J. MATER. CHEM., 1994, 4(8), 1201-1204 Electrochromic Behaviour and X-Ray Structure Analysis of a Pechmann Dye, (€)=5,5’=Diphenyl-3,3’-bifuranylidene-2,2’=dione Jack Silver,*” Mustafa T. Ahmet,” Keith Bowden,a John R. Miller,” Shahdah Rahmat,” Christopher A. Reynolds,” Alan Bashall: Mary McPartlin*’ and Jill Trottee a Department of Chemistry and Biological Chemistry, University of Essex, Wivenhoe Park, Colchester, UK C04 3SQ School of Applied Chemistry, University of North London, Holloway Road, London, UK N7 8DB Studies on the title compound have established that, although it could be oxidised with a concomitant change in its visible spectrum, this process is not reversible. This is discussed in the light of the crystal structure which IS also reported in this paper.The title compound crystallises in the monoclinic space group P2,/n, a =27.033(6) A, b =4.988(2) A, c =5.372(2)A, p =94.59(2)”. The molecules have close intermolecular contacts along stacks packing in the c direction. The packing is compared to that found in two other Pechmann dyes. Electrochromic behaviour has been observed in four different types of chemical material: inorganic, organic, organometallic and chlathrate compounds.’ Most electrochromics offer only a monochromatic colour change, though the lanthanide bisph- thalocyanines offer polychromic changes.2 A monochromatic electrochromic material which produces a good red colour, even for a limited number of colour changes (cycles), has yet to be reported in the literature.Red dyes are not rare in the literature, but, in order to be sublimed to form useful films, they must be relatively thermally stable. One such group of materials is a family of dyes known as ‘Pechmann dyes’. The 1.2 h 07 -.-C 0.9 2 (dv 5 0.62s:Llparent compound is (E)-5,5’-diphenyl-3,3’-bifuranylidene-2,2’-(ddione (1).3 We report here its electrochromic properties and its crystal structure. Results and Discussion Fig. l(u) shows the electronic absorption spectrum of the parent Pechmann dye (1) in chloroform solution. This solution is pink-orange and fluorescent. Subliming the dye onto an optically transparent electrode [indium-doped tin oxide glass (ITO)] produced an orange-red non-fluorescent film with the electronic absorption spectrum presented in Fig.2(a). This spectrum is substantially different from that in Fig. l(a) and indicates either that the conformation of the molecule changes between the solid state and solution, or that there are signifi- cant intermolecular interactions in the solid state. The IR spectrum of the film is identical to the Nujol mull spectrum previously reported3 and confirms that no structural change is induced by the sublimation process. Oxidation of the Pechmann dye film on the IT0 at +1.0 V changed the visible spectrum, giving a loss of intensity across the entire visible absorption band between 390 and 620nm, and a slight increase between 750 and 850 nm. The shape of the band between 400 and 550nm also changes.Reduction of the oxidised form back towards the neutral form suggested only a slight recovery of the original material [Fig. 2(b)]. IR spectra of the sublimed films are presented in Fig. 3 for 1 0.3 0 0.40 400 600 800 400 600 800 Wnm Fig. 1 (a) Electronic absorption spectrum of 1 in chloroform solution. (b)Electronic absorption spectrum of 1 dissolved in polystyrene film. the neutral and oxidised states. On oxidation there are several changes in the IR spectrum, particularly in the catbonyl- stretching frequency region. The absorption area of the band at 1745 cm-’ is rcduced (by about one half) and a new band appears at 1781 cm-’. Several explanations are possible: (i) only a fraction of the molecules are oxidised giving rise to a spectrum which is a superposition of the original spectrum and the oxidiseti form, the latter having one band at 1781 cm-l and possibly a second band near 1745 cm-l; (ii) oxidation (complete) is accompanied by a structural change involving loss of the centre of symmetry giving rise to two IR-allowed C-0 stretching frequencies; (iii) oxidation occurs in only one half of the molecule and the oxidised molecule is asymmetrical, with two non-equivalent environments for the ca rbonyl groups; (iv) intermolecular interactions change on oxidation, a distinct possibility because of the close intermolecular approach of the C-0 bonds (see the crystal sti-ucture reported herein).The third possibility fits in with previous findings that it is possible for a functional group on one ring to react indepen- dently of the ~ther.~,~ There is also a change in the IR spectrum near 1100 cm- ’.The nature of the vibration cannot be assigned here but the general change is similar to the carbonyl-stretching region, J. MATER. CHEM., 1994, VOL. 4 1.88 1.32 0.7E h cu).- C3 4 v Q)0 0.2c 500 700 1 IB investigated using the AM l5 semi-empirical method, as implemented within the MOPAC 6.0 program,6 for both the neutral molecule and the radical cation. We note that the latter is not directly relevant as the radical cation would be short-lived and would be an intermediate to a more stable neutral molecule of unknown structure. The calculations predict both modelled molecules to be planar, as confirmed by second-derivative calculations which show that in both cases the Hessian is positive definite.This result suggests that the asymmetry observed between the two carbonyl groups is a result of inter- rather than intra-molecular interactions. The highest occupied molecular orbital, HOMO, of the neutral molecule is spread over the entire molecule (see Fig. 4). Moreover, the change in Coul~on~.~charge distributions between the neutral molecule and the radical cation is in accord with the electron being removed from a highly delocal- ised orbital. As stated above, the solid-state electronic absorption spec- trum [Fig. 2(u)]was clearly different from the solution spec- trum [Fig. l(u)].The origins of this difference are important. To try to resolve the reasons for this, the dye was dissolved in a chloroform-polystyrene solution and the solvent was allowed to evaporate.The resulting pink fluorescent film gave the visible absorption which is presented as Fig. l(b). Clearly, this spectrum is closer to the solution spectrum [Fig. l(u)] than to the spectrum of the sublimed solid film [Fig. 2(u)]. This finding supports the earlier premise that the molecules of the Pechmann dye interact in the solid state. Indeed, such an interaction is found in the crystal structure reported below. 0.20 1The films did not undergo reduction at -1.0 V either in 500 700 Vnm Fig. 2 A, Electronic absorption spectra of a thin sublimed film of 1: (a) as sublimed, (b)oxidised.B, Electronic absorption spectrum of the sublimed film of Fig. 2A (b) reduced back to neutral [little change is seen, compare Fig. 2A (b)]. 59.56r h Y VI 1800 1760 1720 iVcm-’ Fig.3 IR spectra of 1 thin sublimed film from Fig. 2A (-) (as sublimed), (---) and after oxidation in the 1850-1660 cm-’ range i.e. loss of intensity of the original spectrum accompanied by the appearance of a new band. From the differences in the electronic absorption spectra [Fig. 2(a), (b)] we feel that explanation (i) can be ruled out. However, we are not able to choose between explanations (ii), (iii) and (iv) on the basis of experimental data alone. For this reason computer-modelling programs were used to study molecular structure of 1 and its radical cation.The geometry, molecular orbitals and charge distribution were the neutral or in any state once oxidised. Therefore, the electrochromic properties of this particular material are not likely to be of practical use. X-Ray Structural Data The molecular structure of 1 is illustrated in Fig. 5 showing it to be almost planar overall. For comparison, the bond lengths in the diphenyl compound 1 are listed in Table l(u), together with those of the two polymorphs of the correspond- ing dimesityl compound, red 2a and black 26;’ the interbond angles for 1 are in Table l(b). Th? central double lactone unit in 1 is planar to within 0.02 A an! the bridging bond length, C(1)-C(l’), of 1.361(11) A is similar to those in 2a and 2b (mean 1.365 A).The phenyl substituent in 1 is almost coplanar with the H H H H H’ H -H H H Fig.4 Semiempirical optimized geometry showing a schematic rep- resentation of the HOMO (a n: orbital). The contributions to the orbital with positive phase are shown in grey; those with negative phase are shown in black. The radii of the circles are roughly proportional to the coefficient of the orbitals in the LCAO expansion. J. MATER. CHEM., 1994, VOL. 4 Fig. 5 Almost planar molecular structure of 1 Table 1 Comparison of bond lengths (A)and angles (degrees) found in the three knowp Pechmann structures (u) Bond lengths/A band 1, reda 2b, red" 2b, blackb 1.361(11-) 1.364 1.365 1.492(8) 1.476 1.480 1.415(8) 1.359(8) 1.427 1.340 1.422 1.343 1.403 (7) 1.393( 7) 1.396 1.382 1.400 1.380 1.191(8) 1.189 1.187 1.433(8) 1.468 1.459 1.406( 8) 1.402 1.407 1.361(8) 1.392( 9) 1.402 1.385 1.383 1.382 1.390(9) 1.386 1.389 1.368(8) 1.384 1.376 1.407(8) 1.408 1.415 (b)Inter-bond angles/degrees for 1 C(6)-C( 1)-C(2) 106.1(5) 0(3)-C(2)-C(l) 132.1(6) 0(4)-C(2)-C( 1) 106.5(5) 0(4)-C(2)-0(3) 121.3(5) C(5)-O(4)- C( 2) 107.8( 4) C( 6)- C( 5)-O(4) 11 1.4( 5) C(7)-C(5)-0(4) 116.4(5) C( 7)-C( 5)-C( 6) 132.3( 5) C(5)-C(6)-C( 1) 108.2(5) C(8)-C(7)-C(5) 121.5(5) C( 12)-c(7)-c(5) 120.3(5) C(12)-C(7)-C(8) 118.2(5) C(9)-C( 8)-C(7) 120.1 (5) C( lO)-C(9)-C(8) 121.3(6) C( 11)-C( 10)-c(9) 118.7(6) C( 12)-C( 11)-C( 10) 120.8(6) C( 11)-C( 12)-c(7) 120.9( 6) lactone, giving a dihedral angle between the two rings of 1.2", in marked contrast to the mesityl substituents in 2a and 2b, which are, respectively, at angles of 56" and 65" to the lactone rings.This observation is reflected in the shorter bond length between !he lactone and phenyl rings in 1, S(5)-C(7) of 1.433(8)A [Table l(a)], cornpaTed with 1.468 A for the equiv- alent length in 2a and 1.459 A in 2b. These differences are consistent with much greater conjugation between lactone and substituent in the almost planar molecule of 1; in 2a and 2b coplanarity is prevented by the two methyl groups at the 2,6-positions of the dimesityl units [i.e. equivalent to C(6) and C( 12) of the diphenyl compound 1, shown in Fig.51. Part of the solid-state structure of 1 is illustrated in Fig. 6(a) and (b) which shows the 'stepped' arrangement of three adjacent molecules along the c axis viewed (a)onto the plane of the molecule and (b) at 75" to (a). The molecules in 1 mainly interact via one carbonyl group at x,y, z of the central molecule (A) and one at -x,-y, -1-z of an adjacent molecule (B), with a second set of identical interactions between the second carbonyl (at -x, -y, -z) of the first molecule (A) and one carbonyl (at x, y, 1+z) of the molecule on the other side (C); both sets correspond to the short contact distances \ \ \ A \ \ \ \ B Fig.6 Arrangement of the molecules along the c axis of the crystal viewed: (a) onto the plane of the molecule, (b)at 75" to the plane of the molecule C(2)..-0(3) 3.19, 0(3)..-0(3) 3.11, and H(6)...0(3) 2.35 A (Table 2).In contrast, the blue-black mesityl analogiie (2b) has no interactions involving the carbonyl groups: the only short contacts being four identical distances of 3.31 A bctween carbon atoms of one molecule and those on either side; the red form (2a) shows no interactions at less than the van der Waals distances and the material may be thought of as composed of essentially non-interacting molecules. The struc- ture of the diphenyl red material (1) is therefore interesting as its solid-state interactions (at less than the relevant 1 an der Waals distances) all involve the oxygen atom of the carbonyl group, 0(3), which may well be involved in the changes encountered in the IR spectrum on oxidation of the thin sublimed films.We note that the blue-black phase of the dimesityl deriva- tive did not have good conducting properties. From our electrochromic investigation, which showed that sublimed films of 1 were slow to oxidise and would not cycle to neutral, we would not expect 1 to be a very good conductor either. Table 2 Selected intermolecular interactions (A)with first atom at x, Y, z equivalent position of second atom 0(3)...0(3) 3.11 -x, -y, -1-z C(2)*..0(3) C( 5)--0( 3) H( 12)-0(3) 3.19 3.30 2.35 -x, -y, -1-z x,1 +y, z x, 1+y, l+z J. MATER. CHEM., 1994, VOL. 4 Conclusions The Pechmann dye sublimed readily and formed good-quality even films. However, it proved to have limited electrochromic potential (one shot use). Experimental (E)-5,5'-Diphenyl-3,3'-bifuranylidene-2,2'-dione(1) was pre-pared as previously de~cribed.~,~ Crystal Data for 1 C20H1204,0M=316.31, monoclinic, spate group P2,/n, a= 27.033(Q) A, b=4.988(2) A,~=5.372(2) A, p=94.59(2)", U= 722.04 A3, F(000)=328, ,u(Mo-Ka)=0.60 cm-l, Z=2, D,= 1.455 g cmP3.A crystal of size 0.38 mm x 0.14 mm x 0.06 mm was used in the data collection. Data were collected in the 8 range 3-23' using a 8-28 scan mode with a scan width of 0.90'. Equivalent reflections were merged to give 684 unique data with I/g(I)>2.0. Structure solution and refinement were carried out using SHELX 76.'' The structure was solved using direct methods. All the hydrogen atoms were inc!uded in the refinement at calculated positions (C-H 1.08 A), assuming idealised sp2 hybridisation, except that bonded to C(6) which was located from a difference-Fourier synthesis and inserted without refinement at that position.All the non-hydrogen atoms of the five-membered ring were assigned anisotropic thermal parameters in the final cycles of the full-matrix refinement (which converged at R =0.0749 and R' =0.0686 with weights of w = l/02Foassigned to the individual reflections). In a final difference-Fourier synthesis there were no residual maxima greater than 0.5 of an electron. The final atomic coordinates are given in the supplementary data.? ~~ t Observed and calculated structure factors fractional atomic coordi- nates and thermal parameters for 1 (1 1 pp.) are deposited with the Cambridge Crystallographic Data Centre.Details available from the Editorial Office. IR Spectroscopy The IR spectra were collected on material sublimed as thin films onto ITO-covered glass. A Perkin-Elmer 1700 Fourier transform infrared spectrometer was used to collect the spec- trum by reflection from the film. Electrochromic Studies The films were oxidised by applying + 1.0 V to the IT0 in an electric cell using 10% KCl-water as the electrolyte and a silver wire as the counter electrode. They were subjected to voltages of +1 both in the neutral and oxidised states. UV-VIS Spectroscopy Spectra were recorded on thin films [deposited under vacuum (lo-' Torr) on IT0 glass using an Edwards 306a vacuum sublimation unit] on a Perkin-Elmer Lambda 5 spectrometer.References 1 J. Silver, New Scientist, 1989, 30th September, p. 48. 2 C. S. Frampton, M. J. O'Connor, J. Peteraon and J. Silver, Displays Techn. Appl., 1988,9, 174, and references therein. 3 E. Klingsberg, Chem. Rev., 1954,54,59. 4 K. Bowden, R. Etemadi and R. J. Ranson, J. Chem. Soc., Perkin Trans. 2,1991,743. 5 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am. Chem. SOC.,1985,107,3902. 6 J. J. P. Stewart, Mopac 6.0, Frank J. Seiler Research Laboratory, US Air Force Academy, Colorado Springs, CO 80840. 7 C. A. Coulson and H. C. Longuet-Higgins, Proc. R. Soc. London, Ser. A, 1947,191,39. 8 B. H. Chiurguin and C. A. Coulson, Proc. R. Soc. London, Ser. A, 1950,201,196. 9 M. J. Begley, L. Crombie, G. L. Griffiths, R. C. F. Jones and M. Rahman, J. Chem. SOC.,Chern. Commun., 1981,823. 10 G. M. Sheldrick, SHELX 76 Program for Crystal Structure Determination, University of Cambridge. Paper 4/00602J; Received 31st January, 1994