tuning the magnitude and the polarity of the …tuning the magnitude and the polarity of the...

Tuning the Magnitude and the Polarity of the PiezoresistiveResponse of Polyaniline through Structural ControlMelda Sezen-Edmonds,† Petr P. Khlyabich,† and Yueh-Lin Loo*,†,‡

†Department of Chemical and Biological Engineering and ‡Andlinger Center for Energy and the Environment, Princeton University,Princeton, New Jersey 08544, United States

*S Supporting Information

ABSTRACT: We demonstrate the tunability of both the polarityand the magnitude of the piezoresistive response of polyaniline thatis template-synthesized on poly(2-acrylamido-2-methyl-1-propane-sulfonic acid), PANI−PAAMPSA, by altering the templatemolecular weight. Piezoresistivity is quantified by gauge factor, aunitless parameter that relates changes in electrical resistance toapplied strain. The gauge factor of PANI−PAAMPSA decreaseslinearly and becomes negative with decreasing PAAMPSAmolecular weight. The polarity of PANI−PAAMPSA’s gauge factoris determined by macroscopic connectivity across thin films. PANI−PAAMPSA thin films comprise electrostatically stabilizedparticles whose size is determined at the onset of synthesis. An increase in the interparticle spacing with applied strain results in apositive gauge factor. The presence of PANI crystallites increases connectivity between particles; these samples instead exhibit anegative gauge factor whereby the resistance decreases with increasing strain. The tunability of the piezoresistive response of theseconducting polymers allows their utilization in a broad range of flexible electronics applications, including thermo- andchemoresistive sensors and strain gauges.

KEYWORDS: flexible electronics, conducting polymers, piezoresistivity, polyaniline, gauge factor, strain gauge, resistive sensors

■ INTRODUCTION

Conducting polymers are widely used in next-generationelectronic devices because they possess the chemical,mechanical, and processing attributes of conventional polymers,yet have electrical conductivities approaching those of metals.1

Further, the mechanical compliance of conducting polymerswith flexible substrates makes them excellent candidates for usein conformal electronic devices, either as electrodes2−4 or assensing units.5,6 Like other conducting materials, however,conducting polymers are inherently piezoresistive; changes intheir electrical resistances due to mechanical deformationsoften limit their use as chemo- or thermoresistive devices.7 Thispiezoresistive effect can either arise from changes in theresistance of the material due to changes in the samplegeometry with strain or be an intrinsic materials propertystemming from changes in resistivity of the material due to thestructural changes with strain.8 Whereas the changes in thesample geometry under tensile strain always cause theresistance of conducting materials with positive Poisson’sratio to increase, changes in the material’s structure can causean increase (e.g., due to the formation of cracks) or decrease(e.g., due to strain-induced chain alignment) in the resistancewith strain.9 Understanding the intrinsic factors that govern thepiezoresistivity of conducting polymers is therefore important,so materials design can be tailored according to specificapplication needs.The relationships between the structure of conducting

polymers and their piezoresistive properties have mostly been

investigated under large strains that inadvertently result inplastic deformation. Such plastic deformation can result eitherin strain-induced morphological changes that can in turnenhance the polymer’s electrical conductivity or in cracks andfractures that are detrimental to the polymer’s electricalconductivity.9−11 The solid-state structure of poly(3,4-ethyl-enedioxythiophene):polystyrenesulfonate, PEDOT:PSS, (Cle-vios pH1000) thin films, for example, is shown to drasticallyimpact how PEDOT:PSS thin films respond to mechanicalstrain. Applying a 60% tensile strain to pristine PEDOT:PSSthin films causes PEDOT-rich domains to coalesce and formconnected pathways for charge transport that results in anincrease in the conductivity. Since the films are permanentlydeformed, the improvement in electrical conductivity isretained even when the strain is removed. On the otherhand, the addition of 5 wt % dimethyl sulfoxide, DMSO, toPEDOT:PSS dispersion before deposition eliminates anyconductivity improvement with subsequent strain. DMSO isthought to uniformly disperse PEDOT chains, so stretchingdoes not further alter domain connectivity.9

In order to utilize conducting polymers in resistive sensingapplications, where reversible changes in the resistance arecorrelated with changes in the environment, their piezoresistiveresponse under elastic deformation must be understood. Strain-

Received: January 3, 2017Accepted: March 20, 2017Published: March 20, 2017

Research Article

www.acsami.org

© 2017 American Chemical Society 12766 DOI: 10.1021/acsami.7b00101ACS Appl. Mater. Interfaces 2017, 9, 12766−12772

induced changes in the resistance of conducting polymers, forexample, can be utilized to detect mechanical deformation ofbuildings or bridges, or for monitoring human motion.12,13 Instrain-sensing applications, the mechanical deformation isquantified through a unitless parameter called the gauge factor,GF, which is defined as

ε=

ΔR RGF

/ 0(1)

where Ro is the resistance at zero strain; ΔR is the change in theresistance; and ε is the strain.14 Maximizing GF optimizes forsensitivity in the strain direction. Conversely, if the conductingpolymers are used for flexible thermoresistive or chemoresistivesensing applications, small changes in resistance in response tomechanical deformationseven reversiblewill cause error insensor output.15,16 Minimizing the GF of conducting polymersis thus important to minimize mechanical deformation-relatedresistance drifts during such sensor use.Polyaniline, PANI, is a conducting polymer that has shown

promise in flexible or conformable electronics applica-tions.5,17−19 Template polymerization of aniline along thepolymeric acid poly(2-acrylamido-2-methyl-1-propane-sulfonicacid), PAAMPSA, results in the formation of a water-dispersibleand electrically conducting polymer complex PANI−PAAMP-SA with its chemical structure shown in Scheme 1.20 With its

biocompatibility and stability at ambient conditions, PANI−PAAMPSA is especially suitable for wearable electronics andhuman health monitoring applications.19,21 As-synthesizedPANI−PAAMPSA forms electrostatically stabilized colloidalparticles, and this particulate nature is retained when PANI−PAAMPSA is cast as thin films from its water dispersions.Previous studies in our group showed that decreasing themolecular weight of PAAMPSA results in the formation ofsmaller PANI−PAAMPSA particles, and increases PANI’scrystallinity concurrently. These structural changes haveresulted in an increase in the conductivity of PANI−PAAMPSAfrom 0.4 S/cm to 1.1 S/cm when a template of 724 kg mol−1

PAAMPSA is replaced with 45 kg mol−1 PAAMPSA.20 In thiswork, we show the tunability of PANI−PAAMPSA’s piezor-esistive response by changing the molecular weight ofPAAMPSA at the onset of synthesis. As we lower the molecularweight of PAAMPSA, the GF of PANI−PAAMPSA decreasesprogressively and even becomes negative for PANI−PAAMPSAthat is synthesized with the two lowest molecular-weightPAAMPSAs (45 and 106 kg mol−1). Although negative andpositive GFs (GF of −4.9 and 22) have been reported for acommercial grade of PANI (PANIPOL T), the source of thisdifference in the polarity of GF was not fully understood.22,23

With PANI−PAAMPSA as our model system, we show howthe structure of PANI affects the polarity and the magnitude oftheir piezoresistive properties. We conclude that the change inthe GF polarity with decreasing PAAMPSA molecular weight isa result of increased crystallinity of PANI that is likely toconnect the individual PANI−PAAMPSA particles, which canthen undergo strain-induced alignment under tensile deforma-tion. Control of the particle size and the extent of crystallizationby altering the molecular weight of PAAMPSA has enabled thetuning of both the magnitude and the polarity of PANI−PAAMPSA’s GF. Such tunability allows us to address the needof high GF for strain sensing and near-zero GF for thermo- orchemoresistive sensing applications with a single class ofmaterial.

Scheme 1. Chemical Structure of Electrically ConductingPANI−PAAMPSA

Figure 1. Relative change in the resistance (black) of as-cast (a) PANI−PAAMPSA-724, (b) PANI−PAAMPSA-255, (c) PANI−PAAMPSA-106,and (d) PANI−PAAMPSA-45 serpentine patterns under cyclic strain (blue) along their strain-sensitive direction.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b00101ACS Appl. Mater. Interfaces 2017, 9, 12766−12772

12767

■ RESULTS AND DISCUSSION

In order to test the piezoresistive response of PANI−PAAMPSA-X, where X represents the molecular weight ofPAAMPSA in kg mol−1, we prepared serpentine patterns ofPANI−PAAMPSA-X on flexible polyimide substrates (FigureS1). A serpentine pattern is used in order to maximizesensitivity in the direction of the applied strain and minimizetransverse strain during our uniaxial tensile measurements.Figure 1 shows the change in resistance of PANI−PAAMPSA-Xpatterns with respect to their zero-strain resistance, Ro (blackline), under uniaxial cyclic strain (blue line). The films werestretched in their strain-sensitive direction. The recovery of Roafter the strain is released indicates that these tests wereperformed in the elastic deformation region of the conductingpolymer patterns. Figure 1a and b shows that the resistances ofPANI−PAAMPSA-724 and PANI−PAAMPSA-255 patternsare in-phase with applied strain. An opposite trend is observedfor PANI−PAAMPSA-106 and PANI−PAAMPSA-45 patterns(Figures 1c and d); resistances of these conducting polymerpatterns are out-of-phase with applied strain. Interestingly, weobserve hysteresis in the piezoresistive response of PANI−PAAMPSA-255 between the loading and unloading cycles.Such hysteresis is commonly observed in the piezoresistiveresponse of carbon nanotube/polymer composites due to theirthermoplastic nature.24 Hysteresis was also reported in thepiezoresistive response of nanoarchitectured TixCuy films dueto structural defects in the crystals.8 We do not believe these tobe the origin of the hysteresis we observe in PANI−PAAMPSA-255. While the exact source of the hysteresisobserved in the piezoresistive response of PANI−PAAMPSA-255 patterns is still under investigation, we have previouslyshown that such hysteresis can be eliminated throughpostdeposition dichloroacetic acid, DCA, treatment.19 Similarly,we are able to eliminate this nonlinearity in the piezoresistiveresponse of PANI−PAAMPSA-255 through structural mod-ification with DCA treatment (Figure S2).We quantified the piezoresistive response of PANI−

PAAMPSA-X patterns by calculating their GFs from theslope of the normalized resistance change with strain (FigureS3) using eq 1. Figure 2a shows the change in GF of PANI−PAAMPSA-X (black data points) depending on the molecularweight of the PAAMPSA template used in the synthesis. TheGF of PANI−PAAMPSA-X decreases and becomes negativewith decreasing PAAMPSA molecular weight, changing from3.7 ± 0.28 to −0.60 ± 0.15 as the PAAMPSA molecular weightis decreased from 724 to 45 kg mol−1. The tunability of the GFenables us to select the appropriate grade of PANI−PAAMPSA-X per application needs. For example, PANI−PAAMPSA-724 should be used among the different PANIgrades for strain sensing. At a GF of 3.7, PANI−PAAMPSA-724 strain gauges are almost twice as sensitive to strain ascommercially available metal-foil strain gauges (GF = 2).25 Inexamples of chemoresistive or thermoresistive sensors in whichany electrical response to mechanical strain should besuppressed to maximize signal-to-noise output, however,PANI−PAAMPSA-106 should instead be used given that ithas the smallest GF among the different PANI grades tested(GF = −0.34). The linear relationship between GF and themolecular weight of PAAMPSA also provides an opportunity tospecify a priori new PANI−PAAMPSA with tailored piezor-esistive responses and synthesize according to these specifica-tions. Our empirical correlation suggests that PANI synthesized

with PAAMPSA having a molecular weight of 140 kg mol−1

should yield conducting patterns having a GF of zero. At GF =0, we can completely suppress errors associated with resistancechange due to mechanical deformation in flexible chemo- orthermoresistive sensors.Unique to PANI−PAAMPSA-X is our easy access to both

negative and positive GFs of different magnitudes. Instead ofsynthesizing a new grade of PANI−PAAMPSA, existing PANI−PAAMPSA-X, having different GF polarities with similarmagnitudes (e.g., PANI−PAAMPSA-45 and PANI−PAAMP-SA-255), can thus be combined as an alternative approach toaccessing near-zero net GFs.19 This approach for creating net-zero GF sensors is much more straightforward compared toprior reports of placing the active component of the sensor atthe sensor’s neutral bending plane or lithographically patterningthe active component in mesh-like structures to reducestrain.16,26,27 These near-zero GF conducting patterns canalso be combined with high-GF PANI−PAAMPSA strain

Figure 2. (a) GF of as-cast (black) and DCA-treated (blue) PANI−PAAMPSA-X serpentine patterns as a function of PAAMPSAmolecular weight. Dashed lines represent linear fits to the data. (b)Azimuthally integrated XRD traces of as-cast PANI−PAAMPSA-X andDCA-treated PANI−PAAMPSA-724 extracted from their 2D-GIXDpatterns. Labels indicate the PAAMPSA template molecular weightused in the synthesis. (c) Fractional crystallinity of as-cast PANI−PAAMPSA-X thin films (black data points) and hydrodynamicdiameter of the as-synthesized PANI−PAAMPSA particles (bluedata points) as a function of template molecular weight. Fractionalcrystallinity is obtained from the azimuthally integrated XRD tracegiven in (b), by calculating the ratios of the crystalline area to totalarea. Hydrodynamic diameter values are adapted from ref 20. Dashedlines connecting the data points are included as a guide to the eye.



12768

sensors in order to account for the non-strain-related drifts inthe electrical resistance of PANI−PAAMPSA. Additionally,PANI−PAAMPSA-X grades having similar magnitudes of GFbut opposite polarities can also be incorporated in Wheatstonebridge circuits to build strain gauges that can quantitativelydecouple thermal, shrinkage, and mechanical strains.28

We previously observed an anticorrelation between theelectrical conductivity of PANI−PAAMPSA-X films and themolecular weight of the PAAMPSA template.29 The improve-ments in the electrical conductivity of PANI−PAAMPSA-Xfilms with decreasing PAAMPSA molecular weight arecorrelated with the improved PANI crystallinity and theparticle connectivity of the resulting films.20,29 The dependenceof PANI−PAAMPSA-X films’ GF on the molecular weight ofPAAMPSA is reminiscent of how the electrical conductivity ofPANI−PAAMPSA-X is correlated with PAAMPSA molecularweight. We thus surmised this trend in the piezoresistiveresponse of PANI−PAAMPSA-X films to also have structuralorigins.A positive GF is known to be a result of increasing separation

between conducting domains under tensile deformation.30,31 Infact, we previously observed a positive GF in PANI−PAAMPSA-724 films and attributed this positive GF toincreased separation between PANI−PAAMPSA-724 particlesupon stretching.19 As synthesized, PANI−PAAMPSA-X formselectrostatically stabilized colloidal particles, and this particulatenature is retained in thin films of PANI−PAAMPSA-X that arecast from its water dispersions.20,29 Figure 2c (blue data points)shows that the hydrodynamic diameters of the as-synthesizedPANI−PAAMPSA-X particles decrease with decreasingPAAMPSA molecular weight. Figure 3a quantifies thecorrelation between GF and the hydrodynamic diameter ofparticles of as-synthesized PANI−PAAMPSA-X. The GF ispositive for the PANI−PAAMPSA-X grades comprising thelargest particles. As the hydrodynamic diameter of as-

synthesized PANI−PAAMPSA-X decreases, the GF becomesprogressively negative. The observation that the GF decreaseswith particle size for PANI−PAAMPSA-724 and PANI−PAAMPSA-255 is consistent with previous reports that theGF of films composed of gold nanoparticles decreases from 250to 50 with decreasing nanoparticle diameter.31 This trend isattributed to smaller particles being more interconnectedcompared to larger particles, and therefore films comprisingsmaller particles are less sensitive to strain-induced interparticleseparation.31,32 The discontinuity in the data correlating the GFto the hydrodynamic diameter of PANI−PAAMPSA-X and theswitch in the GF polarity suggest another mechanism to alsocontribute to PANI−PAAMPSA-X films’ piezoresistive re-sponse.We also tested PANI that is doped with polystyrenesulfonate,

PSS. Dynamic light scattering experiments revealed thatPANI:PSS forms particles with Dh = 205 ± 14 nm (FigureS5). When cast from its aqueous dispersions, PANI:PSS formsfilms that are topographically similar to PANI−PAAMPSA-45films.20,33 If particle size is the only determining factor in thepiezoresistive response of PANI films, we would expectPANI:PSS, like PANI−PAAMPSA-45, to exhibit a negativeGF per Figure 3a data. Figure 4 shows the relative change in the

resistance of a PANI:PSS serpentine pattern under cyclic strainin its strain-sensitive direction. PANI:PSS patterns instead havea resistance that is in-phase with applied strain, yielding apositive GF of 3.0 ± 0.1 (Figure S6). The observation thatPANI:PSS exhibits a positive GF despite its small particle size isfurther evidence that factors other than the particle size are atplay in determining the piezoresistive response of PANI films.Interestingly, powder X-ray diffraction, XRD, studies show thatPANI:PSS, unlike PANI−PAAMPSA-45 but similar to PANI−PAAMPSA-724, is mostly amorphous (Figure S7). Wetherefore hypothesize that PANI crystallinity could also playa role in determining the piezoresistive response of PANI films.In order to examine the differences in the crystallinity of

different PANI−PAAMPSA-X grades, we performed grazing-incidence X-ray diffraction, GIXD, on PANI−PAAMPSA-Xthin films. Figure 2b shows the azimuthally integrated XRDtraces of as-cast PANI−PAAMPSA-X films. All traces show abroad halo centered around q = 1.5 Å−1 that stems fromamorphous PAAMPSA and a reflection at q = 1.84 Å−1 thatstems from PANI subchain alignment.29 Additional reflectionsassociated with PANI appear in PANI−PAAMPSA-X that weresynthesized with lower molecular weight PAAMPSA. Wequantified the fractional crystallinity of PANI−PAAMPSA-Xfilms through the ratio of the area under the reflectionsassociated with crystalline PANI to the total area under the

Figure 3. Change in the GF of PANI−PAAMPSA-X serpentinepatterns with (a) hydrodynamic diameter of the as-synthesized PANI−PAAMPSA particles and (b) fractional crystallinity of PANI−PAAMPSA-X films (red data point represents DCA-treated PANI−PAAMPSA-724). Linear fits to the data (dashed lines) are includedshowing the different slopes in the negative and the positive gaugefactor regions.

Figure 4. Relative change in the resistance (black) of as-castPANI:PSS serpentine pattern under cyclic strain (blue) along itsstrain-sensitive direction.



12769

XRD trace. Figure 2c (black data points) shows that PANIbecomes more crystalline as the template PAAMPSA molecularweight decreases.Figure 3b shows the dependence of GF on the fractional

crystallinity of PANI−PAAMPSA-X thin films. A negative GF isobserved in the more crystalline PANI−PAAMPSA-X grades(PANI−PAAMPSA-45 and PANI−PAAMPSA-106). Not un-like the data shown in Figure 3a, we observe a discontinuity inthe data that correlates the GF of PANI−PAAMPSA-X films tothe fractional crystallinity; the GF switches to positive as PANIcrystallinity decreases. We believe the negative GF in thesehigher crystallinity samples is a result of further alignment ofcrystalline PANI upon stretching that in turn promotesconduction in the direction of applied strain. Such strain-induced chain alignment has previously been observed in PANIthat is doped with volatile acids under large strains (>400%strain), and plastic deformation of the polymer was shown toresult in a decrease in the resistance of PANI.34 Here, weobserve a similar decrease in resistance with stretching, even inthe elastic deformation regime of the polymer.Taken together, our data suggest that PANI−PAAMPSA’s

piezoresistive response is dominated by PANI’s crystallinitywhen the samples are sufficiently crystalline. In less crystallinesamples, PANI−PAAMPSA’s piezoresistive response is domi-nated by the interparticle separation, which is in turn related tothe particle size. However, having crystalline PANI and theability for PANI crystals to align under strain alone cannot besufficient to obtain negative GFs. Supporting this circum-spection is the fact that the GF of gold nanoparticles, which arecrystalline, can be entirely predicted based on its particlediameter alone.31 This observation indicates that the macro-scopic piezoresistive response of gold nanoparticles isindependent of the crystallinity within individual particles.We are thus left to surmise that the PANI crystals in the morecrystalline PANI−PAAMPSA-X films must bridge neighboringparticles in order for preferential alignment of PANI understrain to influence its piezoresistive response. This hypothesis isconsistent with our previous report that the conducting PANI ispreferentially localized on the exterior of the PANI−PAAMPSAparticles.20 It should thus not be surprising that crystallinePANI enhances the particle connectivity in PANI−PAAMPSAthin films.Consistent with the notion that negative GF is accessed

when PANI is sufficiently crystalline and when the conductingdomains are connected, solvent annealing of PANI−PAAMP-SA-X in dichloroacetic acid, DCA, results in patterns havinguniformly negative GF (blue data points in Figures 2a) that isindependent of the molecular weight of PAAMPSA used in thesynthesis. DCA treatment eliminates the particulate nature andinduces further crystallization in PANI−PAAMPSA films.19,33

This structural change is confirmed by the data in Figures 2b(black line) and Figure 3b (red data point). The simultaneouselimination of the particulate nature of PANI−PAAMPSA andincrease in PANI crystallinity with DCA treatment enablepreferential alignment of PANI during strain to dominate theirpiezoresistive response. Postdeposition DCA annealing has thusprovided access to negative GFs, even in PANI−PAAMPSA-724 and PANI−PAAMPSA-255, that previously exhibitedpositive GFs in their as-cast states.

■ CONCLUSIONSWe demonstrated systematic tunability of both the polarity andthe magnitude of the piezoresistive response of PANI−

PAAMPSA-X by controlling its thin-film morphology. Byaltering a single parameter, the molecular weight of PAAMPSA,at the onset of synthesis, we change the particle size of theresulting polymer complex and the crystallinity of PANI. Theinterplay between these parameters coupled with the PANIcrystals’ ability to bridge the individual PANI−PAAMPSA-Xparticles determine the piezoresistive response of PANI−PAAMPSA-X films. The tunability of the piezoresistiveresponse of these water-dispersible, electrically conductingpolymers, especially accessing negative GFs, allows us to utilizethem in a broad range of flexible and conformable electronicsapplications. Optimizing for the GF enables utility of PANI−PAAMPSA as sensitive strain sensors, and minimizingeveneliminatingGF enables the use of PANI−PAAMPSA asflexible resistive-temperature or chemical sensors that workaccurately under mechanical deformations.

■ EXPERIMENTAL SECTIONPolymer Synthesis. In order to synthesize PANI−PAAMPSA-X

(where X represents the poly(ethylene oxide) equivalent number-average molecular weight of PAAMPSA used in the synthesis), weused aniline (Sigma-Aldrich, 99.5%), ammonium peroxydisulfate(98.9%, Fisher Scientific), and PAAMPSA (Scientific PolymerProducts, 10.36 wt % in water, reported MW 800 kg mol−1) aspurchased. We determined the poly(ethylene oxide), PEO, equivalentmolecular weight of the as-purchased PAAMPSA to be 724 kg/mol.Lower molecular weight PAAMPSA (PAAMPSA-45, PAAMPSA-106,and PAAMPSA-255) was synthesized by conventional free-radicalpolymerization of 2-acrylamido-2-methyl-1-propanesulfonic acid,AAMPSA (Aldrich. 99%), according to previously publishedprocedures.29 PANI−PAAMPSA-X was synthesized by templatepolymerization of aniline along PAAMPSA-X at 1:1 monomer-to-acid molar ratio in deionized water as described by Yoo et al.29

PANI:PSS was synthesized by template polymerization of aniline alongPSS according to published procedures.33 5 wt % aqueous dispersionsof PANI−PAAMPSA-X and PANI:PSS were prepared and stirred for 2weeks prior to deposition.

Tensile Tests. Serpentine patterns of PANI−PAAMPSA-X orPANI:PSS on polyimide substrate were prepared as described in ourprevious work.19 Dichloroacetic acid, DCA (Sigma-Aldrich, 99%),treatment was performed by vigorously shaking the samples at 100 °CDCA for 3 min. Samples were annealed at 170 °C for 30 min and keptunder vacuum for at least 3 h to remove any residual DCA.33 Cyclictensile tests were performed by stretching the samples uniaxially in thestrain-sensitive direction of the serpentine pattern using an Instronelectromechanical universal testing machine (model 5969) followingpreviously published procedures.19 We performed cyclic tensile tests inthe elastic deformation region of the conducting polymers. Theresistances of the samples were recorded during the tensile tests usinga four-probe geometry with Agilent 4145B Semiconductor ParameterAnalyzer. At least three different films were tested for each sample, andeach film was exposed to at least 10 strain cycles. Gauge factors weredetermined from linear fits to the normalized change in the resistanceof the serpentine patterns with applied strain.

Structural Characterization. GIXD experiments were conductedat the G1 station (9.82 ± 0.05 keV) of the Cornell High EnergySynchrotron Source. The beam was chosen to be 0.05 mm tall and 1mm wide. PANI−PAAMPSA-X aqueous dispersions were spin-coatedon precleaned glass substrate at 1000 rpm for 60 s, and DCAtreatment was performed as described above. The width of eachsample was 5 mm. The X-ray beam was aligned above the conductingpolymer film’s critical angle but below that of the substrate, at a 0.17°incident angle with the substrate. X-ray scattering was collected with a2D CCD detector, positioned 113.9 mm from the sample. All GIXDimages have been background subtracted. The fractional crystallinity ofPANI−PAAMPSA-X films was calculated from their 2D-GIXDpatterns by dividing the crystalline area by the total area under theazimuthally integrated traces. In order to calculate the total crystalline



12770

area, the area under the broad amorphous halo was subtracted fromthe total area. Powder XRD data were collected using a RigakuMiniflex X-ray diffractometer (with Cu Kα radiation at 30 kV and 15mA) at a scan rate of 0.6° min−1 by loading 100 mg of PANI−PAAMPSA-45, PANI−PAAMPSA-724, and PANI:PSS. Dynamic lightscattering, DLS, experiments were performed using a ZetaSizerMalvern DLS instrument (Malvern Instruments, Malvern, UK) witha 633 nm laser and backscatter detection angle of 173° on 0.001 wt %of PANI:PSS in 0.1 M NaCl aqueous solutions at 25 °C as describedby Yoo et al.20 0.1 M NaCl was added in order to screen interparticleinteractions. Hydrodynamic diameters of as-synthesized PANI−PAAMPSA-X are obtained from previous publications.20

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.7b00101.

Additional figures as described in main text (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: [email protected].

ORCIDYueh-Lin Loo: 0000-0002-4284-0847Author ContributionsThe manuscript was written through contributions of allauthors. All authors have given approval to the final version ofthe manuscript.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSA portion of this work was conducted at the Cornell HighEnergy Synchrotron Source (CHESS), which is supported bythe National Science Foundation and the National Institutes ofHealth/National Institute of General Medical Sciences underNSF award DMR-1332208. The authors acknowledge Prof.Craig B. Arnold (Princeton University) for access to theInstron, Brian K. Wilson and Prof. Robert K. Prud’homme(Princeton University) for access to Dynamic Light Scatteringinstrument, and Dr. Joung-Eun Yoo for help in synthesizing thelow-molecular weight PANI−PAAMPSA-X grades. This workwas partially supported by the Anonymous Fund, sponsored bythe School of Engineering and Applied Science at PrincetonUniversity and by Princeton Center for Complex Materials thatwas funded by NSF-MRSEC under NSF award DMR-1420541.

■ REFERENCES(1) Epstein, A. J. Electrically Conducting Polymers: Science andTechnology. MRS Bull. 1997, 22, 16−23.(2) Tait, J. G.; Worfolk, B. J.; Maloney, S. A.; Hauger, T. C.; Elias, A.L.; Buriak, J. M.; Harris, K. D. Spray-Coated High-ConductivityPEDOT:PSS Transparent Electrodes for Stretchable and Mechan-ically-Robust Organic Solar Cells. Sol. Energy Mater. Sol. Cells 2013,110, 98−106.(3) Khodagholy, D.; Doublet, T.; Gurfinkel, M.; Quilichini, P.;Ismailova, E.; Leleux, P.; Herve, T.; Sanaur, S.; Bernard, C.; Malliaras,G. G. Highly Conformable Conducting Polymer Electrodes for in VivoRecordings. Adv. Mater. 2011, 23, 268−272.(4) Takamatsu, S.; Lonjaret, T.; Ismailova, E.; Masuda, A.; Itoh, T.;Malliaras, G. G. Wearable Keyboard Using Conducting PolymerElectrodes on Textiles. Adv. Mater. 2016, 28, 4485−4488.

(5) Jang, J.; Ha, J.; Cho, J. Fabrication of Water-DispersiblePolyaniline-Poly(4-Styrenesulfonate) Nanoparticles for Inkjet-PrintedChemical-Sensor Applications. Adv. Mater. 2007, 19, 1772−1775.(6) Freund, M. S.; Lewis, N. S. A Chemically Diverse ConductingPolymer-Based “Electronic Nose”. Proc. Natl. Acad. Sci. U. S. A. 1995,92, 2652−2656.(7) Vijay, V.; Rao, A. D.; Narayan, K. S. In Situ Studies of StrainDependent Transport Properties of Conducting Polymers onElastomeric Substrates. J. Appl. Phys. 2011, 109, 084525.(8) Ferreira, A.; Borges, J.; Lopes, C.; Martin, N.; Lanceros-Mendez,S.; Vaz, F. Piezoresistive Response of Nano-Architectured TixCuy ThinFilms for Sensor Applications. Sens. Actuators, A 2016, 247, 105−114.(9) Lee, Y.-Y.; Lee, J.-H.; Cho, J.-Y.; Kim, N.-R.; Nam, D.-H.; Choi,I.-S.; Nam, K. T.; Joo, Y.-C. Stretching-Induced Growth of PEDOT-Rich Cores: A New Mechanism for Strain-Dependent ResistivityChange in PEDOT:PSS Films. Adv. Funct. Mater. 2013, 23, 4020−4027.(10) Pouget, J. P.; Jozefowicz, M. E.; Epstein, A. J.; Tang, X.;MacDiarmid, A. G. X-Ray Structure of Polyaniline. Macromolecules1991, 24, 779−789.(11) Lee, Y.-Y.; Choi, G. M.; Lim, S.-M.; Cho, J.-Y.; Choi, I.-S.; Nam,K. T.; Joo, Y.-C. Growth Mechanism of Strain-DependentMorphological Change in PEDOT:PSS Films. Sci. Rep. 2016, 6, 25332.(12) Liu, N.; Fang, G.; Wan, J.; Zhou, H.; Long, H.; Zhao, X.Electrospun PEDOT:PSS−PVA Nanofiber Based Ultrahigh-StrainSensors with Controllable Electrical Conductivity. J. Mater. Chem.2011, 21, 18962−18966.(13) Mateiu, R.; Lillemose, M.; Hansen, T. S.; Boisen, A.; Geschke,O. Reliability of Poly 3,4-Ethylenedioxythiophene Strain Gauge.Microelectron. Eng. 2007, 84, 1270−1273.(14) Tung, S.-T.; Yao, Y.; Glisic, B. Sensing Sheet: The Sensitivity ofThin-Film Full-Bridge Strain Sensors for Crack Detection andCharacterization. Meas. Sci. Technol. 2014, 25, 075602.(15) Zhang, F.; Zang, Y.; Huang, D.; Di, C.; Zhu, D. Flexible andSelf-Powered Temperature−Pressure Dual-Parameter Sensors UsingMicrostructure-Frame-Supported Organic Thermoelectric Materials.Nat. Commun. 2015, 6, 8356.(16) Hattori, Y.; Falgout, L.; Lee, W.; Jung, S.-Y.; Poon, E.; Lee, J.W.; Na, I.; Geisler, A.; Sadhwani, D.; Zhang, Y.; Su, Y.; Wang, X.; Liu,Z.; Xia, J.; Cheng, H.; Webb, R. C.; Bonifas, A. P.; Won, P.; Jeong, J.-W.; Jang, K.-I.; Song, Y. M.; Nardone, B.; Nodzenski, M.; Fan, J. A.;Huang, Y.; West, D. P.; Paller, A. S.; Alam, M.; Yeo, W.-H.; Rogers, J.A. Multifunctional Skin-Like Electronics for Quantitative, ClinicalMonitoring of Cutaneous Wound Healing. Adv. Healthcare Mater.2014, 3, 1597−1607.(17) Hong, S. Y.; Lee, Y. H.; Park, H.; Jin, S. W.; Jeong, Y. R.; Yun, J.;You, I.; Zi, G.; Ha, J. S. Stretchable Active Matrix Temperature SensorArray of Polyaniline Nanofibers for Electronic Skin. Adv. Mater. 2016,28, 930−935.(18) Du, P.; Liu, H. C.; Yi, C.; Wang, K.; Gong, X. Polyaniline-Modified Oriented Graphene Hydrogel Film as the Free-StandingElectrode for Flexible Solid-State Supercapacitors. ACS Appl. Mater.Interfaces 2015, 7, 23932−23940.(19) Sezen, M.; Register, J. T.; Yao, Y.; Glisic, B.; Loo, Y.-L.Eliminating Piezoresistivity in Flexible Conducting Polymers forAccurate Temperature Sensing under Dynamic Mechanical Deforma-tions. Small 2016, 12, 2832−2838.(20) Yoo, J. E.; Krekelberg, W. P.; Sun, Y.; Tarver, J. D.; Truskett, T.M.; Loo, Y.-L. Polymer Conductivity through Particle Connectivity.Chem. Mater. 2009, 21, 1948−1954.(21) Mano, N.; Yoo, J. E.; Tarver, J.; Loo, Y.-L.; Heller, A. AnElectron-Conducting Crosslinked Polyaniline-Based Redox Hydrogel,Formed in One Step at pH 7.2, Wires Glucose Oxidase. J. Am. Chem.Soc. 2007, 129, 7006−7007.(22) Pereira, J. N.; Vieira, P.; Ferreira, A.; Paleo, A. J.; Rocha, J. G.;Lanceros-Mendez, S. Piezoresistive Effect in Spin-Coated PolyanilineThin Films. J. Polym. Res. 2012, 19, 9815.



12771

(23) Lillemose, M.; Spieser, M.; Christiansen, N. O.; Christensen, A.;Boisen, A. Intrinsically Conductive Polymer Thin Film Piezoresistors.Microelectron. Eng. 2008, 85, 969−971.(24) Costa, P.; Ribeiro, S.; Lanceros-Mendez, S. Mechanical vs.Electrical Hysteresis of Carbon Nanotube/Styrene-Butadiene-StyreneComposites and Their Influence in the Electromechanical Response.Compos. Sci. Technol. 2015, 109, 1−5.(25) Herrmann, J.; Muller, K. H.; Reda, T.; Baxter, G. R.; Raguse, B.;De Groot, G. J. J. B.; Chai, R.; Roberts, M.; Wieczorek, L. NanoparticleFilms as Sensitive Strain Gauges. Appl. Phys. Lett. 2007, 91, 183105.(26) Sekitani, T.; Iba, S.; Kato, Y.; Noguchi, Y.; Someya, T.; Sakurai,T. Ultraflexible Organic Field-Effect Transistors Embedded at aNeutral Strain Position. Appl. Phys. Lett. 2005, 87, 173502.(27) Widlund, T.; Yang, S.; Hsu, Y.-Y.; Lu, N. Stretchability andCompliance of Freestanding Serpentine-Shaped Ribbons. Int. J. SolidsStruct. 2014, 51, 4026−4037.(28) Glisic, B.; Verma, N. In Innovative Developments of AdvancedMultifunctional Nanocomposites in Civil and Structural Engineering; Loh,K. J., Nagarajaiah, S., Eds.; Woodhead Publishing Ltd. (Elsevier):2016; Chapter 12, pp 281−302.(29) Yoo, J. E.; Cross, J. L.; Bucholz, T. L.; Lee, K. S.; Espe, M. P.;Loo, Y.-L. Improving the Electrical Conductivity of Polymer Acid-Doped Polyaniline by Controlling the Template Molecular Weight. J.Mater. Chem. 2007, 17, 1268.(30) Lin, Y.; Liu, S.; Chen, S.; Wei, Y.; Dong, X.; Liu, L. A HighlyStretchable and Sensitive Strain Sensor Based on Graphene−elastomerComposites with a Novel Double-Interconnected Network. J. Mater.Chem. C 2016, 4, 4304−4311.(31) Segev-Bar, M.; Ukrainsky, B.; Konvalina, G.; Haick, H. In Situand Real-Time Inspection of Nanoparticle Average Size in FlexiblePrinted Sensors. J. Phys. Chem. C 2015, 119, 27521−27528.(32) Han, B. G.; Han, B. Z.; Yu, X. Effects of the Content Level andParticle Size of Nickel Powder on the Piezoresistivity of Cement-BasedComposites/Sensors. Smart Mater. Struct. 2010, 19, 65012.(33) Yoo, J. E.; Lee, K. S.; Garcia, A.; Tarver, J.; Gomez, E. D.;Baldwin, K.; Sun, Y.; Meng, H.; Nguyen, T.-Q.; Loo, Y.-L. DirectlyPatternable, Highly Conducting Polymers for Broad Applications inOrganic Electronics. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 5712−5717.(34) Fischer, J. E.; Tang, X.; Scherr, E. M.; Cajipe, V. B.;MacDiarmid, A. G. Polyaniline Fibers and Films: Stretch-InducedOrientation and Crystallization, Morphology, and the Nature of theAmorphous Phase. Synth. Met. 1991, 41, 661−664.



12772

tuning the magnitude and the polarity of the …tuning the magnitude and the polarity of the...

Documents