18578
J. Phys. Chem. C 2008, 112, 18578–18583
Electrocatalysis of Template-Electrosynthesized Cobalt-Porphyrin/Polyaniline Nanocomposite for Oxygen Reduction Qin Zhou,† Chang Ming Li,*,† Jun Li,† and Juntao Lu‡ School of Chemical and Biomedical Engineering & Center for AdVanced Bionanosystems, Nanyang Technological UniVersity, 70 Nanyang DriVe, Singapore 637457, Singapore, and College of Chemistry and Molecular Sciences, Wuhan UniVersity, Wuhan 430072, China ReceiVed: August 30, 2008; ReVised Manuscript ReceiVed: October 6, 2008
Nanocomposite cobalt-porphyrin/polyaniline was one step-synthesized on a glassy carbon electrode in a solution containing aniline and cobalt-porphyrin, in which J-aggregates were assembled from cobalt-porphyrin and functioned as templates for polyaniline electropolymerization. The nanocomposite thus formed showed porous structures composed of nanorods that were 30∼50 nm in diameter and about a 0.5 µm in length. The nanocomposite -coated electrode was able to catalyze oxygen reduction with the number of electrons involved being close to 4. In 1 M HCl, the nanocomposite electrode performed comparably to a platinum disk electrode, demonstrating its potential as a non-novel metal electrocatalyst for oxygen reduction reaction. Introduction Oxygen reduction has been intensively studied in the history of chemistry.1 In recent decades, the research activities on the catalysts for oxygen reduction were mainly driven by the urgent demands of fuel cells which are believed to become a key technique in the future “hydrogen era”.2 The proton exchange membrane based fuel cells (PEMFCs) are the most advanced low temperature fuel cell today and have been tested and demonstrated on various vehicles and other applications. In PEMFC, platinum (Pt) is used as the catalyst on the cathode (oxygen electrode) with a much higher loading than that on the anode (hydrogen electrode) because of the very sluggish kinetics of oxygen reduction in comparison to the hydrogen oxidation. It is well-known that Pt is not only expensive but also very scarce in resources. As a result, the so-far exclusive dependence of PEMFC on Pt has been one of the major obstacles to the commercialization of this attractive new energy technique.2,3 Promoting Pt utilization and improving Pt catalyst stability will certainly reduce the cost of fuel cells and important progresses have been achieved in past decades.4 However, the ultimate and ideal solution would be replacing Pt with non-Pt catalysts which has attracted increasing research interests.5,6 In addition to fuel cells, oxygen reduction electrodes also have found many other applications, such as in metal-air (oxygen) batteries7 and sensors.8 In these applications, the electrochemical devices are often disposable and Pt should be excluded. Pt is usually considered to be the best catalyst for oxygen reduction, but it, in fact, does not perform as well as expected in many practical situations. Pt has been found to be easily contaminated and to quickly lose activity for oxygen reduction. It is generally observed that the current of oxygen reduction on Pt electrodes at an applied potential in sulfuric acid or potassium hydroxide solutions decays very fast with time, unless the solutions have been specially purified. All of the factors discussed above fuel up the studies of non-Pt catalysts for oxygen reduction. * To whom correspondence should be addressed. Tel: 65 67904485. Fax: 65 67911761. E-mail:
[email protected]. † Nanyang Technological University. ‡ Wuhan University.
Transition metal macrocyclic compounds,5,6,9-14 especially metalloporphyrins,5,11-14 form a major category of non-Pt catalysts for oxygen reduction. Most metallomacrocyclics catalyze only the so-called two-electron (2e) oxygen reduction to hydrogen peroxide,11,12 some specially designed porphyrins containing two properly arranged cobalt centers are able to catalyze the so-called four-electron (4e) oxygen reduction for directly producing water (or OH- in alkaline solutions).13,14 On the other hand, conducting polymers are widely tested as supports of different catalysts15,16 or as the catalysts for oxygen reduction.17 Inourpreviouswork,18 J-aggregatesofcobalt-porphyrin were used as templates in electropolymerization of pyrrole to form cobalt-porphyrin/polypyrrole nanocomposites with special morphologies. One of them showed good activity for 4e reduction of oxygen in neutral solution. However, further exploration in applications indicates that the cobalt-porphyrin/ polypyrrole nanocomposite suffers from its instability, which will be discussed later. In the present paper, we report an extension of the J-aggregates templated conducting polymers for oxygen reduction, i.e., the cobalt-porphyrin/polyaniline nanocomposite in HCl solutions. Experimental Section Chemicals. All chemicals were of analytical grade. 5,10,15,20Tetrakis(4-sulfophenyl)-21H,23H-porphine (TPPS) was purchased from TCI Co., Japan. Aniline and cobalt(II) acetate were purchased from Aldrich Co., USA, and used as received. Deionized water was produced by a Millipore water purification system (Q-Grad 1, Millipore Co., USA). TPPS-Co Synthesis and Characterization. Cobalt-5,10,15,20tetrakis(4-sulfophenyl)-21H,23H-porphine (TPPS-Co) was prepared in our laboratory according to the method described in literature.18 The product was characterized with UV-vis and AFM. In UV-vis measurements, 1 mM TPPS-Co in 1 M HCl aqueous solution was tested on a Hitachi U-2800 spectroscope equipped with a quartz cell of 1 cm optical length. For AFM observation, a drop of TPPS-Co solution (1 mM in 1 M HCl) was applied onto a freshly cleaved mica surface and dried in ambient air. The deposited TPPS-Co was then characterized by AFM (Dimension 3100 SPM, Veeco, USA).
10.1021/jp8077375 CCC: $40.75 2008 American Chemical Society Published on Web 11/04/2008
Cobalt-Porphyrin/Polyaniline Nanocomposite The TPPS-Co/PANI composite was one-step electrochemically synthesized on a glassy carbon (GC) electrode (diameter 3 mm) by cyclic voltammetry (CV) from -0.2 to +1.0 V (vs SCE) at 50 mV/s in a three-electrode electrochemical cell containing 0.1 M aniline + 1 mM TPPS-Co + 1 M HCl. For morphology observation, a piece of gold-coated silicon wafer was used as the substrate for electropolymerization and the specimen was examined under a field emission scanning electron microscope (FESEM, JEOL JSM-6700F, Japan). For comparison, pure PANI was also electrochemically polymerized under the same conditions in the absence of TPPS-Co. Electrochemical Measurements. The synthesized materials were characterized by cyclic voltammetry (CV) in the background electrolyte (1 M HCl) and tested for catalytic activity toward oxygen reduction reaction (ORR) in oxygen saturated solution. All of the electrochemical measurements were performed using a CHI 760B bipotentiostat (CH Instruments, USA) in a three-electrode electrochemical cell. The working electrode (WE) was a GC disk electrode. A Pt wire and a saturated calomel electrode (SCE) were used as the counter electrode (CE) and the reference electrode (RE), respectively. Before each experiment, the WE surface was polished sequentially with fine grade alumina powders, from 1 to 0.3 to 0.05 µm, to obtain a mirror-like surface. Then the WE was sonicated to remove traces of alumina from the surface, washed with distilled water and acetone, and dried in air. Prior to CV characterization, the solution was deaerated by bubbling nitrogen (99.99%) for at least 15 min, and a gentle nitrogen stream was maintained over the solution during the measurement to prevent the solution from being contaminated by oxygen. For evaluating the catalytic activity for oxygen reduction, oxygen (99.99%) was bubbled into the cell for 15 min before measurement and flushed over the solution during measurement. In order to estimate the electron number involved, the oxygen reduction was studied with rotating Pt ring-GC disk electrode (RRDE, Pine Model AFMT28) with collection efficiency 22%. The potential of the disk electrode was scanned while the ring potential was kept at 1.0V (vs SCE) at which hydrogen peroxide can be easily detected by electrochemical oxidation. All measurements were carried out at room temperature (20 ( 2 °C).
J. Phys. Chem. C, Vol. 112, No. 47, 2008 18579
Figure 1. UV-vis absorption spectrum of TPPS-Co (1 mM in 1 M HCl).
Results and Discussion Characterizations of TPPS-Co J-Aggregates. The UV-vis absorption spectrum of the synthesized TPPS-Co (1 mM in 1 M HCl) is shown in Figure 1. As can be seen, in addition to the intense band around 430 nm (Soret band) characteristic of porphyrin ring, there are two peaks around 490 and 710 nm, respectively, attributable to the formation of J-aggregates.18-22 In order to provide more direct evidence of the existence of J-aggregates, we applied a small amount of the TPPS-Co solution (1 mM in 1 M HCl) onto a freshly cleaved mica surface and examined the dried deposit under AFM. The rod-like structures with a diameter of about 4∼5 nm similar to the reported J-aggregates of porphyrins22 are clearly seen in the AFM image (Figure 2). Thus, it was confirmed that the J-aggregates of TPPS-Co were formed. The spontaneous formation of the J-aggregates may be attributed to the electrostatic interactions between the negative sulfonic groups of a porphyrin molecule and the positively charged Co(II) ions of neighboring porphyrin molecules as proposed in our previous work.18 Formation, Morphology and Stability of TPPS-Co/PANI Composite. Since TPPS-Co is soluble in water, it could be conveniently incorporated into PANI by electrochemical polymerization in aqueous solutions and the spontaneously formed J-aggregates of TPPS-Co could act as templates to form
Figure 2. AFM image of TPPS-Co aggregates on mica (a) and the height profile along the line shown in the image (b).
nanostructured TPPS-Co/PANI composites. In this work, the TPPS-Co/PANI composite was one-step electrochemically synthesized on a GC electrode by CV method (details are described in the Experimental Section). Figure 3 shows the evolution of CV during the repeated potential scans. There is a prominent anodic peak at positive potentials in the first cycle, indicative of the oxidation of the aniline monomers as initiation of the polymerization. After that, three pairs of redox peaks appear during the successive cycles and the peak intensities increase with time, indicating the formation and growth of a conducting polymer film on the electrode surface. In order to confirm the template effect of TPPS-Co, the TPPSCo/PANI composite electrodeposited on a piece of Au-coated silicon wafer was examined by FESEM. As shown in Figure 4, the composite presents porous structures composed of cemented
18580 J. Phys. Chem. C, Vol. 112, No. 47, 2008
Figure 3. Cyclic voltammogram of TPPS-Co/PANI composite electrodeposition, scan rate is 50 mV/s.
nanorods with diameters of 30∼50 nm, which is completely different from the structures of pure PANI synthesized under the same conditions in the absence of TPPS-Co. The increase in thickness of the rod in the composite in comparison to that in the templates (Figure 2a) can mainly be attributed to the deposited polymer on the rod surfaces. However, further assembly of J-aggregates during electropolymerization was also possible and contributed to the increased rod thickness to some extent. The stability of the TPPS-Co/PANI nanocomposite was evaluated by CV. Figure 5 shows the CVs of the nanocompositemodified GC electrode in 1 M HCl right after the polymerization and after being stored for 72 h in the solution at open circuit potential. It can be seen that the peak area of the redox current changed only slightly. This is in contrast to the situation of similar composite TPPS-Co/PPY (ref 18) which lost its redox activity after being stored overnight. The result may indicate that the active centers, TPPS-Co can be secured in the PANI matrix but may leach out from the PPY film. Oxygen Reduction on TPPS-Co/PANI Nanocomposite Modified Electrode. The electrocatalytic activity of the TPPSCo/PANI nanocomposite for oxygen reduction was evaluated in 1 M HCl. Figure 6a shows the linear sweep voltammograms of the TPPS-Co/PANI nanocomposite modified GC electrode in N2- and O2-saturated 1 M HCl, respectively. The current-potential curves were obtained during the negative-going linear potential scan. In N2-saturated HCl, the current was caused by the reduction of the oxidized TPPS-Co/PANI nanocomposite. When the solution was saturated with oxygen, a larger cathodic current appeared, corresponding to the reduction of both oxidized TPPS-Co/PANI itself and oxygen in the solution. In order
Zhou et al.
Figure 5. Cyclic voltammograms of the TPPS-Co/PANI nanocomposite modified GC electrode in 1 M HCl right after the polymerization (solid line) and after being stored for 72 h (dashed line).
to see the current of oxygen reduction more clearly, a correction for the polymer reduction current was performed by subtracting curve i from curve ii. The resulting curve iii represents approximately the net current of oxygen reduction alone and shows a peak potential round 0.16 V (vs SCE), which is remarkably more positive than the corresponding potential -0.33 V for PANI modified electrode (Figure 6b). On the other hand, the bare GC electrode showed a peak potential around -0.4 V for oxygen reduction in the same solution. These results prove that TPPS-Co/PANI nanocomposite possessed a rather high activity toward oxygen reduction in 1 M HCl. Oxygen reduction is a multielectron complicated reaction and its mechanism changes depending on a number of factors. These mechanisms are usually classified into two main categories, i.e., two-electron (2e) and direct four-electron (4e) mechanisms.1 In the 2e mechanism, oxygen is reduced by a transfer of 2 electrons to form hydrogen peroxide without breaking the O-O bond. In contrast, in the direct 4e mechanism, the O-O bond is broken down and 4 electrons are transferred to form water molecules (in acid solutions). There is another possibility that the peroxide produced by the 2e mechanism can be further reduced to water. In this case, the peroxide is an intermediate and the total number of electrons for oxygen reduction is also 4. This mechanism may be called “indirect 4e mechanism” to distinguish it from the “direct 4e mechanism” in which no peroxide intermediate is involved. In real cases, the above-mentioned mechanisms often proceed in parallel at a certain ratio which changes with experimental conditions, including electrode potential. As a
Figure 4. FESEM images of TPPS-Co/PANI nanocomposite (left) and pure PANI (right).
Cobalt-Porphyrin/Polyaniline Nanocomposite
J. Phys. Chem. C, Vol. 112, No. 47, 2008 18581
Figure 6. Linear sweep voltammograms of (a) TPPS-Co/PANI nanocomposite and (b) pure PANI modified GC electrodes. Curve i and ii are taken in N2- and O2-saturated 1 M HCl, respectively (scan rate is 50 mV/s), curve iii is obtained by subtracting curve i from curve ii.
Figure 7. Koutecky-Levich plot of the plateau current in rotating disk voltammograms for oxygen reduction on the TPPS-Co/PANI nanocomposite modified electrode in O2-saturated 1 M HCl, dash lines are calculated for mass-transport controlled 2-electron and 4-electron reduction of oxygen.
result, the whole reaction is governed by mixed mechanisms and the measured electron number varies between 2 and 4. The n value is usually estimated by two ways. The simpler one is the so-called Koutecky-Levich plot using a rotating disk electrode. In this method, the reciprocal plateau current (ID-1) is plotted against the reciprocal of the rotating rate (ω-1) and the n value is estimated from the slope. Figure 7 shows the Koutecky-Levich plot of oxygen reduction at the TPPS-Co/ PANI nanocomposite coated GC disk electrode. By comparing the experiment line to the two calculated lines for n ) 2 and 4, the n value for the electrode under study was found to be very close to 4, indicating that the oxygen reduction at the composite coated electrode proceeded mainly through the 4e mechanism. A more sophisticated method to estimate n is the rotating ring-disk electrode (RRDE) technique, in which the peroxide produced at the disk electrode is detected by the ring electrode and the n value is calculated from the ratio of the ring current (IR) to the disk current (ID). In the literature, two equations have been reported for the calculation: n ) 4 - 2(IR/NID)13,23 and n ) 4/(1 + IR/NID).24,25 The former equation may be regarded as a linear approximation to the latter, which is strict. Figure 8a shows the RRDE data for the nanocomposite coated GC disk electrode and the change of n with disk potential. The n value
increases as the potential moving to more negative and levels off at 3.7 in reasonable agreement with the result of KouteckyLevich plot (Figure 7). The observed near 4e oxygen reduction indicated that most oxygen molecules participating in the reaction were reduced to water with only a small fraction of hydrogen peroxide production (ca. 6%). However, the n value alone cannot serve as evidence for either the presence or absence of the direct 4e mechanism. The water production might proceed via the formation of hydrogen peroxide as an intermediate, i.e., through an indirect 4e mechanism. As recently pointed by Dahn et al.,26 at least for some catalysts, the n value calculated from RRDE data would increase with catalyst loading from 2 at low loadings to near 4 at high loadings. This is because on these catalysts the peroxide can be further reduced or decomposed to water and oxygen if it has enough time staying on the electrode surface before escaping to the bulk solution. According to the FESEM image of TPPS-Co/PANI nanocomposite (Figure 4), it seems highly possible for the peroxide to stay in the nanostructures for certain duration and, therefore, to be further reduced to water before leaving the electrode to the bulk solution. On the other hand, the direct 4e mechanism cannot be ruled out either. Cobalt-porphyrins, like most macrocyclic transition metal compounds, mainly catalyze 2e oxygen reduction to hydrogen peroxide.27 However, when two Co centers in porphyrin are arranged at an appropriate distance so as to adsorb an oxygen molecule in the bridge mode, direct 4e reduction becomes possible.13,14 In this work, the distance between two adjacent Co centers in the J-aggregate of TPPS-Co is certainly too large to allow the bridge mode of oxygen adsorption. However, if we assume that oxygen molecules can be adsorbed in a bridge mode between a Co center in porphyrin and a site in the neighboring PANI, direct 4e reduction might become possible. This assumption remains for further study. However, the nanocomposite catalyst does produce the efficient 4e reduction effect regardless of the “direct” or “indirect” path. For oxygen reduction reaction, the number of electrons is important but the working potential at which ORR proceeds at a reasonable rate is as important. In fuel cells and air batteries, the oxygen cathode should have a high working potential, otherwise the energy conversion efficiency will be low, even with 4e oxygen reduction. It is therefore significant to compare the potential of oxygen reduction for different catalysts. A
18582 J. Phys. Chem. C, Vol. 112, No. 47, 2008
Zhou et al.
Figure 8. (a) Rotating ring-disk electrode voltammogram for oxygen reduction at TPPS-Co/PANI nanocomposite coated disk electrode in O2saturated 1 M HCl (rotation rate 1000 rpm, scan rate 50 mV/s) and n value as a function of disk potential. (b) Rotating ring-disk electrode voltammogram for oxygen reduction at Pt disk electrode (all the experimental conditions are same as TPPS-Co/PANI) and n value as a function of disk potential.
convenient way to compare the aspect of working potential for different catalysts is to compare the so-called half-wave potential, a potential at which the current is a half of the limiting current. For a meaningful comparison of half-wave potential, the reversible hydrogen electrode (RHE) in the same solution as the test solution should be adopted to cancel out the thermodynamic effect caused by the pH difference among different reported works. It is interesting to see from Figure 8 that the half-wave potential of a Pt disk electrode in 1 M HCl was slightly lower than that of the nanocomposite modified electrode, 0.43V versus 0.45V (vs RHE), though the n value of the Pt electrode was slightly higher, 3.9 versus 3.7. In sulfuric acid or perchloric acid, the half-wave potential of ORR on Pt RDE is around 0.8 V (vs RHE).28,29 The unusually low halfwave potential found in HCl for Pt should be attributed to the strong adsorption of chloride anions. The anion effect on ORR on Pt can also be seen in a reported activity sequence in different electrolytes: KOH > H2SO4 ≈ CF3SO3H > H3PO4 (The lowest activity in HClO4 reported in the same reference seems questionable and is omitted).30 Thus, the activity of the TPPSCo/PANI nanocomposite modified electrode for ORR in 1 M HCl is comparable to that of a Pt disk electrode at least. Conclusions For the first time, our work demonstrated that J-aggregates of TPPS-Co could be spontaneously formed in an acidic aqueous solution and be used as templates in electropolymerization of PANI. The one-step electrosynthesized TPPS-Co/PANI nanocomposite showed porous structures consisting of cemented nanorods 30∼50nm in diameter and about a half-micrometer in length. The nanocomposite modified electrode showed significant electrocatalytic activity for the oxygen reduction in acidic solution with the number of electrons close to 4. In 1 M HCl, the activity of the nanocomposite modified GC electrode is comparable to that of a Pt disk electrode in terms of both the number of electrons involved and the electrode potential. These features make the nanocomposite attractive for some applications, especially as an oxygen reduction electrode in HCl
solutions. The working mechanism of this special material is worth further studying. Acknowledgment. The authors are grateful to U.S.A. Army Research Office under Contract No. W911NF-05-1-0303 for financial support to this work. References and Notes (1) Tarasevich, M. R.; Sadkowski, A.; Yeager, E. In ComprehensiVe Treatise of Electrochemistry; Conway, B. E., Bockris, J. O., Yeager, E., Khan, S. U. M., White, R. E., Eds.; Plenum Press: New York, 1983; Vol. 7, p 301. (2) Vielstich, W., Lamm, A., Gasteiger, H. A., Eds.; Handbook of Fuel Cells: Fundamentals, Technology and Application; Wiley: West Sussex, 2003. (3) Gottesfeld, S.; Zawodzinski, T. A. In AdVances in Electrochemical Science and Engineering; Alkire, R. C., Kolb, D. M., Eds.; Wiley-VCH: Weinheim, Germany, 1997; Vol. 5, p 195. (4) Zhang, J.; Sasaki, K.; Sutter, E.; Adzic, R. R. Science 2007, 315, 220. (5) Wang, B. J. Power Souces 2005, 152, 1. (6) Zhang, L.; Zhang, J.; Wilkinson, D. P.; Wang, H. J. Power Souces 2006, 156, 171. (7) Dewi, E. L.; Oyaizu, K.; Nishide, H.; Tsuchida, E. J. Power Sources 2004, 130, 286. (8) Stetter, J. R.; Li, J. Chem. ReV. 2008, 108, 352. (9) Li, C. M.; Wang, Z.; Cha, C. S. J. Wuhan UniV. Technol. 1983, 4, 129. (10) Lin, A. S.; Huang, J. C. J. Electroanal. Chem. 2003, 541, 147. (11) Yuasa, M.; Nagaiwa, T.; Kato, M.; Sekine, I.; Hayashi, S. J. Electrochem. Soc. 1995, 142, 2612. (12) Liu, H.; Zhang, L.; Zhang, J.; Ghosh, D.; Jung, J.; Downing, B. W.; Whittemore, E. J. Power Sources 2006, 161, 743. (13) Collman, J. P.; Denisevich, P.; Konai, Y.; Marrocco, M.; Koval, C.; Anson, F. C. J. Am. Chem. Soc. 1980, 102, 6027. (14) Chang, C. J.; Deng, Y.; Shi, C.; Chang, C. K.; Anson, F. C.; Nocera, D. G. Chem. Commun. 2000, 1355, n/a. (15) Coutanceau, C.; Hourch, A. El.; Crouigneau, P.; Leger, J. M.; Lamy, C. Electrochim. Acta 1995, 40, 2739. (16) Coutanceau, C.; Croissant, M. J.; Napporn, T.; Lamy, C. Electrochim. Acta 2000, 46, 579. (17) Khomenko, V. G.; Barsukov, V. Z.; Katashinskii, A. S. Electrochim. Acta 2005, 50, 1675. (18) Zhou, Q.; Li, C. M.; Li, J.; Cui, X. J. Phys. Chem. C 2007, 111, 11216.
Cobalt-Porphyrin/Polyaniline Nanocomposite (19) Kvarnstrom, C.; Neugebauer, H.; Blomquist, S.; Ahonen, H. J.; Kankare, J.; Ivaska, A. Electrochim. Acta 1999, 44, 2739. (20) Cosnier, S.; Gondran, C.; Wessel, R.; Montforts, F. P.; Wedel, M. J. Electroanal. Chem. 2000, 488, 83. (21) Cong, H. N.; Abbassi, K. El.; Chartier, P. Electrochem. Solid-State Lett. 2000, 3, 192. (22) Rotomskis, R.; Augulis, R.; Snitka, V.; Valiokas, R.; Liedberg, B. J. Phys. Chem. B 2004, 108, 2833. (23) Kingsborough, R. P.; Swager, T. M. Chem. Mater. 2000, 12, 872. (24) Jakobs, R. C. M.; Janssen, L. J. J.; Barendrecht, E. Electrochim. Acta 1985, 30, 1085. (25) Antoine, O.; Durand, R. J. Appl. Electrochem. 2000, 30, 839.
J. Phys. Chem. C, Vol. 112, No. 47, 2008 18583 (26) Bonakdarpour, A.; Lefevre, M.; Yang, R.; Jaouen, F.; Dahn, T.; Dodelet, J.; Dahn, J. R. Electrochem. Solid State Lett. 2008, 11, B105. (27) Durand, R. R., Jr.; Anson, F. C. J. Electroanal. Chem. Interfacial Elecrrochem. 1982, 134, 273. (28) Halseid, R.; Bystron, T.; Tunold, R. Electrochim. Acta 2006, 51, 2737. (29) Markovic, N. M.; Adzic, R. R.; Cahan, B. D.; Yeager, E. B. J. Electroanal. Chem. 1994, 377, 249. (30) Hsueh, K. L.; Gonzales, E. R.; Srinivasan, S. Electrochim. Acta 1983, 28, 691.
JP8077375
