A

pptpdopbevw©

K

1

oeaTltaAp

C2

(

0d

Progress in Organic Coatings 59 (2007) 115–121

Transparent, conductive polymer blend coatingsfrom latex-based dispersions

Jiakuan Sun ∗, William W. Gerberich, Lorraine F. Francis 1

Department of Chemical Engineering and Materials Science, University of Minnesota,151 Amundson Hall, 421 Washington Avenue SE, Minneapolis, MN 55455, USA

Received 9 November 2005; received in revised form 20 October 2006; accepted 19 January 2007

bstract

Flexible, transparent and conductive polymer blend coatings were prepared from aqueous dispersions of poly(3,4-ethylenedixoythiophene)/oly(styrenesulfonate) [PEDOT/PSS] gel particles (∼80 nm) and latex (∼300 nm). The stable dispersions were deposited as wet coatings ontooly(ethylene terephthalate) substrates and dried at 80 ◦C. Microstructure studies using tapping mode atomic force microscopy (TMAFM) indicatehat a network-like microstructure formed during drying at 0.03 volume fraction PEDOT/PSS loading. In this network-like structure, the PEDOT/PSShase was forced into the boundary regions between latex. In addition, migration of the PEDOT/PSS particles towards coating surface is likelyuring drying of the aqueous dispersions. The addition of a small amount of dimethyl sulfoxide (DMSO) in dispersions altered the distributionf the PEDOT/PSS phase. As PEDOT/PSS concentration increases to 0.15 volume fraction, the coating surface is dominated by the PEDOT/PSShase. The effect of DMSO on microstructure becomes less apparent as PEDOT/PSS concentration increases. The conductivity of the polymer

lend coatings increases in a percolation-like fashion with a threshold of ∼0.02 volume fraction PEDOT/PSS. The addition of DMSO in dispersionsnhanced the coating conductivity beyond the threshold by more than two orders of magnitude. The highest conductivity, ∼3 S/cm, occurs at 0.20olume fraction PEDOT/PSS concentration. The polymer blend coatings have good transparency with only a weak dependence of transparency onavelength due to the small refractive index difference between filler and matrix.2007 Elsevier B.V. All rights reserved.

ductiv

ilfdovptt

eywords: Transparent and conductive coatings; Polymer blend; Electrical con

. Introduction

Transparent and conductive coatings have many applications,ften serving as transparent electrodes. Currently, most transpar-nt electrodes are based on coatings of conductive oxides, suchs indium tin oxide (ITO) and antimony-doped tin oxide (ATO).hese coatings are often produced at elevated temperature and

ack mechanical flexibility, both of which limits their applica-ion on flexible substrates. Intrinsically conductive polymers aren alternative; however, they are not stable or are expensive.

nother route to transparent and conductive coatings is to useolymeric nanocomposite coatings.

∗ Corresponding author. Present address: Rohm and Haas (China) Holdingo., LTD. 1077 Zhang Heng Road, Zhang Jiang Hi Tech Park, Shanghai, China01203. Tel.: +86 21 39628577; fax: +86 21 5895 9865.

E-mail addresses: jiakuansun@rohmhaas.com (J. Sun), lfrancis@umn.eduL.F. Francis).1 Tel.: +1 612 625 0559; fax: +1 612 626 7246.

sot[cpi

rm

300-9440/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.porgcoat.2007.01.019

ity; Optical transparency; Microstructure; Percolation

Flexible, transparent and conductive polymer nanocompos-te coatings were prepared from stable aqueous dispersions ofatex (∼300 nm) and nanosized ATO particles (∼15 nm) [1] byorming a segregated microstructure [1–4]. Since electrical con-uctivity is achieved at relatively low filler content and the sizef ATO nanoparticles is much smaller than the wavelength ofisible light, the nanocomposite coatings are optically trans-arent as well [1]. The above-mentioned ATO/latex system hashree limitations: (1) relatively low achievable conductivity onhe order of 10−2 S/cm [1] primarily attributed to the low intrin-ic conductivity of the ATO nanoparticles [1] and the intrusionf polymer into the interstices between ATO particles and clus-ers [1,5]; (2) strong transparency dependence on wavelength1]; (3) low mechanical flexibility of the resulting compositeoatings [6]. In an attempt to overcome these limitations, thisaper explores the use of intrinsically conductive nanoparticles

n place of ATO.

Improving the conductivity of nanocomposite coatingsequires attention to the properties of the conductive filler andatrix, as well as the microstructure and interface structure.

1 anic C

Tfiibta

R

wtclclittbclatc

mspia[pdp

ci(atoi

2

2

mBgb∼u–

(

dpats

2

fiPPoaasaatp0dw3titst

spasd(ac

2

mccmpUcsst

16 J. Sun et al. / Progress in Org

he conductivity of polymer composites loaded with conductivellers beyond percolation is increased by using fillers with lower

ntrinsic resistivity and by taking steps to decrease the resistanceetween fillers [5]. The resistance between conductive filler par-icles (Rc) consists of two contributions: constriction resistancend tunneling resistance [5]:

c = ρi

d(constriction) + ρt

a(tunneling) (1)

here ρi the intrinsic filler resistivity, d the diameter of the con-act spot, ρt the tunneling resistivity, and a is the area of theontact spot. From Eq. (1), Rc is lowered by using fillers withower intrinsic resistivity, by increasing the dimensions of theontact between particles, and by avoiding the intrusion of insu-ating polymer between conductive filler particle, which wouldncrease the tunneling resistivity. Genetti et al. [7] demonstratedhe importance of limiting the contact resistance; they enhancedhe conductivity of Ni particle/polyethylene (PE) compositesy nearly three orders of magnitude by synthesizing a layer ofonductive polypyrrole (PPy) on Ni particle surfaces. The PPyayer on particle surface likely reduces the tunneling resistancend increases the contact area between Ni particles. One expectshat deformable, conductive particles will have an even greaterhance of forming composites with low contact resistance.

The transparency of conductive polymer composites is deter-ined by the transparencies of the matrix and filler, the optical

cattering at the interface between the two phases, and the com-osite thickness. Optical scattering decreases as the refractivendex difference between matrix and filler, filler dimension,nd filler concentration decrease for a given coating thickness8]. The refractive index difference between ATO and matrixolymer (PVAc-co-acrylic) is ∼0.5 [9,10]. Since polymer con-uctive fillers have nearly matching refractive index with matrixolymer, improvements on transparency are expected.

This paper explores the potential of using intrinsicallyonductive polymer particles as the filler in a latex-based coat-ng. Poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate)PEDOT/PSS) gel particles [11] were chosen for this study. Inddition to exploring the effect of filler loading on microstruc-ure, conductivity and transparency, the effect of polar solventsn the morphology and conductivity of PEDOT/latex compos-tes was also investigated.

. Experimental

.1. Materials

An aqueous dispersion of PEDOT/PSS gel particles, com-ercially called Baytron P, was provided by the Bayer Company.aytron P is an aqueous dispersion of PEDOT and PSS blendel particles with an average size of ∼80 nm, as specifiedy the manufacturer. The PEDOT/PSS gel particles consist of95 vol.% water and∼5 vol.% polymer, as specified by the man-

facturer. PEDOT is a p-type semiconductor due to doping withSO3

− groups.Flexbond 325 (Air Products) is an aqueous dispersion

55 wt% solids) of PVAc-co-acylic polymer particles. The size

pKup

oatings 59 (2007) 115–121

istribution of latex particles, determined by a light scatteringarticle size analyzer (Coulter), ranges from 50 to 600 nm withvolume average particle size of 333 nm. The glass transition

emperature (Tg) of the PVC-co-acrylic copolymer is 19 ◦C aspecified by the manufacturer.

.2. Preparation of coatings

In this paper, polymer composites composed of PEDOT/PSSller in a PVAc-co-acrylic matrix are referred to asEDOT/PVAc-co-acrylic. Two steps were used to fabricate theEDOT/PVAc-co-acrylic coatings: (1) preparation of an aque-us dispersion containing latex, conductive filler, and otherdditives; and (2) deposition of the dispersion on the substratend drying. To produce the PEDOT/PVAc-co-acrylic disper-ions, the PEDOT/PSS dispersion was diluted in deionized waternd stirred on a magnetic stir plate for 10 min; the latex was thendded, the pH was adjusted to 5.5 using NH4OH solution, andhe mixture was stirred for 2 h. Dispersions were prepared torovide coatings with a range of filler loadings from 0.05 to.50 volume fraction (dry polymer) in the dried coating. Eachispersion contained 2 vol.% solids. The pH of the dispersionsas adjusted to ∼5.5 to achieve colloidal stability (see Section.1). The composite dispersions were coated onto polyethyleneerepthalate (PET) substrates by a wire-wound rod with the coat-ng rate manually controlled to be ∼2 cm/s. The wet coating washen transferred into an oven and dried at 80 ◦C for 10 min. Twoets of coatings were prepared for each dispersion compositiono verify experimental reproducibility.

A Dupont Zonyl FSO surfactant (0.1 vol.% relative to disper-ion volume) was added to all the aqueous composite dispersionsrepared here as a wetting agent. Zonyl FSO is a nonionicnd fluorocarbon-based surfactant, which aids in spreading. Totudy the effect of polar solvents on the microstructure and con-uctivity of PEDOT/latex coatings, a small amount of DMSO∼1 vol.% relative to total dispersion volume) was added into thequeous dispersions of PEDOT/PVAc-co-acrylic latex beforeoating process.

.3. Characterization of microstructure and properties

Several instruments were used to characterize the thickness,icrostructure, conductivity, and transparency of the polymer

omposite coatings. Coating thickness was measured by a Ten-or P-10 profilometer. A DI multimode Nanoscope IIIa tappingode AFM (TMAFM) was used to investigate the surface mor-

hology of the composite coatings in air at room temperature.V–vis transmission and absorbance spectra of the composite

oatings were obtained using a Varian Cary 5-E UV–vis–NIRpectrometer. During the measurements, the PET substrate wascanned as a blank sample first, and then its spectrum was sub-racted from the data.

For the measurement of coating conductivity, a four-point

robe apparatus comprised of a Keithley 220 current source, a&S four-point probe and a Keithley 6517A electrometer wassed. The current was applied between two outer probes of four-oint probe and the voltage across the two inner probes was

J. Sun et al. / Progress in Organic C

F

m[tre

ρ

wdtph

3

3

toppf

sfp

ePt[cPPAca

pstFctas

3

3

PsAeagbPsg

F1

ig. 1. Zeta potential of PVAc-co-acrylic particles in water as a function of pH.

easured by the electrometer. Details were described elsewhere1]. The specimens were 2 cm × 7 cm rectangular coatings onhe plastic substrate. The bulk resistivity of the coating is thenelated to the current and voltage drop according to the followingquation:

= KhV

I(2)

here ρ is the bulk resistivity of the coatings, K a constantepending on the geometry of sample and probe spacing (inhis study, K = 4.24), V the voltage drop between the inner tworobes, I the current applied through the outer two probes, andis the coating thickness.

. Results and discussion

.1. Colloidal stability

Fig. 1 shows the zeta potential of the PVAc-co-acrylic par-icles in water as a function of pH. The zeta potential data was

btained from our previous study [12]. The PVAc-co-acrylicarticles have an isoelectric point (IEP) of ∼pH 2.0. The zetaotential becomes more negative as dispersion pH increasesrom 2.0 to 12.0. For polymeric latex particles, the negative

c

Pc

ig. 2. Tapping mode AFM phase images of the surface of PEDOT/PSS coatings pr�m × 1 �m.

oatings 59 (2007) 115–121 117

urface charges mainly originate from the dissociation of acidicunctional groups and the ionization of the residual groups fromolymerization initiator [13].

Baytron P is synthesized by polymerization of 3,4-thylenedioxythiophene (EDOT) in an aqueous solution ofSS using S2O8

2− as the oxidant. During the polymeriza-ion, PEDOT oligomers form and attach onto the PSS chain11]. Since PEDOT is not soluble in water, the PEDOT/PSShains form gel particles, which are swollen by water. In theEDOT/PSS complex chains, SS units are in excess [11]. FreeSS chains may also exist at the surface of gel particles [14,15].s a result, the PEDOT/PSS gel particles have net negative

harges, consistent with the electrophoresis results of Ghoshnd Inganas [16].

The experimental observation confirmed that an aqueous dis-ersion of PEDOT/PSS and PVAc-co-acrylic at pH 5.5 was verytable with no apparent sedimentation in several days. At pH 5.5,he surface of PVAc-co-acrylic latex is negatively charged (seeig. 1) and the PEDOT/PSS gel particles also have net negativeharges, resulting in electrostatic repulsions between all the par-icles. In addition, the surfactant molecules on the latex surfacend the PSS chains in the Baytron P gel particles may provideteric repulsion.

.2. Microstructure

.2.1. PEDOT/PSS coatingsFig. 2a and b compare the surface microstructures of the

EDOT/PSS coatings prepared without and with using dimethylulfoxide (DMSO) in dispersions by showing the tapping modeFM phase images. From Fig. 2a, a granular structure is appar-

nt with two phases present: (1) regions with high phase contrastnd (2) regions with low phase contrast. This observation ofranular structure is consistent with other results [14,15,17]. Theright granular regions in the phase image originate from theEDOT/PSS gel particles [17]. According to the height image,ome of the dark regions in the phase image are voids betweenel particles; some of them may originate from the free PSS

hains that were dissolved in water [11,14,15,17].

Although the distribution of the PSS chains on theEDOT/PSS gel particles and in the aqueous medium cannot belearly identified, it is important to discuss their effects on coat-

epared (a) without and (b) with using DMSO in dispersions. The scan area is

118 J. Sun et al. / Progress in Organic Coatings 59 (2007) 115–121

F -co-aT

iooocfrrobtoDgs

PDguitoacS

tttnaocvisnenroi

3

PuFswfolrTobP

PtAlcbfA

Piuaci

PuTprnA

ig. 3. Tapping mode AFM phase images of the surface of 0.03 PEDOT/PVAche scan area is 3 �m × 3 �m.

ng morphology based on our hypotheses and previous works byther researchers. The PSS chains affected the microstructuref PEDOT/PSS coatings. As water evaporated during the dryingf aqueous PEDOT/PSS dispersions, PEDOT/PSS gel particlesompacted and deformed (likely similar to latex film formation),orcing the free PSS chains into the boundary regions. As aesult, the outer layer of the grains in the dried coatings is PSS-ich and the interior is PEDOT-rich [14,15,17]. The thicknessf the outer PSS-rich layer is usually ∼4 nm thick as evaluatedy Salaneck and co-workers [14,15] using photoelectron spec-roscopy. This insulating PSS layer may lower the conductivityf the dried PEDOT/PSS coatings, which is discussed later.ue to the conjugated structure of PEDOT, the PEDOT/PSSel particles may not be deformed enough to eliminate all voidpaces.

Fig. 2b shows the tapping mode AFM phase images of theEDOT/PSS coatings prepared with using a small amount ofMSO in the dispersion. As compared with Fig. 2a, the fineranular features are absent and the phase contrast becomes moreniform in the AFM phase image. The uniform phase contrastn Fig. 2b implies that the segregation of the free PSS chains tohe boundary regions between gel particles became less apparentn coating surface due to the addition of DMSO. These resultsbout the effect of DMSO on the morphology of PEDOT/PSSoatings are consistent with the observations and hypotheses byalaneck and co-workers [14,15].

DMSO and other polar organic solvents have been utilizedo enhance the conductivity of the dried PEDOT/PSS coatingshrough altering coating morphology [14,15,17] or affectinghe interaction between PEDOT and PSS chains [18]. Sala-eck and co-workers [14,15] found that the addition of a smallmount of DMSO in dispersion altered the surface compositionf dried PEDOT/PSS coatings. The surface of the PEDOT/PSSoatings prepared with using DMSO or other polar organic sol-ents has higher molar ratio of EDOT to SS than the coatingnterior [14,15]. This increase in EDOT/SS ratio on coatingurface is ascribed to the reduction in the outer PSS layer thick-ess of PEDOT/PSS grains. The PSS chains in the interior areither uniformly distributed or formed PSS grains [14,15]. This

ew morphology persists even after the solvents are completelyemoved [14,15]. During the film formation process, this layerf insulating PSS with reduced thickness may rupture, resultingn a relatively uniform phase contrast in Fig. 2b.

cp

c

crylic coatings prepared (a) without and (b) with using DMSO in dispersions.

.2.2. PEDOT/latex coatingsFig. 3a and b compare the surface microstructures of 0.03

EDOT/PVAc-co-acrylic coatings prepared without and withsing DMSO in dispersions by showing the AFM phase images.rom Fig. 3a, a complex microstructure is apparent on theurface of a 0.03 PEDOT/PVAc-co-acrylic coating preparedithout using DMSO in dispersions. Three regions with dif-

erent phase contrasts are present. Based on our understandingf the microstructure formation from aqueous dispersions ofatex and other conductive particles [1–4,12], the lower contrastegions likely originate from latex (PVAc-co-acrylic) particles.he surrounding region with the highest phase contrast likelyriginates from PEDOT/PSS. This phase was segregated into theoundaries between latex regions, resulting in an interconnectedEDOT/PSS-rich phase network [1–4,12].

As discussed previously, the aqueous dispersions ofEDOT/PSS gel particles and latex were stable at pH 5.5 due to

he electrostatic repulsion between negatively charged colloids.s water evaporates during drying of the stable dispersions,

atex particles begin to consolidate, compact, and deform underapillary forces, forcing the PEDOT/PSS gel particles into theoundaries between latex regions and thereby resulting in theormation of a segregated microstructure in a similar way to theTO/PVAc-co-acrylic system in the previous study [1,12].

Fig. 3b shows the AFM phase image of the surface of a 0.03EDOT/PVAc-co-acrylic coating prepared with using DMSO

n dispersions. As compared with the coating prepared withoutsing DMSO, the three regions with different phase contrastsre still present but their distributions are much different. Theonnectivity of the PEDOT/PSS-rich phase cannot be readilydentified in Fig. 3b.

Fig. 4a and b compare the surface microstructures of 0.15EDOT/PVAc-co-acrylic coatings prepared without and withsing DMSO in dispersions by showing the AFM phase images.he coating surfaces are dominated by the PEDOT/PSS-richhase and are decorated with a few isolated PVAc-co-acrylic-ich regions. The effect of DMSO on coating morphology isot apparent in the coatings with 0.15 volume fraction filler.s filler concentration increases further, the effect of DMSO on

omposite morphology becomes negligible according to AFMhase images.

From Fig. 4, the percentage of the PEDOT/PSS phase onoating surface appears to be much more than one would expect

J. Sun et al. / Progress in Organic Coatings 59 (2007) 115–121 119

F atings prepared (a) without and (b) with using DMSO in dispersions. The scan areai

fe[mspdc

3

((egt[osri[stttc

ooe(

adipflpfc

Table 1Experimental conductivity and transparency data for the PEDOT/PVAc-co-acrylic coatings prepared with DMSO for several filler concentrations

PEDOT/PSSvolume fraction

Coating thickness(�m)a

Conductivity(S/cm)b

Transparencyd

(%)c

0.03 1.6 ∼10−4 94.4 ± 1.40.05 1.97 ∼10−3 94.7 ± 2.20.10 1.0 1.35 ± 0.11 91.2 ± 1.80.15 1.56 2.28 ± 0.11 79.5 ± 0.80.20 1.13 2.84 ± 0.45 87.4 ± 10.00.25 1.35 1.37 ± 0.01 80.3 ± 7.01.0 1.2 16.0 ± 2.9 75.5 ± 5.0ATOe ∼0.1 ∼100 ∼80ITOf ∼0.1 ∼1000 ∼90

a The data is the average coating thickness on two samples and the range iswithin ±0.5 �m.

b The errors are the range of conductivity based on two samples.c The errors are the range of transparency based on two samples.

space of one gel particle is occupied by water [11]), they are moreeffective in reducing the percolation threshold as compared tosolid ceramic particles with smooth surface (like ATO nanopar-ticles). For the same PVAc-co-acrylic latex, an interconnected

ig. 4. Tapping mode AFM phase images of the surface of 0.15 PEDOT/PSS cos 3 �m × 3 �m.

rom 0.15 volume fraction. This result may be due to the prefer-ntial migration of the PEDOT/PSS gel particles during drying19]. During drying of the latex and PEDOT/PSS dispersions,enisci formed on coating surface at some point [19]. The pres-

ure underneath these menisci is lower than the atmosphericressure, drawing up water towards the surface [19]. Thus, lowensity and water-swollen gel particles could be carried to theoating surface [19].

.3. Conductivity

Polar organic solvents such as sorbitol, N-methylpyrrolidoneNMP), isopropanol [20], DMSO, N,N-dimethyl formamideDMF), and tetrahydrofuran (THF) [18] have been used tonhance the conductivity of the dried PEDOT/PSS coatings. Ineneral, two types of mechanisms were proposed for the conduc-ivity increase, usually as large as several orders of magnitude14,15,18]. Kim et al. [18] hypothesized that the enhancementn the conductivity of PEDOT/PSS coatings may be due to thecreening effect offered by the polar solvent molecules withelatively high dielectric constant, leading to a higher mobil-ty for charge carriers [18]. However, Salaneck and co-workers14,15,17,20] found that the conductivity enhancement still per-isted after the solvents were completely removed. As a result,hey hypothesized based on photoelectron spectroscopy resultshat the increases in coating conductivity is due to the reduc-ion in the thickness of the insulating PSS layer surrounding theonducting PEDOT/PSS grains [14,15,17,20].

The conductivity of the coatings prepared with the aque-us dispersion of PEDOT/PSS gel particles alone is on therder of 10−1 S/cm. The addition of a small amount of DMSOnhanced the conductivity of PEDOT/PSS coatings to ∼16 S/cmsee Table 1). This result is consistent with past work [14,15,18].

Fig. 5 shows the dc conductivity of the PEDOT/PVAc-co-crylic coatings prepared with and without using DMSO inispersions as a function of PEDOT/PSS loading. For the coat-ngs prepared without DMSO, coating conduction behavior isercolation-like with a percolation threshold of ∼0.02 volumeraction. However, the results are not consistent with the perco-

ation power law equation. This inconsistency may be due to thereferential location of the PEDOT/PSS phase on coating sur-ace as described before. As the PEDOT/PSS gel particles areomposed of polymeric chains and swelled by waters (∼95%

Fp

d The transparency is the data at a wavelength of 600 nm.e ATO thin film deposited by electron beam evaporation [37].f ITO thin film deposited by dc arc-discharge ion plating [38].

ig. 5. Electrical conductivity of the PEDOT/PVAc-co-acrylic coatings pre-ared without and with using DMSO as a function of filler concentration.

1 anic Coatings 59 (2007) 115–121

Avglca(coife[

bD[tocD∼

atcootPolssscott

3

PPPrfctpotTatd

Fi

o[[ibo

mcvIicslaad

4

tmciblcDohigh conductivity can be achieved at relatively low filler con-

20 J. Sun et al. / Progress in Org

TO (∼15 nm) network formed at a loading greater than 0.06olume fraction [1,12]. However, for the polymeric PEDOT/PSSel particles (∼80 nm) this critical concentration appeared to beess than 0.03 volume fraction. Similarly, low critical filler con-entrations for the formation of an interconnected network werelso achieved by using other intrinsically conductive polymerICP) emulsions [16,21–33] or solutions [21,28,29,32–35] as theonductive filler and an insulating latex [21–24,28,29,31–33,35]r polymer solution [16,21,25–30,32–34] as the matrix start-ng material. The ICP filler networks in these studies wereormed at relatively low filler concentration due to volumexclusion effect [21–24,27–29,31–33,35] or phase separation21,25,26,28–30,32–34] or other mechanisms [16].

The plateau conductivity beyond percolation is enhancedy more than two orders of magnitude by the addition ofMSO, consistent with the results by several other researchers

14,15,18]. Above the percolation threshold, the conductivity ofhe PEDOT/PVAc-co-acrylic coatings prepared with DMSO isn the order of 1 S/cm, which is even higher than that of thearbon black/PVAc-co-acrylic coatings [2–4]. The addition ofMSO did not change the percolation threshold, which is still0.02 volume fraction.The increase in the conductivity of the PEDOT/PVAc-co-

crylic composite coatings with addition of DMSO is likely dueo the physical and morphological changes of the PEDOT/PSSomponents. From the AFM results, the morphological changesf the PEDOT/PSS coatings are partially verified. The additionf DMSO also resulted in a relatively uniform distribution ofhe excessive PSS chains, indicating the reduction in the outerSS layer thickness of the PEDOT/PSS grains. The additionf DMSO may also make the PEDOT/PSS gel particles coa-esce well, leaving less void spaces between them. However, thecreening mechanism as described before cannot be excluded orupported based on our results here. Unfortunately, the discus-ion above about the relationship between microstructures andonductivity is jeopardized by the fact that AFM phase imagesnly identified morphological changes up to 0.03 volume frac-ion but the conductivity increases are observed only well abovehat volume fraction.

.4. Transparency

Fig. 6 shows the UV–vis spectra for the PEDOT/PSS andEDOT/PVAc-co-acrylic coatings (no DMSO) with variousEDOT/PSS concentrations in the range of visible light. TheEDOT/PSS coating displays adsorption in the wavelengthange of orange light, resulting in its dark blue tint [11]. However,or the PEDOT/PVAc-co-acrylic composite coatings with fillerontents lower than or equal to 0.35 volume fraction, the adsorp-ion in the wavelength range of orange light is diluted by matrixolymer. In addition, due to this dilution effect, the transparencyf the PEDOT/PVAc-co-acrylic coatings with low filler concen-rations is improved as compared to the PEDOT/PSS coating.

he small refractive index difference between PEDOT/PSSnd PVAc-co-acrylic (equal to ∼0.05) also contributes to theransparency improvement [8,9,11]. This small refractive indexifference also results in a weak dependence of the transparency

cdtt

ig. 6. UV–vis spectra for the PEDOT/PSS and PEDOT/PVAc-co-acrylic coat-ngs (no DMSO) with various PEDOT/PSS concentrations.

n wavelength for the coatings with low PEDOT/PSS contents8]. For the ATO/PVAc-co-acrylic coatings prepared previously1,12], the dependence of coating transparency on wavelengths strong due to a relatively large refractive index differenceetween ATO and PVAc-co-acrylic (>0.2) [8,9,11]. The additionf DMSO in dispersions did not affect transparency.

The PEDOT/PSS dispersion (Baytron P) is an expensiveaterial, which costs at least ∼US$ 50/kg [36]. Electrical per-

olation occurs at very low PEDOT/PSS loading level (0.02olume fraction) using latex as the matrix starting materials.ncorporating PEDOT/PSS into latex to make composite coat-ngs is an effective way to reduce material cost and improveoating transparency. A relatively high coating conductivity cantill be maintained even after incorporation of latex. Table 1ists the electrical conductivity and optical transparency (at

wavelength of 600 nm) of some (PEDOT/PSS)/PVAc-co-crylic coatings prepared with using DMSO in the startingispersions.

. Conclusion

Intrinsically conductive PEDOT/PSS gel particles are effec-ive conductive fillers in latex coatings because they have special

orphology (water-swelled gel particle structure), high intrinsiconductivity (after modification by DMSO), and nearly match-ng refractive index with matrix polymer. The conductivityehavior of the PEDOT/latex composite coatings is percolation-ike with a percolation threshold of ∼0.02 volume fraction. Theonductivity beyond percolation threshold of the (PEDOT/PSS,MSO-modified)/latex is ∼3 S/cm, which is higher than thatf the ATO/latex [1] by more than two orders of magnitude. A

entration (e.g. 0.10 volume fraction). As the refractive indexifference between PEDOT/PSS and matrix polymer is small,he polymer blend coatings have good transparency and weakransparency dependence on wavelength.

anic C

A

PIvVtmd

R

[[

[

[

[

[

[[

[[[

[

[[[[[[[[[[

[[

[

[

[36] Baytron Price Schedule, Application Notes, H.C. Starck, Bayer Company,

J. Sun et al. / Progress in Org

cknowledgements

The authors thank the industrial supporters of the Coatingrocess Fundamentals Program of the Industrial Partnership innterfacial and Materials Engineering (IPRIME) at the Uni-ersity of Minnesota for financial support, and Dr. Bhaskarelamakanni for help with obtaining zeta potential data. J. Sun

hanks Mr. Hong Wei and Ms. Lijun Zu for help with the trans-ittance spectra collection and Mr. Xun Yu for some stimulating

iscussions about conductivity improvement mechanisms.

eferences

[1] J. Sun, W.W. Gerberich, L.F. Francis, J. Poly. Sci. B Poly. Phys. 41 (2003)1744.

[2] J.C. Grunlan, W.W. Gerberich, L.F. Francis, J. Appl. Polym. Sci. 80 (2001)692.

[3] J.C. Grunlan, Y. Ma, M.A. Grunlan, W.W. Gerberich, L.F. Francis, Polymer42 (2001) 6913.

[4] J.C. Grunlan, W.W. Gerberich, L.F. Francis, Polym. Eng. Sci. 41 (2001)1947.

[5] G.R. Ruschau, S. Yoshikawa, R.E. Newnham, J. Appl. Phys. 72 (1992)953.

[6] J. Sun, W.W. Gerberich, L.F. Francis, Polym. Eng. Sci. 45 (2005) 207.[7] W.B. Genetti, W.L. Yuan, B.P. Grady, E.A. O’Rear, C.L. Lai, D.T.

Glaztzhofer, J. Mater. Sci. 33 (1998) 3085.[8] C.F. Bohren, V. Huffman, Absorption and Scattering of Light by Small

Particles, John Wiley & Sons, New York, 1983.[9] J. Wen, Polymer Data Handbook, Oxford University Press, Oxford, 1999.10] B. Stjerna, E. Olsson, C.G. Granqvist, J. Appl. Phys. 76 (1994) 3739.11] L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Adv.

Mater. 12 (2000) 481.12] J. Sun, B.V. Velamakanni, W.W. Gerberich, L.F. Francis, J. Colloid Inter.

Sci. 280 (2004) 387.13] R.J. Hunter, Foundations of Colloid Science, Oxford University Press,

Oxford, 2001.

[

[

oatings 59 (2007) 115–121 121

14] G. Greczynski, Th. Kugler, M. Keil, W. Osikowicz, M. Fahlman, W.R.Salaneck, J. Electron Sprectrosc. Relat. Phenom. 121 (2001) 1.

15] X. Crispin, S. Marciniak, W. Osikowicz, G. Zotti, A.W. Denier Van DerGon, F. Louwet, M. Fahlman, L. Groenendaal, F. De Schryver, W.R. Sala-neck, J. Polym. Sci. B Polym. Phys. 41 (2003) 2561.

16] S. Ghosh, O. Inganas, Synth. Met. 101 (2001) 413.17] G. Greczynski, Th. Kugler, W.R. Salaneck, Thin Solid Films 354 (1999)

129.18] J.Y. Kim, J.H. Jung, D.E. Lee, J. Joo, Synth. Met. 126 (2002) 311.19] Y. Ma, H.T. Davis, L.E. Scriven, Prog. Org. Coat. 52 (2005) 993.20] S.K.M. Jonsson, J. Birgerson, X. Crispin, G. Greczynski, W. Osikowicz,

A.W. Denier Van Der Gon, W.R. Slaneck, M. Fahlman, Synth. Met. 139(2003) 1.

21] A. Pud, N. Ogurtsov, A. Korzhenko, G. Shapoval, Prog. Polym. Sci. 28(2003) 1701.

22] S. Yang, E. Ruckenstein, Synth. Met. 59 (1993) 1.23] E. Ruckenstein, S. Yang, Synth. Met. 53 (1993) 283.24] E. Ruckenstein, Y. Sun, Synth. Met. 74 (1995) 107.25] P. Banerjee, B.M. Mandal, Synth. Met. 74 (1995) 257.26] P. Banerjee, B.M. Mandal, Macromolecules 28 (1995) 3940.27] A.R. Hopkins, J.R. Reynolds, Macromolecules 33 (2000) 5221.28] H.Q. Xie, Y.M. Ma, J. Appl. Polym. Sci. 77 (2000) 2156.29] H.O. Xie, Y.M. Ma, J. Appl. Polym. Sci. 76 (2000) 845.30] B.H. Jeon, S. Kim, M.H. Choi, I.J. Chung, Synth. Met. 104 (1999) 95.31] Y. Haba, E. Segal, M. Narkis, G.I. Titelman, A. Siegmann, Synth. Met. 110

(2000) 189.32] H.Q. Xie, Y.M. Ma, J.S. Guo, Polymer 40 (1998) 261.33] J. Anand, S. Palaniappan, D.N. Sathyanarayana, Prog. Polym. Sci. 23

(1998) 993.34] M. Reghu, C.O. Yoon, C.Y. Yang, D. Moses, P. Smith, A.J. Heeger, Phys.

Rev. B. 50 (1994) 13931.35] P. Beadle, S.P. Armes, S. Gottesfeld, C. Mombourquette, R. Houlton, W.D.

Andrews, S.F. Agnew, Macromolecules 25 (1992) 2516.

2003.37] E. Kh. Shokr, M.M. Wakkad, H.A. Abd El-Ghanny, H.M. Ali, J. Phys.

Chem. Solids 61 (2000) 75.38] F. Niino, H. Hirasawa, K. Kondo, Thin Solid Films 411 (2002) 28.