lable at ScienceDirect

Polymer 55 (2014) 4405e4419

Contents lists avai

Polymer

journal homepage: www.elsevier .com/locate/polymer

Feature article

Electrical transport and magnetoresistance in advanced polyanilinenanostructures and nanocomposites

Hongbo Gu a, Jiang Guo a, Xingru Yan a, b, Huige Wei a, Xi Zhang a, Jiurong Liu c,Yudong Huang d, Suying Wei a, b, **, Zhanhu Guo a, *

a Integrated Composites Lab (ICL), Dan F. Smith Department of Chemical Engineering, Lamar University, Beaumont, TX 77710, USAb Department of Chemistry and Biochemistry, Lamar University, Beaumont, TX 77710, USAc Key Laboratory for Liquid�Solid Structural Evolution and Processing of Materials, Ministry of Education and School of Materials Science and Engineering,Shandong University, Jinan, Shandong 250061, Chinad School of Chemical Engineering and Technology, Harbin Institute of Technology, Harbin 150001, Heilongjiang, China

a r t i c l e i n f o

Article history:Received 12 April 2014Received in revised form1 May 2014Accepted 7 May 2014Available online 17 May 2014

Keywords:Polyaniline nanostructuresPolyaniline nanocompositesElectrical transportGiant magnetoresistance

* Corresponding author. Tel.: þ1 409 880 7654.** Corresponding author. Integrated Composites Labment of Chemical Engineering, Lamar University,Tel.: þ1 409 880 7976.

E-mail addresses: suying.wei@lamar.edu (S. W(Z. Guo).

http://dx.doi.org/10.1016/j.polymer.2014.05.0240032-3861/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

In this paper, the electrical transport mechanism and giant magnetoresistance (GMR) in advanced pol-yaniline (PANI) nanostructures and the related PANI nanocomposite systems are critically reviewed. TheGMR phenomena in these materials have been theoretically interpreted by numerical models includingforward interference model and wave-function shrinkage model. This knowledge could serve as aguideline for optimal design and manufacturing of conductive polyaniline-based GMR devices.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Magnetoresistance (MR, a resistance change upon applying amagnetic field) is often defined as MR ¼ (R(H)�R(0))/R(0), whereR(0) is the resistance without a magnetic field and R(H) is theresistance under the magnetic field of H. The Nobel Prize in Physicsof 2007 was awarded to Drs. Fert and Grünberg for their discoveryof giant magnetoresistance (GMR) phenomenon [1,2]. Since its firstdiscovery in a multilayered thin-film structural metallic nano-composite consisting of a couple of ferromagnetic Fe layers sepa-rated by a nonmagnetic Cr layer in 1988 [3,4], much more scientificand technological efforts have been dedicated to this topic. On onehand, scientists are working to disclose the knowledge of GMReffects and to understand its behind physics, aiming to guide thedesign and creation of new structural GMR devices [5e12]. On theother hand, engineers are optimizing the design of GMR baseddevices for wide practical applications including rotation/speed

(ICL), Dan F. Smith Depart-Beaumont, TX 77710, USA.

ei), zhanhu.guo@lamar.edu

sensing, angular position sensing, biosensors and biochips, mag-netic field sensors, hard disc drivers and magnetic random accessmemory (MRAM) in computers [13e18]. Recently, a new technol-ogy “spintronics” has emerged in the condensed-matter physicsarea. Spintronics utilizes the electron spins rather than the chargesto store the information and operates the logic in the solid-statesystems, which can decrease the power consumption and in-crease the data processing speed compared to the conventionalsemiconductor devices [19]. More recently, organic spintronics as aclass of promising electronics have gained more attentions [20]owing to their light-weight, easy processing, low-cost, chemicalstability and biocompatible capability compared to the traditionalmetal matrix materials [21e24].

Especially, the GMReffect in the conducting polymer systems hasattracted intensive attentions due to their easy synthesis, low cost,high conductivity and relatively high GMR signals comparedwith other organic spintronic materials [25,26]. Conducting poly-mers, such as polyacetylene (PA), polyaniline (PANI), polypyrrole(PPy), polythiophenes (PTs), poly(DNTD) and poly(p-phenylenebenzobisthiazole) (PBZT), have been widely used aselectronics, electrodes for electodeposition [27] and supercapacitors[28e30], substrate to form composite nano-catalysts for hydrocar-bon fuel oxidation [31], membrane for chemical separation [32],

H. Gu et al. / Polymer 55 (2014) 4405e44194406

electrochromic (a reversible color change upon applying an elec-trical field) devices [33e36], coupling agents for epoxy nano-composites [37,38] and fire retardants for polymer nanocomposites[39e41]. The conductivity of conducting polymers can be increasedby several orders of magnitude through doping processes [42].Among these conducting polymers, PANI is one of the most studiedconducting polymers over the last few decades due to its easy syn-thesis, low cost, environmental stability, simple doping/dedopingprocess, high conductivity and high pseudocapacitance [43e45]. Ithas been widely applied in many fields including supercapacitors[46e48], electrochromic materials [28], hydrogen photoproduction[49], gas sensors [50], and environmental remediation [51,52]. PANIis distinct from other conducting polymers, in which the nitrogenatom is contributed to the p-band formation and conductionmechanism [53]. PANI consists of alternating reduced (aminegroups) and oxidized (imine groups) repeating units and has threedifferent oxidation states. Normally, thefirst state gives a completelyreduced state; the second state stands for the “half-oxidized” poly-mer state; and the third state represents the fully oxidized state [54].These states are termed as “leucoemeraldine base” (LEB), “emer-aldine base” (EB) and “pernigraniline base” (PB) forms, respectively[55]. Generally, LEB and PB forms are insulating in nature, even afterdoping; only EB form can become conductive after doping/proton-ation process and be converted to emeraldine salt (ES) form. Inter-estingly, different states of PANI have different colors. The LEB formis colorless or sometimes light yellow, the EB base form is blue, theES form is green (which is conductive), and the PB form is violet orblack [56]. The exhibition of different colors at different statesmakesPANI serve as the electrochromic materials possible for the smartwindows and displays [28,56].

There are some reviews focusing on the preparation, processingand applications of PANI [57], negative permittivity of PANI nano-composites [53], PANIbasedbiosensors [58], andPANImembranes forseparation and purification of gases, liquids, and electrolyte solutions

Fig. 1. (a) High resolution transmission electron microscope (HRTEM) of the crystalline zonfrom (c) atomic carbonenitrogen bond; and possible rotation of the carbonenitrogen bondfrom Ref. [61].

[59]. However, the review on the GMR effects in the PANI and itsnanocomposites does not exist in the literature yet. In this article, theelectrical transport in the disordered PANI nanostructures andnanocomposite systems is critically presented, and the recent re-searches on the observedGMReffects in the PANI nanostructures andnanocomposites are critically reported. Meanwhile, the theoreticalanalysis of the GMReffects by numerical models including thewave-function shrinkage model and forward interference model is pre-sented together with some prospective on this topic.

2. Fundamentals of GMR in disordered polyanilinenanostructures and nanocomposites

2.1. Electrical transport mechanism

PANI is a heterogeneous system composed of partially orderedcrystalline regions and disordered amorphous regions upon doping[60]. The crystylline domain is metallic in nature (called metallicislands), Fig. 1(a) [61]. The conduction in this region occurs throughelectron delocalization or hopping of the charge carriers after theformation of polarons [57]. The polarons are formed after theproton doping of the emeraldine base form [62]. The metallicdomain is surrounded by the amorphous region that has disorderedor folded chains and the existence of these disordered chains canreduce the conductivity of PANI, Fig. 1(bef). Tunneling or hoppingof the charge carriers can occur between two metallic islandsthrough amorphous regions [57] and the charge transport is gov-erned by the disorder presented in the system [60]. The disorder inthe conducting polymer systems can arise from the variations inthe conjugation length, rotations and kinking of the polymerchains, van der Waals interactions with neighboring conjugatedmolecules, impurities [63], dipole moments of the neighboringdopant molecules and dipole moments of the molecules of thepolymer matrix [64]. Soukoulis et al. [65] mentioned that the

e in PANI; (b) HRTEM of amorphous zone in PANI; the disordered PANI chain structurecan form (d) zigzag, (e) helical chains, and (f) buckle chains. Reprinted with permission

Fig. 3. Mechanism of hopping conduction. There are two types of hops, one is from A(an occupied state) to B; another one is from B to C. EF is the Fermi energy. Reprintedwith permission from Ref. [73].

H. Gu et al. / Polymer 55 (2014) 4405e4419 4407

disorder would cause tails to appear at the band edges in thedensity of states (DOS), which is firstly documented by Urbach andcalled Urbach tail, Fig. 2(a). When the disorder became stronger, theamplitude fluctuations of the wave-functions (j(x)) at scales star-ted to become larger than the atomic size. Eventually, the eigen-states (the measured state having quantifiable characteristics, suchas position, momentum et al. [66]), in which the amplitudesdecayed away from a center (called localized states), wereappeared, Fig. 2(b). The appeared decay was exponential andbecame zero over a very large distance. Since the mean free path(the distance, where the phase of the wave-function deviatesevidently from that of the plane wave) is comparable to thewavelength, the eigenstates became localized. This phenomenon,first proposed by Philip Anderson at Bell labs in 1958, was calledAnderson localization [67]. Physically, the mechanism of localiza-tion is the suppression of electron tunneling over a very large dis-tance due to the de-coherent effect induced by the randompotential. The parameter of localization length (a0, nm) is used tocharacterize this distance and defined as the logarithmic average ofthe absolute value of the wave-functions jjðxÞj [65].

Generally, the disorder is at microscopic scale and induces thelocalization of electronic wave-functions [68]. As the density oflocalized states near the Fermi level for the charge carriers is finite,the stronger temperature dependent conduction via “hopping” ofthe charge carriers with variable activation energy (ε, related to thetemperature) and hoping length (Rhop) is called variable rangehopping (VRH) [69]. The best experimental support of the locali-zation is the observation of VRH. The concept of VRH was originallygiven by Nevil F. Mott (a Nobel Prize Laureate in Physics in 1977)[70]. In the disordered systems, the charge carriers can hop fromone site to other sites separated spatially by a certain distance(hopping length), which increases with decreasing temperature.This is the origin of the “VRH” [71].

Meanwhile, the charge transport in the VRH system is a com-bination of tunneling and thermal activation provided by thephonons [72]. Thermal activated hopping conduction by electronsin states near the Fermi energy is shown in Fig. 3, which illustratestwo possible hops: one is from A (an occupied state, below theFermi energy) to B state (above the Fermi energy) and another oneis from B state to C state (both B and C are above the Fermi energy).The hopping process from A state to B state is the rate-determiningprocess for these two types of hops [73]. The hopping probability(p) per unit time of this process is expressed asnph exp �2aRhop �W=

�kBTÞ, where nph is the optical phonon fre-

quency and depends on the phonon spectrum (phonon is thequantized normal modes of atomic vibrations in the solids [74].), kBis Boltzmann constant, a is inverse of a0 of the wave-function, Rhopis the hopping length, and W is the hopping energy between twosites and can be expressed as Equation (1):

W ¼ 34pR3hopNðEFÞ

(1)

Fig. 2. DOS of an electron in a random potential. (a) For weak disorder, an Urbach tail of localis shifted downwards slightly by the disorder, and (b) when the disorder is increased, thestates. Reprinted from the permission Ref. [65].

where N(EF) is the density of states at the Fermi level [73]. Thus, theelectrical transport process requires the charge carriers to hopbetween the localized states with the hopping energy W as low aspossible. This means that the charge carriers will hop to the site,which is above the Fermi level and has the minimum energy dif-ference with respect to its original site irrespective of the spatialdistance [72,75].

In the VRH regime, at moderate temperatures, the Mott's lawdescribes the conduction caused by the phonon-assisted tunneling(hopping thermal energy dependent kBT) between the electroniclocalized states centered at different positions and the electricalconductivity (s) can be presented as Equation (2):

s ¼ s0 exp

2664�

�T0T

�1=nþ1

3775; n ¼ 1;2;3 (2)

where the pre-exponential factor s0 is a constant and representsthe conductivity at infinite low temperature limit, T is the Kelvintemperature, T0 (the Mott characteristic temperature) is related tothe energy needed for the hopping conduction of the charge car-riers and can be expressed as Equation (3) [76]:

T0 ¼ 24.h

pkBNðEFÞa30i

(3)

where a0 (nm) is the localization length of thewave function for thelocalized charge carrier. Normally, the defects (e.g., dopants orimpurities) can perturb the electronic structure of the semi-conductors and act as scattering centers to affect the charge carriertransport in the semiconductors [77]. A higher T0 indicates astronger charge carrier scattering. The n value of 3, 2 and 1 inEquation (2) represents three-, two- and one-dimensional system,respectively. The s0 and T0 can be obtained from the intercept and

ized states below the positive-energy square-root continuum of extended states, whichmobility edge eventually moves into the positive-energy regime. r(E) is the density of

H. Gu et al. / Polymer 55 (2014) 4405e44194408

slope of the linear fitting ln(s) ～ T�1/(nþ1). The a0 is expected to bediverged to infinity as the metal-insulator transition (MIT) isapproached (T0 / 0 K). The Rhop must be greater than the a0 inorder to make the Mott's law valid, since the Rhop is described asEquation (4) [78].

Rhop ¼ ð3=8ÞðT0=TÞ1 =

4a0 (4)

In the Mott's VRH mechanism, T0, a0, Rhop and W are importantparameters to characterize VRH transport. The T0 is reported to bestrongly dependent on the disorder presented in the samples [79].Generally, the disorder in the PANI systems can be evaluated by theresistivity ratio (rr ¼ rlow/rhigh, where rlow is the resistivitymeasured at the lowest temperature; and rhigh is the resistivitymeasured at the highest temperature, normally, around 290 K).Normally, the higher the rr, the stronger disorder in the sampleswill be [80]. Sarkar et al. [81] studied the influence of the rr on theT0 and a0 in the oxalic acid (OX) doped PANI samples and revealedthe relations as: T0 f rr and a0 f rr

�1. Meanwhile, the magneto-conductivity (MC) was found to be related to rr. The magnitude ofMC in the PANI samples increases with increasing rr due to theinterface and grain boundary effects.

Meanwhile, after considering the Coulomb interaction betweenthe localized electrons, which creates a “soft” gap in the N(EF), Efrosand Shklovskii showed that the N(EF) should quadratically vanish atthe Fermi level and the conductivity (Equation (2)) was modified toEquation (5) in all dimensions (called ES's law) [82]:

sðTÞfexp

"��TES

T

�1=2#

(5)

where TES is the characteristic temperature for the ES's law. It'sobserved that the Coulomb interaction can be neglected at tem-peratures above certain critical temperature TC (shown in Equation(6)) and the conductivity will follow the Mott's law.

TC ¼ e4a0NðEFÞkBð4pεε0Þ

(6)

where e is the electron charge, ε is the dielectric constant, ε0 is thedielectric constant of vacuum.

At temperatures below the TC, the conductivity will obey ES'slaw [83].

Fig. 4. The wave-function of the impurity electrons in the condition (a) withoutmagnetic field and (b) with magnetic field.

2.2. Theoretical analysis of GMR effect

The mechanisms of the GMR effect in the organic systemsincluding conducting polymers are still open for discussion. Thecommon models to explain the GMR effects in the organic systemsare excitonic pair mechanism model [84], electronehole (eeh)recombination model [85], bipolaron model [86], forward inter-ference model (also called orbital magnetoconductivity theory) andwave-function shrinkage model [78]. The basic idea of the excitonicpairmechanismmodel is from the spin-dependent effect relating toa negative polaron and a positive polaron. The injected electronsand holes from the cathode and anode into the material form thenegative and positive polarons, respectively. They will form boundpolaron pairs and the pair may further become exciton. The formeddifferent kinds of excitons change with time due to the spin dy-namics induced by the hyperfine interaction (which comes from theatomic physics for the interaction of the electronmagnetic momentwith the nuclear magnetic moment [87]). The transition rate con-stant between different kinds of excitons changes under the appliedmagnetic field, causing the current change. This current change is

related to the square of the transition rate constant ratio, whichfinally leads to a positiveMR in thematerials [84]. Normally, the eehrecombination model is used to explain the large anomalous MRunder a small magnetic field [85]. In this model, the electrons (e)and holes (h) can form the eeh pairs. The eeh recombination pro-cess includes the formation of eeh pairs and the annihilation of eehpairs with different spin interconversion of eeh pairs. The formedeeh pairs include one singlet and three triplets and these singletand triplet eeh pairs may either dissociate or recombine. In the eehrecombination model, the MR is related to the square of thedissociation rates of both the singlet and triplets. However, afterassuming that the dissociation rate constant does not showa strongspin dependence and the fastest process is the singlet recombina-tion, the MR is only described by the recombination rate anddissociation rate of triplets. Then the MR will be simplified to anegative value as �(1/2)[qt/(kt þ qt)], where qt and kt is the disso-ciation and recombination rate of the triplets, respectively [85].Bipolaron means two positive or two negative charges combinedtogether. In the bipolaron model, the positive MR is associated withthe blocking of electron transport through the bipolaron states; thenegative MR is caused by the increase in polaron population at theexpense of bipolarons with increasing magnetic field [88]. Thesethree models discussed the GMRmechanism from the viewpoint ofexciton pairs, polaron of electronehole, and bipolaron.

However, the wave-function shrinkage model and forwardinterference model are two numerical electrical transport models,which have been successfully adopted to describe the MR behaviorof the highly disorderedly localized systems in the VRH regime [89].The forward interference model, originally proposed by Nguyen,Spivak, and Shklovskii (NSS) [90], is often used to predict thenegative MR. NSS considered the effect of interference among allpossible hopping paths, which might substantially change thehopping probability between two sites [89]. The conductivity isproportional to the sum of all the possible paths between two sites.As a result of forward interference under a magnetic field, anincreased conductivity (or decreased resistivity) is envisioned, i.e.negative MR [91]. Strong positive MR with a magnetic field wasfirstly predicted by Tokumoto et al. [92] and Shklovskii [72,93], andit was developed further by Shklovskii and Efros using the wave-function shrinkage model [78,94]. Normally, the wave-function ofthe impurity electrons is squeezed in the transverse direction un-der the strong magnetic field. This means that the spherical sym-metric wave-functions without a magnetic field will become cigar-shaped (spindle-shaped and tapering at each end) with the mag-netic field [94]. In the wave-function shrinkage model, the cigar-shaped wave functions in the magnetic field leads to a sharpdecrease in the overlap of the wave function “tails” for an averagepair of neighboring impurities and a reduction in the hoppingprobability between two sites, causing an increased resistivity withincreasing magnetic field and a positive MR [94]. Fig. 4 shows therelated information for the wave-function shrinkage model.

Fig. 5. NSS proposed model for a hop. c is the microscopic length, i.e. typical distancebetween impurities; rif is the distance between i site and f site. Reprinted withpermission from Ref. [91].

H. Gu et al. / Polymer 55 (2014) 4405e4419 4409

2.2.1. Forward interference modelIn the forward interference model, the hopping paths include

the scattering sequence of tunneling electrons by the impuritieslocated within the cigar-shaped domain of length Rhop and width(Rha0)1/2 between the hopping sites. NSS [91] proposed the forwardinterference model for a hop as shown in Fig. 5. The electronhopping from site i to site f experiences various possible hoppingpaths and the conductivity is proportional to the sum of all thepossible paths from site i to site f. After the numerically logarithmaveraging the conductivity over many different possible paths inthe presence of a magnetic field, a linear negative MR was obtainedin the lowmagnetic field limit. Later on, a quadratic negative MR atsmall magnetic fields, which was saturated at high magnetic fields,was observed [91]. Actually, the linear MR (LMR) was oftenobserved in the low magnetic field and the quadratic MR was onlyoccasionally observed in some samples at veryweakmagnetic field.Thus, the resistance ratio caused by forward interference effects,rforward ¼ R(H)/R(0), is expressed approximately by empiricalEquation (7) (which neglects the quadratic term in H) [95]:

rforwardz1=f1þ Csat½H=Hsat�=½1þ H=Hsat�g (7)

where the fitting parameter Csat is the saturation constant andtemperature independent. The fitting Hsat is the effective saturationmagnetic field. Equation (7) is observed to be saturated at highmagnetic fields to a value of 1þ Csat and yields a linear dependenceon H at intermediate fields.

For the Mott VRH system, Hsat is given by Equation (8) [89]:

Hsatz0:7�83

�3=2

1a20

!�he

��TT0

�3=8(8)

where h is the Planck's constant. When the H/Hsat and Csat are small,Equation (7) becomes Equation (9):

rforwardz1� Csat½H=Hsat� (9)

i.e., Equation (10).

MR ¼ DRðH; TÞRð0; TÞ ¼ RðH; TÞ � Rð0; TÞ

Rð0; TÞ z� Csat½H=Hsat� (10)

The MR value in Equation (10) is always negative and associatedwith the H/Hsat and Csat.

2.2.2. Wave-function shrinkage modelIn the wave-function shrinkage model, by means of the perco-

lation theory (which is a quantitative model to describe the clus-tering, diffusion, phase transition, and electrical property indisordered systems for studying the more complicated criticalphenomenon (the critical point in physics) [96]), the magnetic fielddependent MR in the VRH regime can be easily calculated. Theresistance ratio rwave ¼ R(H)/R(0) in terms of the critical percolationparameter xC(H) (also known as the optimum hopping probabilityparameter) is described as Equation (11) [89]:

rwave ¼ expfxCð0Þ½xCðHÞ=xCð0Þ � 1�g (11)

where xC(0) ¼ (T0/T)1/4 for the 3-D Mott VRH system [97], xC(H)/xC(0) is the normalized hopping probability parameter and is afunction of H/PC for the Mott system. PC is fitting parameter for theMott VRH system:

Setting DEx to be an energy interval around Fermi energy whoseextension depends on the percolation parameter (x, a function ofvector r which connects two different sites. x can reach percolationthreshold to determine the electrical conductivity in a random

disordered system), and setting Vx to be the corresponding volumearound one site [97], with E0 being the binding energy of thelocalized state, m is the effective mass of the electron and thecyclotron frequency (u) is defined as u ¼ eH/m, the Vx can bedescribed as Equation (12):

Vx ¼ 8p�2jE0jmu2

�3=2F�xZu

4jE0j�

(12)

where Z ¼ h=2p, F is the wave-function that is the distributionfunction of x [94]:

FðsÞ ¼ZstðsÞ

hth�x2

�$ðshxþ x� 2sÞ þ ðs� xÞx

i

� ðshxþ x� 2sÞ1=2$x$shxðshx� xÞ5=2

dx (13)

with t(s) being the solution of sht þ t � 2s ¼ 0, where x is theposition, s ¼ xZu/(4|E0|), shx and thx are the hyperbolic sine(shx¼ (e x� e (�x))/2) and hyperbolic tangent (thx¼ (e x� e (�x))/(e x þ e (�x))) functions, respectively. The Equation (12) describesthe gradual change of Vx from the sphere to the double paraboloidas the magnetic field increases.

Thus, in point view of percolation theory, a lower bound xC(H)was defined as the critical percolation parameter for percolation tooccur for the appropriate Vx, and this xC(H) can be determined byintroducing the reduced variables, Equation (14):

x*≡xCðHÞ=xCð0Þ (14)

and the normalized magnetic field ratio H* can be expressed asEquation (15):

H*≡ H=PC (15)

where PC is defined as Equation (16):

Fig. 6. MR (Dr/r) vs H2 for the PANI sample at different temperatures. Reprinted withpermission from Ref. [99].

Fig. 7. MR of the PANI-NSA microspheres at 90 and 260 K. Reprinted with permissionfrom Ref. [100].

H. Gu et al. / Polymer 55 (2014) 4405e44194410

PC≡ 6Z.ea20xCð0Þ (16)

For the 3-D VRH system with a relationship of xC(0) ¼ (T0/T)1/4,the PC can be described as Equation (17):

PC ¼ 6Z.h

ea20ðT0=TÞ1=4i

(17)

In the low-field region, Equation (11) is simplified to Equation(18):

rwavez1þ t2H2

P2C

�T0T

�1=4

(18)

i.e., Equation (19):

DRðH; TÞRð0; TÞ zt2

H2

P2C

�T0T

�1=4

(19)

where the numerical constant t2 z 5/2016 [94]. Thus, the wave-function shrinkage model can be expressed as Equation (20):

MR ¼ DRðH; TÞRð0; TÞ zt2

H2

P2C

�T0T

�1=4¼ t2

e2a4036Z2

�T0T

�3 =

4

H2 (20)

Meanwhile, the wave-function shrinkage model has anotherformula under weak magnetic fields [94,98]:

lnrðHÞrð0Þzt2

�a0LH

�4�T0T

�3=4(21)

where LHð¼ ðZ=eHÞ1=2Þ is the magnetic length (also called Laudaumagnetic length, which is related to the Larmor radius, meaningthat the radius of the classical cyclotron orbit with the zero-pointenergy).

The forward interfere model and wave-function shrinkagemodel are often used together to explain the MR transition fromnegative to positive or from positive to negative, inwhich the MR iscontributed from both forward interference effect and wave-function shrinkage effect. Then the total MR ratio should be writ-ten as Equation (22):

rtotal ¼ expfxCð0Þ½xCðHÞ=xCð0Þ � 1�gþ 1=f1þ Csat½H=Hsat�=½1þ H=Hsat�g � 1

(22)

The last term, �1, is to assure that the rtotal is equal to 1 whenH ¼ 0.

3. Giant magnetoresistance in advanced polyanilinenanostructures

In the last decades, there are some researches focusing on theGMR effect in the PANI systems. For example, Reghu et al. [68] re-ported a very large negative GMR around �100% in the PANI filmprotonated by camphor sulfonic acid (CSA) in the weak magneticfield range (0 ~ 0.2 T) at 1.2 K in 1993. This negative GMR wasassociated with a quite large localization length of the electronicwavefunctions, which was similar to the negative MR in the VRHsystem of the inorganic semiconductors that were doped to be nearthe metal-insulator transition. This is the earliest report on theGMR in the PANI systems. In 1998, Ghosh et al. [99] tested the GMRproperty of the HCl doped PANI synthesized in an aqueous ethanolmedium (30% by volume) in the 3-D VRH conduction system. AGMR around 30% was reported in this PANI system at 1.9 K, Fig. 6.This positive MR for the VRH conduction systemwas believed to be

due to the shrinkage in the overlap of the electronic wave functionsof the localized states after applying a magnetic field. The calcu-lated localization length a0 was about 5.6 nm for the PANI with rr of460 at a constantmagnetic field of 8 T using Equation (21) viawave-function shrinkage model. However, the MR at higher temperaturewas not reported.

In 2004, Long et al. [100] observed a negative MR around �12%in a b-naphthalene sulfonic acid (NSA) doped PANI hollow micro-sphere synthesized by a self-assembly method at 90 K andaround�0.6% was obtained in the same PANI microsphere at 260 K,Fig. 7. The absolute MR value increased with either increasing themagnetic field or decreasing the temperature. The temperaturedependent MR property was assigned to the fact that the appliedmagnetic field helped the conductive carriers to conduct from onePANI microsphere to another across the contact barriers. In 2006,this group also reported an MR value up to 91% in the PANI nano-tube pellet doped with NSA at 3 K [101]. Fig. 8 shows the MRproperty of the PANI nanotube pellet between 3 and 250 K. The MR

Fig. 8. MR of the PANI nanotube pellet at different temperatures. The bottom is theenlarged MR from 50 to 250 K. Reprinted with permission from Ref. [101].

H. Gu et al. / Polymer 55 (2014) 4405e4419 4411

decreased with increasing the temperature significantly and an MRtransition from positive to negative between 50 and 100 K wasobserved. Meanwhile, the obtained negative MR was very smallless than �0.2%. The wave-function shrinkage effect and quantuminterference effect on the hopping conduction were used to illus-trate the positive and negative MR, respectively. The change of theMR sign indicated that both effects coexist and compete. The wave-function shrinkage effect made the main contribution at low tem-peratures. However, as the temperature increased, the localizedstates became less localized and the wave-function shrinkage effectbecame weaker. In this situation, the negative MR contributionfrom quantum interference effect was dominated. Thus, an MRvalue was changed from positive to negative. Interestingly, in orderto confirm the origin of the MR effect in these pellet samples, theMR in a single nanotube attached on platinum (Pt) microleads wasmeasured. The results indicated that the MR in a single PANI

Fig. 9. MR of the PANI nanofibers doped with CSA at different temperatu

nanotube was quite smaller than that of the pellet at a magneticfield of 10 T: for the single PANI nanotube, MR ~2.5% (2 K); for thepellet samples, MR ~91% (3 K). This difference suggested that theMR in the bulk pellet sample of PANI nanotubes mainly originatedfrom a random network of inter-fibril contacts.

In 2010, Bozdag et al. [102] reported a large positiveMR (~55% in8 T at 2.5 K) in the PANI nanofibers doped with camphor sulfonicacid (CSA), which were fabricated by dilute chemical oxidativepolymerization method. This MR property was observed to betemperature and magnetic field dependent, Fig. 9. The MR wasobserved to decrease with increasing temperature and decreasingmagnetic field. A crossover of MR value from positive to negativewas observed with increasing temperature at around 87.5 K.Interestingly, the MR property of the PANI nanofibers was observedto change with applying an electrical field, Fig. 10. A higher elec-trical field enhanced the conductance by reducing the activationenergy and hopping length. Thus, the effect of wave-functionshrinkage arising from the magnetic field decreased withincreasing the electrical field. The electrical field is a new parameteron the effect of MR property in the PANI systems so far. However,the electrical field dependent MR property at different magneticfields and temperatures was not reported.

Generally, the electrical conductivity of PANI depends on themolecular weight, oxidation level [30,103], molecular arrangement,crystallinity, inter-chain separation, degree of doping [57] anddopant [104]. Depending on the dopant, the electrical conductivityof the doped PANI can be 8e11 orders of magnitude(10�1 ~ 102 S cm�1) higher than that of the PANI base (10�9 S cm�1)[105]. Recently, Gu et al. [26,44,79] studied the effects of dopant,oxidant and dose of oxidants on the GMR property of pure PANI.Specifically, 53% of GMR value at room temperaturewas obtained inthe p-toluene sulfonic acid (PTSA) doped pure PANI synthesized byammonium persulfate (APS), Fig. 11(a) [26]. A room temperatureGMR value of 65.5% was observed in the phosphoric acid (H3PO4)doped pure PANI oxidized by APS, Fig. 11(b) [79].

Gu et al. [44] also used the hexavalent chromium (Cr(VI)) asoxidant to prepare the PANI samples and the MR property of thesynthetic PANIwas reported. Only around 5% room temperatureMRvalue was reported in these prepared PANI samples, Fig. 11(c).Meanwhile, the dose level of oxidants is found to affect the MRproperty of PANI, Fig. 11(c). These phenomena were due to the factthat the molecular arrangement and crystallinity of PANI could beaffected by the dopants andoxidants. For example, the cystallinity ofPANI-0.75 (K2Cr2O7:0.75 mmol vs PTSA 15 mmol, aniline: 18 mmol)and PANI-3 (K2Cr2O7:3 mmol vs PTSA 15 mmol, aniline: 18 mmol)was 48.6 and 57.6%, respectively. This resulted in different disor-dered structures in the synthesized PANI as aforementioned, whichled to a different distribution of themetallic islands and amorphousregions in the PANI system. Thus, different charge carrier hopping

res and magnetic fields. Reprinted with permission from Ref. [102].

Fig. 10. IeV behavior of the PANI nanofiber network. Inset shows the current depen-dent MR property. Reprinted with permission from Ref. [102].

H. Gu et al. / Polymer 55 (2014) 4405e44194412

probability and hopping length were obtained to yield different MRbehaviors in the PANI systems under different polymerizationconditions.

4. Giant magnetoresistance in polyaniline nanocomposites

Recently, a great progress on the MR effect was observed in thePANI nanocomposites. For example, Long et al. [106] obtained aseries of MR ¼ 4.15% (2 K, 6 T), 1.9% (4 K, 5.4 T) and 1.6% (6 K, 6 T) inthe Au/PANI microfibers with a coreeshell structure synthesized byreacting aniline monomers with HAuCl4 via a self-assembly process

Fig. 11. GMR of the PANI doped with (a) PTSA and (b) H3PO4; and (c) PANI-0.75 and PANI-3 sand PANI-3:K2Cr2O7 was ~3 mol. Reprinted with permission from Refs. [26,44,79].

with CSA as dopant. In the PANI nanocomposite systems, besidestemperature and magnetic field strength, GMR can also be affectedby the loading level of nanofillers. For example, negative GMR ofthe carbon nanotubes (CNTs)/PANI nanocomposites was observedat 10 K and decreased with increasing the CNT loading, the reducedGMR effect was thought to be associated with the long 1-D locali-zation length of CNTs [107]. Gupta et al. [108] fabricated the PANInanotubes/La0.67Sr0.33MnO3 (LSMO) composites using a two-stepmethod including pyrophoric reaction process (an aqueous solu-tion of stoichiometric amount of high-purity lanthanum oxide(La2O3), manganese acetate (Mn(CH3COO)2, and strontium nitrateSr(NO3)2) was stirred with triethanolamine in a hot plate to formthe black, flurry powders, then was calcined at 650 �C) to prepareLSMO and in-situ chemical oxidative polymerization to synthesizePANI. The low field regimeMR about 57, 67 and 73%was reported inthese PANI nanocomposites with an LSMO loading of 10, 20 and30 mg at a magnetic field of 3 kG (¼0.3 T) and 77 K, respectively[108].

Rakhimov et al. [109] reported a new GMR system consisting ofPANI, polystyrene (PS), manganese (II) acetylacetonate (Mn(Acac)2),lanthanum (III) chloride hexahydrate (LaCl3٠6H2O), praseodymium(III) chloride hexahydrate (PrCl3٠6H2O) and element sulfur (S) poly-meric composites with a component ratio of (LaCl3٠6H2O):(PrCl3٠6H2O):(Mn(Acac)2):(PANI):(PS):(S)¼0.1:0.1:0.15:0.35:0.1:0.2.Theappliedcurrent isa transient current,which is changedwith time.The obtained GMR value of this system was changed from 300 to1800% within various time (0e250 s) under a magnetic field of6000 Oe at room temperature, Fig. 12. The GMR effect in this com-posite system strongly depended on the element S and was notobserved in the composite without element S. The element S mighthave the ability to cross-linkwith the aromatic chains of PS and PANIduring themechanochemical treatment. The obtained composites inthe presence ofmanganese ions could provide a physical basis for the

amples at room temperature. PANI-0.75 represents that the used K2Cr2O7 was 0.75 mol

Fig. 12. Room temperature MR v time for the PANI composites of (LaCl3٠6-H2O):(PrCl3٠6H2O):(Mn(Acac)2):(PANI):(PS):(S) ¼ 0.1:0.1:0.15:0.35:0.1:0.2. (The cur-rent is transient current, which changes with time.) Reprinted with permission fromRef. [109].

Fig. 13. MR of (a) pure PANI, and PANI/ZnFe2O4 nanocomposites with a ZnFe2O4

loading of (b) 20, (c) 40, (d) 60 and (e) 80 wt% at H ¼ 2 T, respectively. Inset shows theMR of PANI at 60 K. Reprinted with permission from Ref. [67].

Fig. 14. GMR of (a) 30 wt% Fe3O4 loading of PANI nanocomposites at room temperaturenanocomposites at room temperature. Reprinted with permission from Ref. [79].

H. Gu et al. / Polymer 55 (2014) 4405e4419 4413

GMR effect, which is like lanthanummanganites. In fact, lanthanummanganites are a typical example for the inorganic MR materials.Manganites, such as LaSrMnO3 and LaCaMnO3, are thematerialswiththe colossal magnetoresistance (CMR) which arises from the inter-play of electron-phonon coupling [110] and ferromagnetic (FM)-to-paramagnetic phase transition [111].

Prasanna et al. [67] reported the temperature dependent MRobserved in the PANI/ZnFe2O4 nanocomposites with differentloadings of ZnFe2O4 nanoparticles at a magnetic field of 2 T, Fig. 13.The MR value of pure PANI was very small. The MR in these PANInanocomposites increased with increasing the ZnFe2O4 nano-particle loading. This phenomenon was believed to arise from theincreased charge carrier scattering with increasing the ZnFe2O4nanoparticle loading due to the presence of magnetic ferrite im-purities. Recently, Guo's group has made a significant progress onthe room temperature GMR effect in the PANI nanocomposites[26,112]. Different nanofillers including magnetic, insulating andsemi-conducting types have been used for the synthesis of thenanocomposites using the surface initiated polymerization (SIP)method. The GMR effects in both the magnetic and the nonmag-netic PANI nanocomposites were reported. Meanwhile, a very largeGMR value at room temperature in the nonmagnetic silica/PANInanocomposites was reported and could be comparable with thatin the magnetic Fe3O4/PANI nanocomposites. For example, around95% GMR was reported in the magnetic PANI nanocomposites witha Fe3O4 nanoparticle loading of 30 wt%, Fig. 14(a) [26]. A GMR valueup to 95.5% was reported in the nonmagnetic PANI nanocompositeswith 20 wt% silica nanoparticle loading, Fig. 14(b) [79]. In addition,a GMR of 20% in the nonmagnetic PANI nanocomposites with aBaTiO3 (~500 nm) loading of 20 wt% synthesized by SIPmethod andaround 35% GMR in the 20 wt% BaTiO3 (~500 nm)/PANI nano-composites prepared by simple physical mixture of PANI andBaTiO3 powders were observed [113]. In the PANI nanocompositeswith different carbon nanostructures fabricated with APS asoxidant, a small negative MR (<1%) was observed at 130 K, Fig. 15(aand c). The significantly larger positive MR (15 ~ 30%) was observedat 290 K, Fig. 15(b and d). The quantum interference effect amongmany possible paths in the magnetic field was used to explain thenegative MR at 130 K [91]. The room temperature GMR obtained inthe PANI nanocomposites with 5 wt% graphene higher than that ofthe PANI nanocomposites with the same loading of 1-D filler (CNTsand carbon nanofibers (CNFs)) was due to the pep stacking-induced efficient electrical transport at the PANI and grapheneinterface, Fig. 15(d) [112].

In contrast, multi-walled carbon nanotubes (MWNTs)/PANInanocomposites were preparedwith Cr(VI) as oxidant and the GMR

, Reprinted with permission from Ref. [26]; and (b) 20.0 wt% silica loading of PANI

Fig. 15. MR behavior of pure PANI and its graphene nanocomposites as a function of Gra-10 loading at (a) 130 and (b) 290 K; MR property of the PANI nanocomposites with 5 wt%loading of different carbon nanostructures at (c) 130 and (d) 290 K. (Gra-10: N008-100-P-10, XY: 5�10 mm, Z: 50�100 nm, BET surface area: ~15 m2/g; Gra-40: N008-100-P-40, XY:�44 mm, Z: 50�100 nm, BET surface area: ~15 m2/g; Gra-P: N006-010-P, XY: �14 mm, Z: <40 nm, BET surface area: ~28 m2/g). Reprinted with permission from Ref. [112].

Fig. 16. (A) Synthesized MWNTs/PANI nanocomposites, and (B) (a) T205, (b) TA05, (c)PANI synthesized by Cr(VI), (d) as-received MWNTs, (e) T5, and (f) T20 at room tem-perature. Reprinted with permission from Ref. [114].

H. Gu et al. / Polymer 55 (2014) 4405e44194414

behaviors of these MWNTs/PANI nanocomposites were investi-gated [114]. Fig. 16(A) shows the MR of the synthesized MWNTs/PANI nanocomposites. The MWNTs/PANI nanocomposites with anMWNT loading of 5.0, 10.0 and 20.0 wt% synthesized with Cr(VI) asoxidant were named as T5, T10, T20, respectively. The aniline wasfurther polymerized on T20 sample with APS as oxidant, which wasnamed as T205. The aniline was further polymerized on T20 byCr(VI), named as TCr205. Meanwhile, the nanocomposites con-taining 5.0 wt% MWNTs were also prepared by oxidizing anilinewith APS, named as TA05. Fig. 16(B) depicts the MR property ofthese nanocomposites. A negative GMR around �2% was observedin the T5 and T20 nanocomposites synthesized by Cr(VI) and pos-itive GMR around 7.9 and 11% in the TA05 and T205 nano-composites synthesized by APS.

A room temperature GMR effect of around 22% in thenonmagnetic PANI nanocomposites with a silicon loading of20.0 wt% under magnetic field of 9 T was reported [115]. Generally,the magnetic field detectors are based on the MR effect and thelinear MR with respect to the magnetic field at the low magneticfield regime can be used to describe the sensitivity of the materialsto the magnetic field [116]. A very high magnetic field sensitivity atlowmagnetic field was observed in these nonmagnetic silicon/PANInanocomposites. A room temperature MR transition from positiveto negative was found in the silicon/PANI nanocomposites with asilicon loading of 10.0 wt% in the VRH regime, Fig. 17 [117].

Meanwhile, the obtained GMR property in the nonmagneticPANI nanocomposites has been theoretically analyzed by usingwave-function shrinkage model and forward interference modelfor positive and negative MR, respectively. The obtained positiveMR in the PANI nanocomposites was investigated bywave-functionshrinkage model through Equation (20). In this model, in the low

Fig. 17. MR transition from positive to negative in the 10.0 wt% silicon PANI nano-composites. Reprinted with permission from Ref. [117].

H. Gu et al. / Polymer 55 (2014) 4405e4419 4415

magnetic field, the applied H decreased the overlap probabilitybetween two sites, resulting in an increased resistance withincreasing the H [78]. In the higher magnetic field, the VRHbetween the localized states having a Coulomb gap in the N(EF)caused the quadratic H dependent behavior [97]. According toEquation (20), the a0 can be calculated from T0, a positive MR valueand H by Equation (23):

a40 ¼ 36Z2MRt2e2

�T0T

��3 =4

H�2 (23)

From the obtained a0, the N(EF) and the Rhop could be calculatedfrom Equations (3) and (4), respectively. The calculated parametersof the PANI nanocomposites by wave-function shrinkage model arelisted in Table 1. It is observed that the a0, N(EF) and Rhop aremagnetic field dependent. The a0 and Rhop decreased withincreasing the magnetic field, the N(EF) increased with increasingthe magnetic field and the average hopping energy had no mag-netic field dependent property [118]. Different parameters wereobtained in the PANI nanocomposites filled with different nano-fillers due to the synergistic interaction between the nanofillersand the PANI matrix as confirmed in the FT-IR test, etc. [79,115],leading to different GMR values in these PANI nanocomposites. Forthe silica/PANI nanocomposites, different GMR values were ob-tained in the nanocomposites with different silica loadings. TheGMR value increased as the silica loading was increased to 20.0 wt%

Table 1Calculated a0, N(EF) and Rhop of the nonmagnetic PANI nanocomposites at different mag

Model Sample

Wave-function shrinkage model(APS as oxidant)

20 wt% loading of silica/PANI [79]

20 wt% loading of silicon/PANI [115]

MWNTs/PANI:TA05 [114]

Forward interference model(Cr(VI) as oxidant)

5 wt% loading of MWNTs [114]

and then decreased as the silica loading was increased further.Different GMR values were observed to be reflected by the DRhop.The more reduced Rhop, the higher value of the MR was obtained.They proposed GMR mechanism in the silica/PANI nanocompositesis shown in Fig. 18. In the silica/PANI nanocomposites, the Rhopdecreased with increasing the magnetic field. Meanwhile, it wasfound that the higher DRhop, the higher GMR was obtained in thesilica/PANI nanocomposites.

In the MWNTs/PANI nanocomposites synthesized by Cr(VI)[114], the negative MR was studied by the forward interferencemodel. In the forward interference model, by non-linear fittingR(H)/R(0) ~ H via Equation (7) using Polymath software, the ob-tained Csat value for T5 and T20 samples (Fig. 16(B)-e and f) was0.0251127 and 0.0317174, respectively. The calculated a0 values forthe T20 via Equations (8) and (10) as shown in Table 1 are 696.5,487.0 and 428.4 nm at H of 0.5, 4 and 9 T, respectively. The obtaineda0 for the negative MR value in the MWNTs/PANI nanocompositesoxidized by Cr(VI) was larger than that of the positive MR value inthe nanocomposites TA05 (Fig.16(B)-b) formed by APS as oxidant atthe same magnetic field. This was due to the fact that in thenegative MR VRH system, the localizationwas relatively weak, thusthe quantum interference effect was dominated.

In the PANI nanocomposites with 10.0 wt% silicon loading, theMR transition from positive to negative at a magnetic field of 5.5 Tdemonstrated that there were two kinds of contributions includingpositive and negative contributions to the GMR value, coexistentand competing [67]. This MR sign was separated by both the for-ward interference model and the wave-function shrinkage modelusing Equation (22) in order to further understand the origin of theGMR in the PANI nanocomposite system [5]. After calculation, boththe forward interference effect and the wave-function shrinkageeffect were observed to contribute to the positive MR part, but onlythe forward interference effect contributed to the negativeMR part.The separated positive MR sign and negative MR sign are shown inFig. 19(a and b), respectively.

Generally, most of the researches focused on the conductingform of PANI in the nanocomposites to investigate the GMR prop-erty. However, Lin et al. [119] obtained room temperature GMR of85.7% at a magnetic field of 0.6 T by using the non-conducting EBform of PANI nanocomposites embedded with iron oxide nano-particles, Fig. 20(a and b). Interestingly, they observed that as themagnetic field gradually decreased from 0.6 to 0 T, the electricalresistance remained approximately constant with small fluctua-tions, Fig. 20(c), which was called “memory effect”. This memoryeffect originated from the used insulating nature of PANI, whichprevented the trapped charges from leaking out of the system afterthe magnetic field was switched off. They also analyzed the ob-tained positive GMR with the wave-function shrinkage model andfurther confirmed that the positive GMR was attributed from the

netic field H.

Parameters Magnetic field H (T)

0.5 4 9

a0 (nm) 32.2 14.7 10.5N(EF) (eV cm3)�1 8.2 � 1013 8.6 � 1014 2.4 � 1015

Rhop (nm) 220.4 100.8 71.9a0 (nm) 39.2 13.6 7.8N(EF) (eV cm3)�1 6.5 � 1013 1.6 � 1015 8.3 � 1015

Rhop(nm) 245.8 84.9 48.8a0 (nm) 31.2 12.5 8.73N(EF) (eV cm3)�1 9.3 � 1014 1.4 � 1016 4.3 � 1016

Rhop(nm) 119.3 47.9 33.3a0 (nm) 97.0 56.0 43.9

Fig. 18. Proposed GMR mechanism in the silica/PANI nanocomposites. Reprinted with permission from Ref. [79].

Fig. 19. (a) MRþ obtained from wave-function shrinkage model; and (b) MR � obtained from forward interference model. Reprinted with permission from Ref. [117].

Fig. 20. (a) Schematic of PANI-iron oxide organic hybrid nanocomposite device; (b) magnetic field dependence GMR (obtained from IeV curve) in PANI-iron oxide nanocompositeswith 10% iron nanoparticles by volume; (c) resistance in 10% loading of PANI-iron oxide nanocomposites remained approximately constant after switching off the magnetic field,exhibiting a memory effect. Reprinted with permission from Ref. [119].

H. Gu et al. / Polymer 55 (2014) 4405e44194416

H. Gu et al. / Polymer 55 (2014) 4405e4419 4417

decreased a0 of the charge carriers. This organic hybrid device waseasily synthesized with a lower power consumption and cost,which has the possibility for practical applications in the magneticfield sensor devices at room temperature.

5. Conclusions and perspectives

In this paper, the up-to-date knowledge of the GMR phenom-enon in the PANI and its nanocomposite systems has beenreviewed. The electrical transport in the PANI systems withoutmagnetic field and the MR mechanism including forward inter-ference model and wave-function shrinkage model for the PANIsystems are discussed as well. The MR effect is a common phe-nomenon in the PANI nanostructures and nanocomposites. Onlythe MR values or property (negative or positive) are different.Different PANI nanocomposites have different MR values arisingfrom the synergistic interaction between PANI and nanofillers.However, the GMR effect in the PANI system is still a crucial andpuzzling problem due to the temperature dependent GMR prop-erty [120], which is normally decreased with increasing the tem-perature [115,119]. The synergistic interactions betweennanofillers and PANI matrix on the electrical transport within thehybrid systems are still open for discussing. Meanwhile, thecrystallinity arising from the dopant type and doping degreeshould play a role in the GMR as well. How to prepare the PANIsystem with high room temperature MR values in the large-scalefor practical applications still remains a challenge.

Acknowledgments

This project is financially supported by Research EnhancementGrant (REG) of Lamar University and the National Science Foun-dation (NSF) Nanoscale Interdisciplinary Research Team and Ma-terials Processing and Manufacturing (CMMI 10-30755), NSF-Nanomanufacturing (CMMI-13-14486), and Chemical and Biolog-ical Separations (CBET 11-37441).

References

[1] Fert A. Angew Chem Int Ed 2008;47:5956e67.[2] Grünberg PA. Rev Mod Phys 2008;80:1531e40.[3] Baibich MN, Broto JM, Fert A, Van Dau FN, Petroff F, Etienne P, et al. Phys Rev

Lett 1988;61:2472e5.[4] Binasch G, Grünberg P, Saurenbach F, Zinn W. Phys Rev B 1989;39:4828e30.[5] Bloom FL, Wagemans W, Kemerink M, Koopmans B. Phys Rev Lett 2007;99:

257201.[6] Guo Z, Park S, Hahn HT, Wei S, Moldovan M, Karki AB, et al. Appl Phys Lett

2007;90:053111.[7] Guo Z, Hahn HT, Lin H, Karki AB, Young DP. J Appl Phys 2008;104:014314.[8] Guo Z, Moldovan M, Young DP, Henry LL, Podlaha EJ. Electrochem Solid St

2007;10:E31e5.[9] Zhang D, Chung R, Karki AB, Li F, Young DP, Guo Z. J Phys Chem C 2009;114:

212e9.[10] Davis D, Zamanpour M, Moldovan M, Young D, Podlaha EJ. J Electrochem Soc

2010;157:D317e22.[11] Zhu J, Wei S, Haldolaarachchige N, He J, Young D, Guo Z. Nanoscale 2012;4:

152e6.[12] Zhu J, Luo Z, Wu S, Haldolaarachchige N, Young D, Wei S, et al. J Mater Chem

2012;22:835e44.[13] Gu H, Zhang X, Wei H, Huang Y, Wei S, Guo Z. Chem Soc Rev; 2013:5907e43.[14] Koets M, van der Wijk T, van Eemeren JTWM, van Amerongen A, Prins MWJ.

Biosens Bioelectron 2009;24:1893e8.[15] Srinivasan B, Li Y, Jing Y, Xu Y, Yao X, Xing C, et al. Angew Chem Int Ed

2009;48:2764e7.[16] Guo J, Gu H, Wei H, Zhang Q, Haldolaarachchige N, Li Y, et al. J Phys Chem C

2013;117:10191e202.[17] Gaster RS, Xu L, Han S-J, Wilson RJ, Hall DA, Osterfeld SJ, et al. Nat Nanotech

2011;6:314e20.[18] Hall DA, Gaster RS, Lin T, Osterfeld SJ, Han S, Murmann B, et al. Biosens

Bioelectron 2010;25:2051e7.[19] Chappert C, Fert A, Van Dau FN. Nat Mater 2007;6:813e23.[20] Bogani L, Wernsdorfer W. Nat Mater 2008;7:179e86.[21] Irimia-Vladu M, Sariciftci NS, Bauer S. J Mater Chem 2011;21:1350e61.

[22] Bobbert PA, Nguyen TD, van Oost FWA, Koopmans B, Wohlgenannt M. PhysRev Lett 2007;99:216801.

[23] Schoonus JJHM, Lumens PGE, Wagemans W, Kohlhepp JT, Bobbert PA,Swagten HJM, et al. Phys Rev Lett 2009;103:146601.

[24] Wagemans W, Schellekens AJ, Kemper M, Bloom FL, Bobbert PA,Koopmans B. Phys Rev Lett 2011;106:196802.

[25] Xiong ZH, Wu D, Valy Vardeny Z, Shi J. Nature 2004;427:821e4.[26] Gu H, Huang Y, Zhang X, Wang Q, Zhu J, Shao L, et al. Polymer 2012;53:

801e9.[27] Wang D, Ye Q, Yu B, Zhou F. J Mater Chem 2010;20:6910e5.[28] Wei H, Zhu J, Wu S, Wei S, Guo Z. Polymer 2013;54:1820e31.[29] Wei H, Yan X, Wu S, Luo Z, Wei S, Guo Z. J Phys Chem C 2012;116:

25052e64.[30] Wei H, Gu H, Guo J, Wei S, Liu J, Guo Z. J Phys Chem C 2013;117:

13000e10.[31] Ding K, Jia H, Wei S, Guo Z. Ind Eng Chem Res 2011;50:7077e82.[32] Shao L, Cheng X, Wang Z, Ma J, Guo Z. J Memb Sci 2014;452:82e9.[33] Zhu J, Wei S, Alexander MJ, Dang TD, Ho TC, Guo Z. Adv Funct Mater 2010;20:

3076e84.[34] Zhu J, Wei S, Alexander MJ, Cocke D, Ho TC, Guo Z. J Mater Chem 2010;20:

568e74.[35] Wei H, Yan X, Li Y, Wu S, Wang A, Wei S, et al. J Phys Chem C 2012;116:

4500e10.[36] Wei H, Yan X, Li Y, Gu H, Wu S, Ding K, et al. J Phys Chem C 2012;116:

16286e93.[37] Gu H, Tadakamalla S, Huang Y, Colorado HA, Luo Z, Haldolaarachchige N,

et al. ACS Appl Mater Interfaces 2012;4:5613e24.[38] Gu H, Tadakamalla S, Zhang X, Huang Y, Jiang Y, Colorado HA, et al. J Mater

Chem C 2013;1:729e43.[39] Zhang X, He Q, Gu H, Colorado HA, Wei S, Guo Z. ACS Appl Mater Interfaces

2013;5:898e910.[40] Zhang X, He Q, Gu H, Wei S, Guo Z. J Mater Chem C 2013;1:2886e99.[41] Gu H, Guo J, He Q, Tadakamalla S, Zhang X, Yan X, et al. Ind Eng Chem Res

2013;52:7718e28.[42] MacDiarmid AG. Angew Chem Int Ed 2001;40:2581e90.[43] Stejskal J, Gilbert RG. Pure Appl Chem 2002;74:857e67.[44] Gu H, Wei H, Guo J, Haldolaarachige N, Young DP, Wei S, et al. Polymer

2013;54:5974e85.[45] Zhang X, Zhu J, Haldolaarachchige N, Ryu J, Young DP, Wei S, et al. Polymer

2012;53:2109e20.[46] Zhu J, Chen M, Qu H, Zhang X, Wei H, Luo Z, et al. Polymer 2012;53:5953e64.[47] Wei H, Gu H, Guo J, Wei S, Guo Z. J Electrochem Soc 2013;160:G3038e45.[48] Qu H, Wei S, Guo Z. J Mater Chem A 2013;1:11513e28.[49] Belabed C, Abdi A, Benabdelghani Z, Rekhila G, Etxeberria A, Trari M. Int J

Hydrogen Energy 2013;38:6593e9.[50] Gong J, Li Y, Hu Z, Zhou Z, Deng Y. J Phys Chem C 2010;114:9970e4.[51] Wang J, Deng B, Chen H, Wang X, Zheng J. Environ Sci Technol 2009;43:

5223e8.[52] Gu H, Rapole S, Sharma J, Huang Y, Cao D, Colorado HA, et al. RSC Adv

2012;2:11007e18.[53] Gu H, Guo J, Wei S, Guo Z. J Appl Polym Sci 2013;130:2238e44.[54] MacDiarmid AG, Manohar SK, Masters JG, Sun Y, Weiss H, Epstein AJ. Synth

Met 1991;41:621e6.[55] Ogoshi T, Hasegawa Y, Aoki T, Ishimori Y, Inagi S, Yamagishi T. Macromol-

ecules 2011;44:7639e44.[56] Mortimer RJ. Chem Soc Rev 1997;26:147e56.[57] Bhadra S, Khastgir D, Singha NK, Lee JH. Prog Polym Sci 2009;34:783e810.[58] Al-Ahmed A, Bahaidarah HM, Mazumder MAJ. Adv Mater Res 2013;810:

173e216.[59] Sairam M, Nataraj S, Aminabhavi TM, Roy S, Madhusoodana C. Sep Purif Rev

2006;35:249e83.[60] Nath C, Kumar A, Syu K-Z, Kuo Y-K. Appl Phys Lett 2013;103:121905.[61] Mazerolles L, Folch S, Colomban P. Macromolecules 1999;32:8504e8.[62] Stafstr€om S, Bredas J, Epstein A, Woo H, Tanner D, Huang W, et al. Phys Rev

Lett 1987;59:1464e7.[63] Blakesley JC, Neher D. Phys Rev B 2011;84:075210.[64] Schein LB, Tyutnev A. J Phys Chem C 2008;112:7295e308.[65] Soukoulis C, Economou E. Wave Random Media 1999;9:255e69.[66] Schwinger J, Englert BG. Quantum mechanics: symbolism of atomic mea-

surements. Springer; 2001.[67] Prasanna GD, Jayanna HS, Prasad V. J Appl Poly Sci 2011;120:2856e62.[68] Reghu M, Cao Y, Moses D, Heeger AJ. Phys Rev B 1993;47:1758e64.[69] Zabrodskii A, Zinov’eva K. Zh Eksp Teor Fiz 1984;86:727e42.[70] Liu H, Pourret A, Guyot-Sionnest P. ACS nano 2010;4:5211e6.[71] Aggarwal M, Khan S, Husain M, Ming T, Tsai M, Perng T, et al. Eur Phys J B

2007;60:319e24.[72] Shklovskii BI. Zh Eksp Teor Fiz 1982;36:43 [JETP Lett 1982;36:287].[73] Mott NF, Davis EA. Electronic processes in non-crystalline materials. Oxford:

OUP; 2012.[74] Chen G, Zeng T, Borca-Tasciuc T, Song D. Mater Sci Eng A 2000;292:155e61.[75] Zhang G, Chen H, Xie Y, Huang F. CrystEngComm 2014;16:1810e6.[76] Zhang L, Tang Z-J. Phys Rev B 2004;70:174306.[77] Lordi V, Erhart P, Åberg D. Phys Rev B 2010;81:235204.[78] Rosenbaum R, Murphy T, Palm E, Hannahs S, Brandt B. Phys Rev B 2001;63:

094426.

H. Gu et al. / Polymer 55 (2014) 4405e44194418

[79] Gu H, Guo J, Zhang X, He Q, Huang Y, Colorado HA, et al. J Phys Chem C2013;117:6426e36.

[80] Gupta K, Jana PC, Meikap AK. Solid State Sci 2012;14:324e9.[81] Sarkar A, Ghosh P, Meikap AK, Chattopadhyay SK, Chatterjee SK, Ghosh M.

Solid State Commun 2007;143:358e63.[82] Efros A, Shklovskii B. J Phys C 1975;8:L49e51.[83] Yu D, Wang C, Wehrenberg BL, Guyot-Sionnest P. Phys Rev Lett 2004;92:

216802.[84] Sheng Y, Nguyen TD, Veeraraghavan G, Mermer €O, Wohlgenannt M, Qiu S,

et al. Phys Rev B 2006;74:045213.[85] Prigodin VN, Bergeson JD, Lincoln DM, Epstein AJ. Synth Met 2006;156:

757e61.[86] Wagemans W, Bloom FL, Bobbert PA, Wohlgenannt M, Koopmans B. J Appl

Phys 2008;103:07F303.[87] Haken H, Wolf H. In: The physics of atoms and quanta. Berlin Heidelberg:

Springer; 1994. pp. 339e63.[88] Bobbert PA, Nguyen TD, Wagemans W, van Oost FWA, Koopmans B,

Wohlgenannt M. Synth Met 2010;160:223e9.[89] Su T-I, Wang C-R, Lin S-T, Rosenbaum R. Phys Rev B 2002;66:054438.[90] Nguyen VI, Spivak BZ, Shklovskii BI. JETP Lett 1985;62:1021.[91] Sivan U, Entin-Wohlman O, Imry Y. Phys Rev Lett 1988;60:1566e9.[92] Tokumoto H, Mansfield R, Lea MJ. Philos Mag B 1982;46:93e113.[93] Shklovskii BI. Fiz Tekh Poluprovodn 1983;17:2055 [Sov Phys Semicond

1983; 2017:1311].[94] Shklovskii BI, Efros AL. Electronic properties of doped semiconductors. Ber-

lin: Springer-Verlag; 1984.[95] Rosenbaum R, Milner A, Hannahs S, Murphy T, Palm E, Brandt B. Phys B

2001;294-295:340e2.[96] Aharony A, Stauffer D. Introduction to percolation theory. Taylor & Francis;

2003.[97] Schoepe W. Zeitschrift für Physik B 1988;71:455e63.[98] Mukherjee AK, Menon RJ. Phys Condens Matter 2005;17:1947e60.[99] Ghosh M, Barman A, De S, Chatterjee S. J Appl Phys 1998;84:806e11.

[100] Long Y, Chen Z, Ma Y, Zhang Z, Jin A, Gu C, et al. Appl Phys Lett 2004;84:2205e7.

[101] Long Y, Chen Z, Shen J, Zhang Z, Zhang L, Huang K, et al. Nanotechnology2006;17:5903e11.

[102] Bozdag KD, Chiou N-R, Prigodin V, Epstein A. Synth Met 2010;160:271e4.[103] Zhu J, Wei S, Zhang L, Mao Y, Ryu J, Haldolaarachige N, et al. J Mater Chem

2011;21:3952e9.[104] Yoo JE, Bucholz TL, Jung S, Loo Y-L. J Mater Chem 2008;18:3129e35.[105] Stejskal J, Sapurina I, Trchov�a M, Proke�s J, K�rivka I, Tobolkov�a E. Macro-

molecules 1998;31:2218e22.[106] Long Y, Huang K, Yuan J, Han D, Niu L, Chen Z, et al. Appl Phys Lett 2006;88:

162113.[107] Long Y, Chen Z, Zhang X, Zhang J, Liu Z. Appl Phys Lett 2004;85:1796e8.[108] Gupta K, Jana PC, Meikap AK, Nath TK. J Appl Phys 2010;107:073704.[109] Rakhimov RR, Amponsah OK, Waller LA, Shevchenko VG, Karmilov AY,

Alexandrov IA, et al. J Appl Phys 2009;105:07E302.[110] Millis AJ, Shraiman BI, Mueller R. Phys Rev Lett 1996;77:175e8.[111] Ramirez AJ. Phys Condens Matter 1999;9:8171e99.[112] Zhu J, Gu H, Luo Z, Haldolaarachige N, Young DP, Wei S, et al. Langmuir

2012;28:10246e55.[113] Zhang X, Wei S, Haldolaarachchige N, Colorado HA, Luo Z, Young DP, et al.

J Phys Chem C 2012;116:15731e40.[114] Gu H, Guo J, He Q, Jiang Y, Huang Y, Haldolaarachige N, et al. Nanoscale

2014;6:181e9.[115] Gu H, Guo J, We H, Huang Y, Zhao C, Li Y, et al. Phys Chem Chem Phys; 2013:

10866e75.[116] de la Cruz GG, Gurevich YG, Kucherenko V, R. de Arellano E. Europhys Lett

2001;53:539e43.[117] Gu H, Guo J, Sadu R, Huang Y, Haldolaarachchige N, Chen D, et al. Appl Phys

Lett 2013;102:212403.[118] Gu H, Guo J, Wei H, Zhang X, Zhu J, Shao L, et al. Polymer 2014;55:944e50.[119] Lin AL, Wu T, Chen W, Wee AT. Appl Phys Lett 2013;103:032408.[120] Bloom FL, Kemerink M, Wagemans W, Koopmans B. Phys Rev Lett 2009;103:

066601.

Dr. Hongbo Gu, received her PhD degree at Harbin Insti-tute of Technology (HIT) in Jan. 2014, China. During hergraduate study, she worked as a joint Chemical Engineer-ing PhD student with Prof. Zhanhu Guo at Lamar Univer-sity sponsored by China Scholarship Council (CSC). Herresearch interests focus on the giant magnetoresistance(GMR) sensors, multifunctional polymer nanocompositesespecially the magnetic and conductive materials for theenvironmental remediation and electronic devices.

Mr. Jiang Guo, currently a PhD candidate in Dan F. SmithDepartment of Chemical Engineering at Lamar University,obtained a Bachelor degree from School of Chemicaland Biological Engineering, Taiyuan University of Scienceand Technology (2012). His research interests includeconductive polymers and multifunctional polymernanocomposites.

Miss Xingru Yan, currently a doctoral student in the In-tegrated Composites Laboratory (ICL), Dan F. SmithDepartment of Chemical Engineering at Lamar Universitysupervised by Prof. Zhanhu Guo, obtained a BE degree inChemical Engineering and Technology from the YangzhouUniversity (2011). Her current research focuses on newsynthetic processes to multifunctional nanocompositesespecially for the magnetic and conductive materials withadvanced applications.

Miss Huige Wei, currently a PhD student in Dan F. SmithDepartment of Chemical Engineering at Lamar University,obtained both her MS (2011) and BE (2009) degrees inChemical Engineering and Technology from Harbin Insti-tute of Technology, Harbin, China. Her research interestsmainly focus on polymer- and ceramic- based nano-composites for electrochromic, energy storage, and anti-corrosion applications.

Miss Xi Zhang, currently a PhD student in Dan F. SmithDepartment of Chemical Engineering at Lamar University,obtained a master's degree of Mechanical and AutomotiveEngineering (2010) and a BS degree in Polymer Scienceand Technology from South China University of Technol-ogy (2006). Her research interests include polymer nano-composites, especially the conductive nanocomposites.

Prof. Jiurong Liu received his PhD degree from Depart-ment of Applied Chemistry, Osaka University, Osaka, Japan,in 2004. After postdoctoral training from Department ofChemical and Biomolecular Engineering, University ofCalifornia at Los Angeles (UCLA), USA, in 2008, he joinedthe School of Materials Science and Engineering, ShandongUniversity, China. His current research focused on thecombining synthesis and fabrication methods with archi-tecture and function of novel materials to meet specificapplication needs.

H. Gu et al. / Polymer 55 (2014) 4405e4419 4419

Prof. Yudong Huang, currently a Professor and Dean ofSchool of Chemical Engineering and Technology at HarbinInstitute of Technology in China, obtained a CompositesMaterial PhD degree from Harbin Institute of Technology(1993), a BS in Polymer Material Engineering from HarbinInstitute of Technology (1988). His research focuses on thefibers reinforced polymer composites and surface treatmentof fibers, fabrication of high performance of organic fibers.

Dr. Suying Wei, currently an Assistant Professor in theDepartment of Chemistry and Biochemistry at LamarUniversity, obtained a PhD degree in chemistry from theLouisiana State University (2006), an MS degree in appliedchemistry from the Beijing University of Chemical Tech-nology (2000) and a BS degree in chemical engineeringfrom the Shandong University of Science and Technology(1996). Her research interests are in analytical and envi-ronmental chemistry involved especially in polymer andcarbon based nanocomposite systems.

Dr. Zhanhu Guo, currently an Associate Professor ofChemical Engineering at Lamar University, obtained aChemical Engineering PhD degree from Louisiana StateUniversity (2005) and received three-year (2005e2008)postdoctoral training in Mechanical and Aerospace Engi-neering Department in University of California Los Angeles.Dr. Guo directs the Integrated Composites Laboratory withmore than 20 members and chaired the Composite Divi-sion of American Institute of Chemical Engineers (AIChE,2010e2011). His current research focuses on fundamentalscience behind multifunctional nanocomposites for energyharvesting, sustainable electronic devices, environmentalremediation, anti-corrosion, fire-retardancy, and electro-magnetic radiation shielding/absorption applications.