Conductive Blends of Polyaniline with PlasticizedPoly(methyl methacrylate)

PASCAL JUVIN,1 MAGDALENA HASIK,1 JEROME FRAYSSE,1 JEROME PLANES,1 ADAM PRON,1

IRENA KULSZEWICZ-BAJER2

1 Laboratoire de Physique des Metaux Synthetiques, UMR 5819, DRFMC/SI3M, CEA Grenoble 38 054 Grenoble, France

2 Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00 664 Warszawa, Poland

Received 31 August 1998; accepted 16 November 1998

ABSTRACT: Polyaniline (PANI) protonated with camphorsulfonic acid (CSA) and threedifferent poly(alkylene phosphates) (PAPs) (where alkylene 5 pentylene, hexylene, ornonylene) was used in the fabrication of conductive polyaniline–poly(methyl methac-rylate) (PMMA) blends. The lowest percolation threshold ( fp 5 0.041 wt %) wasobtained for the PANI(CSA)0.5–PMMA blend plasticized with 35 wt % of dibutylphtalate (DBPh). This blend is also very resistant against the deprotonation of itsconductive phase in basic solutions of pH 5 9. In the case of blends prepared with theuse of PAPs as PANI dopants, the percolation threshold strongly depends on the lengthof the hydrophobic spacer (alkylene group) in the dopant. The percolation thresholddecreases in the order PPP . PHP . PNP, whereas the resistance against deprotona-tion in basic solutions decreases in the following inverse order: PNP . PHP . PPP.This last observation can be rationalized by increasing contribution of hydrophobicsegments in the polymeric dopant, when going from PPP to PNP, which renderspolyaniline more resistance toward the penetration by aqueous basic solutions. © 1999John Wiley & Sons, Inc. J Appl Polym Sci 74: 471–479, 1999

Key words: PANI; CSA; PAPs; PMMA

INTRODUCTION

The progress in the processibility of polyaniline(PANI) observed since 19921 has resulted in thefabrication of several types of conductive polyani-line blends with such industrial polymers as poly-(vinyl chloride),2,3 polystyrene,4,5 and poly-amides6 to name a few. In modern technology,there exists a strong demand for materials thatcombine flexibility of plastics with high transpar-ency in the visible part of the spectrum and highelectronic conductivity.7 In our previous article,8

we have reported the preparation of flexible andhighly transparent, conductive blends of polyani-line with plasticized cellulose acetate.

Plastification is a convenient method for im-proving the flexibility of blends. In addition tocellulose derivatives, several industrial polymerscan be used in the plasticized state, among thempoly(methyl methacrylate) (PMMA). Poly(methylmethacrylate) is very well suited for the prepara-tion of transparent conductors by blending withpolyaniline due to its low extinction coefficient inthe visible. Blends of PANI with nonplasticizedPMMA have been prepared by several au-thors.9–12 Some of them report, for these materi-als, a very low percolation threshold, significantlybelow 1 wt %.2,12 In view of the transparency ofthe conductive blends, the value of the percolation

Correspondence to: J. Fraysse.Contract grant sponsor: KBN, Poland; contract grant num-

ber: 3T09B07111.Journal of Applied Polymer Science, Vol. 74, 471–479 (1999)© 1999 John Wiley & Sons, Inc. CCC 0021-8995/99/030471-09

471

threshold is of crucial importance. Because of ahigh value of the extinction coefficients in thevisible measured for the conductive form of poly-aniline, the transparency of its blends with con-ventional polymers can be assured only for ex-tremely low values of the percolation threshold.

In this article, we demonstrate that the use ofan appropriate plasticizer does not only improvethe flexibility of PANI–PMMA blends but alsoleads to a significant increase of their conductiv-ity and to much lower percolation thresholds ascompared to nonplasticized blends.

In addition to the studies of the preparation ofextremely low percolation threshold PANI–PMMA blends, we have investigated their resis-tance against deprotonation. The weakest point ofPANI is its tendency to deprotonate in basic me-dia. Of course, this deprotonation reaction has adisastrous effect on the conductivity of PANI-based blends because only the protonated form ofPANI is conductive. In the majority of potentialapplications, improved resistance against depro-tonation is required for PANI blends since suchmaterials can be exposed to basic media and, de-spite this exposure, they should retain their highconductivity.

EXPERIMENTAL

Reagents

Polyaniline in the oxidation state of emeraldinewas synthesized chemically by oxidation of ani-line with ammonium persulfate at 225°C. Thedetails of the preparation can be found else-where.13 The resulting emeraldine hydrochloridewas then transformed into the free base state bydeprotonation with an excess of 0.3M aqueoussolution of NH3. The inherent viscosity of thepolymer was equal to 1.49 dL/g at 25°C (0.1 wt %of emeraldine in H2SO4).

The following four protonating agents wereused: poly(pentylene phosphate) (PPP), poly-(hexylene phosphate) (PHP), poly(nonylene phos-phate) (PNP), and camphorsulfonic acid (CSA).Poly(alkylene phosphates) of the general formula

where synthesized according to the method de-scribed in the literature.14,15 First poly(phospho-nates) were obtained by polycondensation of di-methyl phosphate with corresponding diols andthen oxidized with N2O4 to poly(alkylene phos-phates). The number-average molecular weightMn varied from 8500 to 14,000. (12) 210-CSA(98%) was purchased from Aldrich and was vac-uum-dried prior to use. Dibutyl phtalate (DBPh)(Aldrich) and m-cresol (MC) (Merck) and PMMA(medium molecular weight, Aldrich) were usedwithout further purification.

Protonation of Emeraldine Base

As already stated, the following four protonatingagents were used: three poly(alkylene phos-phates) (PPP, PHP, and PNP) and CSA. Prior tothe protonation reaction emeraldine base (EB)and the protonating acids were dried in a vacuumline at 80°C for 2 h. The molar ratio of poly(alky-lene phosphate) mer to PANI mer or CSA mole-cule to PANI mer was 0.5. It is assumed here thatPANI mer has the formula C6H4NH0.5. The num-ber 0.5, denoting hydrogen bonded to nitrogen,originates from the fact that in emeraldine imineand amine, nitrogens are in equal numbers,which gives, on the average, 0.5 H per N. EB andthe protonating agent were then transferred tom-cresol to give a suspension of 0.5 wt % (calcu-lated with respect to EB). The suspension wasthen stirred typically for 1 week. During thistime, its ultraviolet–visible—near-infrared (UV–vis—NIR) spectrum was periodically registered.The reaction was stopped when two consecutiveregistrations gave the same spectra. Then, thesuspension was centrifuged at 5000 rpm for 15min. The following two fractions of protonatedPANI were separated upon centrifugation: thefraction that sedimented and the one that re-mained in m-cresol. Only nonsedimented fractionwas used for the preparation of PANI–PMMAblends. The concentration of nonsedimentedPANI in m-cresol was determined by measuringthe difference between the initial EB mass takenfor the protonation and the mass of sedimentedPANI after its deprotonation. It turned out that inthe case of the protonation with CSA, 94% ofinitial PANI does not sediment upon the centrif-ugation. In the case of the protonation with PPP,PHP, and PNP, these values are 22, 22, and 44%,respectively. Whether the obtained solutions arereal or colloidal solutions is still the matter of ascientific debate.16 Nevertheless, they can be con-

472 JUVIN ET AL.

veniently used in the solution processing of PANI.It should be noted here that the amount of non-sedimented fraction strongly depends on theamount of water in EB, the protonating acid, andthe solvent. The amount of this fraction quicklydecreases with the increase of water content. Forthis reason, careful drying of all components isrecommended.

Preparation of PANI–PMMA Blends

Three series of blends were prepared with 0, 25, and35 wt % of plasticizer (DBPh). First, 10 wt % solu-tions of PMMA in m-cresol were prepared (with orwithout the plasticizer). Then, these solutions weremixed with appropriate amounts of m-cresol solu-tions of protonated PANI and additionally stirredfor 2 h. From this homogeneous, by eye solution,freestanding films were cast by slow evaporation ofm-cresol at 40°C. Typical thickness of the sampleswas between 60 and 100 mm.

Deprotonation Studies

The prepared blends were kept in a large excess ofpH 5 9 buffer solution (Carlo Erba; boric acid,potassium chloride, and sodium hydroxide), andtheir UV–vis—NIR spectra were periodically reg-istered. For parallel samples kept in the samebuffer solution, direct current (DC) conductivitywas measured for increasing exposure time to thebasic medium.

Measurements

UV–vis—NIR absorption spectra of solutions andfilms were measured on a Lambda 900 Perkin–Elmer spectrometer. For the solution, spectraquartz infrasil cells of 1-mm optical path were used.

DC conductivity of the blends was measuredunder room conditions using a four-probe tech-nique with parallel gold contacts. The ohmic be-havior was checked in each case.

RESULTS AND DISCUSSION

Electrical transport properties of protonatedPANI films, among other factors, depend stronglyon the protonating agent and the processing me-dium. There exist dopant/solvent couples, whichfavor high conductivity of PANI films obtained bycasting, such as, for example, CSA–m-cresol1 andCSA–hexafluoro-2-propanol.17 In the case of athree-components system PANI–dopant–m-

cresol, it is possible to predict electrical transportproperties of the processed films by the investiga-tion of UV–vis—NIR spectra of the solution usedfor casting.18 If due to specific solvent–dopantinteractions,19,20 a conformation favoring the de-localization of charge carriers (polarons) isachieved, the spectrum shows one localized peakat 440 nm and a monotonically increasing absorp-tion, which starts at approximately 500 nm andextends into NIR. Chain conformation, leading tolocalized polarons, gives rise to three well-definedabsorption peaks at 360, 430, and approximately900 nm. Of course, upon casting, better conduc-tors are obtained in the former case. In Figure 1,the spectra of PANI protonated with poly(alky-lene phosphates) (PAPs) and with CSA are pre-sented. It is clear that the desired conformation isachieved for CSA and PNP-protonated PANI. Pro-tonation with PPP and PHP results in a smallcontribution of the peaks due to localized pol-arons, which are superimposed on the increasingNIR absorption. The removal of the solvent wors-ens the spectra, which is manifested by more pro-nounced contribution of the localized polaronspeaks in the spectra of free standing film of PANI–PMMA blends (compare Figs. 1 and 2). Neverthe-less, similarly as previously studied CSA, PAPsseem to be very good protonating agents, suitablefor the preparation of highly conductive blends.

In Table I, the results of electrical conductivitymeasurements of PANI(PAP)0.5–PMMA blendsare collected. At this point, the role of the plasti-cizer (DBPh) in the improvement of the electricalproperties of the blends should be underlined. Fornonplasticized samples, the percolation thresholdis well above 4 wt %, and none of the preparedsamples was conductive. The addition of 25 wt %of DBPh efficiently lowers the percolation thresh-old, which is now , 4 wt % for PANI(PPP)0.5–PMMA, , 2 wt % for PANI(PHP)0.5–PMMA, and, 1 wt % for PANI(PNP)0.5–PMMA. A similareffect of the plasticizer was observed for PANI–cellulose acetate blends.8

Even lower percolation thresholds are obtainedfor PANI(CSA)0.5–PMMA blends. In Figure 3, theconductivity of the blend versus the PANI contentis plotted for different contents of the plasticizer.From the inset of Figure 3, which presents thedata obtained for PANI contents below 0.3 wt %,it is clear that the addition of the plasticizer hastwo effects. First, for a given content of PANI, itcauses approximately a two–three-fold increasein the conductivity of the blend. Second, it effi-ciently lowers the percolation threshold.

CONDUCTIVE PANI–PLASTICIZED PMMA BLEND 473

In order to determine the percolation thresholdmore precisely, we have fitted the data presentedin Figure 3 to the scaling law of percolation the-ory,21 as follows:

s~f! 5 c~f 2 fp!t (1)

where c is a constant, t is the critical exponent, fis the volume fraction of the conductive phase,and fp is the volume fraction at the percolationthreshold.

In our calculations, for practical reasons, wehave used the mass fraction instead of the volume

Figure 1 UV–vis—NIR spectra of polyaniline protonated with poly(alkylene phos-phates) and camphor sulphonic acid registered for m-cresol solutions: (a) PANI(PPP)0.5;(b) PANI(PHP)0.5; (c) PANI(PNP)0.5; (d) PANI(CSA)0.5.

474 JUVIN ET AL.

fraction. Since the densities of all components ofthe blend are very close, this change is of minimalsignificance. Also, for comparative reasons, weexpressed the PANI content as EB wt % to make

it independent of the mass of the dopant, whichvaries depending on its type. In all three series ofstudies (0, 25, and 35 wt % of the plasticizer), weget an excellent fit to the scaling law of percola-

Figure 2 UV–vis—NIR spectra of free standing films of polyaniline–poly(methylmethacrylate) blends: (a) PANI(PPP)0.5–PMMA; (b) PANI(PHP)0.5–PMMA;(c) PANI(PNP)0.5–PMMA; (d) PANI(CSA)0.5–PMMA.

CONDUCTIVE PANI–PLASTICIZED PMMA BLEND 475

tion. The calculated parameters of eq. (1) arelisted in Table II. For nonplasticized samples, weget the critical exponent 5 1.6, which is slightlyhigher than reported in Reghu et al.12 for thesame blend. These authors give the value of 1.33,which is smaller than the value predicted by theuniversal law (t 5 2.0) and attribute it to ther-mally induced hopping between disconnected (orweakly connected) part of the percolating net-work. Upon lowering of the temperature, t ap-

proaches the theoretical value. Our fp value isalso lower than reported in Reghu et al.12 Itshould be noted that in Reghu et al.,12 the fp isexpressed as PANI(CSA)0.5 wt % fraction. Forcorrect comparison, it should be expressed as EBwt % fraction, as in our article. The recalculationgives fp 5 0.13 for the blend studied in Reghu etal.12 and 0.054 in our nonplasticized samples.

The addition of the plasticizer does not onlylower the fp value but also increases the value oft. For 35 wt % of DBPh, the percolation thresholdis as low as 0.041, and the t value reaches, withinthe experimental error, the universal value pre-dicted by the percolation theory. The interpreta-tion of the variation of these two parameters isvery difficult unless clearly resolved transmissionelectron microscopy (TEM) micrographs are avail-able. However, one is tempted to propose the fol-lowing interpretation. The increase of the t valuefrom 1.6 to 2 upon addition of the plasticizer canbe treated as a manifestation of the increasingramification of the conducting network; t 5 2corresponds to the exponent obtained for the in-cipient percolation cluster, whereas t 5 1 is ob-tained for the effective medium approximation;the decrease of fp indicates that an infinite cluster

Table I Conductivities of Nonplasticized and DBPh Plasticized Blends ofPoly(alkylene phosphate)-Doped Polyaniline with Poly(methylmethacrylate), Expressed in S/cm

Dopant

% PANI Base

% DBPh4 Wt % 2 Wt % 1 Wt %

PPP 9.0 p 1028 5.7 p 1029 1.3 p 1029 0 wt %PHP 2.4 p 1027 1.7 p 1027 6.7 p 1028 0 wt %PNP 2.8 p 1027 3.6 p 1028 1.0 p 1028 0 wt %PPP 2.8 p 1024 9.9 p 1027 3.0 p 1028 25 wt %PHP 1.1 p 1022 1.0 p 1023 4.1 p 1028 25 wt %PNP 6.7 p 1022 2.6 p 1022 3.5 p 1023 25 wt %

Figure 3 Conductivity versus PANI content in PANI-(CSA)0.5–PMMA blends containing varying amounts ofDBPh plasticizer. The inset shows, in detail, the resultsfor PANI content , 0.3 wt %.

Table II Scaling Law Parameters Calculatedfor PANI(CSA)0.5–PMMA Blends ContainingDifferent Amounts of DBPh Plasticizer

Amount pc t

R(CorrelationCoefficient)

0 wt % DBPh 0.054 6 0.005 1.6 6 0.2 0.99725 wt % DBPh 0.046 6 0.001 1.9 6 0.1 0.99635 wt % DBPh 0.041 6 0.001 2.0 6 0.1 0.997

476 JUVIN ET AL.

spans over the entire sample for even lower PANIcontents. Based on the above arguments, it maybe postulated that the plasticizer plays the role ofan interphase builder (compatibilizer). Morpho-logical studies of the PANI–PMMA–plasticizersystem are not totally conclusive at the presenttime; however, TEM observations of a similar sys-tem of PANI–cellulose acetate–plasticizer un-equivocally show ramified morphology.22

The weakest point of conductive polyaniline isits tendency to deprotonate. Since only the pro-tonated form of PANI is conductive, its deproto-nation results in a significant conductivity drop.Polyaniline used in the fabrication of blendsshould therefore exhibit improved resistanceagainst deprotonation because these materialsare sometimes exposed to basic media. The idea ofthe use of macromolecular dopants (polyalkylenephosphates) for the improvement of PANI resis-tance against deprotonation was based on theassumption that this reaction may be inhibiteddue to kinetic reasons. Since the deprotonationproduct is polymeric (polyalkylene phosphate in aform of a salt), it cannot easily diffuse out fromthe PANI matrix, and, for this reason, the depro-tonation process can be slowed down.

For the studies of the deprotonation reaction ofPANI in four types of blends prepared in this

research, we have used UV–vis—NIR spectros-copy and conductivity measurements. Blends con-taining 4 wt % of PANI were kept in a large excessof a buffer solution of pH 5 9, and their spectraand conductivity were periodically measured.In Figure 4, the evolution of the spectrum forPANI(PHP)0.5–PMMA blend, with increasingtime of the exposure to the solution of pH 5 9, ispresented. The spectrum significantly changes,even for very short exposure times. Initially, theintensity of the broad absorption tail, associatedwith the presence of delocalized polarons, quicklydecreases; and the peak at approximately 800–850 nm due to localized polarons becomes morepronounced. For longer exposure times, a newpeak, characteristic of the totally deprotonatedPANI–emeraldine base, appears at 630 nm. Suchevolution implies strong heterogeneity of the dep-rotonation process with zones of localized anddelocalized polarons coexisting with zones of to-tally deprotonated PANI (i.e., emeraldine base).For two other macromolecular dopants, PPP andPNP, the behavior is qualitatively the same. How-ever, as judged from the evolution of the spectra,the deprotonation rate decreases in the followingorder: PPP . PHP . PNP.

The above results are confirmed by the conduc-tivity measurements. In Figures 5 and 6, the nor-malized conductivity s/s0 (where s0 is the conduc-tivity for the time of exposure 5 0) is plotted versus

Figure 4 Evolution of the UV–vis—NIR spectrum ofPANI(PHP)0.5–PMMA blend upon exposure to a pH 5 9buffer solution (plasticizer content 25 wt %).

Figure 5 Evolution of the electrical conductivity ofPANI–PMMA blends upon exposure to a pH 5 9 buffersolution for short exposure times (plasticizer content 25wt %).

CONDUCTIVE PANI–PLASTICIZED PMMA BLEND 477

the time of exposure. From the evolution of theconductivity, it is clear that PANI(PPP)0.5 andPANI(PHP)0.5 deprotonate quicker than PANI(PNP)0.5 and PANI(CSA)0.5. Up to 24 h of expo-sure to the buffer solution of pH 5 9 PANI(PNP)0.5 and PANI(CSA)0.5 deprotonate with asimilar rate, then the deprotonation of PANI(PNP)0.5 is more accelerated (compare Figures 5and 6). Evidently, PANI(CSA)0.5 is the most re-sistant against deprotonation and shows a rela-tively small conductivity drop, even after 10 daysof exposure to solutions of pH 5 9.

For three dopants of the macromolecular na-ture, PPP, PHP, and PNP, one can notice that thedeprotonation rate measured by the decrease ofthe conductivity depends on the length of thehydrophobic spacer between the phosphate hy-drophilic groups. In agreement with spectroscopicstudies, the deprotonation rate decreases in thefollowing order: PPP . PHP . PNP. Thus, thedeprotonation rate can be correlated with the hy-drophobicity of poly(alkylene phosphate) anioncreated upon the protonation of PANI. In thepolyanion originating from PNP, the contributionof the hydrophobic segments is larger than in thecase of the polyanion originating from PHP,which, in turn, is more hydrophobic than thepolyanion formed from PPP. Thus, more hydro-phobic dopant anions prevent more efficientlypolyaniline matrix from the penetration of aque-

ous deprotonation solution, and the deprotona-tion reaction is slowed down. A similar effect ofthe length of hydrophobic spacer on the ability toabsorb water or aqueous solutions is well knownfor polyamides of the following formula:

O[COONHO(CH2)n]Ox

In this case, the maximum amount of absorbedwater decreases with increasing n. The analogywith polyaniline doped with poly(alkylene phos-phates) is evident in this case.

As demonstrated in this research, the macro-molecular nature of the dopant does not preventpolyaniline from deprotonation. Since, in the caseof poly(alkylene phosphate)-protonated polyani-line, the deprotonation must create polyalkylenephosphate salt, which is of limited mobility due toits macromolecular nature, the deprotonationmust proceed via microphase separation. This mi-crophase separation is indirectly confirmed byUV–vis—NIR spectra, which indicate the coexist-ence of the zones of undoped PANI together withtwo types of zones of protonated PANI. Thus, adopant that improves PANI stability against dep-rotonation does not need to be of polymeric type;but, in addition to strong Bronsted acid centers, itshould possess hydrophobic groups built, whichassure sufficient hydrophobicity of the dopedPANI.

At the end, it should be noted that the Bron-sted acidity of the dopant molecule also plays animportant role in PANI resistance against thedeprotonation. PANI(CSA)0.5 deprotonates muchslower at pH 5 9 than PANI(PNP)0.5 does, despitethe fact that PNP is more hydrophobic than CSA.However, CSA is a stronger Bronsted acid thanPNP and, for this reason, forms stronger adductwith the polymer.

CONCLUSIONS

To summarize, we have prepared polyaniline–poly(methyl methacrylate) blends using four dif-ferent PANI protonating agents. In all cases, wehave demonstrated that the addition of a plasti-cizer–dibutyl phtalate (DBPh) improves the con-ductivity of the blend and lowers the percolationthreshold. For PANI(CSA)0.5–PMMA blends con-taining 35 wt % of DBPh, we obtained the perco-lation threshold as low as 0.041 wt %. In addition,the deprotonation reaction at pH 5 9, responsible

Figure 6 Evolution of electrical conductivity ofPANI–PMMA blends upon exposure to a pH 5 9 buffersolution for long exposure times (plasticizer content 25wt %).

478 JUVIN ET AL.

for the conductivity drop, is extremely slow in thismaterial.

In the case of the use of macromolecular PANIdopants of the poly(alkylene phosphate) type, weobserve a clear correlation between the length of thehydrophobic spacer, the value of the percolationthreshold, and the resistance against deprotona-tion. With increasing length of the alkylene spacer,the value of the percolation threshold decreases,and the resistance against deprotonation increases.

The contribution of one of the authors (I.K.B.) wasfinanced through a grant 3T09B07111 provided byKBN (Poland).

REFERENCES

1. Cao, Y.; Smith, P.; Heeger, A. J. Synth Met 1992,48, 91.

2. Banerjee, P.; Mandal, B. M. Macromolecules 1995,28, 3940.

3. Conn, C.; Booth, N.; Unsworth, J. Adv Mater 1995,7, 790.

4. Ruckenstein, E.; Yang, S. Synth Met 1993, 53, 283.5. Jousseaume, V.; Morsli, M.; Bonnet, A.; Lefrant, S.

J Appl Polym Sci 1998, 67, 1209.6. Im, S. S.; Byun, S. W. J Appl Polym Sci 1994, 51,

1221.7. Colaneri, N. F.; Shacklette, L. W. IEEE Trans In-

strum Meas 1992, 41, 291.8. Pron, A.; Nicolau, Y.; Genoud, F.; Nechtschein, M.

J Appl Polym Sci 1997, 63, 971.

9. Yang, S.; Ruckenstein, E. Synth Met 1993, 59, 1.10. Yang, C. Y.; Cao, Y.; Smith, P.; Heeger, A. J. Synth

Met 1993, 53, 293.11. Reghu, M.; Yoon, C. O.; Yang, C. Y.; Moses, D.;

Heeger, A. J.; Cao, Y. Macromolecules 1993, 26,7245.

12. Reghu, M.; Yoon, C. O.; Yang, C. Y.; Moses, D.;Smith, P.; Heeger, A. J.; Cao, Y. Phys Rev B: Con-dens Matter 1994, 50, 13931.

13. Beadle, P. M.; Nicolau, Y. F.; Banka, E.; Rannou,P.; Djurado, D. Synth Met 1998, 95, 29.

14. Pretula, J.; Penczek, S. Makromol Chem 1990, 191,671.

15. Pretula, J.; Penczek, S. Macromolecules 1993, 26,2228.

16. Wessling, B. in Handbook of Organic ConductiveMolecules and Polymers, Vol. 3; Nalwa, H. S., Ed.;J. Wiley and Sons: New York, 1997; p. 497.

17. Rannou, P.; Gawlicka, A.; Berner, D.; Pron, A.;Nechtschein, M.; Djurado, D. Macromolecules1998, 31, 3007.

18. Xia, Y.; Wiesinger, J. M.; MacDiarmid, A. G. ChemMater 1995, 7, 443.

19. Ikkala, O. T.; Pietila, L. O.; Ahjopalo, L.; Oster-holm, H.; Passiniemi, P. J. J Chem Phys 1995, 103,9855.

20. Vikki, T.; Pietila, L. O.; Osterholm, H.; Ahjopalo,L.; Takkala, A.; Toivo, A.; Levon, K.; Passiniemi,P.; Ikkala, O. Macromolecules 1996, 29, 2945.

21. Aharony, A.; Stauffer, D. Introduction to Percola-tion Theory, 2nd ed.; Taylor and Francis: London,UK, 1993.

22. Planes, J.; Cheguettine, Y.; Samson, Y. Synth Metto appear.

CONDUCTIVE PANI–PLASTICIZED PMMA BLEND 479