Spectra of the 1: 1 and 3: 1 solid complexes of coronene-TCNQ

KIM DOAN TRUONG A N D ANDRE D. BANDRAUK DPpartement de chimie, et Grolipe de recherche sur les serniconducteurs et les diilectriques,

UniversitP de Sherbrooke, Sherbrooke (Que'.), Canada JIK 2R1 Received May 31, 1977

KIM DOAN TRUONG and ANDRE D . BANDRAUK. Can. J. Chem. 55,3712 (1977). Two new solid TCNQ complexes have been isolated, coronene-TCNQ 1 : 1 and 3 : 1. The

infrared and electronic absorption spectra are presented for the two different stoichiometries. From these spectra we infer that the complexes are covalent in the ground state with a charge transfer band appearing at 730 nm. The out of plane vibrations of the individual molecules are noticeably perturbed upon complexation.

KIM DOAN TRUONG et ANDRE D. BANDRAUK. Can. J. Chem. 55,3712 (1977). Deux nouveaux complexes de TCNQ ont CtC isoles a l'etat solide, coronene-TCNQ 1 : 1 et

3: 1. Les spectres d'absorption infrarouge et electronique sont prksentes pour les deux stoi- chiometries. Nous avons Ctabli que ces con~plexes sont covalents a l'etat fondamental avec une bande de transfert de charge situee a 730 nm. Les vibrations de deformation des molecules sont perceptiblement perturbkes par la complexation.

Introduction nection with the nature of their charge transfer M ~ ~ - , attention is presently being focused excitons (6). These are thought to be responsible

upon charge transfer con~plexes as these offer for the photoconductor properties of charge alternative possibilities for semiconductor ma- transfer systems terials (1). In particular, complexes with TCNQ In this we present the synthesis and as an acceptor have given rise to interesting ~pectroscopic characterization of a system systems, ranging from insulators to one- similar to the ~er~lene-TCNQ system. We have dimensional conductors. These latter complexes obtained crystals of coronene-TCNQ in 1 1 and fall in the category ofn-molecular charge transfer 3 : 1 stoichiometric ratios. In Fig. 1 we illustrate crystals and have unusual electrical properties the structure of the individual molecules. Of

(1, 2) . note is the fact that perylene and coronene have we have recently discovered a new stoichio- very similar dimensions. In addition, TCNQ fits

metric form of perylene-TCNQ, i.e., its 3 : 1 right into the middle of perylene. It is not sur- modification (3). ~ ~ ~ ~ t h ~ ~ with the 1: 1 form, prising therefore that the crystal structure of they present interesting examples of solid state P ~ ~ Y ~ ~ ~ ~ - ~ ~ ~ Q (I : I) an alternation of solutions of organic molecules. Of note was the dollor-acceptor in linear stacks (7). Further- fact that the vibrational structure of the absorp- more, ~ e r ~ l e n e in its P form (8) and coronene (9) tion band of pure solid perylene (4) remained have similar crystal structures. Both crystals are intact in the 3: 1 complex but was completely broadened in the 1 : 1 complex. Furthermore, the N N charge transfer band of both complexes showed no room temperature structure. To our know- ledge there is only one confirmed observation of vibrational structure in the organic solid state charge transfer band (5).' As it is generally believed that the mechanism of charge carrier

C formation in these crystals is associated with the charge transfer interaction between donor and Perylene C / 'C & NN Coronene

acceptor molecules (I), the study of crystal N spectra of these complexes is interesting in con- Tetracyanoquinodirnethane

'Recently, Haarer (5) has observed a zero phonon (TCNQ)

transition for a charge transfer crystal. We thank the FIG. 1 . Structures of perylene, tetracyanoquinodi- referee for this observation. methane (TCNQ), and coronene.

Can

. J. C

hem

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

TA

SMA

NIA

on

11/1

3/14

For

pers

onal

use

onl

y.

TRUONG AND BANDRAUK

FIG. 2. Infrared spectra of: (a) 3: 1 complex; (b) 1 : 1 complex.

monoclinic with nearly perpendicular linear stacks of molecules. Considering the similar structure both for the isolated molecules and their crystals, it is not surprising these form analogous complexes with TCNQ. We should add that we were unable to synthesize 3 : 1 crystals of pyrene with TCNQ. Although pyrene has physical similarities to perylene and coronene, its crystal structure (lo), although monoclinic, shows four molecules per unit cell as opposed to two molecules for perylene and coronene. Thus a slight geometrical disparity seems to be important in controlling the stoi- chiometries of these charge transfer systems.

became green. The solution was cooled overnight at room temperature. Green needles were found to preci- pitate. The same crystals were found when prepared by the diffusion method as described in ref. 11. Elemental analysis gives the following result (theoretical values in parentheses) for the 3 : l complex: C 91.28 (91.30), H 3.92 (3.62), N 4.87 (5.08).

The 1: 1 complex was prepared as above but using equimolar ratios. Deep blue-black crystals were found after crystallization. Elemental analysis gives the following: C 85.72 (85.71), H 3.16 (3.17), N 10.9 (11.12).

Electronic absorption spectra were taken in KBr with a Varian Techtron 635 spectrometer. Infrared spectra were recorded on a Perkin-Elmer 621 model with a range of 200-4000 cm-'. Spectra were taken in both Nujol mulls and KBr. Little difference was observed between the two media.

Experimental Infrared Spectra Commercial coronene was recrystallized from ben- In Fig. 2 we present the infrared spectra of the

zene whereas pure TCNQ was obtained from aceto- two different stoichiometric complexes. Table 1 nitrile solution. The purified coronene (0.001 mol) was also gives a r~sumi of the more important fie- dissolved in hot benzene. Purified TCNQ (0.003 mol) was also dissolved in hot benzene in a different flask. The quencies including measurements on pure two solutions were then mixed hot. An exothermic TCNQ and coronene. The interpretation of reaction was observed immediately and the solution these results will be based on the infrared and

Can

. J. C

hem

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

TA

SMA

NIA

on

11/1

3/14

For

pers

onal

use

onl

y.

CAN. J . CHEM. VOL. 5 5 , 1977

TABLE 1. Infrared frequencies (cm-I)"

Coronene-TCNQ Coronene-TCNQ 3: l TCNQ Coronene 1 : 1

as = strong, m = medium, w = weak.

Raman study of TCNQ by Takenaka (12) where detailed calculations and assignments of TCNQ vibrational modes have been made. The vibra- tional spectra of coronene are less complete and here we rely on the partial assignments of Babkov et al. (13, 14).

A perusal of Table 1 for the high frequency modes reveals that these modes are little per- turbed when compared to the isolated molecular frequencies. Thus the characteristic C=C vibra- tion of TCNQ at 1540 cm-' is essentially un- displaced with respect to pure TCNQ in the 1 : 1 and 3 : 1 complexes. Similarly the very strong absorption of coronene at 1314 cm-' also re- mains unchanged in the complexes. The one slight difference between the conlplexes is the accentuation of an absorption at 1190 cm-I in the 3 : l con~plex. This line occurs in pure TCNQ (1200 cm-I) albeit weak (12) but is also accentuated in the 3 : 1 perylene TCNQ complex (3). Thus in the region of 800-1600 cm-I, one can infer that one has superpositions of un- perturbed TCNQ and coronene molecules.

The low frequency region of Fig. 2, 1000-700 cm-l, shows more clearly the effect of the for- mation of the complexes. In particular the band in the region 900-800 cn1-' is quite different for the two stoichiometries. Thus the TCNQ and coronene C-H out of plane bending vibrations with corresponding frequencies at 856 and 848 cm-I (12, 14) are visibly perturbed in the 3: 1

complex. Thus the shoulder at 850 cm-I in the 1 : 1 complex separates off in the 3 : 1 complex from the vibrations at 865 and 831 cm-I and seems to be split into two components. We attribute the 865 cm-' vibration to TCNQ, whereas the 831 cm-' vibration would seem to belong to complexed coronene. The doublet at 854 cm-I probably comes from a Davydov splitting occurring between two equivalent un- complexed coronenes. This interpretation is sup- ported by the fact that in the 1: 1 complex the 865 and 831 cm-' vibrations appear isolated. Of note is the fact that the same sort of behaviour occurs in the C-H region of the perylene- TCNQ complexes (3). As indicated in the Intro- duction, this is not surprising in view of the physical similarities between the two donors. We add, finally, that whereas the 962 cm-' band of TCNQ is absent in the 1 : 1 complex, it re- appears in the 3 : 1 complex.

Electronic Spectra The infrared spectra in the previous section

indicate that the complex is weak and therefore that one has hardly any charge transfer in the ground state. The charge transfer in fact occurs in the first excited state as shown in Fig. 3. A maximum occurs for both stoichiometries at 730 nm. This transition also appears in the solu- tion spectrum of coronene with TCNQ ob- served at 740 nm by Beukers and Szent-Gyorgyi

Can

. J. C

hem

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

TA

SMA

NIA

on

11/1

3/14

For

pers

onal

use

onl

y.

TRUONG AN1 D BANDRAUK 3715

I I I , I , I

X X ) 350 400 450 480 580 68) 7EO nrn

FIG. 3. Electronic spectra: (-) coronene; (---) 3 : l complex; (-.-) 1 : 1 complex.

(15). This corresponds to the charge transfer band. The maximum at 400 nm corresponds to TCNQ (16) whereas the maxima at 300 and 350 nm correspond to the characteristic P and p bands of coronene (17). In this region, the spec- tra of the individual moieties are well separated. Of interest is the fact that the p band of coro- nene remains virtually unchanged in pure coronene in the 1 : 1 and 3: 1 complexes, as seen clearly in Fig. 3. This is in contrast to the pery- lene-TCNQ complexes, (3), where the vibra- tional structure of the perylene absorption persisted in the 3 : 1 complex but was completely broadened in the 1 : 1 complex. This aspect of the coronene spectra reinforces the suggestion that the p band is indeed a new electronic transition (17).

Returning once more to the charge transfer band at 730 nm, it must be noted that no vibra- tional structure is observed in the case of both 1 : 1 and 3 : 1 complexes. In the analogous pery- lene-TCNQ system this unstructured band appears at 930 nm. The difference in energy between the charge transfer bands of these two systems amounts to 0.36 eV. This corresponds precisely to the difference in ionization poten- tials of the last occupied molecular orbitals of the isolated perylene and coronene molecules as reported by Boschi et al. (18). Inasmuch as these orbitals are of a, symmetry for perylene and of e,, symmetry for coronene, one would expect some Jahn-Teller distortion in the charge transfer band of the coronene complexes. It would therefore be of interest to examine these bands at low temperature in order to resolve

the vibrational structure and thus elucidate further .the true nature of these charge transfer states.

Discussion Our spectroscopic studies of the new 1 : 1 and

3: 1 coronene-TCNQ complexes in the solid state show that both stoichiometric forms are weak complexes with a charge transfer band occurring at 730 nm. The infrared spectra of these complexes confirm the weak interaction in the ground state. In addition, it is observed that the C-H bending vibrations are perturbed in both stoichiometries. This phenomenon was also observed in the perylene-TCNQ analogous system (3). The crystal structure of the 3: 1 modification of .this system is now being investi- gated by HansonZ and this should help ascertain the cause of the splitting of these vibrations for that stoichiometry.

The coronene, perylene, and TCNQ systems deserve further study in view of the many simi- larities alluded to in the Introduction. In par- ticular, there are phase transitions which occur in pure coronene (17,19) and also in perylene (8). It will be interesting to see what is the effect of an acceptor such as TCNQ on these phase transi- tions. We are presently studying these by reso- nance Raman scattering experiments and plan to report these results shortly (20). Another out- standing problem in these systems is the vibronic couplings in the charge transfer bands. As men- tioned in the previous section, the coronene cation should exhibit a Jahn-Teller distortion in the charge transfer state. This aspect and other problems will be examined at low temperatures to establish the true nature of charge transfer states in the solid state.

Acknowledgments We thank the Defence Research Board for a

grant to GRSD which financed this research. We further thank Dr. L. Caron, director of GRSD, for encouraging studies of the organic solid state. 1: H. MEIER. Monographs in modern chemistry. Vol. 2.

Organic semiconductors. Verlag Chemie, Weinheim. 1974; R. FOSTER. Organic charge transfer complexes. Academic Press, New York, NY. 1969.

2. A. I. GARITO and A. J. HEEGER. ACC. Chem. Res. 7, 232 (1974).

3. K. D. TRUONG and A. D. BANDRAUK. Chem. Phys. Lett. 44,232 (1976).

2A. W. Hanson. Private communication.

Can

. J. C

hem

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

TA

SMA

NIA

on

11/1

3/14

For

pers

onal

use

onl

y.

3716 CAN. J. CHEM. VOL. 5 5 , 1977

4. K. FUKE, K . KAYA, T. KAJIWARA, and S. NAGA- KURA. J. Mol. Spectrosc. 63,98 (1976).

5. H. KLTRODA, T. KUNII , S. HIROMA, and H. AKAMATU. J. Mol. Spectrosc. 22, 60 (1967); D. HAARER. Chem. Phys. Lett. 27,91 (1974).

6. R. L. BECKMAN, J. M. HAYES, and G. J . SMALL. Chem. Phys. 21, 135 (1977).

7. I . J. TICKLE and C. K. PROUT. J. Chem. Soc. Perkin Trans. 11,726 (1973).

8. J . TANAKA. Bull. Chem. Soc. Jpn. 36, 1237 (1963). 9. J . M. ROBERTSON and J . G. WHITE. J. Chem. Soc. 607

(1947). 10. J. M. ROBERTSON and J. G. WHITE. J . Chem. Soc. 358

(1947). 11. G. T. POTT and J . KOMMANDEUR. Mol. Phys. 13,373

(1967). 12. T. TAKENAKA. Spectrochim. Acta, Part A , 27, 1753

(1971).

13. L. M. BABKOV et al . Opt. Spectrosc. (USSR), 34, 38 (1973).

14. L. M. BABKOV and M. A. KOVNER. Opt. Spectrosc. (USSR), 35,214 (1973).

15. R. BEUKERS and A. SZENT-GYORGYI. Recl. Trav. Chim. Pays-Bas, 81,255 (1962).

16. R. R. PENNELLY and C. J. ECKHARDT. Chem. Phys. 12, 89 (1976).

17. K. OHNO, T . KAJIWARA, and H. INOGUCHI. Bull. Chem. Soc. Jpn. 45,996 (1972).

18. R. B o s c ~ r , E. CLAR, and W. SCHMIDT. J. Chem. Phys. 60,4406 (1974).

19. C. A. FYFE, B. A. DUNELL, and J. RIPMEESTER. Can. J. Chem. 49,3332 (1971).

20. A. D. BANDRAUK, K . D. TRUONG, H. J. BERNSTEIN, and V. SALARES. TO be published.

Can

. J. C

hem

. Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

TA

SMA

NIA

on

11/1

3/14

For

pers

onal

use

onl

y.