Incorporation of Homogeneous, Nanoscale MnO2 within Ultraporous

Incorporation of Homogeneous, Nanoscale MnO2 within Ultraporous Carbon Structures via Self-Limiting ..... The Journal of Physical Chemistry C 0 (proof...
0 downloads 0 Views 469KB Size
Download PDF
NANO LETTERS

Incorporation of Homogeneous, Nanoscale MnO2 within Ultraporous Carbon Structures via Self-Limiting Electroless Deposition: Implications for Electrochemical Capacitors

2007 Vol. 7, No. 2 281-286

Anne E. Fischer,† Katherine A. Pettigrew,† Debra R. Rolison,† Rhonda M. Stroud,‡ and Jeffrey W. Long*,† Surface Chemistry Branch (Code 6170) and Materials and Sensors Branch (Code 6360), NaVal Research Laboratory, Washington, DC 20375 Received September 26, 2006

ABSTRACT The self-limiting reaction of aqueous permanganate with carbon nanofoams produces conformal, nanoscopic deposits of birnessite ribbons and amorphous MnO2 throughout the ultraporous carbon structure. The MnO2 coating contributes additional capacitance to the carbon nanofoam while maintaining the favorable high-rate electrochemical performance inherent to the ultraporous carbon structure of the nanofoam. Such a three-dimensional design exploits the benefits of a nanoscopic MnO2−carbon interface to produce an exceptionally high area-normalized capacitance (1.5 F cm-2), as well as high volumetric capacitance (90 F cm-3).

In this report, we describe the electroless deposition of manganese oxide (denoted here as MnO2) under controlled pH conditions within and onto commercially available carbon aerogel substrates (also called “nanofoams”). In this general method, the carbon substrate serves as a sacrificial reductant and converts aqueous permanganate (MnO4-) to insoluble MnO2, as described previously for MnO2 deposition on planar graphite,1 acetylene black powders,2,3 templated mesoporous carbon powders,4 and carbon nanotubes.5 Although this approach offers a straightforward processing protocol, careful control of the reaction is required in order to achieve nanoscale MnO2 deposits throughout a three-dimensional monolithic carbon architecture. Achieving such non-line-ofsight control is particularly challenging when modifying macroscopically thick porous carbon substrates because the supply and flux of permanganate is likely to be inhomogeneous between the electrode exterior and interior, thereby favoring deposition at the exterior boundary. Moreover, preservation of the native carbon template pore structure is vital for electrochemical performance in order to facilitate electrolyte infiltration and ion transport throughout the internal volume of the hybrid electrode. Such hybrid electrode structures can be produced by using deposition * To whom correspondence should be addressed. E-mail: jeffrey.long@ nrl.navy.mil. Telephone: (202) 404-8697. Fax: (202) 767-3321. † Surface Chemistry Branch (Code 6170). ‡ Materials and Sensors Branch (Code 6360). 10.1021/nl062263i CCC: $37.00 Published on Web 01/13/2007

© 2007 American Chemical Society

methods that are inherently self-limiting, as recently demonstrated for the electrodeposition of conformal, ultrathin poly(o-methoxyaniline) coatings on carbon nanofoam templates.6 Manganese oxides are an important and well-studied class of electrode materials for batteries,7,8 and have more recently been investigated as electrochemical capacitors (ECs)9-26 with the anticipation that MnO2 will serve as a low-cost replacement for hydrous RuO2, the state-of-the-art EC metal oxide.27 The cumulative evidence published thus far establishes that the electrochemical performance of MnO2, both in terms of capacitance and power characteristics, critically depends on the electrode architecture. When prepared as micrometer-thick deposits10,11,19 or in composite electrode forms containing carbon and binders,9 MnO2 delivers a specific capacitance of ∼150-250 F g-1, which is competitive with carbon supercapacitors, but falls far short of the 720 F g-1 obtained with hydrous RuO2.27 By contrast, ultrathin MnO2 deposits (tens to hundreds of nanometers thick) deliver specific capacitances ranging from ∼700 F g-1 28,29 to 1380 F g-1.30 The capacitance for thick MnO2 films or conventional composite electrodes is ultimately limited by the poor electrical conductivity of MnO2. In turn, EC device performance using the planar ultrathin configuration28-30 is restricted because of low mass loading. In view of these design considerations and performance limitations, a hybrid electrode architecture that incorporates

nanoscopic MnO2 at the surface of a porous, high surface area, and electronically conducting structure (e.g., carbon nanofoams, templated mesoporous carbon, and nanotube/ nanofiber assemblies) should optimize both the electrochemical performance and mass-loading of an ultrathin MnO2 phase. In such a configuration, the carbon backbone acts as a highly conductive current collector, the through-connected porosity serves as a continuous pathway for electrolyte transport, and the nanoscopic MnO2 phase minimizes solidstate transport distances for ions into the oxide. Recent efforts to integrate nanoscopic carbon and MnO2 have primarily focused on incorporating nanoscale MnO2 deposits onto carbon nanotubes by using a variety of approaches, including physical mixing of the components,31 thermal decomposition,32 chemical deposition using precursors such as permanganate,33,34 and electrochemical deposition.35,36 In all cases, the incorporated MnO2 improves the capacitance of the carbon nanotube assembly, however, the overall specific capacitance remains typically less than 200 F g-1, even for electrodes with high mass loadings of MnO2. One exception was reported by Lee et al.,35 who demonstrated specific capacitances of up to 415 F g-1, normalized to the MnO2 in the composite structure; these results were achieved only for micrometer-thick MnO2-carbon nanotube electrode structures. Nanoscale MnO2 incorporated into templated mesoporous carbons (pore size < 5 nm) yielded even greater enhancement,4 with an MnO2-normalized capacitance of ∼600 F g-1. However, the overall specific capacitance was limited to 200 F g-1 due to the relatively low weight loading (up to 26%) of MnO2. Although the electrochemical performance of such MnO2carbon structures is largely affected by the morphology and distribution of the MnO2 phase, the structure of the carbon substrate is also an important consideration. For example, physical properties of the substrate such as morphology and microstructure, surface area, and pore size affect the nature of the deposited MnO2 as well as the electrochemical properties of the final hybrid device. Carbon nanofoams and related sol-gel-derived nanoarchitectures are particularly attractive for electrochemical capacitor applications due to their inherent structural characteristics that include high specific surface areas, through-connected networks of mesopores and/or macropores, tunable pore sizes ranging from nanometers to micrometers, durable monolithic, moldable forms, and reasonable electronic conductivity (10-40 S cm-1).37 These same characteristics make carbon nanofoams a useful platform for modification with MnO2. For example, the surface area, porosity, and pore sizes of carbon nanofoams can be tuned in order to maximize the weight loadings of incorporated MnO2. In this way, the hybrid architecture combines the desirable structural and electronic characteristics of the native nanofoam with the built-in electrochemical capacitance of MnO2. Our work reported here demonstrates that by simply controlling solution pH to promote self-limiting deposition, homogeneous, nanoscopic MnO2 deposits can be achieved throughout macroscopically thick porous carbon nanofoams, while the through-connected pore network of the nanofoam 282

Figure 1. Scanning electron micrographs of (a and b) 240 min acid-deposited MnO2-carbon, (c and d) 240 min neutral-deposited MnO2-carbon, and (e and f) bare carbon nanofoam.

is simultaneously retained. Carbon-paper-supported resorcinol-formaldehyde nanofoams (MarkeTech, Int., 170 µm thick) were pyrolyzed at 1000 °C under flowing Ar to produce the conductive carbon nanofoam. The pyrolyzed substrates (∼40 mg) were vacuum infiltrated with aqueous 0.1 M H2SO4 or 0.1 M Na2SO4 and then soaked individually in 25 mL of either 0.1 M NaMnO4 + 0.1 M H2SO4 (aciddeposited) or 0.1 M NaMnO4 + 0.1 M Na2SO4 (neutraldeposited) for 5-240 min; samples were then rinsed thoroughly with ultrapure water and dried at ∼50 °C under N2 for 8 h and then under vacuum overnight. The MnO2 mass uptake at short times (5-60 min) depends on solution pH and deposition time, with faster mass uptake occurring in acidic solution (see Supporting Information, Figure S1). At longer times, the rate of mass uptake slows considerably, reaching ∼40-45% mass increase for either pH. There is no visible decoloration of the permanganate solution during the deposition. Despite the similarity in the total MnO2 mass uptake for long exposure times under acidic and neutral conditions, the morphology and distribution of MnO2 within the carbon nanofoam differs markedly, as evidenced by scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS). Thick crusts of MnO2 deposit on the carbon electrode exterior under acidic pH conditions, as shown in Figure 1a and b for a 240 min deposition; a crust is visible even on samples immersed for only 5 min. The aciddeposited MnO2 crust is ∼4 µm thick for a 240 min deposition, as determined by cross-sectional microscopy analysis (see inset of Figure 1a). The rapid MnO2 mass uptake at short times for the acid-deposited sample, as well as the MnO2 crust formation, suggests that the deposition process occurs primarily at the exterior of the structure, where much of the permanganate is consumed before it can diffuse into the internal volume of the carbon nanofoam. Preferential Nano Lett., Vol. 7, No. 2, 2007

Figure 3. Transmission electron micrographs of (a) 240 min neutral-deposited MnO2-carbon, (b) 240 min acid-deposited MnO2-carbon, and (c) STEM of 240 min neutral-deposited MnO2carbon. Figure 2. Scanning electron micrographs with elemental mapping of C and Mn for a 240 min neutral-deposited MnO2-carbon nanofoam.

deposition of MnO2 on the exterior is expected for this sample due to the autocatalytic nature of permanganate decomposition in acidic solutions.38,39 An estimation of the MnO2 mass on the electrode exterior using the density of MnO2 (4.8 g cm-3) and the geometric area and thickness of the crust accounts for ∼80% of the MnO2 mass. The remaining MnO2 is incorporated within the interior of the nanofoam structure, as evidenced by EDS of a sample crosssection (data not presented). By contrast, the SEM images of the 240 min neutraldeposited MnO2-carbon nanofoam structure (Figure 1c and d) are nearly indistinguishable from the bare carbon nanofoam (Figure 1e and f) and have no exterior MnO2 crust. Moreover, the porous texture of the initial carbon nanofoam is largely retained following MnO2 deposition (Figure 1d). Analysis of a cross-section of the hybrid structure by SEM and EDS elemental mapping (Figure 2) confirms, however, that within the spatial resolution limits of the SEM/EDS measurement, Mn is present and evenly distributed throughout the thickness of the electrode structure. The nanoscale structure of the neutral-deposited composite architecture, as seen by transmission electron microscopy (TEM), reveals the presence of layered birnessite MnO2 ribbons and rods that permeate the carbon nanofoam substrate (Figure 3a). A similar ribbon morphology was observed by Ma et al. for spontaneous MnO2 deposition on acetylene black.2 In contrast to the ribbons and rods observed in the neutral-deposited nanofoam, the acid-deposited MnO2carbon nanofoam contains agglomerated nanoparticles of birnessite-type MnO2 that are not as evenly dispersed within the carbon nanofoam structure. In addition to the crystalline manganese oxide structures in both samples, elemental mapping in scanning TEM mode confirms additional homogeneous, nanoscale distribution of an amorphous manganese oxide phase throughout the carbon nanofoam (see Nano Lett., Vol. 7, No. 2, 2007

Figure 4. Nitrogen-sorption porosimetry of bare carbon nanofoam, 240 min acid-deposited MnO2-carbon nanofoam, and 240 min neutral-deposited MnO2-carbon nanofoam.

Figure 3c for neutral-deposited MnO2-carbon nanofoam and Supporting Information, Figure S2, for acid-deposited MnO2carbon nanofoam). Although homogeneous distribution of MnO2 throughout the interior of the electrode is important in order to optimize MnO2 mass loading and subsequent EC energy density, it is also critical that the high-quality pore network of the native nanofoam be preserved to facilitate high-rate EC performance. Changes in the pore volume and distribution as a result of MnO2 incorporation were tracked by nitrogen porosimetry. The cumulative pore volume of the carbon nanofoam decreases from 0.62 cm3 g-1 (normalized to the total mass of the nanofoam, which is dominated by the mass of the carbon paper support) to 0.31 cm3 g-1 (normalized to the total mass of the nanofoam and incorporated MnO2) for a 240 min neutral-deposited MnO2-carbon nanofoam. Importantly, as shown in Figure 4, the pore size distribution for the neutral-deposited sample shifts toward smaller pore sizes, providing further evidence that MnO2 incorporates within the void volume of the nanofoam without occluding the desirable pore structure. The nitrogen-sorption porosimetry data also reveal that pores sized at 10-60 nm contain most of the free volume in the structure; this size range is 283

Figure 5. Cyclic voltammograms for bare carbon nanofoam, 240 min acid-deposited MnO2-carbon nanofoam, and 240 min neutraldeposited MnO2-carbon nanofoam at (a) 2 mV s-1 and (b) 20 mV s-1 in 1 M Na2SO4.

small relative to the MnO2 coating thickness that we are targeting (tens of nanometers), likely limiting the mass loadings of MnO2 that can be achieved in this particular class of nanofoam. The crustlike acid-deposited MnO2-carbon also exhibits a decrease in pore volume (to 0.50 cm3 g-1) and a shift to smaller pore sizes with respect to the native carbon nanofoam. However, the pore volume loss is not as large as that for the neutral-deposited sample, consistent with less MnO2 being incorporated within the pore structure of the carbon. Further interpretation of the porosimetry data for the aciddeposited MnO2-carbon nanofoam is complicated by the fact that the MnO2 crust exhibits its own porous texture, as observed by SEM in Figure 1b. Representative cyclic voltammograms of the bare carbon nanofoam, 240 min acid-deposited, and 240 min neutraldeposited MnO2-carbon hybrid electrodes in 1 M Na2SO4 at 2 and 20 mV s-1 are presented in Figure 5a and b, respectively. At 2 mV s-1, all voltammetric curves exhibit a nearly symmetrical rectangular shape, indicative of an ideal capacitor, with relatively low uncompensated electrode or solution resistance. Both the total gravimetric and volumetric capacitance values increase for the acid- and neutraldeposited samples with respect to the bare carbon nanofoam (see Table 1), but the gravimetric capacitance of the neutraldeposited sample increases by a factor of 2, while the volumetric capacitance is over 4 times higher. It is important to note that homogeneous, nanoscopic MnO2 deposits contribute capacitance without increasing the bulk volume of the electrode structure. When pulse power is required in a footprint- or area-limited configuration, as in microelectromechanical (MEMS) based and on-chip devices, the areanormalized energy storage capacity must also be considered.40 Although the area-normalized capacitance is often not reported for MnO2-carbon composites, it is usually around 10-50 mF cm-2. In contrast, our hybrid electrode design maintains the advantages of a nanoscopic electrode284

electrolyte interface while projecting the electrode structure in three dimensions with a limited footprint such that the area-normalized capacitance for the acid- and neutraldeposited MnO2-carbon hybrid electrodes is orders of magnitude greater at 1.4 and 1.5 F cm-2, respectively. The cycling stability of the MnO2-carbon hybrid electrodes was evaluated by galvanostatic charge-discharge measurements from 0 to 0.8 V at a relatively high chargedischarge rate of 0.25 A g-1. The gravimetric capacitance on cycle 50 for both the acid- (75 F g-1) and neutraldeposited (102 F g-1) hybrid electrodes is higher than that of the bare carbon nanofoam (53 F g-1). The capacitance for cycle 100 is within 10% of the initial cycle for all three porous electrodes, although the stability of the bare nanofoam (6.3% decay) and neutral-deposited MnO2-carbon hybrid (7.9% decay) are better than the acid-deposited sample (9.5% decay; Figure S3). X-ray photoelectron spectroscopy (XPS) was used to ensure that the observed capacitance enhancement for the MnO2-carbon hybrid electrodes was due to electrochemical activity of the MnO2 phase. Neutral-deposited MnO2-carbon electrodes were polarized for 30 min in 1 M Na2SO4 at either the high (0.8 V) or low (0 V) potential limits used in this work, then subsequently rinsed and dried prior to XPS measurement. The electrodes were examined in their intact form, with the XPS primarily sampling the exterior of the MnO2-carbon structure. The binding energies of the Mn 2p and O 1s peaks confirm the presence of a Mn(III/IV) oxide coating, with no residual MnO4- detectable. Although the Mn 2p peaks are relatively insensitive to changes in oxidation state, the O 1s spectra can be deconvolved, with peaks representing MndO and MnOH/MnOH2 chemical states that correspond to the Mn(IV) and Mn(III) oxidation states, respectively.41,42 For our samples, four peaks were used to deconvolve the O 1s region (see Supporting Information, Figure S4): Mnd O at 529.7 eV, MnOH/MnOH2 at 531.2 eV, adsorbed water at 532.7 eV, and carbonaceous oxygen species at ∼534.5 eV. The relative areas of the MnO and MnOH/MnOH2 peaks30 were used to calculate the average Mn oxidation state in the polarized neutral-deposited MnO2-carbon hybrid and yield 3.6 at 0 V and 3.7 at +0.8 V. This relatively small change in the Mn oxidation state for a 0.8 V potential window corresponds to an electrochemical capacitance for the neutral-deposited MnO2 of ∼140 F g-1, which is consistent with our electrochemical measurements, but far lower than the 700 F g-1 reported for ultrathin MnO2 coatings.28-30 Thus, factors other than the thickness of the MnO2 deposit are important for achieving such high specific capacitance. The crystalline phase (or lack thereof) of the MnO2 will also be a critical parameter, particularly if electron-ion insertion reactions involving large hydrated cations, such as the Na+ used here, are responsible for the electrochemical capacitance of MnO2. Currently, there are conflicting observations on the relative participation of protons and supporting electrolyte cations in the pseudocapacitance mechanism for MnO2,9,28,30,43-46 which may be due to the different MnO2 polymorphs used in each study.47 Nano Lett., Vol. 7, No. 2, 2007

Table 1. Capacitance Values for Carbon Nanofoam and MnO2-Carbon Nanofoam Electrodes (240 min Deposition) Derived from Cyclic Voltammetry at a Scan Rate of 2 mV s-1 a

bare nanofoam acid-deposited neutral-deposited

specific capacitance (Fg-1C+MnO2)

MnO2-specific capacitance range (Fg-1MnO2)b

volumetric capacitance (Fcm-3)

area-normalized capacitance (Fcm-2)

53 92 110

150-220 170-230

20 81 90

0.56 1.4 1.5

a A saturated calomel reference electrode (SCE) and reticulated vitreous carbon auxiliary electrode were used in all electrochemical measurements. b The upper and lower limits of capacitance attributable to MnO2 are estimated using one of two limiting assumptions: (1) all measured capacitance arises from the MnO2 phase (upper limit) or (2) the total sample capacitance is the sum of the carbon double-layer capacitance and the MnO2 capacitance (lower limit).

Despite the similar capacitance enhancement for both acidand neutral-deposited MnO2-carbon hybrid electrodes with respect to the bare carbon nanofoam at slow voltammetric scan rates (2 mV s-1), the advantages of introducing MnO2 as homogeneous, nanoscale deposits on the carbon nanofoam become more evident at higher scan rates. For example, the slope of the voltammogram for the acid-deposited MnO2carbon electrode, indicative of polarization resistance, becomes more pronounced as the scan rate is increased to 20 mV s-1 (Figure 5b) with an overall decrease in capacitive current. Although the neutral-deposited curve also exhibits slight polarization resistance at this scan rate, the change in slope and curve shape is much less severe than that for the acid-deposited sample, owing to the significant polarization resistance originating from its thick MnO2 crust. The resistance introduced by the thick MnO2 coatings is also evident when the electrodes are analyzed by electrochemical impedance spectroscopy, using a frequency range of 5-10 mHz and a DC bias of +0.2 V. The Nyquist plot in Figure 6a reveals that, at high frequencies, the uncompensated solution resistance (RΩ) of each electrode is comparable. However, the additional large semicircle observed for the acid-deposited MnO2-carbon electrode is indicative of polarization or charge-transfer resistance (Rp) of ∼15 Ω. We attribute this resistance to the poor electrical conductivity of the 4 µm thick MnO2 crust. An estimation of the crust resistance, assuming a resistivity of 105 Ω cm previously reported for micrometer-thick birnessite MnO2,48 yields a resistance of 40 Ω, which, given the assumptions of this estimation, is consistent with the experimentally measured resistance. Such high charge-transfer resistance would obviously be detrimental for the high-rate operation that is required of electrochemical capacitors. In contrast, the impedance spectra for the neutral-deposited MnO2carbon electrode is very similar to that of the bare carbon nanofoam, with an Rp of only ∼1 Ω. The low-frequency branches of the impedance spectra are typical for the capacitance response of a porous electrode structure.49 Bode plots for the real part of the capacitance (normalized to electrode geometric area) are shown in Figure 6b. For frequencies from 0.01 to ∼2 Hz, the MnO2 component of the neutral-deposited MnO2-carbon significantly increases the capacitance of the bare carbon nanofoam, indicating that the MnO2 domains within the tortuous interior of the carbon nanofoam are electrochemically accessible over a reasonable frequency range. At higher frequencies, the capacitance Nano Lett., Vol. 7, No. 2, 2007

Figure 6. (a) Nyquist plot and (b) Bode plot from impedance spectroscopic analysis of bare carbon nanofoam, 240 min aciddeposited MnO2-carbon nanofoam, and 240 min neutral-deposited MnO2-carbon nanofoam.

curves of these two electrode structures converge. The aciddeposited MnO2-carbon also enhances the capacitance of the nanofoam at low frequencies, but at frequencies >2 Hz, the bare carbon nanofoam actually exhibits higher capacitance. The poor performance of the acid-deposited MnO2carbon at higher frequencies is likely due to the thick MnO2 crust that forms on the electrode exterior, hindering electron and ion transport, while the more ideal homogeneous distribution of the neutral-deposited MnO2-carbon results in electrochemical and rate-critical characteristics consistent with the bare nanofoam. In conclusion, we have demonstrated a very simple, costeffective approach for incorporating homogeneous, nanoscale MnO2 deposits within the tortuous interior of carbon nano285

foam electrodes. The distribution of the MnO2 coating is critically important in order to increase the electrochemical capacitance of the carbon electrode without increasing charge transfer or contact resistance. We predict even greater enhancements in electrochemical performance for these hybrids with further optimization of the electroless deposition conditions, thereby improving MnO2 utilization. For example, varying the carbon template pore structure, particularly targeting larger pore sizes (100-200 nm) and higher overall porosity, should result in higher mass loadings of MnO2 and utilization of more internal pore volume. Similar changes in deposition conditions such as solution precursor concentration, deposition time, or subsequent heat treatment may influence the structural and electrochemical properties of the MnO2 phase. Strategies to tune the crystal structure of these nanoscale deposits may also improve the achievable specific capacitance for such electrode structures.47 In addition to applicability in EC devices, this methodology could be extended to design self-limiting deposition methods for other electrochemical applications such as fuel cells. Acknowledgment. Financial support for this research was provided by the Office of Naval Research. A.E.F. acknowledges a National Research Council postdoctoral fellowship (2006-2007). Supporting Information Available: Figures showing MnO2 mass uptake, scanning transmission electron microscopic elemental maps of Mn, C, and O, galvanostatic charge-discharge measurements, and X-ray photoelectron spectra of the O 1s region for neutral-deposited MnO2carbon nanofoam after polarization. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Wu, M.; Snook, G. A.; Chen, G. Z.; Fray, D. J. Electrochem. Commun. 2004, 6, 499-504. (2) Ma, S.-B.; Lee, Y.-H.; Ahn, K.-Y.; Kim, C.-M.; Oh, K.-H.; Kim, K.-B. J. Electrochem. Soc. 2006, 153, C27-C32. (3) Huang, X.; Yue, H.; Attia, A.; Yang, Y. J. Electrochem. Soc. 2007, 154, A26-A33. (4) Dong, X.; Shen, W.; Jinlou, G.; Xiong, L.; Zhu, Y.; Lu, H.; Shi, J. J. Phys. Chem. B 2006, 110, 6015-6019. (5) Ma, S.-B.; Ahn, K.-Y.; Lee, E.-S.; Oh, K.-H.; Kim, K.-B. Carbon 2007, 45, 375-382. (6) Long, J. W.; Dening, B. M.; McEvoy, T. M.; Rolison, D. R. J. NonCryst. Solids 2004, 350, 97-106. (7) Chabre, Y.; Pannetier, J. Prog. Solid State Chem. 1995, 23, 1-130. (8) Thackeray, M. Prog. Solid State Chem. 1997, 25, 1-71. (9) Lee, H. Y.; Goodenough, J. B. J. Solid State Chem. 1999, 144, 220223. (10) Long, J. W.; Young, A. L.; Rolison, D. R. J. Electrochem. Soc. 2003, 150, A1161-A1165. (11) Nam, K.-W.; Kim, K.-B. J. Electrochem. Soc. 2006, 153, A81-A88. (12) Hu, C.-C.; Wang, C.-C. J. Electrochem Soc. 2003, 150, A1079A1084. (13) Reddy, R. N.; Reddy, R. G. J. Power Sources 2004, 132, 315-320. (14) Nagarajan, N.; Humandi, H.; Zhitomirsky, I. Electrochim. Acta 2006, 51, 3039-3045.

286

(15) Ye, C.; Lin, Z. M.; Hui, S. Z. J. Electrochem. Soc. 2005, 152, A1272-A1278. (16) Wu, M.-S. Appl. Phys. Lett. 2005, 87, 153102-153102. (17) Subramanian, V.; Zhu, H.; Wei, B. J. Power Sources 2006, 159, 361-364. (18) Hu, C.-C.; Tsou, T.-W. J. Power Sources 2003, 115, 179-186. (19) Chang, J.-K.; Tsai, W.-T. J. Electrochem. Soc. 2003, 150, A1333A1338. (20) Pang, S.-C.; Anderson, M. A. J. Mater. Res. 2000, 15, 2096-2106. (21) Djurfors, B.; Broughton, J. N.; Brett, M. J.; Ivey, D. G. J. Power Sources 2006, 156, 741-747. (22) Shinomiya, T.; Gupta, V.; Muira, N. Electrochim. Acta 2006, 51, 4412-4419. (23) Zhou, Y. K.; Toupin, M.; Be´langer, D.; Brousse, T.; Favier, F. J. Phys. Chem. Solids 2006, 67, 1351-1354. (24) Chin, S.-F.; Pang, S.-C.; Anderson, M. A. J. Electrochem. Soc. 2002, 149, A379-A384. (25) Devaraj, S.; Munichandraiah, N. Electrochem. Solid-State Lett. 2005, 8, A373-A377. (26) Chang, J.-K.; Tsai, W.-T. J. Electrochem. Soc. 2005, 152, A2063A2068. (27) Zheng, J. P.; Cygan, P. J.; Jow, T. R. J. Electrochem. Soc. 1995, 142, 2699-2703. (28) Pang, S.-C.; Anderson, M. A.; Chapman, T. W. J. Electrochem. Soc. 2000, 147, 444-450. (29) Broughton, J. N.; Brett, M. J. Electrochim. Acta 2004, 49, 44394446. (30) Toupin, M.; Brousse, T.; Be´langer, D. Chem. Mater. 2004, 16, 31843190. (31) Wang, G.-X.; Zhang, B.-L.; Yu, Z.-L.; Qu, M.-Z. Solid State Ionics 2005, 176, 1169-1174. (32) Fan, Z.; Chen, J.; Wang, M.; Cui, K.; Zhou, H.; Kuang, Y. Diamond Relat. Mater. 2006, 15, 1478-1483. (33) Subramanian, V.; Zhu, H.; Wei, B. Electrochem. Commun. 2006, 8, 827-832. (34) Raymundo-Pin˜ero, E.; Khomenko, V.; Frackowiak, E., Be´guin. F. J. Electrochem. Soc. 2005, 152, A229-A235. (35) Lee, C. Y.; Tsai, H. M.; Chuang, H. J.; Li, S. Y.; Lin, P.; Tseng, Y. T. J. Electrochem. Soc. 2005, 152, A716-A720. (36) Wu, Y.-T.; Hu, C.-C. J. Electrochem. Soc. 2004, 151, A2060-A2066. (37) Pekala, R. W.; Farmer, J. C.; Alivaso, C. T.; Tran, T. D.; Mayer, S. T.; Miller, J. M.; Dunn, B. J. Non-Cryst. Solids 1998, 225, 74-80. (38) Perez-Bentio, J. F.; Arias, C.; Brillas, E. Int. J. Chem. Kinet. 1990, 22, 261-287. (39) Kova´cs, K. A.; Gro´f, P.; Burai, L.; Riedel, M. J. Phys. Chem. A 2004, 108, 11026-11031. (40) Long, J. W.; Dunn, B.; Rolison, D. R.; White, H. S. Chem. ReV. 2004, 104, 4463-4492. (41) Lopez de Mishima, B. A.; Ohtsuka, T.; Konno, H.; Soto, N. Electrochim. Acta 1991, 36, 1485-1489. (42) Lopez de Mishima, B. A.; Ohtsuka, T.; Soto, N. Electrochim. Acta 1993, 38, 341-347. (43) Kanoh, H.; Tang, W.; Makita, Y.; Ooi, K. Langmuir 1997, 13, 68456849. (44) Abou-El-Sherbini, K. S.; Askar, M. H. J. Solid State Electrochem. 2003, 7, 435-441. (45) Kuo, S.-L.; Wu, N.-L. J. Electrochem. Soc. 2006, 153, A1317A1324. (46) Chun, S.-E.; Pyun, S.-I.; Lee, G.-J. Electrochim. Acta 2006, 51, 6479-6486. (47) Brousse, T.; Toupin, M.; Dugas, R.; Athoue¨l, L.; Crosnier, O.; Be´langer, D. J. Electrochem. Soc. 2006, 153, A2171-A2180. (48) McEvoy, T. M.; Long, J. W.; Smith, T. J.; Stevenson, K. J. Langmuir 2006, 22, 4462-4466. (49) Song, H.-K.; Jung, Y.-H.; Lee, K.-H.; Dao, L. H. Electrochim. Acta 1999, 44, 3513-3519.

NL062263I

Nano Lett., Vol. 7, No. 2, 2007