Article pubs.acs.org/JPCC
Electropolymerized Polypyrrole Nanowires for Hydrogen Gas Sensing Laith Al-Mashat,*,† Catherine Debiemme-Chouvy,*,‡ Stephan Borensztajn,‡ and Wojtek Wlodarski† †
School of Electrical and Computer Engineering, RMIT University, Melbourne Vic 3001, Australia Laboratory of Interfaces and Electrochemical Systems (LISE, UPR15), National Center for Scientific Research (CNRS), Pierre and Marie Curie University (UPMC), Paris 75005, France
‡
ABSTRACT: In this report, polypyrrole nanowires have been successfully deposited on interdigitated transducers through template-free electropolymerization. The nanowires were 40−90 nm in diameter according to scanning electron microscopy analysis, and some of them were bridging the insulating gaps between gold electrodes. An X-ray photoelectron spectroscopy study has been conducted to determine the chemical composition of the synthesized nanomaterial. The developed sensors were tested toward five concentrations of hydrogen gas at room temperature, and their sensitivities were compared. Due to the very high surface area of the deposited sensitive films, these sensors provided faster response compared to other polypyrrolebased gas sensors. Moreover, it is shown that the sensors’ sensitivities are related to the amount of the deposited PPY nanowires.
■
INTRODUCTION Conductive polymers, alternatively referred to as conjugated polymers, can be considered as organic semiconductors. Depending on their synthetic conditions, their electrical and optical properties can become comparable to that of metals and nonorganic semiconductors. Polypyrrole (PPY) is a heteroaromatic conductive polymer that attracted a great deal of research attention due to its high conductivity and environmental stability.1 It can be easily synthesized in conjunction with other nonconjugated polymers such as poly(vinyl alcohol) (PVA), polystyrene (PS), poly(vinyl phosphate) (PVP), etc., to form conductive polymer blends that have the mechanical properties of regular plastics.1 PPY is usually prepared through chemical or electrochemical polymerization. Through chemical oxidative polymerization, nanostructured PPY has been synthesized in a number of studies. In recent years, a template-free synthetic approach has been increasingly adopted by researchers for mass production of PPY nanostructures. Through this approach, PPY nanofibers have been synthesized by utilizing bipyrrole to speed the polymerization process.2 Also, pure PPY nanoparticles have been synthesized via unstirred oxidative polymerization of pyrrole monomer in an acidic aqueous media at 0 °C without any template.3 Similarly, self-stabilized copolymer nanoparticles were produced through the chemical polymerization of pyrrole and 2-hydroxy-5-sulfonic aniline (HS) in an HCl solution.4 Chemically synthesized nanostructured conducting polymers are usually suspended in polar or nonpolar solvents and can be deposited on a substrate by either drop casting or airbrushing. However, these deposition techniques are not adequate for accurate deposition on a small area in the development of miniature devices. © 2012 American Chemical Society
On the contrary, electropolymerization allows a controllable polymer deposition on a conductive substrate utilized as the working electrode in an electrochemical cell. In earlier reports, polyaniline and polypyrrole nanowires were synthesized through template-free electropolymerization.5,6 For polyaniline, controlling the nucleation and growth was achieved by a stepwise decrease in the polymerization current density 0.08− 0.02 mA/cm2. In the case of polypyrrole, weak acidic anions (monohydrogeno phosphate) and nonacidic anions (perchlorate) were both included in the electrolyte. Debiemme-Chouvy proposed that due to the low anion concentration and pH value at the electrode/solution interface, hydroxyl radicals and O2 nanosized bubbles were formed in the first stages of the process.6 The bubbles can prevent the PPY overoxidation due to hydroxyl radicals providing a structural support for the nanowires’ growth. Soon after the first report on PPY electropolymerization,7 Kanazawa et al. reported that the conductivity of doped PPY decreased by a factor of 10 during the exposure to ammonia gas.8 The original PPY conductivity value was restored after removing the NH3 gas leading to the conclusion that physisorption was the dominant adsorption mechanism. In another report, Nylander et al. found that the exposure of PPY impregnated filter paper to NH3 gas with concentrations of (0.5−5%) diluted in either argon or oxygen−argon mixture increased the polymer resistance.9 Linear response was observed with a 30% resistance change per 1% of NH3 concentration. It was proposed that NH3 gas causes depletion Received: February 17, 2012 Revised: May 26, 2012 Published: May 31, 2012 13388
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
of the charge carriers in PPY or restriction to their mobility leading to a decrease in the overall film conductivity.10 Alternatively Blanc et al. suggested that the conductivity changes of PPY film during the interaction with NH3 gas consisted of two subsequent events, which are an initial fast decrease in conductivity due to a superficial adsorption followed by a relatively slower diffusion process.11 In recent studies, it has been reported that concentrations as low as 5 ppm of NH3 gas can be detected with conductometric gas sensors based on PPY nanoparticles.12,13 Kwon et al. investigated the gas sensing properties of PPY nanoparticles with diameters of 20, 60, and 100 nm.12 They have found that decreasing the nanoparticles size produced noticeable increase in the sensitivity toward NH3 gas. It was anticipated that the increased sensitivity can be attributed to the higher surface-tovolume ratio of the 20 nm nanoparticles compared to the 60 and 100 nm nanoparticles. However, the sensitivity of pure PPY films toward H2 gas has not been widely investigated. In a recent work, a composite of PPY and Pt was chemically synthesized and deposited in situ on a flexible conductometric transducer.14 It was found that the highest sensitivity toward H2 gas was obtained when the concentration of Pt nanoparticles in the sensitive film was 20 000 ppm. However, the response of the Pt/PPY based gas sensor was slow. A response time of 5.5 min was observed toward 1000 ppm of H2 gas, which can be attributed to the slow chemisorption of H2 molecules. Therefore, this sensor may not be practical in terms of real life situations where fast detection of any H2 gas leak is of utmost importance in protecting lives and properties from the devastating explosions that can be caused by the highly flammable H2 gas. Here we report for the first time, to the best of authors’ knowledge, room temperature hydrogen gas sensors based on electropolymerized polypyrrole nanowires without utilizing any templates. The developed sensors produced reversible responses toward different concentrations of hydrogen gas.
Figure 1. (A) Schematic diagram of the sensor before the electrodeposition; (B) image of a developed gas sensor.
at 0.85 V/SCE, and the current (I) vs time (t) was recorded allowing the determination of the anodic charge. Three sensors were developed with different anodic charges: 13, 55, and 90 mC, and the sensors were named as A, B, and C, respectively. In this work, either an Ultra55 Zeiss or an FEI Nova NanoSEM field emission scanning electron microscope was employed to analyze the deposited film’s morphology. The operation voltages were 3 and 5 kV for the Ultra55 Zeiss and the FEI Nova NanoSEM, respectively. For the XPS measurements, a Thermo Scientific K-Alpha spectrometer with an Al− Kα source and a spot size of 400 μm was utilized. Charging was minimized by a low-energy electron flood gun. Survey and high-resolution spectra were recorded with a pass energy of 200 and 50 eV, respectively. A gas sensing experiment was conducted using an enclosed environmental cell. The sensor was exposed to different concentrations of hydrogen gas diluted in synthetic air at room temperature. A computerized gas calibration system was used to vary the concentration of hydrogen gas in the synthetic air. The sensor was connected to a Keithley 2001 multimeter in order to measure the variations in film resistance during the interaction with the target gas. Certified gas bottles were used in this study. Any external humidity from the room environment introduced during the placement of the sensor inside the gas chamber was removed by purging the gas chamber with dry synthetic air for 1.5 h before starting the tests. The synthetic air used consisted of 79% nitrogen and 21% oxygen as specified by the supplier, while the hydrogen gas bottle contained 1% of hydrogen gas balanced in synthetic air. The synthetic air was used as the washing gas. All gas sensing experiments were conducted at room temperature, which was 24 °C.
■
EXPERIMENTAL SECTION The fabrication of conductometric transducers started by the deposition of metal thin films on a 3 in. quartz substrate using a BalzerTM electron beam evaporator. A 50 nm thick chromium film was deposited first to provide an adhesive layer for gold particles. Then a 100 nm gold layer was deposited over the chromium thin film. Interdigital electrodes (IDEs) were patterned by using a typical microfabrication procedure of photolithography and etching with positive photoresist (AZ1512). The spacing between the gold electrodes and the width of each electrode were 50 μm. Then the quartz substrate was diced to 8 × 12 mm2 pieces. The transducers were shortcircuited during the electrodeposition in order to obtain a closed loop for the current to flow between electrodes. After the electrodeposition the short-circuit wire was removed. A schematic diagram of a microfabricated conductometric transducer before PPY deposition is shown in Figure 1A. Figure 1B shows an image of a developed gas sensor after the electrochemical deposition of PPY nanowires. The electropolymerization of pyrrole monomer was conducted in a three-electrode electrochemical cell. The conductometric transducer was used as the working electrode. The counter electrode was a platinum wire. Saturated calomel electrode (SCE) was used for the reference electrode. The polymerization solution contained 0.15 M pyrrole, 0.2 M Na2HPO4, and 0.002 M LiClO4. The anodic potential was fixed
■
RESULTS AND DISCUSSIONS Characterization Results. SEM study of the PPY sensitive film confirmed the deposition of one-dimensional structures with a nanowire morphology on the surfaces of conductometric transducers. PPY nanowires deposition started at high density on the gold electrodes as shown in Figure 2A. The nanowires grow vertically from the PPY zones which were protected from the hydroxyl radicals by the O 2
13389
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
Figure 2. SEM images of PPY nanowires deposited on a conductometric transducer’s surface: (A) 65° tilted view (scale bar = 1 μm) and (B) top view showing PPY nanowires bridging the insulating gap between gold electrodes (scale bar = 10 μm).
Figure 3. High-magnification SEM images of the electrodeposited PPY nanowires: (A) on sensor A (scale bar = 1 μm), (B) on sensor B (scale bar = 1 μm), (C) on sensor C (scale bar = 1 μm), and (D) magnified SEM image of the PPY nanowires on sensor C (scale bar = 300 nm).
nanobubbles that prevented overoxidization.6 After packing of the gold electrodes with nanowires, further growth of the PPY chains leads to some horizontal growth of the nanowires. This type of growth allows nanowires initiated from adjacent gold electrodes to intersect with each other and form a bridge over the insulating gap between the gold electrodes, which is the quartz substrate for these sensors as shown in Figure 2B. Obviously, without the formation of these PPY nanowires’ bridges, the sensor’s resistance value would remain infinite. High-magnification SEM images for the PPY nanowires deposited on the surfaces of sensors A, B, and C are shown in Figure 3A, B, and C, respectively. Extra magnification of the PPY nanowires on sensor C is shown in Figure 3D.
The nanowires diameters are in the range of 40−90 nm and lengths in the order of several tens of micrometers. Similar sensitive films’ morphologies consisting of nanowires were observed for the three developed sensors as shown in Figure 3. However, the SEM image in Figure 3A shows that the sensitive film of sensor A consisted of a mixture of short and long PPY nanowires that tend to merge with each other to form irregularly shaped lumps. This observation can be attributed to the low value of charge (13 mC) passing through the working electrode during electropolymerization. It was suggested in earlier reports that the electrostatic repulsive force of the charged groups in the reaction solution was responsible for selfstabilization and homogeneous nucleation of PPY chains to 13390
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
Figure 4. XPS spectra of PPY nanowires: (A) survey spectrum; (B) C 1s spectrum is deconvoluted into aromatic C peak (green), C−N peak (brown), and CO peak (black); (C) N 1s spectrum is deconvoluted into the following: neutral N peak (green) and imine N peak (brown); (D) O 1s spectrum is deconvoluted into perchlorate anion peak (green) and CO peak (brown).
into two peaks. A large peak is centered at binding energy ∼400 eV, and a smaller peak at a binding energy ∼398 eV can be assigned to the neutral N atom and the imine nitrogen ( N−), respectively.17 The O 1s spectrum shown in Figure 4D can be deconvoluted into a peak centered at binding energy ∼533 eV that is assigned to perchlorate anion and a peak at a binding energy ∼532 eV which is attributed to CO.17 Hydrogen Gas Sensing Results. The developed gas sensors’ performances were analyzed during the exposure toward five concentrations of hydrogen gas balanced in synthetic air at room temperature. The dynamic responses and the sensitivities of these sensors are shown in Figure 5. The sensors sensitivities (S) were calculated according to the following relation
produce nanoparticles of pure PPY and its composites.3,4 Therefore, as the amount of charge passing between the electrodes increases, homogeneous nucleation becomes more dominant over heterogeneous nucleation promoting the formation of one-dimensional nanostructures as shown in Figure 3B, C, and D. Due to the uniform shape of the PPY nanowires, the deposited thin film is highly porous. From a gas sensing point of view, highly porous thin films are favorable since they permit large penetration of gas molecules into the whole volume of the sensitive film. The PPY nanowires’ chemical composition was investigated by analyzing the XPS spectra of the deposited films. Figure 4 displays the XPS spectra for PPY nanowires. A survey photoelectron spectrum is provided in Figure 4A showing the elemental composition. The presence of the incorporated perchlorate anions into the PPY matrix is evident from the Cl 2p peak at binding energies 208 eV.15 However, phosphorus was not detected in the PPY nanowires since the P 2p peak at a binding energy of 132 eV did not appear in the survey spectrum. This observation can be attributed to two reasons. First, the sensitivity factor for P 2p photoelectron is low in our measurements. Furthermore, the pH level at the PPY/solution interface was low due to proton release during the oxidation of pyrrole monomer.16 Hence the phosphate was under its nonionized form (H 3 PO 4 ), and therefore it was not incorporated in the PPY matrix. Deconvolution of C1s spectrum shown in Figure 4B consists of three component peaks. The main peak at a binding energy of 284.8 eV is assigned to the aromatic C, the peak at 286.3 eV is assigned to C−N, and the peak at 288 eV is assigned to C O.15,16 The N 1s spectrum shown in Figure 4C is deconvoluted
S=
R − R0 × 100% R0
where R and R0 are the film resistances during the exposure to H2 gas and under synthetic air, respectively. The baseline resistances (R0) for sensors A, B, and C were 13 287, 562, and 9430 Ω, respectively. The differences in the sensors’ baseline resistances are indicative of a different number of PPY bridges crossing the insulating gaps between the gold electrodes of these sensors. The number of PPY bridges is affected by all the electrochemical conditions governing the formation and precipitation of polymer colloids on the working electrode.18 Hence the history of the electrolyte and the number of monomer units available for oxidation become of significant importance that can affect the sensors’ baseline resistance values in addition to the amount of anodic charge 13391
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
Figure 5. Dynamic responses and sensitivities of the H2 gas sensors.
slight morphological difference between them as shown in Figure 3. Therefore the high sensitivity of sensor C compared to the other sensors can be attributed to the high surface-tovolume ratio for its sensitive film that consists of long and uniform PPY nanowires as shown in Figure 3C and D. It is also worth to mention here that the response of sensor A exhibited in Figure 5 is rather noisy compared to that of other sensors. This difference can be caused by the presence of a mixture of long and short nanowires in the sensitive film of sensor A that tend to tangle with each other as shown in Figure 3A. Experimentally, the lowest detectable gas concentration is restricted by the gas calibration system limitations.19 In our measurement setup, the lowest H2 gas concentration that can be obtained from the mass flow controller is 600 ppm. The sensor’s detection limit can be estimated by extrapolating the normalized resistance values during gas exposure after calculating the noise level of the baseline resistance (3σ/R0), where σ is the standard deviation R0 values. 20,21 By implementation of this method, the detection limit of the H2 gas sensors was calculated and found to be 12 ppm of H2 gas. The linear detection range of the gas sensors can be extracted from the linear part of the sensitivity curves which is ∼600− 2500 ppm of H2 gas. Nowadays, there is a global interest in minimizing the use of fossil fuel to improve air quality. Hydrogen technology is playing a major role in that endeavor. In that perspective, gas sensors with a fast response to a hydrogen gas leak are critical. Although a highly sensitive H2 gas sensor was reported by
during electropolymerization. Therefore the sensitivity values of the sensors were adopted as the basis for comparison in this work since they rely on normalized resistance as evident from the above equation. Also the sensitive films’ morphologies were compared to elucidate the difference between the developed sensors’ sensitivities. The three sensors produced reversible responses toward H2 gas as demonstrated in Figure 5. Table 1 summarizes the Table 1. Shifts in Gas Sensors’ Resistances during the Exposure to H2 gas H2 gas concentration (ppm)
sensor A ΔR (Ω)
sensor B ΔR (Ω)
sensor C ΔR (Ω)
600 1200 2500 5000 10000
−99.5 −139 −204.9 −256 −234.8
−3.8 −6 −8.5 −12.2 −15.2
−84.7 −166.9 −248 −316.3 −366.6
downshifts in sensors’ resistances during the interaction with five concentrations of H2 gas balanced in synthetic air. It is evident from Figure 5 that sensor C had the highest sensitivity and the fastest response toward H2 gas. For instance, during the exposure to 10 000 ppm of H2 gas, the drop in resistance of sensor C reached saturation after 54 s, whereas sensors A and B responded in 90 and 86 s toward the same concentration of H2 gas, respectively. Despite the fact that the three PPY sensitive films consist of structures in the form of nanowires, there is a 13392
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
decorating PPY thin film with Pt nanoparticles, the sensor response was slow (330 s toward 1000 ppm of H2 gas),14 whereas in this report, the response time of sensor C was 72 s during the exposure to 600 ppm of H2 gas. This improvement can be attributed to the high surface-to-volume ratio of the sensitive film based on PPY nanowires compared to that of the Pt-decorated PPY thin film and therefore allowing for fast adsorption/desorption kinetics. Since hydrogen technology is becoming more popular in the car manufacturing industry everyday, a gas sensor’s selectivity toward H2 gas among other interfering gases in the ambient becomes an important parameter. Due to our experimental setup limitation, we could assess the H2 gas sensor selectivity in the presence of one interfering gas only. We have chosen CO as the interfering gas since it is a toxic gas produced as a result of incomplete combustion of petrol in conventional automobiles. We have chosen the sensor with the lowest sensitivity toward H2 gas in this study, sensor A, in order to evaluate the highest interference impact. Figure 6 shows the response of the gas sensor toward 10 000 ppm of H2 gas with and without the interference of 500 ppm of CO gas.
to decrease during the interaction with H2 gas which agrees with our experimental results.
■
CONCLUSIONS Hydrogen gas sensors have been fabricated based on polypyrrole nanowires synthesized through template-free electrooxidation of pyrrole. SEM analysis confirmed the deposition of nanowires with diameters of 40−90 nm on conductometric transducers. It was found that the PPY nanowires were forming bridges to cross the quartz gap between adjacent gold electrodes. It was evident that the sensors’ baseline resistance and the surface-to-volume ratio of the deposited sensitive films were related to the charges flowing during electropolymerization. The gas sensors have been investigated for their sensitivities toward five concentrations of hydrogen gas at room temperature. Compared to previously reported sensors, the developed sensors herein featured fast and reversible responses due to the high porosity of the PPY nanowires-based sensitive films.
■
AUTHOR INFORMATION
Corresponding Author
*Tel. +613-9925-3690, fax +613-9925-2007, e-mail laith.
[email protected] (L.A-M.); tel. +331-4427-4149, fax +3314427-4074, e-mail
[email protected] (C.DC.). Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Wang, L. X.; Li, X. G.; Yang, Y. L. React. Funct. Polym. 2001, 47, 125−139. (2) Tran, H. D.; Shin, K.; Hong, W. G.; D’Arcy, J. M.; Kojima, R. W.; Weiller, B. H.; Kaner, R. B. Macromol. Rapid Commun. 2007, 28, 2289−2293. (3) Li, X. G.; Li, A.; Huang, M. R.; Liao, Y.; Lu, Y. G. J. Phys. Chem. C 2010, 114, 19244−19255. (4) Li, X. G.; Hou, Z. Z.; Huang, M. R.; Moloney, M. G. J. Phys. Chem. C 2009, 113, 21586−21595. (5) Liu, J.; Lin, Y.; Liang, L.; Voigt, J. A.; Huber, D. L.; Tian, Z. R.; Coker, E.; McKenzie, B.; McDermott, M. J. Chem.Eur. J. 2003, 9, 604−611. (6) (a) Debiemme-Chouvy, C. Electrochem. Commun. 2009, 11, 298− 301. (b) Debiemme-Chouvy, C.; Tran, T. T. M. Electrochem. Commun. 2008, 10, 947−950. (7) Diaz, A. F.; Kanazawa, K. K.; Gardini, G. P. J. Chem. Soc., Chem. Commun. 1979, 635−636. (8) Kanazawa, K. K.; Diaz, A. F.; Geiss, R. H.; Gill, W. D.; Kwak, J. F.; Logan, J. A.; Rabolt, J. F.; Street, G. B. J. Chem. Soc., Chem. Commun. 1979, 854−855. (9) Nylander, C.; Armgarth, M.; Lundström, I. Proc. Int. Mtg. Chem. Sens. 1983, 17, 203−207. (10) Gustafsson, G.; Lundström, I.; Liedberg, B.; Wu, C. R.; Inganäs, O.; Wennerström, O. Synth. Met. 1989, 31, 163−179. (11) Blanc, J. P.; Derouiche, N.; El-Hadri, A.; Germain, J. P.; Maleysson, C.; Robert, H. Sens. Actuators, B 1990, 1, 130−133. (12) Kwon, O. S.; Hong, J. Y.; Park, S. J.; Jang, Y.; Jang, J. J. Phys. Chem. C 2010, 114, 18874−18879. (13) Chougule, M. A.; Pawar, S. G.; Patil, S. L.; Raut, B. T.; Godse, P. R.; Sen, S.; Patil, V. B. IEEE Sens. J. 2011, 11, 2137−2141. (14) Su, P. G.; Shiu, C. C. Sens. Actuators, B 2011, 157, 275−281. (15) Carquigny, S.; Sanchez, J. B.; Berger, F.; Lakard, B.; Lallemand, F. Talanta 2009, 78, 199−206. (16) Debiemme-Chouvy, C. Electrochem. Solid-State Lett. 2007, 10, E24−E26.
Figure 6. H2 gas sensor performance with and without CO gas interference.
Despite the presence of CO gas, the sensor produced a reversible decrease in its sensitive film resistance during the exposure to 10 000 ppm of H2 gas. However, the shift in the sensitive film resistance toward the H2/CO gas mixture was less by 89 Ω compared to that toward pure H2 gas. CO gas molecules can withdraw electrons from the nitrogen atoms in the pyrrole aromatic rings leading to the formation of polarons.22 This reaction can decrease the number of adsorption sites available for interaction with H2 gas in the bulk of the polymer film contributing to a loss in sensitivity of the H2 gas sensor during CO gas interference. The gas sensing mechanism of PPY is still under further investigation. The interaction between PPY sensitive films and the adsorbed gas molecules can lead to a heterogeneous charge transfer. Therefore a chemical modulation of the polymer doping level can occur leading to changes to the Fermi level of the organic semiconductor.23 This effect produces variations to the electronic conductivity or the work function, Φ, of the polymer. The polarity of the response relies on the charge density exchange kinetics between the diffusing gas molecules and the polymer matrix through either oxidation (ΔΦ > 0) or reduction (ΔΦ < 0). Hence the gas molecules can be regarded as electron acceptors or donors, respectively. Since the H2 gas is a reducing gas, the resistance of the PPY nanowires is expected 13393
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394
The Journal of Physical Chemistry C
Article
(17) Tongol, B. J. V.; Binag, C. A.; Sevilla, F. B., III. Sens. Actuators, B 2003, 93, 187−196. (18) Heinze, J.; Frontana-Uribe, B. A.; Ludwigs, S. Chem. Rev. 2010, 110, 4724−4771. (19) Li, J.; Lu, Y.; Ye, Q.; Cinke, M.; Han, J.; Meyyappan, M. Nano Lett. 2003, 3, 929−933. (20) Kim, I. D.; Rothschild, A.; Lee, B. H.; Kim, D. Y.; Jo, S. M.; Tuller, H. L. Nano Lett. 2006, 6, 2009−2013. (21) Huang, J.; Wan, Q. Sensors 2009, 9, 9903−9924. (22) Radhakrishnan, S.; Paul, S. Sens. Actuators, B 2007, 125, 60−65. (23) Janata, J. Anal. Chem. 1991, 63, 2546−2550.
13394
dx.doi.org/10.1021/jp3015854 | J. Phys. Chem. C 2012, 116, 13388−13394