Synthesis and characterization of polypyrrole/Sn-doped TiO2 nanocomposites (NCs) as a protective pigment

Page 1

S(

MD

a

ARRAA

KTNPCC

1

sTePtc

appoFsgtpoai

0d

Applied Surface Science 257 (2011) 8317– 8325

Contents lists available at ScienceDirect

Applied Surface Science

j our nal ho me p age: www.elsev ier .com/ loc ate /apsusc

ynthesis and characterization of polypyrrole/Sn-doped TiO2 nanocompositesNCs) as a protective pigment

.R. Mahmoudian ∗, W.J. Basirun, Y. Alias, M. Ebadiepartment of Chemistry, University of Malaya, Kuala Lumpur 50603, Malaysia

r t i c l e i n f o

rticle history:eceived 1 January 2011eceived in revised form 14 March 2011ccepted 14 March 2011vailable online 28 April 2011

eywords:itanium dioxide

a b s t r a c t

We have chemically polymerized pyrrole in the presence of Sn-doped TiO2 nanoparticles (NPs) and TiO2

(NPs) which act as a protective pigment. Field emission scanning electron microscopy (FESEM) and trans-mission electron microscopy (TEM) results show a core–shell structure of pigments in which TiO2 andSn-doped TiO2 NPs have a nucleus effect and caused a homogenous PPy core–shell type morphologyleading to coverage of the TiO2 and Sn-doped TiO2 NPs by PPy deposit. The XRD results indicate thatthe crystalline size of polypyrrole/TiO2 NCs and polypyrrole/Sn-doped TiO2 NCs were approximately93.46 ± 0.06 and 23.36 ± 0.06 nm respectively. The electrochemical impedance spectroscopy (EIS) results

anocompositeolypyrroleorrosionoating

show that the performance of polypyrrole/Sn-doped TiO2 NCs is better than polypyrrole/TiO2 NCs. Theresults indicate that increasing the area of synthesized polypyrrole in the presence of Sn-doped TiO2 NPscan increase its ability to interact with the ions liberated during the corrosion reaction of steel in thepresence of NaCl. The UV–vis results show that the band gap of TiO2 NPs increases with doped of Sn inlattice of TiO2. The increase of the band gap of TiO2 with doping of Sn can decrease the charge transferthrough the coating.

. Introduction

In recent years conducting polymers have attracted con-iderable interest for the development of advanced materials.hese compounds are organic materials that generally possess anxtended conjugated �-electron system along a polymer backbone.olypyrrole (PPy) is conducting polymer which due to its high elec-rical conductivity, has been suggested to be used as protectiveoatings on oxidizable metals [1–7].

Conducting polymer/inorganic oxide nanocomposites havettracted great attention due to their unique microstructure,hysiochemical and electro-optical properties, and wide range ofotential uses as a battery cathode and also in the constructionf nanoscopic assemblies in sensors and microelectronics [8–10].erreira et al. [11] codeposited polypyrrole/TiO2 on AISI 1010 steelubstrates in oxalic acid medium. They reported that PPy/TiO2 filmsive a more anodic corrosion potential than PPy or steel substrates,herefore, the presence of the pigment improves the protectiveroperties of the films. Armes et al. [12] synthesized PPy-tin (VI)

xide colloidal NCs and reported that certain tin (IV) oxide sols canct as effective particulate dispersants for PPy in addition to sil-ca sols. Subasri et al. [13] revealed that the composite electrode

∗ Corresponding author. Tel.: +60 173928320; fax: +60 79674193.E-mail address: M R [email protected] (M.R. Mahmoudian).

169-4332/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.apsusc.2011.03.075

© 2011 Elsevier B.V. All rights reserved.

bearing Sn: Ti in the molar ratio 1:1 seems to be suited for cor-rosion protection of copper under UV illumination. Cao et al. [14]reported the synthesis of Sn-doped TiO2 nanoparticles by sol–gelmethod and annealed them at different temperatures. They showedthat Sn4+ ions can be doped into TiO2 lattice in substitutional modeand/or exist in the form of SnClx or tin oxide, which depends onthe annealing temperature. Dai et al. [15] reported that Sn-dopingresults in a blue-shift of the optical absorption edge of anatase TiO2.They showed that in Sn-doped TiO2 system, most Sn 5s states arelocated at the bottom of the conduction band and are mixed withthe Ti 3d states.

In this work, PPy/Sn-doped TiO2 nanocomposites (NCs) was syn-thesized and used as a protective pigment in organic coatings. Thecorrosion resistance properties of PPy/Sn-doped TiO2 NCs wereevaluated by incorporating in an epoxy-polyamide polymer andthen it was applied onto steel substrates and tested in sodiumchloride electrolyte.

2. Experimental

2.1. Synthesis of Sn-doped TiO2 and TiO2 NPs

Nanoparticles of Sn-doped TiO2 were prepared by sol–gelmethod with titanium tetraisopropoxide (Ti[OCH(CH3)2]4, Aldrich,A.C.S. Reagent) and tin acetate (Aldrich) was used as source of tita-nium and tin respectively. Firstly, titanium (IV) isopropoxide and

Page 2

8 Surface Science 257 (2011) 8317– 8325

tiadartpS

2c

twr2pfpwat

eTlsi

w1sdsIiiidorwtS

2

ct4Fdemmu3T1stbi

3.1.1. XRDFig. 1(a1) and (b1) displays the XRD patterns of TiO2 and Sn-

doped TiO2 NPs respectively. Both of them exhibit the anatasestructure of TiO2. From the comparison of both diffractograms, the

318 M.R. Mahmoudian et al. / Applied

in acetate were dissolved in glacial acetic acid. Secondly deion-zed water was added to the prepared solution for hydrolysisnd poly-condensation reaction. The molar ratio of composite Sn-oped (9.0 mol%) was 1:10:200 of [Ti(OCH(CH3)2)]4:glacial aceticcid:H2O. The mixture was stirred for 6 h at room temperature. Theesultant homogeneous solution was maintained at 75 ◦C for gela-ion process. The gel was calcined at 500 ◦C for 5 h. In addition, werepared TiO2 NPs by using the same method for comparison withn-doped TiO2 NPs [16].

.2. Synthesis of PPy/Sn-doped TiO2 NCs and preparation ofoating

PPy/Sn-doped TiO2 NCs were prepared by chemical polymeriza-ion. In a typical procedure, 0.24 g of Sn-doped TiO2 nanoparticlesere added to 35 mL of 0.1 M HCl solution containing 1 mL of pyr-

ole monomer and stirred for 30 min. Then 3.51 g of (NH4)2S2O8 in0 mL of 0.1 M HCl solution was slowly added drop-wise to the sus-ension mixture with constant stirring. The stirring was continuedor 12 h at room temperature in order to completely polymerize theyrrole monomers. The PPy/Sn-doped TiO2 NCs was filtered andashed with distilled water repeatedly and dried in vacuum oven

t 40 ◦C for 24 h. The same method was used for the preparation ofhe PPy/TiO2 NCs.

The synthesized PPy/Sn-doped TiO2 NCs powder was added topoxy resin solution (Epikote 1001 solution 70 wt.% in toluene).he hardener, a polyamide (Crayamid 115) was mixed with theiquid substance with the dispersed PPy/Sn-doped TiO2 NCs attoichiometric ratio. The PPy/Sn-doped TiO2 NCs content in thencorporated epoxy-polyamide coating was 1 wt.%.

The elemental composition analysis of the steel plate substrateas: (weight%: 2.71% C, 0.49% Si and 94.79% Fe). Steel panels of

0 cm × 5 cm × 0.2 cm sizes were sand-blasted to get a near whiteurface profile as per Swedish specification SA 2.5 [17]. The PPy/Sn-oped TiO2 NCs incorporated epoxy-polyamide was applied overand-blasted steel surfaces by brushing and drying for 4 days.n addition, a set of panels was coated with epoxy-polyamidencluding PPy and PPy/TiO2 NCs separately and used as control tonvestigate the effect of the existence of TiO2 and Sn-doped TiO2n the performance of pigment respectively. A glass tube of 2 cmiameter of length 4 cm was fixed on the coated steel. The specimenf coated steel was used as working electrode (WE) and graphiteod as counter electrode (CE). A saturated calomel electrode (SCE)as used as the reference; all the potential values were referred to

his electrode. The experimental cell used in this work is shown incheme 1.

.3. Characterization

The thickness of the coating was measured using Mechani-al Profiler (KLA-Tencor, P-6) equipment. The panels with coatinghickness 60 ± 5 �m were selected for corrosion studies. Spectrum00 (FT-IR/FT-FIR spectrometer) equipment was used to obtain theT-IR of TiO2 NPs, Sn-doped TiO2 NPs, PPy/TiO2 NCs and PPy/Sn-oped TiO2 NCs powder. The X-ray diffraction (XRD), transmissionlectron microscopy (TEM) and field emission scanning electronicroscopy (FESEM) were used to calculate the crystalline size andorphology of nanoparticles and nanocomposites. The EDX was

sed to detect the presence of Sn in the TiO2 lattice. A iCamscope-05A (magnification 40×) was used to take the sample images.he impedance spectra were obtained over the frequency range of00 kHz–10 mHz, with acquisition of 10 points per decade, with a

ignal amplitude of 5 mV around the open circuit potential. A Poten-iostat/Galvanostat Model PGSTAT-302N from Autolab, controlledy a USB IF030 interface and by the FRA.EXE software both installed

n a PC computer was used to perform these experiments. For cor-

Scheme 1. Experimental cell setup: (1) coated steel, (2) coated removed area formaking working electrode contact, (3) reference electrode, and (4) counter elec-trode.

rosion measurements each surface was immersed into a 3.5% NaClaqueous solution. The analysis of the impedance spectra was doneby fitting the experimental results to equivalent circuits using thenon-linear least-square fitting procedure. The quality of fitting tothe equivalent circuit was judged firstly by the x2-value (i.e. the sumof the square of the difference between theoretical and experimentpoints) and secondly by limiting the relative error in the value ofeach element in the equivalent circuit to 5%.

3. Result and discussion

3.1. Characterization of TiO2 NPs

Fig. 1. XRD patterns of: (a1) the TiO2 NPs and (b1) the Sn-doped TiO2 NPs heat-treated at 500 ◦C for 5 h, EDX of: (a2) the TiO2 NPs and (b2) the Sn-doped TiO2 NPs.

Page 3

M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325 8319

Table 1XRD parameters and average particle sizes of the TiO2 NPs and Sn-doped TiO2 NPs calculated by Scherrer’s equation.

Sample 2� h k l dh k l (nm) Structure Lattice parameter (nm) Cos ϕ Daverage (nm) V (nm)3

TiO2 38.100 (0 0 4) 0.2513 Tetragonal a = 0.3784 0 27.85 ± 0.05 0.143

al

ioccwFfXrta

(tti(ttidp(d

3

((TwSTXpT

3

NPsbo

TEF

was used to calculate the crystalline size in this work. The crys-talline size was estimated about 93.46 ± 0.06 and 23.36 ± 0.06 nm

48.179 (2 0 0) 0.1886

Sn-doped TiO2 37.660 (0 0 4) 0.2385 Tetragon48.012 (2 0 0) 0.1892

ntensity of the diffraction peak (1 0 1) decreases with the presencef Sn in the lattice of TiO2 NPs. The reason for this phenomenonould be due to the decrease in crystallinity. The crystallite size (D)an be calculated using the Scherrer’s equation, D = 0.9�/ cos �,here � is the wavelength of X-ray radiation (� = 0.154 nm), is

WHM (full width half maximum) of the peak (radians) correctedor instrumental broadening (FWHM (deg) of the peak (1 0 1) of theRD patterns of TiO2 and Sn-doped TiO2 NPs are 0.510 and 0.776espectively), � is the Bragg angle. The average crystalline size forhe TiO2 and Sn-doped TiO2 prepared in this study was calculatedbout 27.85 ± 0.05 and 18.30 ± 0.06 nm respectively.

The values of the distance (d) between adjacent planes in theh k l) were calculated from the Bragg equation � = 2d sin �; the lat-ice constants a, b, c, the interplanar angle, the angle ϕ betweenhe plane (h1 k1 l1), of spacing d1, and the plane (h2 k2 l2), of spac-ng d2 were calculated from the Lattice Geometry equation [18]. The0 0 4) and (2 0 0) peaks of TiO2 and Sn-doped TiO2 NPs were usedo calculate the structural parameters (Table 1). It is well knownhat the shift of diffraction peak supports the doping of metal ionsnto lattice structure. The larger the difference of ion radii betweenoped guest-metal ion and host-metal ion, the more obvious is theattern shift. Since Sn4+ ion radius (0.71 A) is similar to that of Ti4+

0.68 A) [19] therefore the shift of diffraction peaks supporting theoping of metal ions into the lattice is small.

.1.2. The EDX of TiO2 NPs and Sn-doped TiO2Fig. 1(a2) and (b2) shows the EDX of TiO2 NPs and Sn-doped TiO2

NPs) respectively. This result shows the existence of elements of Tifrom TiO2), O (from TiO2) and Sn (from doped Sn in TiO2 lattice).he weight percentage of each element is given in Table 2. Theeight percentage of doped Sn shows that Ti was substituted with

n so the weight percentage of Ti decreased in the EDX of Sn-dopediO2 NPs. While the EDX result proved the existence of Sn and theRD pattern of Sn-doped TiO2 NPs did not show any extra peaks ofure Sn or SnO2, therefore Sn-doping into the lattices of TiO2 andi was substituted with Sn.

.1.3. The FT-IR of TiO2 NPs and Sn-doped TiO2 NPsFig. 2 shows the FT-IR spectrum of TiO2 NPs and Sn-doped TiO2

Ps prepared by the sol–gel method in the range of 250–4000 cm−1.

eaks appearing at around 3300 cm−1 are assigned to fundamentaltretching vibration of O–H hydroxyl groups [20]. The absorptionands at around 1600 cm−1 are caused by the bending vibrationf coordinated H2O as well as from the Ti–OH. Peaks located at

able 2DX data of synthesized TiO2 NPs (from Fig. 1(a2)) and Sn-doped TiO2 NPs (fromig. 1(b2)).

Weight%

From Fig. 1(a2)Element

Ti 71.71O 28.29

From Fig. 1(b2)Element

Ti 62.94O 29.02Sn 8.04

c/a = 2.5134a = 0.3785 0 18.30 ± 0.06 0.136c/a = 2.5204

400 cm−1 and 760 cm−1 in the FT-IR spectrum of TiO2 NPs are likelydue to the vibration of the Ti–O bond and the O–O stretching modefor TiOOH on the surface [21]. Since the Ti–O bond is shorter thanthe Sn–O bond [22], therefore, the doping of Sn in TiO2 NPs maylead to a shift of the wavenumber of Ti–O lattice vibration.

3.1.4. The TEM of TiO2 NPs and Sn-doped TiO2 NPsTo determine the morphology and size distribution of TiO2 NPs

and Sn-doped TiO2 NPs, the powder sample calcined at 500 ◦C wassuspended in acetone under sonication for 1 h. Fig. 3(a) and (b)shows the TEM of TiO2 NPs and Sn-doped TiO2 NPs respectively.As can be seen, the shape of nanoparticles is random (spherical andcubic). This is due to the fact that when the reaction time and tem-perature increase, the nanoparticles become random in shape [23].When the size of the particle is very small, the ratio of the atoms onthe surface to all of the atoms in the particle increases. In this sit-uation, the surface atoms can affect the morphology of the particle[24]. When the particle size grows, ratio of the atoms on the sur-face to all of the atoms in the particle decreases. At a certain point,the effect of the surface atoms is negligible. The size histograms ofthe TiO2 NPs and Sn-doped TiO2 NPs are shown below the corre-sponding TEM images. The histograms show that the mean particlesize of the TiO2 NPs and Sn-doped TiO2 NPs are about 28 ± 7 and24 ± 4 nm, respectively.

3.2. Characterization of PPy/TiO2 NCs and PPy/Sn-doped TiO2 NCs

3.2.1. The XRD of PPy/TiO2 NCs and PPy/Sn-doped TiO2 NCsFig. 4 displays the XRD patterns of the PPy, PPy/TiO2 NCs and

PPy/Sn-doped TiO2 NCs. The broad amorphous diffraction peakwhich appears at 2� = 15–35◦ range in the XRD diffractogram ofthe PPy can be ascribed to the scattering of the bare polymer chainsat the interplanar spacing [25,26] (Fig. 4(a)). This broad diffractionpeak of PPy in the PPy/TiO2 NCs is replaced with peak (1 0 1) of TiO2,indicating that the crystallinity of PPy for the PPy/TiO2 (Fig. 4(b))is much lower than that of the pristine PPy (Fig. 4(a)). The samephenomenon occurred in XRD of PPy/Sn-doped TiO2 (Fig. 4(c)).The (1 0 1) peak is attributed to the anatase TiO2 phase [27] and

for PPy/TiO2 NCs and PPy/Sn-doped TiO2 NCs respectively from theScherrer equation.

Fig. 2. FT-IR spectra of: the TiO2 NPs and the Sn-doped TiO2 NPs heat treated at500 ◦C for 5 h.

Page 4

8320 M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325

rams

3N

atnpbm(PsoscrN

3T

ptNitv

Fig. 3. Transmission electron microscopy (TEM) image and the size histog

.2.2. The FESEM ad TEM of PPy/TiO2 NCs and PPy/Sn-doped TiO2Cs

The scanning electron micrographs (FESEM) of PPy/TiO2 NCsnd PPy/Sn-doped TiO2 NCs are shown in Fig. 5(a) and (b) respec-ively. The FESEM indicates that TiO2 and Sn-doped TiO2 NPs have aucleus effect and caused a homogenous PPy core–shell type mor-hology leading to the coverage of TiO2 and Sn-doped TiO2 NPsy PPy deposit. This result was confirmed by TEM and elementalapping of PPy/Sn-doped TiO2 NCs. As can be seen in Fig. 6(a) and

b) the nanocomposite structure of PPy/Sn-doped TiO2 NCs andPy/TiO2 NCs show a nucleus of Sn-doped TiO2 and TiO2 and ahell of polymer. Fig. 6(c) and (d) shows the elemental mappingf Ti and Sn in PPy/Sn-doped TiO2 NCs respectively. This resulthows a distribution of Sn-doped in lattice of TiO2 and Ti as theore in core–shell structure of PPy/Sn-doped TiO2 NCs. The TEMesults show that the nanocomposite size of PPy/Sn-doped TiO2Cs is smaller than PPy/TiO2 NPs.

.2.3. FT-IR spectroscopy of PPy, PPy/TiO2 NCs and PPy/Sn-dopediO2 NCs

The FT-IR spectrum of polymer coatings is shown in Fig. 7. Theeaks at 3624 cm−1, 3630 cm−1 and 3738 cm−1 in the FT-IR spec-ra of the synthesized PPy, PPy/TiO2 NCs and PPy/Sn-doped TiO2

Cs respectively can be attributed to N–H bond. Peaks observed

n the range of 1400 cm−1 to 1600 cm−1 in the FT-IR spectra ofhe synthesized polymers can be attributed to the fundamentalibrations of the pyrrole rings. Peaks at 1659 cm−1, 1654 cm−1 and

of the TiO2 NPs and the Sn-doped TiO2 NPs heat treated at 500 ◦C for 5 h.

1698 cm−1 are related to C–N C bond in the FT-IR spectra of PPy(Fig. 7(a)), PPy/TiO2 NCs (Fig. 7(b)) and PPy/Sn-doped TiO2 NCs(Fig. 7(c)) respectively. While the FT-IR spectra of PPy/TiO2 NCs(Fig. 7(b)) and PPy/Sn-doped TiO2 (Fig. 7(c)) show the character-istic peaks of PPy, but these peaks in the nanocomposites have ashift, indicating a certain interaction between polymer backboneand TiO2 and Sn-doped TiO2 nanoparticles [28]. The strong peakat 351 cm−1 and 343 cm−1 in Fig. 7(b) and (c) for the spectrum ofPPy/TiO2 NCs and PPy/Sn-doped TiO2 NCs respectively are relatedto the Ti–O bond [21]. The peak of Ti–O in spectrum of PPy/Sn-doped TiO2 NCs shifted to lower wavenumber. The reason of thisphenomenon could be related to the replacement of some Ti withSn ions. The bonding energy of Sn–O is lower than Ti–O so it isreasonable that the shift is observed.

3.3. Corrosion study

Fig. 8 shows the effect of 3.5% NaCl solution on the surface ofsteel coated with; PPy (a), PPy/TiO2 NCs (b) and PPy/Sn-doped TiO2NCs (c) incorporated epoxy-polyamide coating and epoxy free ofpigment (d); before (up) and after 30 days of corrosion test (down).The elemental mapping of prepared coatings with nanocomp-site pigments indicates the homogeneity and distribution of Ti

and Sn-doped TiO2 incorporated epoxy-polyamide. The increase incorrosion resistance of steel coated with PPy/TiO2 NCs and PPy/Sn-doped TiO2 NCs incorporated epoxy-polyamide coating could berelated to the lower permeability of the coatings against corro-

Page 5

M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325 8321

sefwcdt

ieidcsoie

Table 3Electrochemical parameters obtained by simulation of the EIS results of the steelcoated with PPy/Sn-doped TiO2 NCs incorporated into epoxy-polyamide coating asa function of immersion time in 3.5% NaCl aqueous solution.

Time (days) Rc (M� cm2) Q/Yo (n�−1 sn cm−2) n Cc (nF cm−2)

1 2.182 1.313 0.825 0.26403 1.577 1.845 0.807 0.31416 1.305 2.859 0.778 0.43439 1.319 2.499 0.789 0.3788

12 1.103 2.961 0.779 0.408115 1.009 3.084 0.778 0.399918 1.012 3.778 0.761 0.470521 0.866 3.855 0.763 0.4366

Fig. 4. XRD of: (a) PPy, (b) PPy/TiO2 NPs and (c) PPy/Sn-doped TiO2 NPs.

ive species, when compared to the steel coated with pigment-freepoxy. In addition, the pitting corrosion can be observed on the sur-ace coated steel with PPy incorporated epoxy-polyamide coatinghile the number of pitting corrosion decreased in PPy/TiO2. The

omparison between all coating images shows that there is lessamage on the steel surface coated with PPy/Sn-doped TiO2 NCs athe end of 30 days immersion time.

The corrosion protection performance of the coating contain-ng PPy, PPy/TiO2 NCs and PPy/Sn-doped TiO2 NCs in incorporatedpoxy-polyamide have been studied by EIS on the coated panelsn 3.5% NaCl solution. Fig. 9(a) shows the Nyquist plots of PPy/Sn-oped TiO2 NCs, PPy/TiO2 NCs, PPy incorporated epoxy-polyamideoating and epoxy only coated steel specimen after 1 day immer-

ion time. The results show that the Zre increased with the presencef TiO2 NPs and Sn-doped TiO2 NPs in the matrix of PPy. Fig. 9(a)ndicates that the presence of Sn in TiO2 lattice can increase thefficiency of performance of PPy/TiO2 NCs as pigment for corrosion

24 0.926 4.741 0.754 0.468927 0.800 5.325 0.752 0.543530 0.677 6.782 0.7394 0.6242

control. Fig. 9(b) shows the Bode plot of coated steel with differentcoatings after 12 days immersion time in NaCl 3.5% solution. Theresults show that the log|Z| of steel coated with PPy/TiO2 NCs andPPy/Sn-doped TiO2 incorporated epoxy-polyamide reached about5.5 and 6.5 respectively after 12 days immersion time in 3.5% NaClsolution. These results confirm that with the increase of immersiontime, the corrosion protection performance of the coating contain-ing PPy/Sn-doped TiO2 NCs is better than other coatings. Fig. 9(c)shows the variation of the log|Z| of coating containing PPy/Sn-doped TiO2 NCs with the increase of immersion time. The resultsshow that the log|Z| decreased with the increase of electrolytepenetration through the coating for longer immersion time. Thevariations of log|Z| became smaller at the end of 30 days immersiontimes where the value is 5.8.

The electrochemical parameters of the prepared coating byPPy/Sn-doped TiO2 NCs/electrolyte were evaluated by employ-ing the ZSimpWin software. We observed an excellent agreementbetween experimental results and the parameters obtained fromthe R (RQ) equivalent circuit model where the chi-squared (x2)minimized at 10−4 value (Fig. 9(c) inset). The R (RQ) equiva-lent circuit model was used in the simulation of the impedancebehaviour of the coating/electrolyte, from the experimentallyobtained impedance data. The model was built using series com-ponents; the first is the bulk solution resistance of the electrolyteRs, the second is the parallel combination of the constant phaseelement (CPE) of the coating capacitance (Cc), and Rc whichis the coating resistance. Instead of pure capacitors, constantphase elements (CPE) were introduced in the fitting procedure toobtain good agreement between the simulated and experimentaldata (Table 3). The impedance of CPE is defined as ZCPE = 1/Yo = 1/Q(jω)n; (1 �−1 sn) where Q is the combination of properties relatedto both the surface and the electroactive species independent offrequency; “n” is related to a slope of the log|Z| vs. log f in the Bodeplot; ω is the angular frequency [25]. CPE is then given as boththe parameter Q and the exponent “n”; it should be stressed thatfor simplicity Q is often considered as a capacitance. The followingequation has been used to convert Q into coating capacitance Cc

from the impedance values [29]:

Cc = Q (ω′′

m)n−1

(1)

where Cc is the coating capacitance, ω′′m is the angular frequency at

which Z imaginary is maximum.An estimation of the change in coating deterioration can be

obtained by considering the variation of Rc and Cc as a functionof immersion time. Fig. 10 indicates the variation of Rc and Cc withthe increase of immersion time. There is an initial decrease of Rc

up to 6th day, and it increased until 9th day immersion time. Thedecrease and minimum of Rc values have been associated with theformation and rupture of blisters and the subsequent increase andmaximum of Rc is due to the deposition of corrosion products in the

Page 6

8322 M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325

(b) PP

bbrCdithTwd

tpr

Fig. 5. FESEM of the synthesized: (a) PPy/TiO2 NCs and

lister [30]. The variation of Cc is another way to understand theehaviour of coating (Fig. 10). As the coating degrades, the coatingesistance decreases and the coating capacitance increase. Initiallyc increased but after 6th days there is a decrease until the 9thay. The increase of the capacitance value is an indication of the

ncrease in the uptake of water (electrolyte) [31]. Comparatively,here is an agreement with the results of Rc and Cc in Fig. 10. Theigher value of Rc is in accordance with the lower capacitance value.he results show that Cc value increases from 0.26 to 0.62 nF cm−2

hile the value of Rc decreased from 2.18 to 0.67 M� cm2 after 30ays immersion time in 3.5% NaCl solution.

The performance of an organic coating as corrosion protec-ion can be described by several factors; barrier effects whichrevent oxygen and moisture from reaching the metal substrate,eaction with the corrosive ions and formation of a passivation

Fig. 6. Transmission electron microscopy (TEM) image of: (a) PPy/Sn-doped TiO2 N

y/Sn-doped TiO2 NCs by using (NH4)2S2O8 as oxidant.

layer [32–34]. There are two reasons for the better performanceof PPy/Sn-doped TiO2 NCs than PPy/TiO2 NCs.

First: the one of the advantages of polymerization of pyrrole inthe presence of nanoparticles is the increase of dispersion of PPy.The dispersion of synthesized PPy in presence of Sn-doped TiO2NPs is more than TiO2 NPs. The increase of PPy dispersion in thereaction medium can increase the efficiency of PPy in corrosioncontrol. The PPy forms around these particles giving a core–shelltype structure: the core being nanoparticles with the shell formedby PPy. This type of structure can increase the surface area for PPywhich can interact with the metal surface. Other authors reported

the effect of the nano-size additives to improve the barrier proper-ties of polymers for diffusion of solvent/gases [35]. Since the XRDresults show that the crystalline size of PPy/Sn-doped TiO2 NCs issmaller than PPy/TiO2 NCs hence the performance of PPy/Sn-doped

Cs, (b) PPy/TiO2 NCs. The elemental mapping of: (c) Ti and (d) Sn-doped TiO2.

Page 7

M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325 8323

FT

TWttpteioiot

Fig. 9. (a) Nyquist plots of steel coated with: (1) epoxy-polyamide + PPy/Sn-dopedTiO2 NCs, (2) epoxy-polyamide + PPy/TiO2 NCs, (3) epoxy-polyamide + PPy as pig-ment and (4) pigment free epoxy, after 1day of immersion time in 3.5% NaClsolution, (b) Bode plots of steel coated with: (1) epoxy-polyamide + PPy/Sn-dopedTiO2 NCs, (2) epoxy-polyamide + PPy/TiO2 NCs, (3) epoxy-polyamide + PPy as pig-ment and (4) pigment free epoxy after 12 days of immersion time in 3.5% NaClsolution and (c) Bode plots of coated steel with PPy/Sn-doped TiO2 NCs incorpo-rated epoxy-polyamide coating in different immersion times in 3.5% NaCl solution

Fi

ig. 7. FT-IR spectra of synthesized: (a) PPy, (b) PPy/TiO2 NCs and (c) PPy/Sn-dopediO2 NCs by using (NH4)2S2O8 as oxidant.

iO2 NCs has more effective corrosion control than PPy/TiO2 NCs.ith the decrease of the nanocomposite size, the area of interac-

ion between PPy and the surface of steel increases. It is knownhat, the redox behaviour of PPy has a special role in its corrosionrotection properties. Radhakrishan et al. [36] reported that withhe increased area of synthesized electroactive polymer in the pres-nce of nanoparticles can increase its ability to interact with theons liberated during the corrosion reaction of steel in the presence

f NaCl, water and oxygen and get doped and liberate the dopantons which form a passivating layer even when there is initiationf the corrosion process at the substrate. It is notable that the crys-alline size of the Sn-doped TiO2 NPs and PPy/Sn-doped TiO2 NCs

(inset equivalent electrical circuit used for the simulation of the EIS results of Bodeplots of coated steel with PPy/Sn-doped TiO2 NCs incorporated epoxy-polyamidecoating).

ig. 8. The sample images steel coated with; PPy (a), PPy/TiO2 NCs (inset Ti-distribution) (b) and PPy/Sn-doped TiO2 NCs (inset Sn-doped TiO2 and Ti-distribution) (c)ncorporated epoxy-polyamide coating and epoxy free of pigment (d) before corrosion test (up) and after 30 days immersion time (down).

Page 8

8324 M.R. Mahmoudian et al. / Applied Surface Science 257 (2011) 8317– 8325

Ftc

idinPw

fsofipttcrpawintst

F

ig. 10. The variation of: the polymer coating capacitance, Cc and the coat resis-ance, Rc of coated steel with PPy/Sn-doped TiO2 NCs incorporated epoxy-polyamideoating as a function of immersion time in 3.5% NaCl aqueous solution.

s smaller than TiO2 NPs and PPy/TiO2 NCs respectively; hence theispersion of synthesized PPy in the presence of Sn-doped TiO2 NPs

s more than TiO2 NPs. In addition, the TEM result showed that theanocomposite size of PPy/Sn-doped TiO2 NCs was smaller thanPy/TiO2 NCs, subsequently the performance of prepared pigmentith PPy/Sn-doped TiO2 NCs is better than PPy/TiO2 NCs.

Second: The mechanism of corrosion shows that charge trans-er is necessary for the formation of rust and dissolution of theteel. So, one of the ways for corrosion control is the preventionf charge transfer. Fig. 11 shows a schematic diagram for the dif-erent energy levels for the materials in contact with one anothern the present system. It is well understood that PPy and TiO2 are-type and n-type respectively. The TiO2 gives hindrance to holeransport across the interface [37]. The tendency for the forma-ion of permanent passive oxide layer on the steel surface by PPyan be explained with different energy levels of PPy and Fe. Otheresearchers reported that PPy has a catalytic behaviour towards theassive oxide layer formation because a charge transfer reaction islready known to take place at the metal/PPy interfaces, as the PPyas reduced (Fig. 11). It is also known that the oxidation process

nvolves a freshly produced water soluble ferrous (Fe2+) oxides to3+

on-soluble ferric (Fe ) oxides. Therefore a very compact, protec-

ive oxide layer was formed on steel [38]. Fig. 12 shows the UV–vispectra of TiO2 NPs and Sn-doped TiO2 NPs. The result shows thathe band gap of TiO2 NPs increases with the doping of Sn in the

ig. 11. Schematic diagram of the energy levels of the PPy/Sn-doped TiO2 NCs.

Fig. 12. Band gap of: the TiO2 NPs and the Sn-doped TiO2 NPs heat-treated at 500 ◦Cfor 5 h.

TiO2 lattice. Dai et al. [15] reported that Sn doping results in a blue-shift of the optical absorption edge of anatase TiO2. The shift in theFermi level towards the conduction band energy is the result of theincrease of band gap. If electrons in PPy are excited to the LUMOlevel, since the conduction band (CB) of TiO2 lies below the low-est unoccupied molecular orbital (LUMO) level of PPy, the excitedelectrons can inject into the CB of TiO2, while the holes are leftin the PPy layer. The electron injection is driven by the potentialdifference between LUMO (PPy) and CB (TiO2) [39]. The potentialdifference in this case is about 370 mV (LUMOPPy–CBTiO2

). The bandgaps of SnO2 and TiO2 are about 3.88 and 3.2 eV, respectively, andthe conduction band of SnO2 is approximately 0.5 V more positivethan that of TiO2 [40]. When the two semiconductor particles arecoupled, the conduction band of SnO2 acts as a sink for electrons.Fig. 11 shows that the electron transfer is more possible to occurin the lower SnO2 energy level compared to the higher O2/OH−

energy level. The corrosion and rust formation on steel involvesseveral steps and oxidation and reduction equations [41]. One ofthe steps is reduction of O2 and H2O to OH−:

O2 (g) + 2H2O + 4e− → 4OH− (R1)

Thus, the increase of the band gap for the Sn-doped in latticeof TiO2 and the action of the conduction band of SnO2 as a sinkfor electrons can decrease the charge transfer through the coatingand decrease the probability for the occurrence of reaction (R1).With the decrease of the charge transfer, the corrosion processesare prevented and the coating becomes more effective for corrosionprevention.

4. Conclusions

Polypyrrole/Sn-doped TiO2 nanocomposite (NC) as a protectivepigment in organic coating was synthesized by chemical oxida-tion in aqueous solution. The TEM and FESEM results showed thatthe Sn-doped TiO2 NPs had a nucleus effect and caused a homoge-nous PPy core–shell type of morphology leading to the coverage ofSn-doped TiO2 NPs by PPy deposit. The XRD result showed thatthe crystalline size of PPy/Sn-doped TiO2 NCs was smaller thanPPy/TiO2 NCs. The EIS results confirmed better performance forcorrosion protection for the PPy/Sn-doped TiO2 NCs in comparisonwith PPy/TiO2 NCs. There are two reasons for the better perfor-mance of PPy/Sn-doped TiO2 NCs:

1. The increase of area of synthesized PPy in the presence of Sn-doped TiO2 NPs can increase its ability to interact with the ionsliberated during the corrosion reaction of steel in the presenceof NaCl.

Page 9

Surfac

2

A

aoF

R

[

[[[[

[[

[

[

[[[

[

[[

[[[

[[

[

[[

[

[

[[

[[

[

M.R. Mahmoudian et al. / Applied

. The increase of the band gap for the Sn-doped TiO2 and theaction of the conduction band of SnO2 as a sink for electrons candecrease the charge transfer through the coating and decreasethe probability of reduction reaction of O2 and H2O.

cknowledgments

One of the authors wishes to thank Mojdeh Yeganeh for valu-ble discussion. This work has been supported by the Universityf Malaya grant No: PS322/2008C, fundamental research grantP086/2007C and PPP grant FS343/200.

eferences

[1] J.I. Martins, M. Bazzaoui, T.C. Reis, S.C. Costa, M.C. Nunes, L. Martins, E.A. Baz-zaoui, Prog. Org. Coat. 65 (2009) 62–70.

[2] D. Sazou, M. Kourozidou, E. Pavlidou, Electrochim. Acta 52 (2007) 4385–4397.[3] M.R. Mahmoudian, Y. Alias, W.J. Basirun, M. Ebadi, Cur. Appl. Phys. 11 (2011)

368–375.[4] H.N. Thi Le, B. Garcia, C. Deslouis, Q. Le Xuan, Electrochim. Acta 46 (2001)

4227–4259.[5] A. Ashrafi, M.A. Golozar, S. Mallakpour, Syn. Met. 156 (2006) 1280–1285.[6] T. Tupen, G. Tansug, B. Yazici, M. Erbil, Surf. Coat. Technol. 202 (2007) 146–154.[7] M.R. Mahmoudian, Y. Alias, W.J. Basirun, Mater. Chem. Phys. 124 (2–3) (2010)

1022–1028.[8] M.R. Mahmoudian, W.J. Basirun, Y. Alias, Appl. Surf. Sci. 257 (2011) 3702–3708.[9] H. Yuvraj, E.J. Park, Y.S. Gal, K.T. Lim, Coll. Surf. A 313–314 (2008) 300–303.10] M.R. Nabid, M. Golbabee, A.B. Moghaddam, R. Dinarvand, R. Sedghi, Int. J. Elec-

trochem. Sci. 3 (2008) 1117–1126.11] C.A. Ferreira, S.C. Domenech, P.C. Lacaze, J. Appl. Electrochem. 31 (2001) 49–56.12] S. Maeda, S.P. Armes, Syn. Met. 73 (1995) 151–155.13] R. Subasri, T. Shinohara, Electrochem. Solid-State Lett. 7 (7) (2004) B17–B20.

14] Y. Cao, T. He, L. Zhao, E. Wang, W. Yang, Y. Cao, J. Phys. Chem. C 113 (2009)

18121–18124.15] R. Long, Y. Dai, B. Huang, J. Phys. Chem. C 113 (2009) 650–653.16] K. Karthik, S.K. Pandian, K.S. Kumar, N.V. Jaya, Appl. Surf. Sci. 256 (2010)

4757–4760.

[[

e Science 257 (2011) 8317– 8325 8325

17] J.R. Venison, Structural Steel Painting: The International Decorative Paints,Allen Devices and Co., Ltd., Bristol, England, 1973, pp. 5–6.

18] B.D. Cullity, Elements of X-ray Diffraction, Addison-Wesley Publishing Com-pany Inc., California, 1956.

19] R. Sui, J.L. Young, C.P. Berlinguette, J. Mater. Chem. 20 (2010) 498–503.20] R. Zhang, L. Gao, Key Eng. Mater. 224–226 (2002) 573–576.21] R. Nakamura, A. Imanishi, K. Murakoshi, Y. Nakato, J. Am. Chem. Soc. 125 (2003)

7443–7450.22] W. Lin, Y.-F. Zhang, Y. Li, K.-N. Ding, J.-Q. Li, J. Chem. Phys. 124 (054704) (2006)

8, doi:10.1063/1.2162896.23] A.R. Rao, V. Dutta, Sol. Energy Mater. Sol. Cells 91 (2007) 1075–1080.24] M. Hosokawa, K. Nogi, M. Naito, T. Yokoyama, Nanoparticle Technology Hand-

book, Elsevier, Amsterdam, 2007.25] J. Ouyang, Y. Li, Polymer 38 (1997) 1971–1976.26] I. Seo, M. Pyo, G. Cho, Langmuir 18 (2002) 7253–7257.27] Y.J. Kwon, K.H. Kima, C.S. Limb, K.B. Shim, J. Ceram, J. Ceram. Process. Res. 3

(2002) 146–149.28] J. Jiang, L. Li, F. Xu, J. Appl. Polym. Sci. 105 (2007) 944–950.29] J. Pan, C. Leygraf, D. Thierry, A.M. Ektessabi, J. Biomed. Mater. Res. 35 (1997)

309–318.30] E.P.M. Van Westing, G.M. Ferrari, J.H.W. De Wit, Corros. Sci. 36 (1994)

979–994.31] B.S. Skerry, C.-T. Chen, C.J. Ray, J. Coat. Technol. 64 (1992) 77–86.32] R.A. Dicky, F.L. Floyd, ACS Symposium Series 322, ACS, Washington, DC, 1986,

p. 1.33] A. Forsgren, P.A. Schweitzer, Corrosion Control Through Organic Coatings, CRC

Press, Boca Raton, 2006.34] Z.W. Wicks, F.N. Jone, S.P. Pappas, Organic Coatings: Science and Technology,

Wiley, New Jersey, 2007.35] Q. Sun, F.J. Schork, Y. Deng, Compos. Sci. Technol. 67 (2007) 1823–1829.36] S. Radhakrishnana, C.R. Sijua, D. Mahantab, S. Patil, G. Madrasc, Electrochim.

Acta 54 (2009) 1249–1254.37] L.E. Umoru, O.O. Ige, J. Min. Mater. Charact. Eng. 7 (2007) 105–113.38] T.L. Nguyen, B. Garicia, C. Deslouis, L.Q. Xuan, J. Appl. Electrochem. 32 (2002)

105–110.39] D. Cahen, G. Hodes, M. Gratzel, J.F. Guillemoles, I. Riess, J. Phys. Chem. B 104

(2000) 2053.40] U. Stafford, K.A. Gray, P.V. Kamat, Chem. Rev. 3 (1996) 77.41] W.S. Tait, An Introduction to Electrochemical Corrosion Testing for Practicing

Engineers and Scientists, Pair O Docs Publications, Racine, Wisconsin, 1994, p.57.