- Email: [email protected]
polyvinylphosphate conducting polymer materials
Synthetic Metals 99 Ž1999. 191–199
Synthesis, characterisation and ion transport studies on polypyrrolerpolyvinylphosphate conducting polymer materials J.M. Davey, S.F. Ralph, C.O. Too, G.G. Wallace
)
Intelligent Polymer Research Institute, Department of Chemistry, UniÕersity of Wollongong, Wollongong, NSW 2522, Australia Received 1 October 1998; accepted 2 October 1998
Abstract A polypyrrolerpolyvinylphosphate ŽPPyrPVP. polymer was prepared electrochemically from aqueous solutions of pyrrole and polyvinylphosphate. The polymer was found to have a ratio of 10:1 for nitrogen to phosphorous, a water content of 28 " 3%, and a conductivity of 0.28 S cmy1. The polymer, as shown by SEM and AFM, has a rough amorphous morphology. Electrochemical studies using electrodes with thin polymer films indicated that oxidationrreduction of the material varied depending on whether potassium, sodium, calcium, lithium, magnesium or hydrogen phosphate ion was present. Transport of each of these ions across membranes composed of the above polymer was achieved using pulsed potential waveforms. q 1999 Elsevier Science S.A. All rights reserved. Keywords: Polypyrrole; Polyvinylphosphate; Ion transport; Membrane
1. Introduction The incorporation of polyelectrolytes as dopants in conducting polymers can be achieved directly via electropolymerisation. For example, pyrrole can be oxidised in the presence of a polyelectrolyte ŽPE u ., leading to polypyrrolerpolyelectrolyte composite materials ŽEq. Ž1...
Ž1. The presence of the polyelectrolyte during polymerisation can inhibit polymer deposition w1,2x. This is believed to be due to excess Žuncompensated. charge on the polyelectrolyte dopant resulting in more soluble oligomeric and polymeric structures being formed. Incorporation of polyelectrolytes into polypyrrole also gives rise to some unique chemical and physical properties in the resultant materials. For example, upon reduction of the polymer ŽEq. Ž2.. charge compensation is usually achieved exclusively via incorporation of cations, Xq, from the electrolyte w3,4x. ) Corresponding author. q61-42-21-3127; Fax: q61-42-21-3114; E-mail: [email protected]
Ž2. This is in contrast to what has been observed with polypyrrole containing low molecular weight dopants. On these occasions the small mobile anions leave the polymer allowing it to retain charge neutrality upon reduction. It has also been found that incorporation of polyelectrolytes into conducting polymers results in materials with more open porous structures, and high water contents w5x. The latter have been reported to be greater than 90% Žwrw. in some instances, and are attributed to uncompensated charge on the polyelectrolyte within the structure. Incorporation of polyelectrolytes into conducting polymers also provides a facile route for introduction of functional groups that are electroactive w6x, capable of complexing metal ions w1x or confer biocompatibility w5x. We have investigated the use of polyvinylphosphate as a dopant in polypyrrole. The effect of polyvinylphosphate on the polymerisationrdeposition process was studied, and the composition of the resultant polymer as well as its electrical, mechanical and ion transport properties examined. The latter are of particular interest for development of novel electrochemically controlled membranes as selec-
0379-6779r99r$ - see front matter q 1999 Elsevier Science S.A. All rights reserved. PII: S 0 3 7 9 - 6 7 7 9 Ž 9 8 . 0 0 1 9 1 - X
192
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
tive metal ion transport systems w7–16x. Prior to this work, there have been only a relatively small number of studies which focused on the metal ion transport properties of polypyrrolerpolyelectrolyte composites w3–5,17–22x.
phate. Determination of the thickness of membranes was performed using digital vernier calipers. Composite materials used in metal transport studies were prepared by depositing PPyrPVP onto platinum coated polyvinylidene fluoride ŽPVDF. membranes ŽMillipore..
2. Experimental
2.3. Techniques and instrumentation
2.1. Reagents
Elemental analyses on membranes were performed by the Microanalysis Unit at The Australian National University. The water content of large membranes Žtypically ; 60 cm2 area. was determined using the following procedure. After preparation the membranes were placed in an atmosphere of constant humidity Ž95%, 23.58C. and allowed to attain constant mass. They were then dried to constant mass in an oven at 608C, and the change in mass expressed as a percentage of the initial wet membrane. UV-visible and reflectance infrared spectra of membranes were obtained using a SHIMADZU Model UV-1601 spectrophotometer, and a BOMEM MB-100 Series spectrophotometer, respectively. Conductivity measurements were performed using large free standing membranes and the four point probe method. Scanning Electron Microscopy ŽSEM. and Atomic Force Microscopy ŽAFM. photographs were obtained using a Leica Cambridge 440 Scanning Electron Microscope and a Digital Multitude Nanoscope 3, respectively.
Pyrrole was obtained from Merck and distilled before use. Polyvinylphosphate Žsodium and ammonium salts; average molecular weight 133,000; average % phosphorylation of vinyl groupss 75%. was obtained from Polysciences. All inorganic reagents used were of AR grade and obtained from Ajax chemicals. All solutions were prepared using deionized Milli-Q water Ž18 M V cm.. 2.2. Preparation of polypyrroler polyÕinylphosphate polymers Electropolymerisation was performed in aqueous solutions containing 0.6 M pyrrole and 0.2% polyvinylphosphate, using an EG and G Princeton Applied Research Model 363 PotentiostatrGalvanostat. All solutions were thoroughly deoxygenated using nitrogen prior to use. A three electrode system consisting of a working electrode, AgrAgCl Ž3 M NaCl. reference electrode, and either a platinum mesh Ž100 linesrin.. or reticulated vitreous carbon ŽRVC. auxiliary electrode, was used in conjunction with a conventional one compartment electrochemical cell. All experimental data were collected using a MacLab ArD and DrA data collection system ŽAD Instruments.. Previous work with polypyrrolerpara-toluenesulfonate ŽPPyrpTS. polymers showed that galvanostatic methods were superior for reproducibly obtaining thin films and membranes with optimal mechanical and electrochemical properties w 10 x . Consequently, polypyrrolerpolyvinylphosphate ŽPPyrPVP. polymers were deposited onto platinum and stainless steel electrodes, gold coated Mylar and ITO coated glass using galvanostatic techniques. During these deposition processes the following resting potentials, 0.69, 0.75, 0.71 and 1.03 V, respectively were attained. Large free standing membranes grown onto stainless steel using low current densities were hard to remove. However, when high current densities were used to grow polymers bubbles formed on the surface of the membrane and overoxidation of the polymer occurred. As a result a current density of 1 mA cmy2 was used for all membrane preparations. After completion of electrodeposition the membranes were thoroughly rinsed with deionized water prior to analysis. Microanalyses, water content measurements, AFM and SEM studies were performed on membranes grown on a stainless steel electrode Ž60 cm2 area. using aqueous solutions containing 0.6 M pyrrole and 0.2% polyvinylphos-
2.4. Cyclic Õoltammetry studies Cyclic voltammetry studies were conducted in a three electrode electrochemical cell using a AgrAgCl Ž3 M NaCl. reference electrode, platinum auxiliary electrode, and PPyrPVP polymer deposited onto a platinum working electrode. The potential scan range was q0.45 V to y0.8 V, and the scan rate 100 mV sy1 . 2.5. Ion Transport Studies Metal ion transport studies were performed using a single membrane transport cell described previously w10x. A pulsed potential waveform with a 50 s pulse width, and a potential range of q0.45 V to y0.80 V was used. During transport studies aliquots for metal ion analysis were removed every 30 min when no potential was applied, and every 15 min when a potential was applied. The volume of sample removed varied from 0.2 ml to 2.0 ml depending on the sensitivity of the technique used to quantify the analyte and the rate of transport. Feed solutions contained 0.1 M metal ion, while the permeate solution contained initially 0.1 M electrolyte. The electrolyte used was NaNO 3 in all cases except when sodium transport was being examined, in which case it was KNO 3 . Atomic absorption spectroscopy ŽAAS. was used to determine metal ion concentration.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
193
Fig. 1. Ža. Cyclic voltammogram of an aqueous solution containing 0.6 M pyrrole and 0.2% PVP at a platinum disc electrode. Scan rate s 50 mV sy1 . Žb. Chronopotentiogram for PPyrPVP grown galvanostatically on a platinum disc electrode using 1 mA cmy2 current density and a solution containing 0.6 M pyrrole and 0.2% PVP. Žc. Chronoamperogram for PPyrPVP grown potentiostatically on a platinum disc electrode at a potential of 0.72 V using a solution containing 0.6 M pyrrole and 0.2% PVP.
194
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Anion transport studies were conducted in exactly the same manner as for metal ion transport. However, to determine hydrogen phosphate ion concentration, a colorimetric method using molybdate and ascorbic acid as reductant was employed w23x.
3. Results and discussion 3.1. Polymer growth The PPyrPVP polymer was grown using potentiodynamic, galvanostatic and potentiostatic techniques. Potentiodynamic methods showed an increase in current as the number of potential cycles was increased, indicating that deposition of the polymer occurred ŽFig. 1a.. From the first scan the minimum oxidation potential of the monomer was determined to be q0.70 V. When galvanostatic methods were used for growing the polymer a steady state potential was attained. For the
typical chronopotentiogram illustrated in Fig. 1b, the steady state potential is q0.70 V. The chronopotentiogram illustrated in Fig. 1b showed a significant decrease in growth potential with time, demonstrating that the polymer grown was conducting. In addition, it also indicated that initiation of polymer growth required a higher potential of q0.80 V, as expected. It is well known that monomer oxidises more readily on the deposited polymer than on a bare metal or carbon surface. This may be due to slow diffusion of the large polyelectrolyte to the electrode surface, or the large number of anionic sites on the polyelectrolyte. The latter would require that more pyrrole monomer is present around the polyelectrolyte before a polymer could be formed. Varying the concentration of pyrrole present in the monomer solution had little effect on the steady state potentials of polymers grown galvanostatically. Fig. 1c illustrates a typical chronoamperogram obtained during polymerisation at constant potential Žq0.72 V., and shows an increase in current with time indicative of polymer deposition.
Fig. 2. Reflectance FTIR spectra of ŽA. free polyvinylphosphate, and ŽB. PPyrPVP polymer galvanostatically deposited Ž0.71 V. onto gold coated mylar from a solution containing 0.6 M pyrrole and 0.2% PVP.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Fig. 3. Scanning electron micrographs of a PPyrPVP composite membrane galvanostatically deposited Ž0.75 V. onto stainless steel from a solution containing 0.6 M pyrrole and 0.2% PVP: Ža. solution side of membrane, Žb. electrode side of membrane.
3.2. Chemical and physical characterisation of composites The results of replicate C, H, N and P analyses performed on membranes grown under identical conditions showed a surprisingly high ratio Ž10:1. of nitrogen to phosphorous. With small anionic dopants the ratio of pyrrole to dopant molecules is typically 3:1 or 4:1. However, much higher values Žup to 9:1. have been reported for polypyrrole doped with bulky anions w24x. This indicates that the size and shape of dopant molecules can have a significant effect on polymer structure. The results of the current study indicate that the positive charge on the polypyrrole chain is spread out over a large number of monomer units. However, this assumes that the phosphate groups present in the dopant are only singly deprotonated, whereas if each of them were doubly deprotonated the positive charge would be spread over only 5 pyrrole units. Although the average p K a2 for polyvinylphosphate is unknown, it is unlikely that under the pH conditions employed for polymerisation in this study ŽpH ; 3. a significant proportion of the phosphate groups present would have been doubly deprotonated. The water content of polymer membranes was found to be 28 " 3%. Even higher water contents have been found previously for composites prepared from polypyrrole and
195
polyelectrolytes such as poly-2-acrylamido-2-methylpropanesulfonic acid w1x. The latter materials occupy large volumes and are able to undergo reversible hydrationrdehydration. The relatively high water content found for the PPyrPVP polymer suggests that this material may also have an open, porous structure which would facilitate movement of ions to and from the polymer. The reflectance infrared spectrum of sodium polyvinylphosphate is illustrated in Fig. 2A, and contains a broad absorption centred at approximately 1650 cmy1 , as well as absorptions at 1400, 1290, 1099 and 927 cmy1 . The three lowest energy absorption bands are assigned to P5O and P–O–C stretching modes of the phosphate group, while the two higher energy bands are attributed to bending andror deformation modes of the carbon and hydrogen atoms in the polymer backbone w25–27x. The most intense absorption in the reflectance infrared spectrum of PPyrPVP illustrated in Fig. 2B is a very broad band centred at 1558 cmy1 . This absorption band, which is absent from Fig. 2A, is assigned to the C5C stretching mode of the pyrrole groups w25–27x. Absorptions at 1034 and 921 cmy1 are again attributed to vibrational modes due to the phosphate groups. The lower frequencies of these absorption bands in the composite polymer reflect electrostatic interactions between the polypyrrole backbone and the dopant. Fig. 2B also contains a broad absorption extending between 1368 cmy1 and 1188 cmy1 , which is assigned to overlapping bands due to vibrational modes of both the phosphate and pyrrole groups. Scanning electron micrographs of the solution and electrode sides of a PPyrPVP membrane are illustrated in Fig. 3a and b, respectively. Both indicate that the membrane surface has a rough, amorphous morphology. Atomic force microscopic ŽAFM. studies of the membrane provided further support for this conclusion. By using AFM, the root mean square surface area of nodules on the membrane surface was determined to be 890 nm2 , and the average peak to valley distance 4.1 mm ŽFig. 4.. The effect of monomer concentration on membrane conductivity and thickness is presented in Table 1. In this
Fig. 4. Atomic force micrograph of the solution side of a PPyrPVP composite membrane galvanostatically deposited Ž0.75 V. onto stainless steel from a solution containing 0.6 M pyrrole and 0.2% PVP.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
196
Table 1 Conductivities and thicknesses of membranes grown galvanostatically using aqueous solutions containing 0.2% polyvinylphosphate and different pyrrole concentrations wPyrrolex ŽM.
Conductivity ŽS cmy1 .
Thickness Žmm.
0.2 0.3 0.4 0.5 0.6
0.04 0.08 0.11 0.17 0.28
3.0"0.3 4.0"0.3 5.5"0.4 7.0"0.4 9.0"0.5
Ž3b. study it was found that, when growing to the same charge Ž3.2 mC. solutions containing the highest pyrrole concentration examined, 0.6 M, gave membranes that were thicker and had the highest conductivities. The overall trend suggests that even at the highest pyrrole concentration examined, 0.6 M, there was insufficient monomer present in solution to optimise polymer formation. Instead, the polymer may have contained significant amounts of low molecular weight pyrrole oligomers. Such species would not be large enough to be doped with polypyrrole, providing an explanation for the low degree of dopant incorporation suggested by microanalysis. This may also explain the relatively low conductivities observed for these polymers. The highest conductivity observed for a PPyrPVP membrane was 0.28 S cmy1 , which is several orders of magnitude less than values typically found for polypyrrole membranes Ž10–100 S cmy1 . w10x. Another possible explanation for the low conductivity may be the very large size of the polyelectrolyte dopant, as it has been shown previously that large anions influence polymer chain packing in conducting polymers, and increase interchain and interparticle hopping distances. w24x Further support for the low conductivity of the PPyrPVP polymer was provided by absorption spectroscopic studies performed on films grown onto IndiumrTinrOxide ŽITO. coated glass. Fig. 5 shows the absorption spectra of PPyrPVP films exposed to solutions with different pH values for 2 h. In each spectrum there is no evidence of a free carrier tail, which is consistent with the low conductivities of these materials determined by the four point probe method.
Ž3c. When the polymer backbone was electrochemically reduced the overall polymer became negatively charged due to the immobile polyvinylphosphate counteranion ŽEq. 3a.. In order to maintain overall charge neutrality, sodium ions from solution were incorporated into the polymer ŽFig. 6, peak A, Ep s y0.53 V.. On re-oxidation of the polymer ŽEq. 3b., the polymer backbone became positively charged, and overall charge neutrality was achieved by both expulsion of sodium ions ŽFig. 6, peak B, Ep s y0.31 V. and incorporation of chloride ions from solution ŽFig. 6, region C.. When the polymer was reduced during subsequent cycles an additional feature was seen in the cyclic voltammogram ŽFig. 6, region D. which corresponded to expulsion of previously incorporated chloride ions ŽEq. 3c.. Polymer oxidation was therefore accompanied by two competing processes: expulsion of cations and incorporation of anions. The latter process may lead to stabilisation of the polymer backbone through electrostatic interactions with incorporated chloride ions, leaving the polyvinylphosphate and sodium ions to interact predominantly with each other. When the supporting electrolyte was varied the potentials of features in the cyclic voltammogram attributable to
3.3. Electrochemical characterisation The cyclic voltammogram of a PPyrPVP coated platinum electrode immersed in 0.1 M NaCl is illustrated in Fig. 6. The possible processes occurring are described below.
Ž3a.
Fig. 5. UV-vis spectra of PPyrPVP composites galvanostatically deposited Ž1 mA cmy2 . onto ITO coated glass from a solution containing 0.6 M pyrrole and 0.2% PVP, and exposed to solutions with different pH values. Solution 1: distilled water; Solution 2: pH 2 ŽHCl.; Solution 3: pH 12 ŽNaOH.; Solution 4: pH 6 Žphosphate buffer.; Solution 5: pH 7 Žphosphate buffer.; Solution 6: pH 8 Žphosphate buffer..
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
197
porting electrolyte used was NaCl or KCl. Furthermore, the cation incorporation peaks seen in cyclic voltammograms obtained using solutions containing divalent metal ions were present at more positive potentials, suggesting that a stronger driving force existed for incorporation of divalent metal ions into the polymer than univalent metal ions. 3.4. Transport studies
Fig. 6. Cyclic voltammogram of a PPyrPVP coated platinum electrode prepared galvanostatically Ž1 mA cmy2 . from a solution containing 0.6 M pyrrole and 0.2% PVP, after immersion in 0.1 M NaCl. Scan rate 100 mV sy1 . Region A s cation insertion. Region Bs cation expulsion. Region C sanion insertion. Region Dsanion expulsion.
incorporation and expulsion of cations changed. Table 2 summarises the results of these experiments. Both the cathodic peak potentials EpŽc. and anodic peak potentials EpŽa. vary depending on the electrolyte used, indicating that the PPyrPVP composite is sensitive to the identity of cations and anions present in solution. Incorporation and expulsion of Liq, Naq and Kq to and from the PPyrPVP polymer was clearly demonstrated by cyclic voltammetry. This was also clearly seen with Cly and NOy 3 , despite the possibility that the large immobile polyvinylphosphate dopant would inhibit incorporation of other negatively charged species. However, anion incorporation and expulsion was not generally observed, as H 2 PO4y, pTSy, and SO42y ions did not appear to interact with the polymer. The cyclic voltammograms of polymer coated electrodes immersed in solutions containing either CaŽNO 3 . 2 or MgŽNO 3 . 2 did not show peaks due to incorporation and expulsion of these cations. Similar behaviour was observed previously for polypyrrole doped with perchlorate from tetrabutylammonium perchlorate, after immersion in solutions containing either BaCl 2 or MgCl 2 w28x. Cyclic voltammograms of the latter solutions showed much smaller peaks due to cation insertion than when the sup-
Metal ion transport studies were undertaken using the PPyrPVP polymer deposited onto PtrPVDF composite membranes using galvanostatic techniques. Before commencing these experiments it was necessary to grow membranes that did not allow water to pass freely through. Fig. 7 illustrates the effect that varying the charge density used to grow membranes had on the flux of water through the membrane. It can be seen that a charge of 1680 mC for each cm2 of membrane resulted in a membrane that was impermeable to water. This charge density was then used to grow all membranes used in transport experiments. The cations potassium, sodium, lithium, calcium and magnesium were all transported across the membrane ŽFig. 8., as was hydrogen phosphate ion. In all cases no transport was observed in the absence of an applied electric potential ŽFig. 8, region A., or on application of a constant negative or positive potential to the membrane. However, immediately after application of the pulsed potential waveform to the membrane was commenced, transport of ions was observed in all cases ŽFig. 8, region B.. After an initial dramatic increase in the concentration of metal ion or phosphate anion in the permeate side, the concentration subsequently increased at a slower rate. After applying the pulsed potential waveform for 2 h the concentration of metal ion or phosphate ion in the permeate side approached steady state values. No further significant transport was observed after application of the pulsed potential waveform was ceased at this point ŽFig. 8, region C..
Table 2 Sensitivity of PPyrPVP coated platinum electrodes to different ions: anodic peak potentials EpŽa . and cathodic peak potentials EpŽc. obtained from cyclic voltammograms of electrodes immersed in solutions containing different supporting electrolytes Supporting electrolyte Ž0.1 M.
EpŽc .1 ŽV.
EpŽa.1 ŽV.
EpŽa.2 ŽV.
EpŽc.2 ŽV.
NaNO 3 KNO 3 LiNO 3 NaCl NapTS NaH 2 PO4 Na 2 SO4
y0.50 y0.45 y0.50 y0.50 y0.35 y0.30 y0.30
y0.45 y0.40 y0.40 y0.30 y0.20 y0.20 y0.25
0.05 0.10 0.05 0.10 – – –
y0.10 y0.05 y0.05 y0.15 – – –
Fig. 7. Effect of charge passed during growth on the permeability to water of PPyrPVP coated PVDFrPt membranes prepared galvanostatically using a solution containing 0.6 M pyrrole and 0.2% PVP.
198
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Fig. 8. Electrochemically controlled transport of Kq, Naq, Ca2q, Liq and Mg 2q across a PPyrPVP composite membrane. Feed solution: 0.1 M metal ion Žas the nitrate salt.. Permeate solution: 0.1 M NaNO 3 or KNO 3 Žfor Naq transport experiments.. Region A: No applied potential. Region B: Applied potential pulsed between y0.8 V to q0.4 V; pulse width 50 s. Region C: No applied potential.
The effect of varying the duration of the pulsed potential waveform on ion transport was investigated using potassium ion. In one experiment, after checking that no potassium transport occurred after 1 h when no potential was applied, the same pulsed potential waveform as used above was applied for 30 min then stopped. During application of the potential waveform the concentration of potassium in the permeate phase increased sharply in an analogous fashion to that shown in Fig. 8. However, immediately after the potential was removed potassium transport ceased and the concentration of potassium ions in the permeate phase remained constant. In a second experiment the pulsed potential waveform was applied for a total of 4 h, again after ensuring that no transport occurred in the absence of electrochemical stimulus. After 2 h the concentration of potassium ion in the permeate phase was the same as that illustrated in Fig. 8. However, during the following 2 h there was no further significant increase in the amount of potassium transported. This indicates that the flux of potassium ions across the membrane reached an equilibrium state after 2 h. In order to achieve further transport of metal ion across the membrane it would be necessary to remove the metal ions present in the permeate phase. The average fluxes observed for each ion examined are presented in Table 3 together with the ionic radius w29x for the hydrated ion. Overall the fluxes observed in this study are comparable to values observed for polypyrrole polymers containing simple dopant molecules w10,12,13,30x. The average fluxes observed for the sodium and potassium ions were comparable to each other, and an order of magnitude greater than that for lithium. This may be due to the significantly greater size of the hydrated lithium ion compared to those of sodium and potassium, which restricts its passage through molecular-sized pores in the membrane. Alternatively, it may be due to stronger interactions between incorporated lithium and the polyvinylphos-
phate dopant. Previously it had been noted that the flux of alkali metal ions through membranes composed of either PPyrpTS or PPyrdodecylsulphate decreased overall in descending order of the periodic table w30x. This trend was attributed to the greater porosity of these membranes to the smaller hydrated ions of the heavier elements. The flux of calcium through the membrane was slightly less than that of lithium. In view of the similar hydrated ionic radii for these metal ions, this also suggests that the rate of transport was determined to a large extent by the size of the hydrated ion relative to that of pores in the membrane. This is supported by the significantly lower flux observed for magnesium, which has the largest hydrated metal ion radius of the metals examined, compared to that for calcium. However, it should also be remembered that the lower fluxes observed for calcium and magnesium may also be due to higher affinities for the dopant as a result of their greater charge. This would retard release of the metal ion upon polymer reoxidation compared to when a univalent metal ion was present. Despite the relatively small size of its hydrated ion, the flux of hydrogen phosphate ion observed was very low. This is perhaps not surprising since for transport to occur the anion must first be incorporated into the re-oxidised polymer, a process not likely to be favoured by the presence of a large, immobile, negatively charged dopant in the structure of the polymer. The PPyrPVP polymer may therefore be considered as functioning predominantly as a cation exchange material. The transport selectivity factor Ž s . is given by:
ss
FluxMq FluxXq
It had been shown previously that the KqrCa2q selectivity factor decreased as the size of the dopant incorporated into polypyrrole increased w13x. This suggested that larger dopants had a lower ability to interact selectively with these metal ions. Further evidence for this is provided by the current study. For example, the KqrCa2q selectivity factor determined in the current work was 13.2, which is seven times less than that observed previously Ž94. for polypyrrolerpara-toluenesulfonate w13x. Furthermore, the KqrNaq selectivity factor for the latter polymer Ž3.1. w12x was also significantly greater than that observed in this
Table 3 Average flux values for ions during transport experiments Ion
Flux Žmol sy1 cmy2 .
Effective ionic ˚ . w29x radius ŽA
Kq Naq Liq Ca2q Mg 2q H 2 PO4y
5.74=10y9 5.63=10y9 4.83=10y10 4.36=10y10 1.03=10y11 5.09=10y12
3 4 6 6 8 4
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
study Ž1.02.. Overall this tends to suggest that polypyrroles containing polyelectrolyte dopants may not necessarily offer any advantages over similar polymers containing simple dopants for selectively transporting metal ions.
4. Conclusion A polypyrrolerpolyvinylphosphate polymer was successfully prepared electrochemically and shown to have a ratio of 10:1 for pyrrole repeat unit to charged phosphate group, a relatively high water content, low conductivity, and a rough amorphous morphology. The use of a large relatively immobile polyelectrolyte dopant gave membranes composed of this material an open porous structure. Despite the low conductivity, electrochemical studies indicated that the polymer was responsive to a number of different cations and phosphate ion. Transport of these ions and HPO42y was achieved by applying a pulsed potential waveform to a single membrane transport cell. The fluxes of metal ions across the membrane followed the sequences: Kq) Naq)) Liq for the group 1 metals, and Ca2q)) Mg 2q for the group 2 cations.
Acknowledgements We thank Dr. Ashton Partridge and Industrial Research Limited, New Zealand, for helpful discussions and financial support of this project.
References w1x G.G. Wallace, H. Zhao, C.O. Too, C.J. Small, Synth. Met. 84 Ž1997. 323. w2x Y.H. Liao, K. Levon, Macromol. Rapid Commun. 16 Ž1995. 393. w3x K. Naoi, M. Lien, W.H. Smyrl, J. Electrochem. Soc. 138 Ž1991. 440.
199
w4x X. Ren, P.G. Pickup, J. Phys. Chem. 97 Ž1993. 5356. w5x A.J. Hodgson, K. Gilmore, C. Small, G.G. Wallace, I.L. MacKenzie, T. Aoki, N. Ogata, Supramolecular Science 1 Ž1994. 77. w6x C. Lee, M.H. Lee, Y.K. Kang, B.G. Moon, S.B. Rhee, Synth. Met. 55 Ž1993. 1119. w7x P. Burgmayer, R.W. Murray, J. Am. Chem. Soc. 104 Ž1982. 6139. w8x P. Burgmayer, R.W. Murray, J Electroanal. Chem. 147 Ž1983. 339. w9x P. Burgmayer, R.W. Murray, J. Phys. Chem. 88 Ž1984. 2515. w10x H. Zhao, W.E. Price, C.O. Too, D. Zhou, G.G. Wallace, J. Membr. Sci. 119 Ž1996. 199. w11x H. Zhao, W.E. Price, G.G. Wallace, J. Electroanal. Chem. 334 Ž1992. 111. w12x A. Mirmohseni, W.E. Price, G.G. Wallace, H. Zhao, J. Int. Mater. Syst. Struct. 4 Ž1993. 43. w13x H. Zhao, W.E. Price, G.G. Wallace, J. Membr. Sci. 87 Ž1994. 47. w14x H. Zhao, W.E. Price, P.R. Teasdale, G.G. Wallace, Reactive Polymers 23 Ž1994. 213. w15x A. Mirmohseni, W.E. Price, G.G. Wallace, J. Membr. Sci. 100 Ž1995. 239. w16x D. Zhou, H. Zhao, W.E. Price, G.G. Wallace, J. Membr. Sci. 98 Ž1995. 173. w17x A. Wojda, K. Maksymiuk, J. Electroanal. Chem. 424 Ž1997. 93. w18x T. Shimidzu, A. Ohtani, K. Honda, J. Electroanal. Chem. 251 Ž1988. 323. w19x Y.-J. Qiu, J.R. Reynolds, Polymer Science and Engineering 31 Ž1991. 417. w20x C.K. Baker, Y.-J. Qiu, J.R. Reynolds, J. Phys. Chem. 95 Ž1991. 4446. w21x M. Lien, W.H. Smyrl, J. Electroanal. Chem. 309 Ž1991. 333. w22x M. Po, J.R. Reynolds, J. Chem. Soc. Chem. Commun., Ž1993. 258. w23x A.E. Greenberg, L.S. Clescen, A.D. Eaton, ŽEds.., Standard Methods For The Examination Of Water And Wastewater, 18th. edn., American Public Health Association, Washington D.C., 1992. w24x K.M. Cheung, D. Bloor, G.C. Stevens, J. Mater. Sci. 25 Ž1990. 3814. w25x K. Nakamoto, Infrared And Raman Spectra Of Inorganic And Coordination Compounds, 3rd. edn., Wiley, New York Ž1978.. w26x C.J. Pouchert, ŽEd.., The Aldrich Library of IR Spectra, Sigma-Aldrich, Milwaukee Ž1981. w27x G. Socrates, Infrared Characteristic Group Frequencies, Wiley, Chichester, 1980. w28x R. John, G.G. Wallace, J. Electroanal. Chem. 354 Ž1993. 145. w29x J.A. Dean, ŽEd.., Lange’s Handbook of Chemistry, 13th edn., McGraw-Hill, 1985, Table 5-2. w30x A.C. Partridge, C. Milestone, C.O. Too, G.G. Wallace, J. Membr. Sci. 132 Ž1997. 245.
Synthesis, characterisation and ion transport studies on polypyrrolerpolyvinylphosphate conducting polymer materials J.M. Davey, S.F. Ralph, C.O. Too, G.G. Wallace
)
Intelligent Polymer Research Institute, Department of Chemistry, UniÕersity of Wollongong, Wollongong, NSW 2522, Australia Received 1 October 1998; accepted 2 October 1998
Abstract A polypyrrolerpolyvinylphosphate ŽPPyrPVP. polymer was prepared electrochemically from aqueous solutions of pyrrole and polyvinylphosphate. The polymer was found to have a ratio of 10:1 for nitrogen to phosphorous, a water content of 28 " 3%, and a conductivity of 0.28 S cmy1. The polymer, as shown by SEM and AFM, has a rough amorphous morphology. Electrochemical studies using electrodes with thin polymer films indicated that oxidationrreduction of the material varied depending on whether potassium, sodium, calcium, lithium, magnesium or hydrogen phosphate ion was present. Transport of each of these ions across membranes composed of the above polymer was achieved using pulsed potential waveforms. q 1999 Elsevier Science S.A. All rights reserved. Keywords: Polypyrrole; Polyvinylphosphate; Ion transport; Membrane
1. Introduction The incorporation of polyelectrolytes as dopants in conducting polymers can be achieved directly via electropolymerisation. For example, pyrrole can be oxidised in the presence of a polyelectrolyte ŽPE u ., leading to polypyrrolerpolyelectrolyte composite materials ŽEq. Ž1...
Ž1. The presence of the polyelectrolyte during polymerisation can inhibit polymer deposition w1,2x. This is believed to be due to excess Žuncompensated. charge on the polyelectrolyte dopant resulting in more soluble oligomeric and polymeric structures being formed. Incorporation of polyelectrolytes into polypyrrole also gives rise to some unique chemical and physical properties in the resultant materials. For example, upon reduction of the polymer ŽEq. Ž2.. charge compensation is usually achieved exclusively via incorporation of cations, Xq, from the electrolyte w3,4x. ) Corresponding author. q61-42-21-3127; Fax: q61-42-21-3114; E-mail: [email protected]
Ž2. This is in contrast to what has been observed with polypyrrole containing low molecular weight dopants. On these occasions the small mobile anions leave the polymer allowing it to retain charge neutrality upon reduction. It has also been found that incorporation of polyelectrolytes into conducting polymers results in materials with more open porous structures, and high water contents w5x. The latter have been reported to be greater than 90% Žwrw. in some instances, and are attributed to uncompensated charge on the polyelectrolyte within the structure. Incorporation of polyelectrolytes into conducting polymers also provides a facile route for introduction of functional groups that are electroactive w6x, capable of complexing metal ions w1x or confer biocompatibility w5x. We have investigated the use of polyvinylphosphate as a dopant in polypyrrole. The effect of polyvinylphosphate on the polymerisationrdeposition process was studied, and the composition of the resultant polymer as well as its electrical, mechanical and ion transport properties examined. The latter are of particular interest for development of novel electrochemically controlled membranes as selec-
0379-6779r99r$ - see front matter q 1999 Elsevier Science S.A. All rights reserved. PII: S 0 3 7 9 - 6 7 7 9 Ž 9 8 . 0 0 1 9 1 - X
192
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
tive metal ion transport systems w7–16x. Prior to this work, there have been only a relatively small number of studies which focused on the metal ion transport properties of polypyrrolerpolyelectrolyte composites w3–5,17–22x.
phate. Determination of the thickness of membranes was performed using digital vernier calipers. Composite materials used in metal transport studies were prepared by depositing PPyrPVP onto platinum coated polyvinylidene fluoride ŽPVDF. membranes ŽMillipore..
2. Experimental
2.3. Techniques and instrumentation
2.1. Reagents
Elemental analyses on membranes were performed by the Microanalysis Unit at The Australian National University. The water content of large membranes Žtypically ; 60 cm2 area. was determined using the following procedure. After preparation the membranes were placed in an atmosphere of constant humidity Ž95%, 23.58C. and allowed to attain constant mass. They were then dried to constant mass in an oven at 608C, and the change in mass expressed as a percentage of the initial wet membrane. UV-visible and reflectance infrared spectra of membranes were obtained using a SHIMADZU Model UV-1601 spectrophotometer, and a BOMEM MB-100 Series spectrophotometer, respectively. Conductivity measurements were performed using large free standing membranes and the four point probe method. Scanning Electron Microscopy ŽSEM. and Atomic Force Microscopy ŽAFM. photographs were obtained using a Leica Cambridge 440 Scanning Electron Microscope and a Digital Multitude Nanoscope 3, respectively.
Pyrrole was obtained from Merck and distilled before use. Polyvinylphosphate Žsodium and ammonium salts; average molecular weight 133,000; average % phosphorylation of vinyl groupss 75%. was obtained from Polysciences. All inorganic reagents used were of AR grade and obtained from Ajax chemicals. All solutions were prepared using deionized Milli-Q water Ž18 M V cm.. 2.2. Preparation of polypyrroler polyÕinylphosphate polymers Electropolymerisation was performed in aqueous solutions containing 0.6 M pyrrole and 0.2% polyvinylphosphate, using an EG and G Princeton Applied Research Model 363 PotentiostatrGalvanostat. All solutions were thoroughly deoxygenated using nitrogen prior to use. A three electrode system consisting of a working electrode, AgrAgCl Ž3 M NaCl. reference electrode, and either a platinum mesh Ž100 linesrin.. or reticulated vitreous carbon ŽRVC. auxiliary electrode, was used in conjunction with a conventional one compartment electrochemical cell. All experimental data were collected using a MacLab ArD and DrA data collection system ŽAD Instruments.. Previous work with polypyrrolerpara-toluenesulfonate ŽPPyrpTS. polymers showed that galvanostatic methods were superior for reproducibly obtaining thin films and membranes with optimal mechanical and electrochemical properties w 10 x . Consequently, polypyrrolerpolyvinylphosphate ŽPPyrPVP. polymers were deposited onto platinum and stainless steel electrodes, gold coated Mylar and ITO coated glass using galvanostatic techniques. During these deposition processes the following resting potentials, 0.69, 0.75, 0.71 and 1.03 V, respectively were attained. Large free standing membranes grown onto stainless steel using low current densities were hard to remove. However, when high current densities were used to grow polymers bubbles formed on the surface of the membrane and overoxidation of the polymer occurred. As a result a current density of 1 mA cmy2 was used for all membrane preparations. After completion of electrodeposition the membranes were thoroughly rinsed with deionized water prior to analysis. Microanalyses, water content measurements, AFM and SEM studies were performed on membranes grown on a stainless steel electrode Ž60 cm2 area. using aqueous solutions containing 0.6 M pyrrole and 0.2% polyvinylphos-
2.4. Cyclic Õoltammetry studies Cyclic voltammetry studies were conducted in a three electrode electrochemical cell using a AgrAgCl Ž3 M NaCl. reference electrode, platinum auxiliary electrode, and PPyrPVP polymer deposited onto a platinum working electrode. The potential scan range was q0.45 V to y0.8 V, and the scan rate 100 mV sy1 . 2.5. Ion Transport Studies Metal ion transport studies were performed using a single membrane transport cell described previously w10x. A pulsed potential waveform with a 50 s pulse width, and a potential range of q0.45 V to y0.80 V was used. During transport studies aliquots for metal ion analysis were removed every 30 min when no potential was applied, and every 15 min when a potential was applied. The volume of sample removed varied from 0.2 ml to 2.0 ml depending on the sensitivity of the technique used to quantify the analyte and the rate of transport. Feed solutions contained 0.1 M metal ion, while the permeate solution contained initially 0.1 M electrolyte. The electrolyte used was NaNO 3 in all cases except when sodium transport was being examined, in which case it was KNO 3 . Atomic absorption spectroscopy ŽAAS. was used to determine metal ion concentration.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
193
Fig. 1. Ža. Cyclic voltammogram of an aqueous solution containing 0.6 M pyrrole and 0.2% PVP at a platinum disc electrode. Scan rate s 50 mV sy1 . Žb. Chronopotentiogram for PPyrPVP grown galvanostatically on a platinum disc electrode using 1 mA cmy2 current density and a solution containing 0.6 M pyrrole and 0.2% PVP. Žc. Chronoamperogram for PPyrPVP grown potentiostatically on a platinum disc electrode at a potential of 0.72 V using a solution containing 0.6 M pyrrole and 0.2% PVP.
194
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Anion transport studies were conducted in exactly the same manner as for metal ion transport. However, to determine hydrogen phosphate ion concentration, a colorimetric method using molybdate and ascorbic acid as reductant was employed w23x.
3. Results and discussion 3.1. Polymer growth The PPyrPVP polymer was grown using potentiodynamic, galvanostatic and potentiostatic techniques. Potentiodynamic methods showed an increase in current as the number of potential cycles was increased, indicating that deposition of the polymer occurred ŽFig. 1a.. From the first scan the minimum oxidation potential of the monomer was determined to be q0.70 V. When galvanostatic methods were used for growing the polymer a steady state potential was attained. For the
typical chronopotentiogram illustrated in Fig. 1b, the steady state potential is q0.70 V. The chronopotentiogram illustrated in Fig. 1b showed a significant decrease in growth potential with time, demonstrating that the polymer grown was conducting. In addition, it also indicated that initiation of polymer growth required a higher potential of q0.80 V, as expected. It is well known that monomer oxidises more readily on the deposited polymer than on a bare metal or carbon surface. This may be due to slow diffusion of the large polyelectrolyte to the electrode surface, or the large number of anionic sites on the polyelectrolyte. The latter would require that more pyrrole monomer is present around the polyelectrolyte before a polymer could be formed. Varying the concentration of pyrrole present in the monomer solution had little effect on the steady state potentials of polymers grown galvanostatically. Fig. 1c illustrates a typical chronoamperogram obtained during polymerisation at constant potential Žq0.72 V., and shows an increase in current with time indicative of polymer deposition.
Fig. 2. Reflectance FTIR spectra of ŽA. free polyvinylphosphate, and ŽB. PPyrPVP polymer galvanostatically deposited Ž0.71 V. onto gold coated mylar from a solution containing 0.6 M pyrrole and 0.2% PVP.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Fig. 3. Scanning electron micrographs of a PPyrPVP composite membrane galvanostatically deposited Ž0.75 V. onto stainless steel from a solution containing 0.6 M pyrrole and 0.2% PVP: Ža. solution side of membrane, Žb. electrode side of membrane.
3.2. Chemical and physical characterisation of composites The results of replicate C, H, N and P analyses performed on membranes grown under identical conditions showed a surprisingly high ratio Ž10:1. of nitrogen to phosphorous. With small anionic dopants the ratio of pyrrole to dopant molecules is typically 3:1 or 4:1. However, much higher values Žup to 9:1. have been reported for polypyrrole doped with bulky anions w24x. This indicates that the size and shape of dopant molecules can have a significant effect on polymer structure. The results of the current study indicate that the positive charge on the polypyrrole chain is spread out over a large number of monomer units. However, this assumes that the phosphate groups present in the dopant are only singly deprotonated, whereas if each of them were doubly deprotonated the positive charge would be spread over only 5 pyrrole units. Although the average p K a2 for polyvinylphosphate is unknown, it is unlikely that under the pH conditions employed for polymerisation in this study ŽpH ; 3. a significant proportion of the phosphate groups present would have been doubly deprotonated. The water content of polymer membranes was found to be 28 " 3%. Even higher water contents have been found previously for composites prepared from polypyrrole and
195
polyelectrolytes such as poly-2-acrylamido-2-methylpropanesulfonic acid w1x. The latter materials occupy large volumes and are able to undergo reversible hydrationrdehydration. The relatively high water content found for the PPyrPVP polymer suggests that this material may also have an open, porous structure which would facilitate movement of ions to and from the polymer. The reflectance infrared spectrum of sodium polyvinylphosphate is illustrated in Fig. 2A, and contains a broad absorption centred at approximately 1650 cmy1 , as well as absorptions at 1400, 1290, 1099 and 927 cmy1 . The three lowest energy absorption bands are assigned to P5O and P–O–C stretching modes of the phosphate group, while the two higher energy bands are attributed to bending andror deformation modes of the carbon and hydrogen atoms in the polymer backbone w25–27x. The most intense absorption in the reflectance infrared spectrum of PPyrPVP illustrated in Fig. 2B is a very broad band centred at 1558 cmy1 . This absorption band, which is absent from Fig. 2A, is assigned to the C5C stretching mode of the pyrrole groups w25–27x. Absorptions at 1034 and 921 cmy1 are again attributed to vibrational modes due to the phosphate groups. The lower frequencies of these absorption bands in the composite polymer reflect electrostatic interactions between the polypyrrole backbone and the dopant. Fig. 2B also contains a broad absorption extending between 1368 cmy1 and 1188 cmy1 , which is assigned to overlapping bands due to vibrational modes of both the phosphate and pyrrole groups. Scanning electron micrographs of the solution and electrode sides of a PPyrPVP membrane are illustrated in Fig. 3a and b, respectively. Both indicate that the membrane surface has a rough, amorphous morphology. Atomic force microscopic ŽAFM. studies of the membrane provided further support for this conclusion. By using AFM, the root mean square surface area of nodules on the membrane surface was determined to be 890 nm2 , and the average peak to valley distance 4.1 mm ŽFig. 4.. The effect of monomer concentration on membrane conductivity and thickness is presented in Table 1. In this
Fig. 4. Atomic force micrograph of the solution side of a PPyrPVP composite membrane galvanostatically deposited Ž0.75 V. onto stainless steel from a solution containing 0.6 M pyrrole and 0.2% PVP.
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
196
Table 1 Conductivities and thicknesses of membranes grown galvanostatically using aqueous solutions containing 0.2% polyvinylphosphate and different pyrrole concentrations wPyrrolex ŽM.
Conductivity ŽS cmy1 .
Thickness Žmm.
0.2 0.3 0.4 0.5 0.6
0.04 0.08 0.11 0.17 0.28
3.0"0.3 4.0"0.3 5.5"0.4 7.0"0.4 9.0"0.5
Ž3b. study it was found that, when growing to the same charge Ž3.2 mC. solutions containing the highest pyrrole concentration examined, 0.6 M, gave membranes that were thicker and had the highest conductivities. The overall trend suggests that even at the highest pyrrole concentration examined, 0.6 M, there was insufficient monomer present in solution to optimise polymer formation. Instead, the polymer may have contained significant amounts of low molecular weight pyrrole oligomers. Such species would not be large enough to be doped with polypyrrole, providing an explanation for the low degree of dopant incorporation suggested by microanalysis. This may also explain the relatively low conductivities observed for these polymers. The highest conductivity observed for a PPyrPVP membrane was 0.28 S cmy1 , which is several orders of magnitude less than values typically found for polypyrrole membranes Ž10–100 S cmy1 . w10x. Another possible explanation for the low conductivity may be the very large size of the polyelectrolyte dopant, as it has been shown previously that large anions influence polymer chain packing in conducting polymers, and increase interchain and interparticle hopping distances. w24x Further support for the low conductivity of the PPyrPVP polymer was provided by absorption spectroscopic studies performed on films grown onto IndiumrTinrOxide ŽITO. coated glass. Fig. 5 shows the absorption spectra of PPyrPVP films exposed to solutions with different pH values for 2 h. In each spectrum there is no evidence of a free carrier tail, which is consistent with the low conductivities of these materials determined by the four point probe method.
Ž3c. When the polymer backbone was electrochemically reduced the overall polymer became negatively charged due to the immobile polyvinylphosphate counteranion ŽEq. 3a.. In order to maintain overall charge neutrality, sodium ions from solution were incorporated into the polymer ŽFig. 6, peak A, Ep s y0.53 V.. On re-oxidation of the polymer ŽEq. 3b., the polymer backbone became positively charged, and overall charge neutrality was achieved by both expulsion of sodium ions ŽFig. 6, peak B, Ep s y0.31 V. and incorporation of chloride ions from solution ŽFig. 6, region C.. When the polymer was reduced during subsequent cycles an additional feature was seen in the cyclic voltammogram ŽFig. 6, region D. which corresponded to expulsion of previously incorporated chloride ions ŽEq. 3c.. Polymer oxidation was therefore accompanied by two competing processes: expulsion of cations and incorporation of anions. The latter process may lead to stabilisation of the polymer backbone through electrostatic interactions with incorporated chloride ions, leaving the polyvinylphosphate and sodium ions to interact predominantly with each other. When the supporting electrolyte was varied the potentials of features in the cyclic voltammogram attributable to
3.3. Electrochemical characterisation The cyclic voltammogram of a PPyrPVP coated platinum electrode immersed in 0.1 M NaCl is illustrated in Fig. 6. The possible processes occurring are described below.
Ž3a.
Fig. 5. UV-vis spectra of PPyrPVP composites galvanostatically deposited Ž1 mA cmy2 . onto ITO coated glass from a solution containing 0.6 M pyrrole and 0.2% PVP, and exposed to solutions with different pH values. Solution 1: distilled water; Solution 2: pH 2 ŽHCl.; Solution 3: pH 12 ŽNaOH.; Solution 4: pH 6 Žphosphate buffer.; Solution 5: pH 7 Žphosphate buffer.; Solution 6: pH 8 Žphosphate buffer..
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
197
porting electrolyte used was NaCl or KCl. Furthermore, the cation incorporation peaks seen in cyclic voltammograms obtained using solutions containing divalent metal ions were present at more positive potentials, suggesting that a stronger driving force existed for incorporation of divalent metal ions into the polymer than univalent metal ions. 3.4. Transport studies
Fig. 6. Cyclic voltammogram of a PPyrPVP coated platinum electrode prepared galvanostatically Ž1 mA cmy2 . from a solution containing 0.6 M pyrrole and 0.2% PVP, after immersion in 0.1 M NaCl. Scan rate 100 mV sy1 . Region A s cation insertion. Region Bs cation expulsion. Region C sanion insertion. Region Dsanion expulsion.
incorporation and expulsion of cations changed. Table 2 summarises the results of these experiments. Both the cathodic peak potentials EpŽc. and anodic peak potentials EpŽa. vary depending on the electrolyte used, indicating that the PPyrPVP composite is sensitive to the identity of cations and anions present in solution. Incorporation and expulsion of Liq, Naq and Kq to and from the PPyrPVP polymer was clearly demonstrated by cyclic voltammetry. This was also clearly seen with Cly and NOy 3 , despite the possibility that the large immobile polyvinylphosphate dopant would inhibit incorporation of other negatively charged species. However, anion incorporation and expulsion was not generally observed, as H 2 PO4y, pTSy, and SO42y ions did not appear to interact with the polymer. The cyclic voltammograms of polymer coated electrodes immersed in solutions containing either CaŽNO 3 . 2 or MgŽNO 3 . 2 did not show peaks due to incorporation and expulsion of these cations. Similar behaviour was observed previously for polypyrrole doped with perchlorate from tetrabutylammonium perchlorate, after immersion in solutions containing either BaCl 2 or MgCl 2 w28x. Cyclic voltammograms of the latter solutions showed much smaller peaks due to cation insertion than when the sup-
Metal ion transport studies were undertaken using the PPyrPVP polymer deposited onto PtrPVDF composite membranes using galvanostatic techniques. Before commencing these experiments it was necessary to grow membranes that did not allow water to pass freely through. Fig. 7 illustrates the effect that varying the charge density used to grow membranes had on the flux of water through the membrane. It can be seen that a charge of 1680 mC for each cm2 of membrane resulted in a membrane that was impermeable to water. This charge density was then used to grow all membranes used in transport experiments. The cations potassium, sodium, lithium, calcium and magnesium were all transported across the membrane ŽFig. 8., as was hydrogen phosphate ion. In all cases no transport was observed in the absence of an applied electric potential ŽFig. 8, region A., or on application of a constant negative or positive potential to the membrane. However, immediately after application of the pulsed potential waveform to the membrane was commenced, transport of ions was observed in all cases ŽFig. 8, region B.. After an initial dramatic increase in the concentration of metal ion or phosphate anion in the permeate side, the concentration subsequently increased at a slower rate. After applying the pulsed potential waveform for 2 h the concentration of metal ion or phosphate ion in the permeate side approached steady state values. No further significant transport was observed after application of the pulsed potential waveform was ceased at this point ŽFig. 8, region C..
Table 2 Sensitivity of PPyrPVP coated platinum electrodes to different ions: anodic peak potentials EpŽa . and cathodic peak potentials EpŽc. obtained from cyclic voltammograms of electrodes immersed in solutions containing different supporting electrolytes Supporting electrolyte Ž0.1 M.
EpŽc .1 ŽV.
EpŽa.1 ŽV.
EpŽa.2 ŽV.
EpŽc.2 ŽV.
NaNO 3 KNO 3 LiNO 3 NaCl NapTS NaH 2 PO4 Na 2 SO4
y0.50 y0.45 y0.50 y0.50 y0.35 y0.30 y0.30
y0.45 y0.40 y0.40 y0.30 y0.20 y0.20 y0.25
0.05 0.10 0.05 0.10 – – –
y0.10 y0.05 y0.05 y0.15 – – –
Fig. 7. Effect of charge passed during growth on the permeability to water of PPyrPVP coated PVDFrPt membranes prepared galvanostatically using a solution containing 0.6 M pyrrole and 0.2% PVP.
198
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
Fig. 8. Electrochemically controlled transport of Kq, Naq, Ca2q, Liq and Mg 2q across a PPyrPVP composite membrane. Feed solution: 0.1 M metal ion Žas the nitrate salt.. Permeate solution: 0.1 M NaNO 3 or KNO 3 Žfor Naq transport experiments.. Region A: No applied potential. Region B: Applied potential pulsed between y0.8 V to q0.4 V; pulse width 50 s. Region C: No applied potential.
The effect of varying the duration of the pulsed potential waveform on ion transport was investigated using potassium ion. In one experiment, after checking that no potassium transport occurred after 1 h when no potential was applied, the same pulsed potential waveform as used above was applied for 30 min then stopped. During application of the potential waveform the concentration of potassium in the permeate phase increased sharply in an analogous fashion to that shown in Fig. 8. However, immediately after the potential was removed potassium transport ceased and the concentration of potassium ions in the permeate phase remained constant. In a second experiment the pulsed potential waveform was applied for a total of 4 h, again after ensuring that no transport occurred in the absence of electrochemical stimulus. After 2 h the concentration of potassium ion in the permeate phase was the same as that illustrated in Fig. 8. However, during the following 2 h there was no further significant increase in the amount of potassium transported. This indicates that the flux of potassium ions across the membrane reached an equilibrium state after 2 h. In order to achieve further transport of metal ion across the membrane it would be necessary to remove the metal ions present in the permeate phase. The average fluxes observed for each ion examined are presented in Table 3 together with the ionic radius w29x for the hydrated ion. Overall the fluxes observed in this study are comparable to values observed for polypyrrole polymers containing simple dopant molecules w10,12,13,30x. The average fluxes observed for the sodium and potassium ions were comparable to each other, and an order of magnitude greater than that for lithium. This may be due to the significantly greater size of the hydrated lithium ion compared to those of sodium and potassium, which restricts its passage through molecular-sized pores in the membrane. Alternatively, it may be due to stronger interactions between incorporated lithium and the polyvinylphos-
phate dopant. Previously it had been noted that the flux of alkali metal ions through membranes composed of either PPyrpTS or PPyrdodecylsulphate decreased overall in descending order of the periodic table w30x. This trend was attributed to the greater porosity of these membranes to the smaller hydrated ions of the heavier elements. The flux of calcium through the membrane was slightly less than that of lithium. In view of the similar hydrated ionic radii for these metal ions, this also suggests that the rate of transport was determined to a large extent by the size of the hydrated ion relative to that of pores in the membrane. This is supported by the significantly lower flux observed for magnesium, which has the largest hydrated metal ion radius of the metals examined, compared to that for calcium. However, it should also be remembered that the lower fluxes observed for calcium and magnesium may also be due to higher affinities for the dopant as a result of their greater charge. This would retard release of the metal ion upon polymer reoxidation compared to when a univalent metal ion was present. Despite the relatively small size of its hydrated ion, the flux of hydrogen phosphate ion observed was very low. This is perhaps not surprising since for transport to occur the anion must first be incorporated into the re-oxidised polymer, a process not likely to be favoured by the presence of a large, immobile, negatively charged dopant in the structure of the polymer. The PPyrPVP polymer may therefore be considered as functioning predominantly as a cation exchange material. The transport selectivity factor Ž s . is given by:
ss
FluxMq FluxXq
It had been shown previously that the KqrCa2q selectivity factor decreased as the size of the dopant incorporated into polypyrrole increased w13x. This suggested that larger dopants had a lower ability to interact selectively with these metal ions. Further evidence for this is provided by the current study. For example, the KqrCa2q selectivity factor determined in the current work was 13.2, which is seven times less than that observed previously Ž94. for polypyrrolerpara-toluenesulfonate w13x. Furthermore, the KqrNaq selectivity factor for the latter polymer Ž3.1. w12x was also significantly greater than that observed in this
Table 3 Average flux values for ions during transport experiments Ion
Flux Žmol sy1 cmy2 .
Effective ionic ˚ . w29x radius ŽA
Kq Naq Liq Ca2q Mg 2q H 2 PO4y
5.74=10y9 5.63=10y9 4.83=10y10 4.36=10y10 1.03=10y11 5.09=10y12
3 4 6 6 8 4
J.M. DaÕey et al.r Synthetic Metals 99 (1999) 191–199
study Ž1.02.. Overall this tends to suggest that polypyrroles containing polyelectrolyte dopants may not necessarily offer any advantages over similar polymers containing simple dopants for selectively transporting metal ions.
4. Conclusion A polypyrrolerpolyvinylphosphate polymer was successfully prepared electrochemically and shown to have a ratio of 10:1 for pyrrole repeat unit to charged phosphate group, a relatively high water content, low conductivity, and a rough amorphous morphology. The use of a large relatively immobile polyelectrolyte dopant gave membranes composed of this material an open porous structure. Despite the low conductivity, electrochemical studies indicated that the polymer was responsive to a number of different cations and phosphate ion. Transport of these ions and HPO42y was achieved by applying a pulsed potential waveform to a single membrane transport cell. The fluxes of metal ions across the membrane followed the sequences: Kq) Naq)) Liq for the group 1 metals, and Ca2q)) Mg 2q for the group 2 cations.
Acknowledgements We thank Dr. Ashton Partridge and Industrial Research Limited, New Zealand, for helpful discussions and financial support of this project.
References w1x G.G. Wallace, H. Zhao, C.O. Too, C.J. Small, Synth. Met. 84 Ž1997. 323. w2x Y.H. Liao, K. Levon, Macromol. Rapid Commun. 16 Ž1995. 393. w3x K. Naoi, M. Lien, W.H. Smyrl, J. Electrochem. Soc. 138 Ž1991. 440.
199
w4x X. Ren, P.G. Pickup, J. Phys. Chem. 97 Ž1993. 5356. w5x A.J. Hodgson, K. Gilmore, C. Small, G.G. Wallace, I.L. MacKenzie, T. Aoki, N. Ogata, Supramolecular Science 1 Ž1994. 77. w6x C. Lee, M.H. Lee, Y.K. Kang, B.G. Moon, S.B. Rhee, Synth. Met. 55 Ž1993. 1119. w7x P. Burgmayer, R.W. Murray, J. Am. Chem. Soc. 104 Ž1982. 6139. w8x P. Burgmayer, R.W. Murray, J Electroanal. Chem. 147 Ž1983. 339. w9x P. Burgmayer, R.W. Murray, J. Phys. Chem. 88 Ž1984. 2515. w10x H. Zhao, W.E. Price, C.O. Too, D. Zhou, G.G. Wallace, J. Membr. Sci. 119 Ž1996. 199. w11x H. Zhao, W.E. Price, G.G. Wallace, J. Electroanal. Chem. 334 Ž1992. 111. w12x A. Mirmohseni, W.E. Price, G.G. Wallace, H. Zhao, J. Int. Mater. Syst. Struct. 4 Ž1993. 43. w13x H. Zhao, W.E. Price, G.G. Wallace, J. Membr. Sci. 87 Ž1994. 47. w14x H. Zhao, W.E. Price, P.R. Teasdale, G.G. Wallace, Reactive Polymers 23 Ž1994. 213. w15x A. Mirmohseni, W.E. Price, G.G. Wallace, J. Membr. Sci. 100 Ž1995. 239. w16x D. Zhou, H. Zhao, W.E. Price, G.G. Wallace, J. Membr. Sci. 98 Ž1995. 173. w17x A. Wojda, K. Maksymiuk, J. Electroanal. Chem. 424 Ž1997. 93. w18x T. Shimidzu, A. Ohtani, K. Honda, J. Electroanal. Chem. 251 Ž1988. 323. w19x Y.-J. Qiu, J.R. Reynolds, Polymer Science and Engineering 31 Ž1991. 417. w20x C.K. Baker, Y.-J. Qiu, J.R. Reynolds, J. Phys. Chem. 95 Ž1991. 4446. w21x M. Lien, W.H. Smyrl, J. Electroanal. Chem. 309 Ž1991. 333. w22x M. Po, J.R. Reynolds, J. Chem. Soc. Chem. Commun., Ž1993. 258. w23x A.E. Greenberg, L.S. Clescen, A.D. Eaton, ŽEds.., Standard Methods For The Examination Of Water And Wastewater, 18th. edn., American Public Health Association, Washington D.C., 1992. w24x K.M. Cheung, D. Bloor, G.C. Stevens, J. Mater. Sci. 25 Ž1990. 3814. w25x K. Nakamoto, Infrared And Raman Spectra Of Inorganic And Coordination Compounds, 3rd. edn., Wiley, New York Ž1978.. w26x C.J. Pouchert, ŽEd.., The Aldrich Library of IR Spectra, Sigma-Aldrich, Milwaukee Ž1981. w27x G. Socrates, Infrared Characteristic Group Frequencies, Wiley, Chichester, 1980. w28x R. John, G.G. Wallace, J. Electroanal. Chem. 354 Ž1993. 145. w29x J.A. Dean, ŽEd.., Lange’s Handbook of Chemistry, 13th edn., McGraw-Hill, 1985, Table 5-2. w30x A.C. Partridge, C. Milestone, C.O. Too, G.G. Wallace, J. Membr. Sci. 132 Ž1997. 245.