If E3 0.2 kcal.,'O then E2 = 1.3 kcal./mole and we may further deduce AH1,20 (6OOoK.) = E1 - E2 = 23.7 kcal./mole, at the middle of the temperature range of the present work. From the thermodynamic data given in Table 11, AC,l,2 = 2.1 cal.,/mole which leads to AHl,zo (25') = 23.1 kcal./'mole. Since the bond dissociation ~ very energy, D0(H-I)z50 = 71.4 k c a l . / m ~ l eis, ~known accurately, we can use the relation D'(i-CsHrH) - Do(H-I)
= AH'1.2
(25')
(27) H . E . O'Neal a n d S . W. Benson, J . Chem. P h y s . . 8 7 , 540 (1962). (28) Cpo for t h e I-CSH?radical is assumed t h e same a s for CsHs, and Cpo (HI) is taken as t h e mean of t h e values a t 25', 7.0 cal./mole O K . , and a t 600°K., 7.23 cal./mole OK. (29) S a t i o n a l Bureau of Standards Circular No. 500, U. S. Govt. Printing Office, Washington, D . C . , 1952.
[CONTRIBUTION FROM
THE
2777
to find D'(i-C3H7-H) = 94.5 kcal./mole The heat of formation, AHf' (25'), for the i-C3H7 radical may also be computed from thermodynamic quantities given in Table I1 from the relation AHro(i-C3H7) = AHf'(C3Hs) AHt'(H)
+ D(i-C3H7-H)
and the value obtained is AHfo(i-C3H7) = 17.6 kcal./ mole with an uncertainty of 5 1 kcal.,!mole. Electron impact m e a s ~ r e m e n t shave ~ ~ led to D ( i C3H7-H) = 94' f 2 kcal./mole. From competitive bromination of pairs of alkanes, Fettis, et ~ 1 derived . ~ the bond dissociation energy, D(i-C3Hi-H) = 93.1 kcal./mole. In arriving at this value they used the results of Kistiakowsky and van Artsdalen3? on the bromination of CH3Br and Polanyi's relation E = CYAH C. Though the agreement of our data with these values is good, it is difficult to reconcile with the high precision claimed by Fettis, et al. It is our own feeling that the scatter in their reported rate constants and the uncertainties associated with both the brornination work and the Polanyi relation introduces an uncertainty of the order of f2 kcal. in their value.
+
(30) D. P. Stevenson, Trans. Faraday Soc., 49, 867 (1953) (31) G . C. Fettis and A. F. Trotman-Dickenson, J . Chem. Sac., 3037 (1961). (32) G . B. Kistiakowsky and E . R. van Artsdalen, J . Chem. P h y s . , l a , 469 (1944).
WHITMORE CHEMICAL LABORATORY, THEPENNSYLVANIA STATE UNIVERSITY, UNIVERSITY PARK, P E S N S Y L V A N I A ]
Polarization of Charge-Transfer Bands BY MIHIR CHOWDHURY AND LIONELGOODMAN RECEIVED SEPTEMBER 13, 1963 T h e polarization of the charge-transfer ( c . t . ) band of a-molecular complexes of tetrachlorophthalic anhydride ( T C P A ) and aromatic hydrocarbons has been determined by a photoselection method correlating the polarized c , t . absorption with polarized triplet emission of the donor. I t is found t h a t the c . t . band is polarized predominantly along the intermolecular axis and perpendicular to the plane of the ring, indicating t h a t a welldefined structure exists even in the excited state, and t h a t the change in the permanent dipole moment, on excitation, is mainly responsible for the intensity of the c.t. band in these complexes. The same technique has been utilized to show t h a t the phosphorescence of quinoline is out-of-plane polarized, hence coafirming Dorr and Gropper's assignment of the phosphorescent state of quinoline as a ( a , T * ) triplet. This kind of experiment should be utilizable for structure determination of c.t. complexes in solution.
Introduction
p =
Xulliken2 has shown that the electric dipole transition moment, p, of the charge-transfer transition frequently appearing in molecular complexes with an excited state predominantly described through $e
=
+
a*$bi(D+A-) - bi*$oiD,A) Zldi *$d,(D*,A) (a* >> b*,
=
u$o(D,A)
Zbi$b,(D+A-)
(U
(la)
>> b i )
(lb)
is given (in a.u.) by (1) Supported by a grant from t h e Air Force O 5 c e of Scientific Research. ( 2 ) K. S. Mulliken, J . A m . Chem. S O L .74, , 811 (1952); J. chim. p h y s . , 61, 20 (1964).
S(a$o
+
bl$bl
C b,$b,)r(a*$b,
- h*$o
IZ1
a*blS($b,r$b, - $or$O)d7 bibi*)S$or$b,d7
+
(UU*
+ Zd,*$d,)d7
-
+ Zad,*S$or$d, d7 -k a*b,S$b,,r$b,dT (2)
di*)
and a ground state described through $g
S$&d7=
i f 1
In the above relations the contributions from (D-A+) have been neglected.* The first term in eq. 2 represents the change in permanent dipole moment of the complex on excitation, and was calculated by hIulliken2 to be of the same order as the transition moments obtained from observed c.t. bands in n-molecular complexes This moment has a predicted direction along the line joining the centroids of the donating and accepting or-
~
~
2778
MIHIRCHOWDHCRY AND LIONEL GOODMAN
bitals. For r-niolecular complexes having a sandwich structure, the charge centroids of D + and X- are expected to generate a moment with a predominant component perpendicular to the plane of the sandwich and hence predominantly perpendicular to both the donor and acceptor molecular planes. This term is predicted to be the leading term? and therefore to give rise to predominantly out-of-plane polarization for the c . t . transitions. The direction of the moment given by the second term in ey. 2 will depend upon t h e symmetry properties of the orbitals. l f the structure o i the complex contains an axis 12f symmetry through the acceptor and donor species and provided t h a t #o and in ey. l a and l b belong to the same spinorbital representation in the symmetry group of t h e complex, then this term clearly should be also directed along the intermolecular axis. In the case that #o and IC/,>, do not belong to the same representation, bl will be zero, so t h a t the contribution from the first term in eq. 2 will be negligible, leading to a transition niunient not restricted to the intermolecular axis. The third and fourth terms in eq. 2 represent the borrowing of intensity by the c . t . band from the donor bands and must be directed in-plane in the case of Pcomplexes. The importance of this term has not been established, though Murre13 pointed out its possible large contribution in the case of contact charge transfer The mixingof a (x,x*)state with the c . t .state may also occur if the complex has low symmetry, or through vibronic interaction. Xn example of this latter possibility was shown by Xlbrecht4 for the case of the intramolecular c . t . band in tetramethyl-p-phenylenediamine. This paper describes an attempt toward experimental determination of the polarization of the c.t. band in sandwich-type complexes. I t is clear t h a t such an experiment could determine the important sources of intensity in these transitions. There have been two such attempts before. In 1952, r\iakamoto5 studied the dichroism of quinhydrone and related complexes, and concluded t h a t the perpendicular component (to the plane) has greater intensity than the in-plane component. Recently, Lower, Hochstrasser, and Reid,6 in a study of the crystal (absorption) spectra of the anthracene-TNB complex, found t h a t the c . t . band is again polarized predominantly out-of-plane. There are, however, serious difficulties in analyzing the polarization of single crystal spectra. In addition to these ditliculties, complexes frequently form only microcrystals. Some complexes do not exist in the solid s t a t e ; of those t h a t do, the crystal structures of a relatively small number have been analyzed fully, specially for cases with unsymmetrical acceptors. There are several examples where the molecules are tilted a t odd angles, with each donor interacting nonuniformly with inore than one acceptor, resulting in crystal absorption spectra differing markedly from those in solution.: These factors make i t desirable to carry out an experiment determining the polarization of the c . t . band in rigid solution. McGlynn,aaBriegleb ,'ir J S S l u r r e l . J Arti ( h i , v r .So E Orgel and R S Mulliken, J . A m Chem. S O L 79, , 4839 (1$2.?7'
July 20, 1964
POLARIZATION OF CHARGE-TRANSFER BANDS
speculation is t h a t the T C P A overlaps two adjacent rings on one side of phenanthrene rather than forming a symmetrical complex with less overlap. X-Ray investigations show several examples of asymmetric organic *-molecular complexes.'@ If this asymmetry is true in these cases, the line joi?ing the centroids of acceptor and donor orbitals will not be quite perpendicular t o the plane of phenanthrene or chrysene (giving p a parallel compone n t ) and hence the degree of polarization is expected t o be somewhat reduced. The parallel polarization of the donor T + S emission relative to the c . t . absorption observed in the above experiments clearly requires for these three polyacene-TCPA complexes a predominantly out-of-plane polarized c.t. band, and therefore these c.t. transitions cannot derive much intensity from the donor or acceptor transitions. Quinoline-TCPA Complex.-The polarization of the quinoline phosphorescence was determined b y Dorr and Gropper,llb using direct excitation, t o be negative relative t o both the IL, and ' L b 7 --L T* excitations; they therefore concluded t h a t the transition moment must be perpendicular t o the plane of the molecule analogous t o the phosphorescence polarization of the parent hydrocarbon.lb Unfortunately, the absorption bands of quinoline are not distinct, and hence the polarization of the quinoline phosphorescence cannot be determined completely unambiguously by direct excitation. An extra band can be created, however, by forming a c . t . complex with T C P A , and the polarization of the T + S emission then determined by extending the hydrocarbon results t o quinoline, ; . e . , assuming the c.t. band does not change polarization in going from naphthalene-TCPA t o quinoline-TCPA. In quinoline-TCPA complex, the quinoline behaves a s a *-donor, as shown by the close similarity of the c.t. band with t h a t of the naphthalene-TCPA complex.2o T h e diffuse r-acceptor orbital of T C P A can better overlap the *-donor orbital than the nitrogen n-orbital. T h u s it is probable t h a t the complex has a sandwich structure, with the c . t . band polarized out-of-plane analogous t o t h a t of naphthalene. Figure 6 shows t h a t the degree of polarization of the phosphorescence is +0.35 with respect t o excitation into the c . t . band of the complex, confirming Dorr and Gropper's direct excitation experiment and showing t h a t the phosphorescence does not arise from a n ( n , T * ) triplet state. In this case t h e experiment does not rule out important contributions from donor ( n -+ T*) transitions t o the c.t. intensity. The above method of determining the polarization of phosphorescence b y excitation into a c.t. band may be useful in cases where a direct excitation procedure is not possible. I n order to utilize the method, the following conditions should be satisfied: ( 1 ) the c.t. band needs t o be distinct from the absorption band of the complex components, ( 2 ) the triplet level of the donor should be below the c . t . triplet level, and (3) the radiationless deactivation rate should be low.
Conclusions and Summary The strongly parallel polarization for c.t. excitation of the phosphorescence of the polyacene-TCPA com(19) S. C. Wallwork, J. Chem. Soc., 494 (1981). ( 2 0 ) M .Chowdhury and S . Basu, Trans. F a r a d a y SOL.,66, 335 (1900).
2781
460 500 t WAVELENGTH (my).
540
Fig. 6.-Polarization of quinoline phosphorescence emission in quinoline-TCPA complex relative to escitation into the c . t . band ( a t 405 mM).
plexes requires a c.t. transition moment predominantly perpendicular to the donor and acceptor molecular planes, and thus strongly supports Mulliken's original ideasZ on intensities of c.t. bands. I t eliminates as major sources of intensification borrowing from the donor or acceptor bands and contact charge-transfer interactions. The kind of experiment described here should be applicable to structure determination of c.t. complexes in rigid solutions. Acknowledgments.-Thanks are due to Professor R. S . Mulliken and Dr. V . G. Krishna for helpful criticisms and discussions and to Dr. R . Shimada for help with the experiments.
