J. MATER. CHEM., 1991, 1(6), 971-976 97 1 Electrochemical Salt Formation in Bis(phthalocyaninato)ytterbium(iii)-Stearic Acid Langmuir-Blodgett Films Michael Petty," David R. Lovett,*" John Millerb and Jack Silverb "Department of Physics, University of Essex, Colchester, UK bDepartment of Chemistry and Biological Chemistry, University of Essex, Colchester, UK The negative deviation of areas per molecule from calculated values and low collapse rates of Langmuir monolayers of bis(phthalocyaninato)ytterbium(m)-stearic acid mixtures is discussed in terms of phase miscibility. Postdeposition stearic acid salt formation in Langmuir-Blodgett films is investigated using X-ray diffraction and infrared absorption studies. It is found that the salt formation occurs upon reduction of the bis(phthalocyaninat0) ytterbium(n1).The redox potentials of pure and mixed films of bis(phthalocyaninato)ytterbium(III) are also investigated. A wetting and/or charge-injection mechanism was observed for mixed films. Keywords: Electrochemical salt formation; Bis(phthalocyaninato)ytterbium(m); Stearic acid; Langmuir-Blodgett film ; Electrochromism In a recent paper,' we showed that bis(phtha1ocyaninato)ytter-bium(~~~)([Yb(pc)(pc.)],where pc is the phthalocyanato 2 -anion and pc. is the phthalocyanato 1 -monoradical anion) can be deposited onto a variety of substrates by the Langmuir- Blodgett (LB) technique. The family of compounds to which this material belongs is known to be electrochromic when in evaporated film formzp4 and this was also found to be the case with LB films."' Unfortunately, monolayers of [Yb(pc)(pc.)] were not very stable with respect to collapse, and deposited films exhibited artifacts of this collapse.An attempt was made to improve this situation by mixing [Yb(pc)(pc.)] with stearic acid.5 The greatest success was achieved with a mixture of 1:5 [Yb(pc)(pc-)]:stearic acid deposited from a mol dm-3 CdC1, subphase. This mono- layer collapsed at a rate of just 1.1% h-' and deposited Y-type with unity transfer ratios. In the course of the work some effects were observed for the mixed materials, on a pure- water subphase, which have since been explored further. Mixed monolayers of stearic acid and [Yb(pc)(pc.)] were noted to collapse at a lower rate than monolayers of the parent compounds.The surface pressure uersus area per molecule (A,) isotherms of these mixtures also displayed a negative deviation from calculated values. These phenomena are described and explained in this paper. X-Ray diffraction analysis of the LB films of these materials suggested that the stearic acid was converted into its potassium salt upon oxidation and/or reduction of the [Yb(pc)(pc.)] in a KCl electrolyte. This work has been corroborated with infrared (IR) measurements and is discussed here. In addition, the redox potentials of the [Yb(pc)(pc.)] in both pure and mixed forms have been investigated. Experimental The two-compartment trough and dipping mechanism employed have been described elsewhere.' Deposition was performed at a speed of 4-10 mm min -from a pure-water subphase (resistivity >18 MR cm; temperature =21 & 1 "C).Films were deposited onto clean, hydrophilic indium-tin-oxide (ITO) coated and gold-coated glass slides. [Yb(pc)(pc.)], synthesized in the Department of Chemistry and Biological Chemistry at the University of Essex,' was mixed in the molar ratios 1:0, 1:1, 1:5, 1:9, 1:99 and 0:l [Yb(pcXpc.)] :stearic acid. These mixtures were spread onto the subphase as solutions in chloroform. Surface pressure us. area per molecule isotherms were recorded by compressing the monolayers to 35 mN m-'. Monolayer collapse-rate tests and deposition were carried out at a surface pressure of 30 mN m- '. Constant pressure was maintained using an analogue electronic feedback system.Low-angle X-ray diffraction analysis was performed on films deposited onto IT0 slides using a Raymax RX3D diffractometer. IR spectra were recorded using a Perkin-Elmer 225 infrared spectrophotometer. In order to increase the effective thickness of the films, a variation of the method of Francis and Ellison6 was used to prepare the LB samples for study. Films deposited onto gold-coated substrates were mounted opposite a parallel gold reflector to form a waveguide. Two concave mirrors directed IR light into the waveguide arrangement. The emerg- ent light was collected by a further two mirrors and focused onto the entrance slit of the spectrophotometer. For electrochemical analysis the IT0 electrodes coated with LB films were placed in a custom-built W-shaped cuvette containing a 5% KC1 solution.The transparent IT0 layer formed the working electrode; the counter and reference electrodes being a 1 cm2 platinum gauze and a saturated calomel electrode (SCE) respectively. An E.G. and G. Prince- ton Applied Research model 264A Polarographic Analyzer/ Stripping Voltmeter was used to control the three-electrode cell. Cyclic voltammograms (CVs) were recorded (scan rate 20 mV s-l) on an XY-recorder. All potentials were recorded with respect to the SCE at room temperature. Results and Discussion Monolayer Behaviour Surface pressure uersus area per molecule isotherms for mono- layers of the mixtures on a pure water subphase have been reported previously.' The isotherms of the pure [Yb(pc)(pc-)] showed large hysteresis, but this was progressively reduced on the increase of the stearic acid content in the mixture.The average area per molecule (A,) for the various mixtures, measured in the usual fashion, plotted against mole fraction of [Yb(pc)(pc.)] is shown in Fig.l(a). Also shown in Fig.l(a) J. MATER. CHEM., 1991, Vol. 1 (b'30-L J ., c Fig. 1 A graph of (a)area per molecule and (b)monolayer collapse rate versus mole fraction of [Yb(pc) (pc.)] in monolayers spread on a pure water subphase (x =experimental data, 0 =calculated data based on the area per molecule of [Yb(pc)(pc.)] and stearic acid being 70 and 20 A', respectively) are the calculated average A, based upon an A, of 20 and 70A2 for the individual molecules of stearic acid and [Yb(pc)(pc.)], respectively, and assuming the two phases to be irnmis~ible.~The negative deviation of the experimental curve from the theoretical plot indicates that the [Yb(pc)(pc.)] and stearic acid are indeed miscible.The stability of the com-pressed monolayers was measured at a constant surface pressure of 30 mN m-' and Fig.l(b) shows the percentage hourly fall in A, averaged over 2 h following compression for each of the mixtures. By this test it was found that the pure [Yb(pc)(pc.)] monolayers were only slightly less stable than those of stearic acid. However, the monolayers of the mixed materials exhibited collapse rates much less than those of the parent compounds. It is suggested that, arising from the miscibility, the [Yb(pc)(pc.)] molecules are aligning on the water surface with interspatial filling by the smaller stearic acid molecules, which prevent slippage and, hence, collapse.X-Ray Diffraction Analysis The d-spacings obtained from the X-ray diffraction spectra for the LB films of the 1 : 5 and 1 :9 mixtures are given in Table 1. None of these films had previously been oxidized or reduced in an electrochemical cell. The first-order maximum of the 1 :9 film is shifted slightly with respect to the second- and third-order maxima because of refractive effects.8 For this reason, whenever possible, peaks other than the first- order were used to calculate d-spacings.The maxima in these spectra are believed to be due to the fatty acid present in the mixed films. The results for the 1 : 5 and 1 :9 films gave an average d-spacing for a stearic acid bilayer of 41.3 f0.4 A. Owing to the difficulties encountered in depositing stearic acid m~ltilayers,~ consistent data for the bilayer spacing in stearic acid LB films are scarce. Reported values range from 40.2f0.3 3 50.9f0.3 3 diffraction maxima. 39.7 k0.2 A'' to 45.8 A," which encompasses the average compounds has already been ~bserved.'~-'~ In these cases an explanation was proposed in which the film underwent a reorganization into small crystallites soon after deposition. These crystallites distributed the diffraction intensity over a wide angular range making individual maxima weak and unobservable in the case of thin films.Such a recrystallization may also occur in LB films of pure [Yb(pc)(pc.)]. Striking changes in the X-ray spectra occurred for films which had been oxidized and reduced in an electrochemical cell. The average d-spacings of 50.65 f0.05 and 50.77 f0.09 I$ for the 1 :5 and 1 :9 films, respectively, (see Table 1) are consistent with values obtained from LB films of stearic acid salts.g.10,15-17 It would therefore appear that upon oxidation and/or reduction of the [Yb(pc)(pc.)] in these films the stearic acid in the film is converted to a salt, in this case potassium stearate because of the KC1 electrolyte employed. In order to accommodate the metal ions the film would have to change dimensions and this is confirmed by the X-ray data.However, no sign of stress in these cycled films was observed under the polarizing optical microscope. Similar d-spacing changes in the spectra of films of a 1 :99 [Yb(pc)(pc.)] tricosanoic acid mixture did not occur when the films were placed in an electrochemical cell and the redox potentials of the [Yb(pc)(pc.)] a~plied.~ Infrared Absorption Analysis The IR spectrum of a 47-layer LB film of cadmium stearate (subphase: 10-4moldm-3 Cd2+, pH 7) obtained by the reflection technique is shown in column (a)of Table 2. In this case, the acid bands of the stearic acid spectrum [see column (f) of Table21 have been replaced by the two carboxylate bands at 1540 and 1430 cm-'.These are respectively due to the asymmetric and symmetric CO stretching of the COT carboxylate group. The IR spectrum of a 51-layer 1 :5 mixture LB film obtained by the reflection technique is shown in column (b)of Table 2. The presence of stearic acid in the sample is best confirmed by the carbonyl stretch band and the CH2 progression bands, whilst the 1260, 1105-1020 and 815 cm-' bands indicate the presence of [Yb(pc)(pc.)]. This film was then placed in a two- J. MATER. CHEM., 1991, Vol. 1 Table 2 Minima wavenumbers (/cm-') in IR transmission spectra of deposited films (a) (b) (4 (4 (4 0 remarks 2960 m 2950 m 2960 m 2960 m 2950 w 2910 s 2910 s 2910 s 2910 s 2910 s 2920 s 2870 m 2860 w 2870 w 2850 vw 2840 s 2840 s 2850 s 2840 m 2840 s 2850 s 2660 vw 2320 m 2330 vw 1700 s 1700 s 1708 m 1710 s 1635 w 1620 w 1595 w 1580 w 1550 w 1540 m 1520 m 1505 w 1470 w 1470 w 1470 m 1472 w 1470 m CH2 6 1460 w 1460 m CH, 6 1430 s 1450 w 1430 m 1450 w 1430 w 1450 m 1432 m 1445 w OH 6, C-0 C0.i Vsymm v 1410 w 1410 w 1405 m 1405 m 1395 m 1395 w band progression.1375 w 1378 w (series of peaks) 1355 w 1360 m 1365 m 1348 w 1340 w 1330 m 1325 m 1320 vw 1320 w 1325 m 1310 w 1310 m 1315 m 1310 w 1310 w 1300 w 1295 m 1295 m 1295 m 1295 m 1295 m OH 6, C-0 v 1280 m 1280 w 1284 w 1275 m 1275 w 1266 m OH 6, C-0 v 1260 m 1260 s 1260 s 1260 s 1255 w 1240 m 1240 m 1240 m 1240 w 1235 w 1220 m 1220 m 1220 m 1224 m 1220 w 1212 w 1200 m 1200 m 1200 m 1196 w 1200 w 1185 m 1185 m 1185 m 1185 w 1170 vw 1165 w ll50vw 1140 vw 1100 m 1105 s 1050 w 1040 w 1020 w 1100s 1080 w 1050 w i:broad $m (broad peak) 1020 1020 peak) 1112m 1115w 1100 w 1070 w 935 w 940 w 890 m OH 6 820 w 815 m 810 m 810 s 785 vw 760 w 725 m 725 m 725 m CH2 doublet in solid 715 m 685 w phase with n L4.640 vw (a) Cadmium stearate, 47 layer LB film. (b)51 layer LB film of 1:5 mixture (c) 51 layer LB film of 1:5 mixture after oxidation. (d) 51 layer LB film of 1 :5 mixture after reduction. (e) Tricosanoic acid, 42 layer LB film. (f) Stearic acid mull.The labelling s (strong), m (medium), w (weak) and vw (very weak) indicate the band intensity on an arbitrary scale. electrode cell containing a 5% KC1 solution and the trum of the cadmium stearate film, column (a) of Table21. [Yb(pc)(pc.)] oxidized by applying +0.9 V. The film was Thus, there is no IR evidence for free acid in the mixed film returned to the neutral state by applying OV to obtain the after the reduction of the [Yb(pc)(pc.)]. spectrum shown in column (c) of Table 2. This spectrum is If salt formation is to occur in the mixed films, then a essentially the same as that of the virgin film. It is important simple application of the theory of electrostatics indicates that to note that the spectrum of the film after oxidation still this cannot happen when the [Yb(pc)(pc.)] is oxidized.Oxi- contains the carbonyl stretch at 1700cm-Upon re-insertion dation requires the film electrode to be positive, hence repel- of the film in the electrochemical cell the [Yb(pc)(pc*)] was ling any positive metal ions. Reducing the [Yb(pc)(pc.)] reduced by applying -0.9 V. After return to the neutral state requires the film electrode to be negative, therefore attracting by applying 0 V, the spectrum in column (d) of Table 2 was positive metal ions into the film. The spectra summarized in measured. The most striking feature of this spectrum is the Table2 do indeed show that conversion of the free acid to complete absence of the carbonyl stretch band [c. the spec- the salt only occurs when positive metal ions are attracted into the film.The COY asymmetric stretch band expected between 1500 and 1600 cm- 'was not observed but the sudden disappearance of the 1700 cm -' band and the appearance of the 1450 cm-' band (CO, symmetric stretch) confirms stear- ate formation. The symmetric stretch was the stronger of the two COT bands in the cadmium stearate film spectrum and would therefore be the easier to detect. The positions of these bands are dependent upon the nature of the positive coun- terion." For comparison, Table 2, column (e),shows the IR spectrum of a 42-layer LB film of tricosanoic acid. This spectrum is similar to that of stearic acid [column (f)] except that the band progression has shifted and the OH bending and CO stretching bands are now found at 1432 and 1266 cm-', respectively.Subsequent to the recording of this spectrum the reduction potential of the [Yb(pc)(pc.)] was applied to this film in a 5% CdCl, electrolyte (note that a CdC12 electrolyte was used in this case only). The IR spectrum of the film after this process overlaid exactly the spectrum of the virgin film, within experimental errors; thus indicating that the salt was not formed. The disappearance of the IR absorption bands of the free acid and the appearance of those of the CO, group upon reduction of the [Yb(pc)(pc.)] is conclusive evidence for potass- ium-ion inclusion into the mixed layer films. These data are further confirmed by the X-ray diffraction spectra, and it is reasonable to conclude that metal ions (only K+ is present) have complexed with the stearic acid in the film.A simple model of the reduction of the [Yb(pc)(pc.)] in a 1 :5 [Yb(pc)(pc.)]:stearic acid film would predict that only the cations required to balance the charges on these molecules would be attracted into the film. In this case 1/5 of the stearic acid protons might be expected to be substituted. The com- plete lack of the 1700 cm-' band in the spectrum of the film that had been reduced is therefore somewhat surprising as this suggests that all the protons have been replaced. A more realistic scenario is that during reduction K+ cations form a diffuse region within the film in order to balance the charge on the electrode. This enables the K+ ions to exchange with the protons of the stearic acid.When the [Yb(pc)(pc.)] is returned to the neutral state all the uncomplexed cations flow out leaving, in this case, potassium stearate as the new main phase. Electrochemical Analysis A typical CV of a 1 :0 film is shown in Fig. 2; here, a positive current is taken to represent reduction. (We have already given a full description of the UV-VIS absorption spectra of the 1 :0 and mixed material films in the various redox state^.^) The cycle was started with the application of +1.20 V, i.e. the film became oxidized. At the current peak RG the red oxidation product of the application of the initial potential was reduced back to the green neutral state. The reduction peak GB is due to the conversion of the neutral material to its blue reduced form.The steeply rising current A indicates the onset of the reduction of the tin oxide electrode to tin. Upon reversal of the voltage sweep the current dropped rapidly to B which represents the re-oxidation of the tin to tin oxide. The oxidation peak BG shows the conversion of the reduced [Yb(pc)(pc.)] to the green neutral state. Oxidation of the material to the red state is shown by the peak GR. These data, along with those for other [Yb(pc)(pc.)]:stearic acid ratios, are presented in Table 3. Nernstian behaviour is ruled out by the shape of the peaks which are broader than the theoretical 90 mV full-width-at- half-rnaxirn~m'~(peak widths are up to 500 mV FWHM). In addition, non-reversible behaviour is indicated for both the J.MATER. CHEM., 1991, Vol. 1 35-A 30-25-20-15-.4 10-L 5-L 3 0--5 --1 0--15--20-GR 1111111111111 -25 c 1.2 0.8 0.4 6.0 -0.4 -0.8 -1.2 potential vs. SCE/V Fig. 2 The cyclic voltammogram of an LB film of pure [Yb(pc)(pc.)] oxidation and reduction processes by the non-unity value of the ratios of the red-green:green-red (0.2) and blue-green :green-blue (0.31) peak currents.20 This irreversibility is further confirmed by the large differences between the red- green (140 mV) and blue-green (240 mV) current summits. For reversible processes with the number of electrons involved in the redox reaction equal to unity the theoretical difference would be expected to be 59.5 mV at 25 "C.The first two reduction scans of a virgin 1 :0 film are shown in Fig. 3. Note that the peak potential of the first reduction peak (-0.63 V) is more cathodic than the subsequent peaks. When the oxidation process was the first to be analysed for a virgin film the converse was found to be true; the first oxidation peak being always more anodic than subsequent oxidation cycles. This is in common with the findings for evaporated films of other bi~phthalocyanines.~*~~~~~It was found that the films could be made to revert partially to the virgin state by scanning the potential to the other redox system, for instance to the red-green transition after several cycles at the blue-green system. Upon cycling around the (in this case) blue-green system again the first reduction peak was more cathodic than subsequent peaks.Again the converse of this was also true. The CV of an LB film of the 1 :9 mixture is shown in Fig.4. The most striking feature of this voltammogram is the reduction peak at -0.53 V which occurred for the first cycle only. A second, smaller reduction peak is also visible in the first scan at ca. -0.7 V. In subsequent cycles only the second peak at ca. -0.7 V was recorded in this region. These CVs are typical of those taken of mixed material films. The redox potentials of the 1 :5 and 1 :9 films show striking differences when compared to those of the pure [Yb(pc)(pc.)] film. The green-blue transition shifts in potential by at least -0.2 V, from ca.-0.5 V to ca. -0.7 V or lower. This is accompanied by a new reduction peak appearing at the same potential as the green-blue process in the pure films. No J. MATER. CHEM., 1991, Vol. 1 Table 3 Peak potentials of pure and mixed material films mixing“ ratio RG~IV IP‘jV GBd/V BG‘/V GRJ~ I :og +0.48 -0.53 -0.25 +0.65 1:5 +0.46 -0.55 -0.80 -0.23 +0.59 I :9 +0.46 -0.53 -0.73 -0.23 +0.59 I :9 +0.46 -0.50 -0.75 -0.22 +0.58 I :9 +0.48 -0.53 -0.75 -0.23 +0.59 “[Yb(pc)(pc.)]:stearic acid. bRed-green transition. ‘Ion penetration peak. dGreen-blue transition. ‘Blue-green transition. ’Green-red transition. 8Pure [Yb(pc)(pc.)] LB film. 50 - 45- 40- 35- 30- 25.25- i! 20- 15- 1& 5- 0- -502 I 0’0 I I -02 I I -04 I I -06 I l-08 l l -10 LB film *Ol OOi:;-5 -104 I,,,,,, IIIIII 1.2 0.8 0.4 O!O -0.4 -0.8 -1.2 potential vs.SCEjV Fig. 4 The cyclic voltammogram of a 1 :9 LB film change in colour of the films was observed coincident with this new reduction peak. It was found that films of the 1 :5 and 1 :9 mixtures could be returned to their virgin states by drying. Upon re-immersing the slides in electrolyte and measuring their CVs the new peak appeared on the first cycle. This process could be repeated apparently indefinitely. Storing films in electrolyte did not produce the additional reduction peak upon re-measuring their CVs. Thus the new peak in the CVs of mixed material films would appear to be associated with a wetting mechanism; injecting ions and/or water into the largely hydro- phobic film.This explanation of ion penetration is reminiscent of that proposed for similar peaks produced by another phthalocyanine compound,22 though this material contained no fatty acid. Why the redox potential of the [Yb(pc)(pc.)J should also be displaced is not yet clear. It has previously been observed that in sublimed films of pure phthalocyanine materials the ion penetration peak is associated with an electrochemically induced phase change, which allows inclusion of the counter ion^.^^ This phase is believed to be similar to that formed by annealing a sublimed film.23,24 We note that in the case of the 1 :0 film such a peak is not observed. This probably indicates that when pure [Yb(pc)(pc.)] is deposited by the LB technique it is already in the correct phase for ion incorporation.The effect of covering an IT0 electrode with a fatty-acid film was investigated. A 49-layer Y-type LB film of tricosanoic acid was deposited onto an IT0 electrode and its CV recorded. The total absence of the redox currents seen for the plain potential reached -1.45 or +2.53 V. This effect has been reported for single monolayers of alkyl mercaptans (C12-C18) deposited onto a gold ele~trode.~’ In this case the monolayer was capable of suppressing gold oxidation by up to five orders of magnitude. Thus, the hydrophobic alkyl chains act as an insulator to the flow of solvated ions. Conclusion The A, and collapse rate data for the mixed material mono- layers given previously’ have been explained here.[Yb(pc)(pc.)] has been shown to be miscible with stearic acid. This accounts for both the negative deviation of the A, of the mixtures from theory and the increase in stability of the mixed monolayers with respect to the parent compounds. Although X-ray diffraction peaks attributable to the [Yb(pc)(pc.)J were not observed in the mixed material films, virgin films did produce the diffraction spectrum of stearic acid. When the [Yb(pc)(pc.)] in these films was oxidized and reduced the fatty-acid spectrum changed to that consistent with a fatty-acid salt. IR analysis showed that the salt was formed upon reduction of the [Yb(pc)(pc.)], attracting, in this case, potassium cations into the films. It has long been known that salts of certain metal ions can be difficult if not impossible to deposit.Blodgett” showed that copper or aluminium ions in the subphase at concentrations of lo-’ mol dm-3 com- pletely prevented film deposition. The postdeposition salt- formation technique described above may prove a viable method of producing LB films of compounds that would otherwise not deposit. The redox potentials of LB films of [Yb(pc)(pc.)] have been measured using cyclic voltammetry. These studies have shown that for mixed material films there exists a wetting or charge- injection mechanism associated with the presence of the fatty acid. The results for a pure fatty acid film indicate that the 976 J.MATER. CHEM., 1991, Vol. 1 fatty acid is non-conducting and electrolyte impermeable over the range of redox potentials of the [Yb(pc)(pc.)]. Earlier work,’ showed that films consisting of 1 :99 mixtures of [Yb(pc)(pc.)]-tricosanoic acid were not electrochromic. It was postulated that electrochromism could occur only if 3 4 5 G.C.S. Collins and D.J. Schiffrin, J. Electroanal. Chem., 1982, 139, 335. C.S. Frampton, J.M. O’Connor, J. Peterson and J. Silver, Dis-plays, 1988, 9, 174. M. Petty, D.R. Lovett, J.M. O’Connor and J. Silver, Thin Solid Films, 1989, 179, 387. regions of [Yb(pc)(pc.)] within the film were sufficiently close together so that conduction from one region to the next was possible and that this was not the case in the 1 :99 films.We are now able to expand on this postulate. It has now been shown using cyclic voltammetry that a pure tricosanoic acid film, which approximates to the 1 :99 mixture, indeed, does 6 7 8 9 S.A. Francis and A.H. Ellison, J. Opt. SOC. Am., 1959, 49, 131. M. Petty, Ph. D. Thesis, University of Essex, 1990. N.B. McKeown, M.J. Cook, A.J. Thompson, K.J. Harrison, M.F. Daniel, R.M. Richardson and S.J. Roser, Thin Solid Films, 1988, 159, 469. G.L. Clark, R.R. Sterrat and P.W. Leppla, J. Am. Chem. SOC., 1935, 57, 330. not conduct and also effectively blocks the electrolyte from 10 G.L. Clark, R.R. Sterrat and P.W. Leppla, J. Am. Chem. SOC., the IT0 electrode. This is further confirmed by the IR analysis of tricosanoic acid films, as these studies show that electro- chemical salt formation does not occur within a pure fatty- acid film.Hence, it is not sufficient for the [Yb(pc)(pc-)] to be present merely in the film; this material must also form a conducting matrix, as the co-matrix of fatty acid is both non- 11 12 13 14 15 1936, 58, 2199. K.B. Blodgett, J. Am. Chem. SOC.,1935, 57, 1007. R.H. Tredgold, A.J. Vickers, A. Hoorfar, P. Hodge and E. Khoshdel, J. Phys. D: Appl. Phys., 1985, 18, 1139. R.H. Tredgold, Rep. Prog. Phys., 1987, 50, 1609. A.J. Vickers, Ph. D. Thesis, University of Lancaster, 1984. K. Mizushima, T. Nakayama and M. Azuma, Jpn. J. Appl. Phys., conducting and water repellent. 16 1987, 26, 772. M. Pomerantz and A. Segrnuller, Thin Solid Films, 1980, 68, 33. 17 M. Prakash, J.B. Peng, J.B. Ketterson and P. Dutta, Thin’Solid Thanks are due to the British Technology Group for continued support on electrochromics to J.S., to SERC for providing a studentship to M.P., to the Royal Signals and Radar Establish- 18 19 Films, 1987, 146, L15. L.J. Bellamy, The Infrared Spectra of Complex Molecules, Chap-man and Hall, London, 3rd edn., 1986. F. Castaneda and V. Plinchon, J. Electronanal. Chem., 1987,236, ment, Malvern, for the loan of equipment, and to Z. Ali Adib at the University of Lancaster for assistance with X-ray analysis. 20 21 163. D.T. Sawyer and J.L. Robert:, Experimental Electrochemistry for Chemists, John Wiley, New York, 1974. F. Castaneda. V. Plinchon, C. Clarisse and M.T. Riou, J. Electroanal. Chem., 1987, 233, 77. 22 J. Silver, P. Lukes, S.D. Howe and B. Howlin, J. Muter. Chem., 1991, 1, 29. References 23 P.J. Lukes, Ph. D. Thesis, University of Essex, 1989. 1 M. Petty, D.R. Lovett, P. Townsend, J.M. O’Connor and J. Silver, J. Phys. D: Appl. Phys., 1989, 22, 1604. 24 25 J. Silver and M. Ahrnet, unpublished results. H.O. Finklea, S. Avery and M. Lynch, Langmuir, 1987, 3, 409. 2 P.N. Moskalev and I.S. Kirin, Opt. Spectrosc., 1970, 29, 220. Paper 1/02212A;Received 10th May, 1991