Sept. 20, 1964
CHARGETRANSFER BANDSI N AROMATIC COMPLEXES
[CONTRIBUTION FROM T H E
DEPARTMENT OF CHEMISTRY,
UNIVERSITY O F
3611
BRITISHCOLUMBIA, \.ANCOUVER 8, CANADA]
Multiple Charge-Transfer Bands in Complexes Involving Aromatic Donors BY
E. M.
VOIGT'
RECEIVED DECEMBER 3, 1963 The charge-transfer spectra of substituted benzene donors with tetracyanoethylene a s acceptor were investigated. Sonietiines single, and other times double charge-transfer bands were observed. These were examined on the basis of Orgel's theory which assumes that the two bands derive from a splitting of the degeneracy of the ground state of the benzene cation on substitution. Thus, the charge-transfer energies are expected t o be proportional to the first two ionization potentials of the donor component of the complex. The results obtained in this work quite generally confirm this. However, a simple approximation using an equation of the type hu = @ID was found inadequate: evaluation of the experimental data with this equation gave variable perturbation parameters, instead of the constants predicted by Orgel's theory.
Introduction Early observations2a'bof two charge-transfer (c.-t.) absorption bands in n-complexes having substituted benzenes as the donor component were attributed by Orge13 to the fact that the transferred electron might come from either of the two levels arising when the degeneracy of the E1 level of benzene is removed by substitution. Orgel used a simple first-order perturbation calculation to evaluate both the shifts in position of the c.-t. absorption from that found in the complex of benzene with a given acceptor and the splitting resulting from substitution of the benzenoid donor in various possible ways. Although Orgel showed his results to be in general agreement with the experimental data available at that time, few additional data have been and no attempt has been made to test the rather crude approximations of the Orgel theo r y . I n this paper some of the deviations are considered and accounted for satisfactorily in a qualitative manner.
Theory In benzene the highest occupied level is doubly degenerate. On ring substitution, the degeneracy is re. moved; the two orbitals now split to an extent depending on the number of substituents, their nature, and relative positions on the ring. The resultant two electronic states have two different ionization energies. Orge13 has treated this problem theoretically by a firstorder perturbation calculation and obtains the following expression for the energy difference between the ionization potential, Z g , of benzene itself and the two ionization potentials, ZIand I,],of the substituted molecule. 11.11
=
I B
- nYS
- pl,IIeS
hVCT = I D v
-
1 ) e p a r t i n e n t oi Chemistry, University VI California, Berkeley 4 , Calif. !a) T S m i t h Ph I ) ilissertatinn, T h e University Rr > I . I f there are several, mutually independent substituents on the ring, Orgel’s expression (2) may be expanded to give It.11
= IB -
C niYsi
-
PI.II~s
i
(4)
where the second term sums over the i substituents and in the last term is a weighted average of the cs for the different substituents. For the charge-transfer energies arising from benzene donors with mixed substituents, the ionization potentials of eq. 4 give an expression analogous to that found for n equal substituents in eq. 3 . is
hv
Ii.11 = I B
niYs. - pr,IIes
-
(5)
i
Furthermore, the two charge-transfer bands deriving thus from ionization of the transferred electron from the two uppermost levels of a substituted benzene donor are associated with two isomeric complex configurations, according to The two donor orbitals differ in the position of their nodal planes. Whiie both these planes pass through the center of the ring and lie perpendicular to the molecular plane. the symmetric orbital has a transverse nodal plane (i.(,., cutting across two opposite C-C bonds o f the ring) and the antisymmetric orbital has a longitudinal nodal plane which passes through the 1,S-C atoms. I n donors with electron-releasing substituents the symmetric orbital which encompasses the substituent(s) is expected to have the lower ionization potential. As regards the acceptor, it is assumed here that the transferred electron will always go into the same lowest enipty acceptor orbital, i.e., excited anion acceptor states are neglected. This appears a reasonable assumption for tetracydnoethylene (TCNE), where Pen(101
K 5. llolliken. Rrr.
w a s . c h r m . . 7 6 , 8.15 (1956)
Fig. 2.-Model
R, of a TCNE-monosubstituted
( X ) benzene
c.-t.complex.
fold and Lipscomb’’ calculated an energy difference of 2.07 e.v. between the lowest and second lowest empty MO’s. Also, they found that the first unoccupied’orbital is but slightly antibonding and has bZgsymmetry (in D d . For stable T-T complexes the most stable configurations have been shown to be those where donor and acceptor lie above each other with their molecular planes more or less parallel. 10312--L4 Thus we expect at most an over-all C,-type complex symmetry. Considering then the symmetries of donor and acceptor orbitals in C, for TCNE-substituted benzene complexes and applying Mulliken’s maximum overlap condition, lo two most favorable complex configurations are obtained. They are depicted to scale in Fig. 1 and 2, in which the molecular dimensions of T C N E were taken from Bekoe and Trueblood,15those for benzene from MullikenIfi (except for X, since the distance C-X varies with X ) . Model Ry (Fig. 1) involves the symmetric donor orbital and model R, (Fig. 2) the antisymmetric orbital. Although Fig. 1 and 2 show two definite molecular configurations, we must emphasize that these are not the only configurations allowed, nor will they be expected as sole contributors to the c.-t. bands, especially a t room temperature and in solution. Mutual rotation of the two components and displacements from the center of the complex either along the x or y axis or obliquely may contribute substantially to the c.-t. absorptions.’7 For particularly large and1 or strongly polar substituents a tilting away of the two planes from each other will also be expected. (11) B. K Penfold and W . N . Lipscomb, A d u C r y s l , 14, 589 (19611 ( 1 2 ) S . C . Wallwork, J . Chem Soc., 494 (1961)
(131 A . W. Hanson. A d a C r y s l . , 16, 1147 ( 1 9 6 3 ) . ( 1 4 ) S. P. McGlynn and J. 11 Boggus, 1. A m Chem. S O L . , 80, 5096 (1958) ( 1 5 ) I). A . Bekoe and I(. N Trueblood. 2 . Krislull , l l S , 1 (1960) (16) R S . Mulliken. J . A m . Chem. Soc.. 74, 811 (19521 (17) 1,. E Orgel and R . S . Mulliken, i b i d . . 1 9 , 4839 (1957)
CHARGE-TRANSFER BANDSIN AROJUTICCOMPLEXES
Sept. 20, 1964
36 13
TABLEI CHARGE-TRANSFER SPECTRA OF METHOXYBENZENB-TCNE COMPLEXES~
--
-Transition imar,
103 cm.-l
Donor
IC-A f l / r , 10'
cm. -1
Acceptor:
-
--Transition ;ma.,
ioacm.-1
11-' A?'/'z, 10' cm.-L
TCNE -Separation Af(Buu-I), p~~
10' ern.-'
from b e n z m t Ac(Benz-11).
io' -.
-1
... ... .. ... 25.8 f O . l 5.8 Benzene (reference) ... 25.8 f 0 . 1 ... 0.83 ... 19.6 f 0 . 1 ... .4nisole, obsd. - 0 . 3 f 0.25 6.5 f 0 . 2 5 26.1 f 0 . 1 5 5 . 5 0.77 6.2 f 0 . 2 19.6 f O . l 5.0 Analyzed 23.2 ... 0.58 ... 16.9 f 0 . 1 ... o-Dimethoxybenzene, obsd. 2 . 5 f 0.3 6.4 f 0 . 3 23.3 f 0 . 2 5.8 0.5 8.9 f 0 . 2 16.9 f 0 . 1 5.3 Analyzed ... 20.8 ... 0.M ... 18.35 ... m-Dimethoxybenzene, obsd 3.1 f 0.6 4 . 8 f 1.1 22.7 f 0 . 5 5.8 0.51 7 . 9 f 0.5 17.9 f 0 . 4 5.2 Analyzed 0 . 5 f 0 . 2 10.6 f 0 . 1 5 26.3 f 0 . 1 6.0 0.56 1 0 . 1 f 0 . 1 5 15.7 f 0.05 5.9 p-Dimethoxybenzene ... ... .. 6 . 4 f 0.25 ... ... 19.4 f 0 . 1 5 6.3 1,2,3-Trimethoxybenzene ... ... ... ... .. 7.7 18.1 ... 1..3.5-Trimethoxvbenzeneb . 1,2,4-Trimethoxybenzene, obsd. 1 4 . 6 ... 22.7 ... 0.33 ... ... ... Analyzed 14.6f0.1 6.1 23.4f0.4 5.7 0.27 11.2f0.2 2.4f0.5 8.8f0.5 1,2,4,5-Tetrameth~xybenzene~1 2 , 5 ... 22.7 ... .. 13.3 3.1 10.2 0 Solvent: CHzClz; 22". C -t. energies from ref. 8. A:/* = width at half-maximum intensity of the absorption band. qI = ratio of optical densities of transition 11-to I : OD, 11/ODrnaxI.
Nevertheless, for charge-transfer complexes as strong as the TCNE-substituted benzenes,6 and in which donor and acceptor have such favorable relative molecular sizes, as well as favorable symmetries of the two interacting MO's, such complex orientations will be preferred where *-overlap between the two components can occur over large molecular regions of like sign and if symmetry rules are obeyed. Experimentally there is some evidence that such two isomeric complexes exist in ~ o l u t i o nl8. ~ ~
Experimental Reagents.-Tetracyanothylene (Eastman Kodak, White Label) was recrystallized twice from. chlorobenzene and twice sublimed. All donors used were obtained commercially and of best available grade. Solid donors were all recrystallized several times from either ether or 95y0 ethanol and subsequently dried under vacuum. Liquid donors, when available as spectrograde or B.D.H. Analar, were used directly; otherwise they were fractionated or redistilled under reduced pressure. Special Data.-Thioanide ( K & K Lab., Inc.) was purified by vapor phase chromatography; 1,2,3-trimethoxybenzene ( K & K Lab.. Inc.), after four recrystallizations from 95Y0 ethanol and subsequent drying under vacuum, gave a m.p. 47-48' (lit.lgm.p. 47'); 1,2,4-trimethoxybenzene (Columbia Chemicals) was found quite impure; after repeated distillations a t reduced pressure, a fraction of b.p. 129-130" (6 mm.) and 7 t P O ~1.5339 was used (lit.*O b.p. 242.5' and @D 1.5309); hexafluorobenzene (Pierce Chemical Co.) was found quite pure by vapor phase chromatography, b.p. 80.5-81O.21 The solvent used was dichloromethane (Eastman Spectrograde) for all measurements. Spectra.--All spectra were taken on a Cary Model 1 4 spectrophotometer. A pair of tared, stoppered quartz cells of 1-cm. path length were used and the measurements made with the cell compartments thermostated a t 22". T h e complex component concentrations (-lo-' to lo-' M TCNE and t o lo-' M for the donors) were adjusted to give c.-t. bands of 2ptical density -!3.4-0.5. -4s a reference we used a donor-dichloromethane solution of the same concentration as that used for complex preparatiw . Under above conditions 1 1 molecular complexes Are formed .s 'The generally broad, overlapping charge-transfer bands were ornpvsrri ' 1 1 t h e manner described in detail elsewhere.22 having single transitionp only, quite generallv ier of tw:i clasws: $'ai proper, singlt c.-t. ahwrptinns )r I b ) t h t y rrprrscnt the sum oi two strorgi:. overlapping '
~~~
> < u s , Z p h y s i k . C h e m . (Fraakfurt), ; I % ) $>.B-ieglcb, J. Czcka:la, a~:rii;, 316 (19fii, 1 1 9 ; I3 i'hapmail C . P e l k i n . ond R. Robinson, J Chcm. S O L ,3028 tl92ij. (201 I i . Susz, Helu. C'him. A c t a . IS, 1359 (1936) 121) Y Desirant, Buil. JOG. ' h i m . Beiges, 41, 759 (1955). (22) E M . Voigt and C. Reid, J. .4m. Chem. S O L . , in press.
so,
.\.
bands. T C N E complexes of the former class were always found to be asymmetric and they followed Briegleb's empirical halfwidths relationship.** Complexes belonging t o the second class, however, did not only have exceptionally large half-widths, they also deviated in their band shapes from those of class a."
Results and Discussion Tables I , 11, and I11 give the observed and analyzed c.-t. spectra of a number of substituted benzene complexes with TCNE. That Orgel's general predictions are borne out is shown by the following observations (see points i to iv under Theory above). (i) Without exception, a maximum of twq c.-t. bands was observed in the region under study as expected if the multiplicity arises from the splitting of the doubly degenerate benzene level. Single bands were observed only if the substituents were of the kind expected to interact only weakly with the ring, or in cases where Orgel's parameters PI and pII were identical. (ii) The second c.-t. band of some substituted benz e n e T C N E complexes occurred a t a position close to that of the benzene-TCNE complex itself. Such complexes had the following donor componentsmonosubstituted : anisole, thioanisole, styrene ; disubstituted : p-dimethoxybenzene, p-methylanisole, ochloroanisole, o-bromoanisole. The two o-complexes in the above list are exceptional in that their second c.-t. bands lie close to that of the benzene complex only by accident, ie., due to the particular values of cs and ys in these cases. Apart from these, the list contains only donors for which P I ] = 0, the very condition that benzene orbital "11" is perturbed only by the small term involving ys (eq. 3 ) . (iii) The band separations, Afimax(II-I), observed on varying the number and positi:w of donor substihent groups (OMe, C1, and Br, respectively) were: paradisubstitated donor > ortho- and meta-disubstituted donor '5 monosubstituted donor. Again. this agreed with what was predicted from the difference Ap(II-: : para-disubstituted > ortho- and meta-disubstituted = monosubstituted donor. (iv) Charge-transfer shifts and band ser,aratior.s ir! complexes with mixed substituent donors were r'ound (23)
(IQfiO,.
f.;
Hrieeleb and J. Czekalla, 2 . physik. C h r m . :Frankfur:> 2 4 , 37
E. M. VOIGT
36 14
Vol. 86
TABLE I1 CHARGE-TRANSFER SPECTRA OF HALOBENZENE-TCNECOMPLEXES~
_ - _ ~ ---Transition Donor
vmSx, 101 cm
Benzene (reference) Chlorobenzene, obsd. Analyzed Bromobenzene, obsd. Analyzed p-Dibromobenzene, obsd. Analyzed o-Dichlorobenzene, obsd. Analyzed m-Dichlorobenzene, obsd. Analyzed p-Dichlorobenzene, obsd. Analyzed 1,3,5-TrichIorobenzene p-Difluorobenzene, obsd. Analyzed Hexafluorobenzene Iodobenzene, obsd. Analyzed p-Diiodobenzene, obsd. Analyzed 0 Solvent: CHZCIZ; 22'
-1
258fO1 264fO1 249f03 254fO2 246fO2 238f01 238fOl 268fO 1 25OfO3 276f01 254fO4 249fO1 245fO1 28 I f 0 1 25 6 254fO3 306fO2 224fO1 223fOl 217f02 217f02
I--Transition A C ~ / S ,101 cm - 1 i,,, 101 c m
58 79 46 82 48 54 79 47 78 46 54 65 48 45 46 55
282fO3 297fO2 295fO2
I---
Benzene (reference) p-Meth ylanisole m-Methylanisole, obsd. Analyzed o-Methylanisole, obsd. Analyzed p-Chloroanisole o-Chloroanisole, obsd. Analyzed p-Bromoanisole o-Bromoanisole, obsd. Analyzed p-Fluoroanisole p-Phenylanisole, analyzed o-Phenylanisole, analyzed Phenetole, obsd. Analyzed Thioanisole, obsd. Analyzed Styrene, obsd. Analyzed p-Chlorotoluene, analyzed o-Chlorotoluene, obsd p-Bromotoluene, analyzed o-Bromotoluene, obsd. 1,2,3-TrimethyIbenzene\ 1,3,5-?'rimethylbenzene p-Iodoanisole p-Iodotoluene a Solvent: CH2C12;22".
1
-1
258f01 178fO1 188fO1 188fO 1 183fO1 183fO1 19 5 f 0 1 203fO1 201f02 193fOl 201fOl 197fO2 199fO1 166fO2 182f02 193fOl 193fO1 17 5 f 0 1 175fO1 20 7 f 0 1 208fO2 223fO4 247f02 218f03 243*02
cm
101 cm.
~
from b e n z e n e -
Ai(Benz-111, 10' c m . - 1
-1
_
_
_
Separation 11-1 AYmax, 10' cm - 1
09
0.9f 0.4
-25f05
34107
59
08
... 1.2f 0.3
-24fO4
36f05
6 0
0 99 0 93
... 2.1 f 0.2
37f03
58f03
... 0.8f 0.3
-29fO5
37f07
0.4f 0.4
-2.8 f 0.7
32f10
1.4f 0.2 -23f02
-3.9 f 0 3
53f03
0.4f 0.4 -4.8f 0.3
-4.0 f 0.5
4.4f 0.7
54
287fO4
54
0 8
286fO6 298fO2 297fO2
54
09 09
60
0 87
28.9 29.8f 0.4
45
0 85
28.1 f 0.2 27.8f 0.3 28.2f 0.3 29.0f 0.5
1 1 10 59
1 0 0 95
3.5f 0.2
-2.0 f 0.4
55f04
6 0
09
4.1f 0.3
-3.2 f 0.6
734x07
TABLE I11 BENZENEDONORS WITH MIXEDSUBSTITUENTS~
WITH
-Transition
AYI/,, 10'
i,.,, 10' cm
Donor
TCVE--Separation Ai(Benz-I),
...
283fO4
CHARGE-TRANSFER SPECTRA OF T C N E -Transition
-1
Acceptor: 11-APl/,. 10' cm -1 a1
-Separation Ai(Benz-I),
11--ACl/z, 101
cm
-1
-1
qI
10' cm
-1
from b e n z e n e AS(Benz-11) 10:cm
-1
Separation 11-1 103
cm.
-1
58 4 8
46 47 57 48 57 48 58 53 48 54 46 49 47 71 46
253fO1 234fOl 240f03 238f01 239fO2 28 1 f 0 15 254fO1 257fO3 27 9 f 0 15 240fOl 254fO2 284fO2 268f03 242f04 2563x01 26 0 f 0 15 265fO1 26OfO2 2523~01 259fO4 27lf04 26.6f 0 . 4
55 55 54 59 57 60 56 60 67 60
07 08 08 08 08 0 65 0 96 09 0 65 10 09 0 7 18
80f02
05f02
75f02
70f02
18f04
52f04
19f03 3 & 0 25
75f02 633x02
-2
56103 8 6 f 0 25
57f03 65f02
01104 -2 1 10 25
56105 863~025
61f03 61102 92103 724103
04f03 -26fO3 -1Of04 12f05
65f02
-0 2 f 0 25
83f02
02f03
57&0 85f0 10 2 f 0 6 Of0
4 3 5 6
50 54
10 08 0 76 03 0 26 08 08 08
50f03 35f05
-0lf05
-13fO5
5 lfO6 48f0 8
54
0 8
40f04
-083~05
4 8fO 7
-1 9 f 0.6 -1.2 f 0 3
9 2fO 6 64f03
57 51
6 7 f 0 25 85103
7 9
21 4 f 0 1
53
185fO1 206fOI
52 46
44f02 277fO5 270f02
to be additive in the manner described by eq. 5 , with certain limitations. The data leave no doubt that Orgel's theoretical treatment reproduces most of the essential features of the charge-transfer spectra of substituted benzenes with TCNE. I t remains to be determined to what ex-
60 56
1 0 0 88
73f02 52f02
tent they do so in detail and whether the results obtained with T C N E are in agreement with data for other acceptors, insofar as such data exist. The Parameters cs and ys.-Table IV gives the substituent parameters eS and ys calculated with eq. 3 from the data in Tables I , 11, and 111. I t is seen
CHARGE-TRANSFER BANDSIN AROMATIC COMPLEXES
Sept. 20, 1964
PARAMETERS w, AND
7 8 FROM THE
Complexes
3615
TABLE IV CHARGE-TRANSFER SPECTRA OF SUBSTITUTED BENZENE DONORSWITH T C N E e8 18, e*, YE*, n
1O’cm
AP
1
102 cm. - 1
1O’cm
-1
1 0 2 cm. -1
( a ) Methoxybenzene-TCNE Anisole 1 1 6 5 f O 2 5 -0.3 f 0 . 2 7 8 f 0 3 -0.35 f 0.25 o-Dimethoxybenzene 2 1 6 4 f 0 3 -0.4 f 0 . 2 -0.5 f0.25 7 7 f O 3 5 m-Dimethoxybenzene 2 1 4 8 f l l -0 4 f 0 . 3 5 8 f 1 3 -0.5 f0.3 p-Dimethoxybenzene 2 2 5 3 * 0 1 -0.25 f 0 . 1 6 4 f 0 1 -0.3 f 0.15 1,2,4-Trimethoxybenzene 1.74 3 5 1 f 0 3 -0.3 f 0 2 6 1 f O 3 5 -0.35 f 0.25 1,3,5-Trixnethoxybenzene” -0.4 0 5 9 3 7 1 -0.5 1,2,4,5-Tetramethoxyben~ene~ 4 2 5 1 -0.5 6 1 -0.6 ( b ) Halogenbenzene-TCNE Chlorobenzene 1 1 3 4 1 0 7 -25 f 0 5 4 2 f 0 8 5 -30 f 0 6 2 o-Dichlorobenzene 1 3 7 f 0 7 -24 f 0 6 4 4 f 0 8 5 -29 * 0 7 2 m-Dichlorobenzene 1 3 2 f 1 0 -22 f 0 9 3 9 f 1 2 -26 f 1 0 2 p-Dichlorobenzene 2 2 7 f 0 15 -1 9 f 0 25 3 3 f 0 2 -23 f 0 3 Bromobenzene 1 1 3 6 * 0 5 -24 * 0 4 4 5 f 0 6 -29 f 0 5 p-Dibromobenzene 2 2 2 9 f 0 1 5 -1 8 f O 15 3 5 f 0 2 -21 1 0 2 1 Iodobenzene 1 -21 f O 4 5 6 f 0 4 6 7 f 0 5 -28 f O 5 p- Diiodobenzene 2 2 3 7 f 0 3 5 -16 1 0 3 4 5 f 0 4 -19 f O 4 2 p- Difluorobenzene 2 2 2 f O 3 5 - 2 0 f O 25 2 7 f 0 4 -24 f 0 3 6 Hexafluorobenzene 0 3 1 -2 7 3 7 -3 2 ( c ) Methylbenzene-TCNEb 1 T oI uen e 1 1 9 3 ~ 0 4 $0.6 f O 4 2 3 f 0 5 +o 75 f 0 5 2 p-Xylene 2 1 7 f 0 3 +0.4 0 3 20f035 $0 5 f 0 35 1,2.3-Trimethylbenzene 3 1 9 0 + 0 . 5 2 3 to 6 1,3,5-TrimethyIbenzene Hexamethylbenzene 6 0 1 5 $0.5 18 t O 6 a Data calculated from ref. 8. The 1,2,3-trimethoxy donor (Table I ) was omitted from this list, since its high c.-t. energy is very likely due to steric interaction between the methoxy groups in the donor.* * The cs and 7 8 values for the methyl substituent were obtamed by an approximate analysis of their composite b a n d s z 2
*
1 1
directly that neither parameter is constant within a series of related donors. As n and pIr - PI increase, t~ and y s show a tendency to decrease, the decrease being dependent upon the nature of the substituent. One finds empirically that the parameters es and ys for para-disubstitution are reduced from the value for the monosubstituted compounds by the factors shown in Table V. VARIATION OF
TABLE V COMPARISON OF MONO-AND DISU DISUBSTITUTED DONORS cs AND y8 ON
( 4 d i __
i18)p-iI
Substituent
(f8)rnOrm
i78)rnono
Methoxy Chloro Bromo Iodo Methyl
0.8 0.8 0.8 0.66 0.85
0.7 0.8 0.75 0.76 0.7
Generally, one may write ts,ys =
fs(P,n)
(6)
However, there are insufficient data to decide how t s and ys behave for values of n > 3. p11 - pl does not increase beyond the value 2 for the para case. The factors in Table V were found useful for the prediction of charge-transfer energies of mixed substituted benzene donors from cS and y s values obtained for the “pure” series in Table IV. Alternatively, approximate values of ts and ys for any substituent may be obtained by measuring the c.-t. energy and band separation of a suitable para isomer selected from the series listed in Table IV and calculating the ts and ys from eq. 4 together with the factors of Table V. When the magnitude and sign of the parameters for the various substituent groups are considered, the follow-
ing trends are established: (i) t~ > ys in all cases; (ii) es was always found positive for all substituent groups studied and increased in the order CHI
< F < Cl < Br < CH=CHz < I < G H S < OCHa 2 O G H , < SCHt
(7a)
(iii) ys was found positive or negative; positive only for CH, as substituent; negative for all the other substituent groups studied here. ys increases in magnitude as SCHI
2
OCH,
< CHI < I < Br < C1 < F
(7b)
Perturbation theory as applied by Orgel in the derivation of eq. l 3 provides the following interpretations for the above observations : (i) The perturbation v i a the u-electron system of the benzene ring by the substituent is in all cases greater than the perturbation v i a the *-electron system. (ii) The energy separation of c.-t. bands I and I1 for varying substituents occupying the same positions on the ring can be related to the perturbation of the Telectron system of the benzene cation by substituent groups. The order of substituent interaction in sequence 7a derives from the fact that ts is a “mixed” parameter, including electrostatic and resonance interaction of substituent with the s-electron system. I t is evident that the substituent interaction described by t~ follows largely the ionization potential order of the substituted donor (excepting CH3 and CH=CHz). I n particular, the halogens show an order of interaction reversed to that observed, from reactivities of substituted ar0matics,~4or from equilibrium constants of c.-t. complexes between iodine and halogen substituted thioanisoles.2b This is probably due to the (24) R. W.Taft. Jr., J . P h y s . Chrm.. 84, 1805 ( 1 9 6 0 ) . (25) J . Van Der Veen and W.Stevens, Rrc. Iron, chim., SI, 287 (1963).
E. M. VOICT
3616
VOl. 86
fact that donor interactions as obtained from c.-t. A ~ V C T= 0 . 8 3 ( 1 ~- I b e n Z ) = -nys - pes (I)) energies relate predominantly to the donor radical Thus, dividing AhvcT, the c.-t. energy differences becations. In that case one expects that (a) any electrontween the substituted benzene-TCN E and benzene releasing mechanism is enhanced over that in the T C N E complexes, through by the factor 0.M gives corresponding neutral donor molecule and (b) the dvalues of t~ and y s essentially independent of the c.-t. acceptor orbital interaction of the halogensz6is strongly resonance i n t e r a ~ t i o n . ? ~ The parameters thus calcusuppressed in the cations. Consequently, the parameI V as t * and t S * , lated are listed in Table ters 6s reflect the net electron release to the s-system of It should be noted that although eq. 9 relates to first the donor cation as the sun1 of halogen-carbon (p-p) ionization potentials, it was also used in calculating overlap and halogen inductive interaction. I n conparameters from the second c.-t. bands. That is, we trast, the data on iodine --halogen-substituted thioanisole complexes cited above involve 7t a - c o m p l e x e ~ . ~ ~assumed that T C N E complexes involving ionization from the second highest filled orbitals in the substituted In this case electron transfer is from the nonbonding benzene donors have similar relative c.-t. resonance sulfur orbital rather than the benzene *-orbital. Thereinteractions as complexes involving the donor cation fore, the halogen substituents do not interact directly ground states. The implications of such an assutnpwith the cation as they do in this study, so that their tion will be commented on below. different order of interaction is not surprising. To correct for the c.-t. resonance interactions in chlor(iii) Perturbation of the a-system by the substituent ani1 complexes we used may be positive or negative. In agreement with all other experimental data on a-interactions, only the ~ U C T= (J.87Zr) - 4.27 e.v. (10) methyl group among the substituents studied here Equation IO was obtained in a manner analogous to (X), ihows a positive effect. Sequence 7b reflects the trend Z . P . , by plotting single, first c.-t. energies of substituted of a-perturbation by substituents and is in the order benzeneechloranil complexes us. donor photoionization that one might reasonably predict. potentials covering a range of -7.8 to 9.3 e.v. ToThus far, c.-t. data are in surprisingly good qualitagether with eq. 3 we thus obtain for chloranil c o n tive agrcwnent with theory. WheIi the quantitative plexes data are examined, however, the variations in t~ and ys throw doubt on the assumption that the interactions of A ~ U C= T o.87(fD- Z,,J = --nys - pep (ilj the substituent with the ring are additive. A first-order perturbation treatment using Hiickel M O ' s does not The parameters ts* and ys* as calculated from eq. $1 allow for the changes occurring in the parameters which and 11 are shown in Table VI. I t is seen that the describe substituent -ring interaction as the number parameter values thus obtained from the chloranil c.-t. of substituents on the ring is increased. I t also neglects spectra agree with those from T C N E spectra in magxionneighbor interactions and both these factors change nitude within the limits of (our) experimental accuracy. the orbital ionization energies as calculated from eq. 3. This leads to the conclusion that the observed variabilThus, it is not surprising to find that parameters t~ ity in the parameters is due mainly to an inadequate and y s vary with number, position, and nature of subapproximation of the donor orbital ionization energies sti tuent . by eq. I . However. it must also be remembered that both On the other hand, Table VI also shows clearly that paraincters were calculated using eq. 3 and 5 which the sign of Ymethoxy as obtained from chloranil spectra is carry the assunlption that all c.-t. interactions are positive, but negative when derived from T C N E iirgligible as compared to the donor ionization potenspectra. tial differences. I n order to see, therefore, whether the Theoretically it is inconceivable that the parameter variations i n the paraineter derive predominantly ys changes sign in different complexes, The a-interfroni an inadequate approximation of the donor orbitals action of a substituent group with a given set of benby theory or from neglect of .c.-t. interactions in their zene orbitals will either be positive or negative, dependcalculatioii from the spectra, a comparison with data ing only on the nature of the substituent. The obserfrom other acceptors is desirable. The only such data vation that ys as obtained from c.-t. energies of one reported and applicable to our study are those by donor with different acceptors and calculation by eq. 3 KuboVania7 on some 'chloranil-methoxybenzene com(or 9 and 11,. respectively) may either be positive or plexes. negative means, therefore, that the quantity calBefore a comparison betweeii corresponding T C N E lated from the c.-t. data does not correspond directly and chloranil complexes can be made, we must corto the theoretical parameter y s . rect the t S and ys values for c.-t. interactions, as far as The methoxy substituent group has a positive T a f t such is possible. The author has shown elsewhere a1 parameter. By definition, its interaction with the that the single. first c.-t. energies of the complexes of benzene ring is then electron attracting relative to H as 'TCN E with substituted benzenes in the ionization a substituent. Consequently, the ys term in eq. 3 is potential range --7.3 to 9.5 e.v. are described fairly positive for methoxy-substituted benzenes. S pecifiaccurately by a linear relationshipz2 cally, for the case pII = 0 , fr,- Z g > 0, and the second c.-t. band of a mono- or para-disubstituted methoxy hvrT = 0.83fD - 4.42 e.v. (8) complex is expected a t higher frequency than the transition of the corresponding benzene complex. where I D is the ionization potential of the donor. This is not true for the anisole and p-dimethoxybenReferring the ionization potentials now to that of zene-chloranil complexes. Therefore, the approximabenzene as i n eq. 3, one obtains for T C N E (?ti) J
K
Hoyland and I . Goodman. J
Phys C h r m . , 64, I816 (1960)
(27) The author wishes to thank a referee f o r r a s i n g this pnlfii
ISOMERS OF SULFUR MONOFLUORIDE
Sept. 20, 1964
3617
TABLE VI
COMPARISON OF CHLORANIL'AND TCNE-MBTHOXYBBNZENB ComPLExEs -Acceptor Chloranil Au (Benz-I),* 10' c m . 7 Cor. by **methaxy, Exptl. eq. 10 10' cm. -1
Donor
7
TCNE-------. y*methoxy, 1 0'
cm-1
c*meti,oxy,
1 0 8 em. -1
y*metbo.r.
10' cm. -1
Anisole 7.3 8.4 7.9 0.5 7 . 8 f 0.3 -0.35 f 0.25 11.8 7.6 0.2 7.7 f 0.35 -0.5 f 0 . 2 5 o- Dimethoxybenzene" 10.3 5.8 f 1 . 3 -0.5 f 0 . 3 m-Dimethoxybenzened 10.2 11.7 5.0 0.7 p - Dimethoxybenzene 11.3 13.0 6.3 0.25 6 . 1 f 0.35 -0.3 f 0 . 1 5 1,3,5-Trimethoxybenzene 9.1 10.5 6.0 (Value 0 . 5 used) 7.1 0.5 ,. .See ref, 7. b A i (Bern-I): separation of first c.-t. energy in methoxybenzene-chloranil complexes from the c.-t. energy ofthe benzene-chloranil complex. e Measured in this work: two c.-t. bands a t 518 and -386 mp overlapping. d Two very strongly overlapping transitions and only approximate values used.
tions hvc,(I,II) = u l ~ ( 1 , I I )which were made in our calculation of ys* are inaccurate, particularly for the one involving the second c . 4 . band, and a more exact expression is required. The most complete description of hvCT is given by eq. 2. Unfortunately, not enough data are available as yet to make use of this equation for a more precise, quantitative evaluation of our results. Qualitatively, however, i t is important to realize that simple approximations which described first c.-t. transitions quite accurately are insuffcient for the description of second transitions. This is also evident from Huckel calculations on the molecular orbital energies of methoxybenzenes and subsequent comparison of their c.-t. energies with a number of acceptors.28 A good linear correlation was found between the energies of the highest filled methoxybenzene orbitals with the first c.-t. energies of their complexes with all acceptors considered. With T C N E , they also compared the energies of the second highest filled donor MO's with the energies of the corresponding second c.-t. absorptions. In this case, no good linear correlation was obtained. Even though this correlation is improved when our analyzed c.-t. values (Table I) are used, the second c.-t. energies still do not fall on the same line with the first. Whether or not this is largely (28) A. Zweig, J. E. Lehnsen, and M. A. Murray, 1.Am. Chcm. Soc., 85, 3933 (1963).
[CONTRIBUTION FROM
THE
the result of poor second MO energies, there can be little doubt that some of the irregularities derive from the c.-t. process. If the complexes have structures similar to those shown in Fig. 1 and 2, then the configurations contributing primarily to the first c.-t. transitions are orthogonal t o those contributing primarily to the second transitions. For benzene donor cations with electron-releasing substituents the ground state is expected to be antisymmetric for mono- and disubstituted isomers, while for tetra- and pentasubstituted isomers the symmetric orbital should be the lowest. Therefore, the first c.-t. band of the former derives principally from structures R, (Fig. l), that of the latter from structures R, (Fig. 2). The opposite applies to configura$ions contributing to the second charge-transfer absorptions. Thus, a change in complex structure with increasing donor substitution is expected for each c.-t. band. Apparently, this theoretically expected changeover in complex geometry does not show up in the experimental first c.-t. energies (at least of the methoxybenzenes). The second c.-t. energies cannot be correlated a t present. Acknowledgments.-This research was supported by the National Research Council of Canada. The writer also acknowledges with pleasure helpful discussions with Professor C. Reid.
DEPARTMENT OF CHEMISTRY, HARVARD UNIVERSITY, CAMBRIDGE, MASACHUSETTS]
The Mass and Microwave Spectra, Structures, and Dipole Moments of the Isomers of Sulfur Monofluoride' BY ROBERTL. KUCZKOWSKI~ RECEIVED MARCH17. 1964 Two isomers of &F2 were prepared by heating AgF and S. The relative mass spectral cracking patterns of both isomers are reported. For one isomer, microwave transitions for the SS2==SS*Fz,S f k S J 4 F zand , S34=S3*Fz species were obtained. The structure was determined to be pyramidal S=SF2with d(SS) = 1.860 -f 0.015 A , , d(SF) = 1.598 =F 0.012 A,, L S S F = 107.5 F 1'. and L F S F = 92.5 f 1'. Thedipolemoment was 1.03 f 0.a7 D. For the second isomer, microwave transitions for the FS3'Sa2Fand FSS2SS4Fspecies were obtained. The structure was determined to be nonplanar chain FSSF with d ( S S ) = 1.888 f 0.01 A., d(SF) = 1.635 f 0.01 A , , LFSS = 108.3 f 0.5', and the dihedral angle = 87.9 F 1.5". The dipole moment was 1.45 f 0.02 D. A discussion of the observed structural parameters for both isomers is given.
Introduction The earlier reports in the literature concerning SzFl presented some conflicting data and gave rise to general (1) This work was supported by a grant extended t o Harvard University by the office of Naval Research. Reproduction in whole or part is permitted for the U. S. Government. (2) (a) Standard Oil of California Predoctoral Fellow 1963-1964; (b) National Bureau of Standords, Washington, D. C.
skepticism concerning this c o m p o ~ n d . ~Because of an interest in this laboratory in the structure of some simiwe were prompted to relar molecules, notably F2OZr4 investigate S2F2 with the hope of unambiguously charactetizing it as well a s determining its structural param(3) See footnotes 2-9 in ref. 5 of this paper. (4) R. H. Jackson. J . Chrm. Soc.. 4585 (1962).
