J. MATER. CHEM., 1991, 1(4), 503-506 Microstructural Studies on Polypyrrole Lars Madsen/ Bogumil N. Zaba,*b Marijke van der Sluijs/, Allan E. Underhill,b Kim Carneiroa a Danish Institute for Fundamental Metrology, Lyngby, Denmark Institute for Molecular and Biomolecular Electronics, University College of North Wales, Bangor, UK Scanning tunnelling microscopy has been used in an attempt to determine the surface roughness of polypyrrole films, prepared by electrodeposition on platinum electrodes. Films of nominal thickness 0.7 and 15 pm were used and compared with bare platinum surfaces using both scanning tunnelling microscopy and conventional scanning electron microscopy. The surface area for the thicker polypyrrole films was found to be ca. 2.5 times that of the underlying platinum surface.This should be contrasted with our previous, much higher, estimate of this parameter (40) from electrochemical measurements (J. Phys. D, 20, 1411). This discrepancy is discussed in terms of the accessibility of internal spaces to electrochemically reactive species. Keywords: Conducting polymer; Scanning tunnelling microscopy; Surface roughness; Modified electrode Conducting polymers are under consideration for a variety of applications in devices such as batteries, sensors and photovoltaic cells. In many of these applications, the polymer is in contact with electrolyte and must function as a charge- transferring electrode. It has previously been pointed out that the properties of conducting polymers are strongly influenced by their relatively high surface areas.’ In our previous work2 we have studied the polymer-electrolyte interface and have shown that electron-transfer reactions are enhanced at a polypyrrole-covered electrode compared with a bare platinum electrode.We ascribed this enhancement to two possible effects: (1) an increase in surface area caused by a greater degree of surface roughness in polypyrrole as compared to bare platinum; (2) an electrocatalytic effect, which enhances the rate of the electron-transfer reaction. By fitting values for model circuit parameters to impedance data over a wide range of frequencies, we were able to estimate the surface area effect. We estimate that the surface area of the polypyrrole is 40 times greater than that of the platinum.In order to evaluate the accuracy of such an estimate it is necessary to use an independent method to measure the surface area (or at least the ratio of the surface area of polypyrrole to that of platinum). In the present paper we present scanning electron micrographs and scanning tunnel- ling microscopy data from which we can estimate indepen- dently surface areas to compare with estimates based on electrochemical parameters. In addition the results presented here show that ‘thin’ polypyrrole films (<1 pm thick) are in fact patchy on the platinum surface, with areas of bare platinum exposed between islands of polymer. STM is a new powerful technique for investigating surfaces of conductors and semiconductor^.^ When a very sharp tip is brought close (ca.1 nm) to a surface there will be a significant overlap of the electronic wavefunctions of the tip and sample. Provided that there is a small voltage applied between the tip and sample a tunnelling current will start to flow. This current, is extremely sensitive to the distance between the tip and sample. This current is used, via a feed- back mechanism to keep the tip-sample distance constant. By using a piezoceramic it is possible to control the motion of the tip to within 0.01 A. It is now possible to scan the tip across the sample, and at the same time measure the vertical displacement, performed by the feedback mechanism, perpen- dicular to the sample. In this way we obtain a topographic map of the surface..~Yang et ~1 have published STM data on polypyrrole formation, but their data, which is at very high resolution and presented in the form of reconstituted images does not allow the type of area calculation which we make below. Fan and Bard’ have also presented some STM data, which is in a similar form to our own, and which is in good agreement with our data, but again, they did not attempt area calcu- lations of the type in which we are interested. Experimental The films prepared for this study were electropolymerised onto a platinum surface. The platinum electrodes used were wires of 0.5 mm diameter which were polished with 0.075 p alumina. The pyrrole monomer was distilled and dissolved in water at a concentration of 0.25 moldm-3 together with tetraethylammonium toluene-p-sulphonate at the same con- centration to act as the counterion during polymerisation.Polymerisation was carried out at constant potential (0.6 V with respect to a saturated calomel electrode) using a EG&G Model 273 potentiostat. A.c. impedance measurements were carried out as described in ref. 2 with potassium ferricyanide as the electroactive species in solution. Nominal polymer film thicknesses were calculated from the empirical relationship between charged passed (Q) in C cm-2 and film thickness (d) in pm first given by Jacobs et aL6 d=2.8Q. In the experiments reported here, the STM used was of our own design, consisting of an inchworm motor (Burleigh) for coarse adjustment and a piezoelectric tubescanner (EBL Com- pany) for fine movement.The tip was of electrochemically etched tungsten. The polypyrrole film samples were examined by STM directly on the Pt-wire surface. Results Fig. 1 shows SEM images and STM scans of the bare Pt-wire surface, which we used as the underlying electrode for polypyr- role film formation. The dislocations seen in the STM scans are reflected in the regular lines seen in the SEM image. The dislocations, which are highly regular, are of ca. 20nm in height, corresponding to some 60-70 platinum atoms. Such features cannot be resolved on the SEM and hence appear only as thin lines even at the highest magnification. Fig. 2 shows similar images and scans for an electrode on which a thin film (of ca.1 pm thickness) had been produced. It is apparent from both imaging techniques that a film of this thickness is in fact discontinuous on the surface. Areas J. MATER. CHEM., 1991, VOL. 1 Fig. 1 (a) Scanning electron micrograph of a clean, polished platinum-wire surface. The area of the specimen in view is 50pmx50p. (b) STM image of the same wire; the scale is in nm Fig. 2 Scanning electron micrographs of a ‘thin film’ of polypyrrole on a platinum surface. (a) Side view; (b) top view. Both views are of 50 pm x 50 pm areas. (c) and (d) show STM images of a similar polypyrrole film; the scale is in nm of bare platinum are seen as flat, featureless regions. Rising seen with a polypyrrole growth rising (within this field of from these flat regions are much rougher areas which in the view) to some 10 nm.These small features are therefore of a SEM images can be seen to have the characteristic features similar size to the dislocations in the underlying platinum. of growing polypyrrole films.7 Fig. 2(a) and 2(b) show these Fig. 2(d) shows an adjacent region in which the height of the features as side and top views respectively. Fig. 2(c) shows the polypyrrole film above the platinum surface has risen to ca. STM scans in which a flat area of bare platinum is clearly 70 nm above the adjacent platinum surface. J. MATER. CHEM., 1991, VOL. 1 In Fig. 3 similar images are again shown, this time for a thicker (10-20 pm) polypyrrole film. Both the SEM images and the STM scans now show no bare platinum surface. It appears that the growing polypyrrole structures have coalesced into a continuous, but very rough sheet.Again Fig. 3(a) and 3(b) show views from the side and top respect- ively. It is apparent that the features growing are as high as 10-15 pm. This is beyond the dynamic range of the STM tip and such gross features cannot be scanned. Nevertheless, Fig. 3(c) shows the surface of a thick polypyrrole film in an area which lacks such outgrowths. The large features that can be seen, which are some 50-60 nm in height, clearly rise from a surface which is much rougher than that seen in Fig. 2(c) or Fig. 2(d).The background from which the larger features rise is characterised by continuous fluctuations on the 10-15 nm scale.It is apparent from Fig. 3(c) and 3(d)that on an atomic scale, the film height can increase very abruptly (by ca. 10-50 nm in this case). In order to quantitate the surface area of the polypyrrole samples from the STM images, the digitised data for tip height were analysed in a portion of the image. Portions were chosen such that for a perfectly flat sample, the area (as defined by the x and y coordinates of the tip movement) would have been 2500 nm2. Fig. 4(a) shows a typical result for a bare platinum surface adjacent to an area in which a polymer film is growing. A portion of the bare platinum area is marked out for area determination. The actual surface-area determined from a point-by-point analysis of the heights was 2754 nm2, giving a roughness ratio value of 1.10. Fig.4(b) shows a similar marked segment from an area in which polypyrrole was obviously growing on the electrode. Here the results of an area analysis showed the surface area to be 6009 nm2, giving a roughness ratio value of 2.40. For 14 such segments taken from two differing electrodes, the average surface area found was 6700 nm2, corresponding to a surface roughness value of 2.68 & 0.26 (standard deviation). Discussion Our previous electrochemical characterisation of the polypyr- role-electrolyte interface suggested that the surface area of a thick film of polypyrrole was 40 times that of a bare platinum surface. This estimate was based on the values obtained for the double-layer capacitance and the Warburg Impedance,* both of which are linearly dependant on the surface area at which the electrochemical reaction takes place.For a thin film (0.7 pm) there was a much smaller apparent increase in surface area. The present SEM and STM experiments support the notion of an increased surface area. It is now apparent that the films which we treated previously as ‘thin’ were in fact partially covering the platinum surface. It is not surprising, therefore, that the increase in surface area for these films was small. There is an obvious discrepancy between the surface area increase as measured electrochemically (x 40) and that found by the STM (x2.2). This is probably due to the fact that the electrochemical ‘probe’, potassium ferricyanide, a small inorganic ion, is able to enter the ‘internal’ water-filled spaces of the polymer, whereas the STM tip is able to measure only the surface features which are directly beneath it (see Fig. 5).There remains an interesting feature of the electrochemical analysis which cannot be explained by surface-area effects alone. This is the effect of the polypyrrole coating on the charge-transfer resistance for the electrochemical reaction, Fig. 3 Scanning electron micrographs of a thick film of polypyrrole on a platinum surface. (a) Side view; (b) top view. Both views are of 50 pm x 50 pm areas. (c) and (d)show STM images of a similar polypyrrole film; the scale is in nm J. MATER. CHEM., 1991,VOL.1 Fig. 4 STM images of electrode surfaces selected for area calculation. The area marked with a heavy line was calculated. (a) Bare platinum surface; (b) polypyrrole-covered surface. In each case a perfectly flat surface would have given a surface area of 2500 nm2 Fig.5 Illustration of STM tip movement over the surface of a polypyrrole film. The tip cannot enter the ‘internal’ surfaces shown at (4,(b)and (4 which falls by some three to four orders of magnitude for a thick film compared with bare platinum, i.e. by a factor much greater than the surface-area increase, however this is meas- ured. This is evidence that the charge-transfer reaction is catalysed by the polypyrrole. It remains to be seen how general this catalytic effect might be. References 1 M. S. Wrighton, Science, 1986, 231, 32. 2 M. J. van der Sluijs, A. E. Underhill and B. N. Zaba, J. Phys. D, 1987, 20, 1411. G. Binning, H. Rohrer, Ch. Gerber and E. Weibel, Phys. Rev. Lett., 1982, 49, 57. 4 R. Yang, D.F. Evans, L. Christensen and W. A. Hendrickson, J. Phys. Chem., 1990,94, 61 17. 5 F-R. Fan and A. J. Bard, J. Electrochem. Soc., 1989,136,3216. 6 R. M. C. Jacobs, L. J. J. Janssen and E. Branderecht, Rec. Trau. Chim. Pays-Bas, 1984,103,275. 7 R. Qian, J. Qiu and D. Shen, Synth. Met., 1987, 18, 13. 8 J. R. Macdonald, Electround. Chem. Interface Electrochem., 1974, 129,115. Paper 0/04572A;Received 1 1 th October, 1990