J. CHEM. SOC. FARADAY TRANS., 1994, 90(14), 2061-2064 2061 Influence of lmmobilising Anions on the Redox Switching of PolyaniIine V. W. Jones, M. Kalaji" and G. Walker Department of Chemistry, University of Wales, Bangor, UK LL57 2UW C. Barbero and R. Kotz Paul Scherrer lnstitut , CH-5232 Villigen PSI, Switzerland The immobilisation of heteropolyacids into polyaniline films has been studied using the probe beam deflection technique, X-ray photoelectron spectroscopy and scanning electron microscopy. The results indicate that the heteropolyanions modify the redox behaviour and morphology of polyaniline. The concept of using conducting polymers to enhance cata- lytic activity continues to arouse much interest.' One strategy is to introduce highly dispersed metal particles such as plati- num, palladium, iridium and ruthenium into the polymer matrix to improve catalytic efficiency of reactions such as hydrogen evolution,' oxygen reduction3 and the reduction of di~xygen.~ Another strategy is to introduce metal-containing ions as the dopants.' Recent reports indicated that heteropolyanions of the Keggin type can be incorporated into conducting polymer films, such as polyaniline, prepared chemically or The idea of trapping such large anions should have a marked effect on the mechanism of redox switching of the conducting polymers.In the case of polyanil- ine, it has been that the first oxidation is accom- panied by proton expulsion and anion insertion in order to maintain charge neutrality. Therefore the trapping of the anions should influence this process in a way which can decrease the intrinsic switching time between the insulating and conducting forms.In this study, we present results on polyaniline (PANI) films containing Keggin-type anions, prepared electrochemi- cally and analysed using cyclic voltammetry, probe beam deflection technique (PBD), X-ray photoelectron spectros- copy (XPS) and scanning electron microscopy (SEM). The data indicate that PANI films prepared with Keggin-type acids exhibit cation exchange in aqueous HCl solutions. Fur- thermore, the doping levels of the PANI films vary according to the heteropolyacid used. Experimental Polyaniline films were grown potentiodynamically on plati- num or indium-doped tin oxide (ITO) electrodes by cycling the potential between 0.4 and 0.95 V us.the saturated calomel electrode (SCE) at a sweep rate of 50 mV s-'. The syntheses were carried out under an argon environment from solutions containing 0.01 mol dm-3 aniline in 0.1 mol dm-3 solution of the Keggin-type acid in acetonitrile. The acids used were phosphomolybdic acid (PM), phosphotungstic acid (PT) and tungstosilicic acid (TS). The electropolymerisation was carried out in acetonitrile because of the poor solubility of the anilinium salt of the heteropolyanions in water. After the polymerisation, the electrodes were rinsed with acetonitrile followed by distilled water before studying their redox behav- iour in aqueous acidic solutions.The films prepared from solutions of Keggin acids will be referred to as PANI/Keggin whereas the polyaniline films prepared in hydrochloric acid will be referred to as PANI/Cl-. Film thicknesses were calcu- lated from the charge-to-thickness ratio measured by * To whom correspondence should be addressed ellip~ornetry'~and compared to thicknesses calculated from SEM. The PBD arrangement has been described previously. l4 The PBD signal, along with the cyclic voltammogram, were recorded on an X-Y recorder (BBC SE780). The electro- chemical cell used was a conventional optical glass cuvette (2 cm x 2 cm x 4 cm). A planar glass (1 cm wide) covered with 200 nm of gold or platinum was used as a working electrode. The counter electrode was a platinum wire and an SCE was used as a reference electrode.The distance between the counter and working electrodes was large enough to prevent any interference due to reactions occurring at the counter electrode with the laser beam. Thin films, less than 300 nm, were used to assure fast establishment of chemical equi- librium inside the film during potential scans. Probe beam deflection is a technique that measures the concentration gradient in front of the electrode by monitor- ing the refractive index gradient with a light beam.14 The electrochemical oxidation-reduction process is accompanied by counterion exchange with the bathing solution to main- tain electroneutrality. The ion concentration in the solution changes, creating a gradient of the refractive index normal to the electrode surface.A beam travelling parallel to the surface undergoes a deviation proportional to the concentration gra- dient; therefore deviation of the beam is proportional to the extent and direction of the ion flux. Positive deflection corre- sponds to insertion of ions in the film, while negative deflec- tion implies release of ions to the solution. Scanning electron micrographs (Hitachi S-520 SEM) were taken of PANI/Cl- and PANI/Keggin films grown on ITO. The films were first cycled in 1 mol dm-3 HC1, washed with triply distilled water and then dried under vacuum at 60°C prior to any measure- ments. The films were sputter coated with gold (5 nm) (Polaron equipment) to prevent charging of the glass.XPS investigations were carried out on polymer films deposited on gold electrodes. The samples were investigated in the pristine state because even slight Ar' sputtering altered the composition significantly. No charging effects were observed. The polymer-coated electrodes were cycled in 1 mol dm-3 HCl prior to recording XPS data to ensure that the ionic composition of the films corresponds to the electro- chemically active material and not to the as-prepared films. The films were examined in the oxidised state. In addition to the overview spectrum, the emission peaks of N Is, C Is, P 2p, W 4f, Si 2p, Mo 3d and 0 1s were investigated. Scattering cross-sections for quantitative analysis were taken from ref. 15. Results and Discussion Fig.1 shows a cyclic voltammogram and the corresponding probe beam deflectogram for a PANI/TS film obtained in 0.1 1000 800 600 400 % 3 200 3 ou -200 -400 I -600I I I I -0.2 0.0 0.2 0.4 0.6 0.8 EfV vs. SCE (a +Fig. 1 Cyclic voltammogram (-) and deflectogram .) o PANI/tungstosilisic recorded in 0.1 mol dm-3 HCl at a sweep rate of 10 mV s-' mol dm-3 HCl. The redox behaviour of PANI/TS resembles that of PANI films grown under 'normal' conditions, i.e. in 0.1 mol dm- HCl.'o.'l However, the corresponding deflec- togram exhibits a negative deflection accompanying the oxi- dation of the polymer which is indicative of the expulsion of ions (protons) from the polymer matrix. A positive deflection on the reverse sweep corresponds to the insertion of cations.This behaviour is in contrast to the deflectograms obtained for PANI/Cl- which exhibit a positive deflection during the first oxidation peak in 0.1 mol dm-3 HC1,16 indicating ion (anion) insertion in the film. PANI/PT (Fig. 2) exhibits similar behaviour to PANIDS in that proton expulsion and insertion accompany the oxida- tion and reduction, respectively. However, the cyclic voltam- mogram exhibits an anodic shift in the first oxidation 2oo 1 400150-100-200 50-J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 101 ,-lo00 500 '4. -500 -1 000 -0.2 0.0 0.2 0.4 0.6 0.8 EfV vs. SCE (aFig. 3 Cyclic voltammogram (-) and deflectogram .) c PANI/phosphomolybdic recorded in 0.1 rnol dm-3 HC1 at a sweep rate of 100 mV s-' potential. The peak separation for the first redox couple is less than that for PANIDS and PANI/Cl-.PANI/PM (Fig. 3), on the other hand, exhibits a remarkable change in the redox behaviour. Whereas PANI/Cl- shows an asymmetry in the shape of the cyclic voltammogram, the peak separation for the first redox couple in PANI/PM is negligible. More- over, the colour of the film in the reduced state is dark blue (Fig. 4), whereas PANI/Cl- is yellow. ''The electrochemistry of heteropolyacids has been previously studied" and no similar observation was noted which excludes the conclusion that what is observed in PANI/PM is solely due to some film formation by the phosphomolybdates. However, it is not pos- sible at this stage to exclude the fact that the redox chemistry observed is due to a combination of PANI switching and electroactivity of trapped heteropolyanions.This is further 1 0.i 0-% --50--200?!5 0 -1 00-400 -1 50-V -600 -200 / I I I I -0.2 0.0 0.2 0.4 0.6 C 3 EfV vs. SCE Fig. 2 Cyclic voltammogram (-) and deflectogram (. . .) of Fig. 4 Absorbance spectra of PANI/phosphomolybdic film on IT0 PANI/phosphotungstic recorded in 0.1 mol dm-3 HCI at a sweep taken at different potentials. Note the absorbance around 650 nm in rate of 100 mV s-' the reduced state which is due to the phosphomolymbdate anion. Table 1 XPS results obtained for the different PANI/Keggin films ~~ ratio PANI/PT PANI/PM PANIPS ~~ ~ W:P 11 (12) 0:P 47 (40) 30 (40)C:N 6 (6) 2.6 (6) N:P 6 30 Mo:P 9.5 (12) Si : W w:o N:W formula (C6HSN)6(PW1 204d3 -doping level, n 0.5 The expected value according to the acid or PANI formula is given in brackets.Cannot be determined. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 (a1 Fig. 5 Scanning electron micrographs of (a) PANI/CI -,(b)PANI/phosphotungstic, (c) PANI/tungstosilicic and (d) PANI/phosphomolybdic.C. All films were grown on IT0 supported by the fact that the blue colour exhibited by the film in the reduced state is most likely to be due to the phos- phomolybdic anion trapped within the films matrix. The reduction of the Mo6+ ions produces a proportion of the Mo5+ ions. Transfer of electrons from Mo5+ to Mo6+ ions is responsible for the intense 'charge-transfer absorption'. The corresponding deflectogram indicates proton expulsion upon oxidation of the polymer followed by anion insertion occurring at potentials above 0.3 V.The interpretation of the PBD results for all the films studied is coherent with the irreversible immobilisation of the heteropolyanion inside the polymer layer. In the reduced state, the negative charge of the immobilised heteropolyanion is compensated by protons. Upon oxidation, the negative charge of the immobilised ion is used to compensate for the positive charge in the polymer backbone, and the protons are expelled from the film. This contrasts with PANI/Cl- films in which the oxidation is accompanied by anion insertion.The difference in behaviour of PANI films prepared in the different heteropolyacids can be partially explained using the XPS data which also provide evidence for the immobilisation of the Keggin-type anions within the polymer matrix. Table 1 shows a summary of the XPS data obtained for PANI films grown in different heteropolyacid solutions, and indicates the ratio of the detected elements, the.deduced formulae of the polymer salt and the doping level (degree of anion charges per polymer ring). The XPS data show that the hetero- polyanions are present after the polymerisation and are retained inside the films even after cycling in an aqueous solution containing 1 mol dm-3 HCl. In the case of phos- photungstic and tungstosilicic acid, the amount of heterpoly- anion present is enough to compensate for the charge in the emeraldine state.In the case of phosphomolybdic acid, the amount is lower than is necessary to compensate for the charge in the half-oxidised state which explains the need for anion insertion as shown in the corresponding deflectogram above 0.3 V. The XPS data for PANI/PM exhibit a marked deviation from the expected values. It has been previously shown, using scanning tunnelling microscopy, that phosphomolybdic acid can undergo an electrochemical decomposition reaction to yield particles of smaller diameter (probably molybdenum oxide).” However, the kinetics of the process and the ratio of products to reactants are still under investigation.2 There-fore, the deviation from the expected values in the XPS mea- surements may be due to the presence of a mixture of phosphomolybdic acid and molybdenum oxide. The origin of the second redox couple in the PANI/PM is still unclear. Our initial studies on phosphomolybdic acid indicate two quasi-reversible reactions in the same potential window used for PANI/PM.However, the peak positions are different to those observed with PANI/PM which may suggest that the redox reactions of the phosphomolybdate anion can be mediated by electron transfer through the polymer network. Furthermore, this is supported by the fact that as the polymer is oxidised, the initial colour change is from blue to green, but changes over a period of time back to blue.22 This suggests that the polymer is initially oxidised, but is then involved in a charge-transfer process with the anions.The possibility of this charge-transfer process is still under investigation. The morphology of the polymer film is also changed by the dopant anion. Scanning electron microscopy (Fig. 5) shows evidence of compact morphological structures consisting of globular microspheroids except for PANI/tungstosilicic, which was found to have a smooth, undulating surface. It is worth noting here that PANI/tungstosilicic exhibited the highest electrical conductivity amongst all the films studied, including PANI/Cl-.22 The results clearly indicate that some anions can be immo- bilked within PANI film and that they can alter the electrical and optical behaviour of the host matrix.The switching time of such polymers is currently being studied using ultra- microelectrodes. The spectroscopic properties and the influ- J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 ence of anion guests in heteropolyanions on the behaviour of PANI films are also under unvestigation. Support from the SERC (M.K.), Pilkington (V.J.), PSI (M.K.) and the University of Wales/Bangor (M.K.) is gratefully acknowledged. The authors would like to thank B. Schnyder for carrying out the XPS measurements and the referees for their helpful comments. References 1 J. Simonet and J. Rault-Bertholet, Prog. Solid State Chem., 1991, 21, 1. 2 K. Kost, D. Bartak, B. Kazec and T. Kuwana, Anal.Chem., 1990,162,151. 3 C. Bose, S. Basak and K. Rajeshwar, 204th ACS Meeting, Wash-ington DC, 1992, Abstract No. 69. 4 T. Vork and E. Barendrecht, Electrochim. Acta, 1990,35, 135. 5 G. Bidan, M. Lapkowski and J. Travers, Synth. Met., 1989, 28, C113. 6 A. Pron, Synth. Met., 1992,46,277. 7 G. Bidan, E. Genies and M. Lapkowski, J. Chem. SOC., Chem. Commun., 1988,533. 8 M. Hasik, A. Pron, I. Kulszewicz-Bajer, J. Pozniczek, A. Biel-anski, Z. Piwowarska and R. Dziemaj, Synth. Met., 1993, 55-57, 972. 9 M. Kalaji, L. M. Peter, L. Abrantes and J. Mesquita, J. Electro-anal. Chem., 1989,274,289. 10 M. Kalaji, L. Nyholm and L. M. Peter, J. Electroanal. Chem., 1991,313,271. 11 M. Kalaji, L. Nyholm and L. M. Peter, J. Electroanal. Chem., 1992,325,269.12 M. Vuki, M. Kalaji, L. Nyholm and L. M. Peter, J. Electroanal. Chem., 1992,332,315. 13 R. Greef, M. Kalaji and L. M. Peter, Faraday Discuss. Chem. SOC.,1989,88, 277. 14 R. Kotz, C. Barbero and 0.Haas, J. Electroanal. Chem., 1990, 2%, 37, and references therein. 15 J. Yeh and I. Lindau, At. Data Nucl. Data Tables, 1985,32, 1. 16 C. Barbero, M. C. Miras, 0.Haas and R. Kotz, J. Electrochem. SOC., 1991, 138, 669. 17 M. Kalaji, V. W. Jones, C. Barbero and R. Kotz, Abstracts of the 44th Meeting of the International Society of Electrochemistry, Germany, 1993, Abst. PI. 7.15. 18 B. Keita and L. Nadjo, J. Electroanal. Chem., 1987,227, 77. 19 N. Greenwood and A. Earnshaw, Chemistry of the Elements, Pergamon Press, Oxford, 1986, p. 1185. 20 A. Kowal, Abstracts of the International Conference on Scanning Tunnelling Microscopy, Interlaken, Switzerland, 1991, Abst. PH/77. 21 A. Kowal, personal communication. 22 M. Kalaji and G. M. Walker, in preparation. Paper 4/00682H; Received 4th February, 1993