Electrocatalysis DOI 10.1007/s12678-015-0255-0
LETTER
The Effect of the Voltage Scan Rate on the Determination of the Oxygen Reduction Activity of Pt/C Fuel Cell Catalyst Nejc Hodnik 1 & Claudio Baldizzone 1 & Serhiy Cherevko 1 & Aleksandar Zeradjanin 1 & Karl J. J. Mayrhofer 1
# Springer Science+Business Media New York 2015
Introduction The oxygen reduction reaction (ORR) is one of the most important chemical reactions. Besides others, it plays a crucial role for the development of a sustainable energy scenario, as it is the decisive reaction in proton exchange membrane fuel cell (PEMFC) [1, 2]. Improving the catalysis of the ORR can lead to a breakthrough in electrochemical energy conversion; therefore, the fundamental understanding of the reaction processes as well as the rapid assessment of the kinetic activity of different catalyst materials is of utmost important [3]. So far, it is well established that the surface coverage of electrosorbed oxygenated species like H2O, OH, and O determines platinum ORR activity, although the true nature of the potential dependent oxygenated species has yet to be resolved [4, 5]. Under the assumption that the Pt surface is almost fully covered at low overpotentials also referred as kinetic region, where Θad is the surface coverage of any adsorbate species, the ORR kinetic current becomes directly proportional to the pre-exponential factor (1-Θad) of the rate expression [6]. This know-how has helped to interpret effects introduced for instance by the particle size of Pt catalysts [7–9] or the
* Nejc Hodnik
[email protected] Karl J. J. Mayrhofer
[email protected] 1
Department of Interface Chemistry and Surface Engineering, Max-Planck-Institut für Eisenforschung GmbH, Max-Planck-Straße 1, 40237 Düsseldorf, Germany
development of catalysts with enhanced activities like Ptalloys with lower Θad compared to pure Pt [5, 6, 10–13]. The advancement in ORR fundamental understanding and performance evaluation has benefited to a great extent from informative half-cell kinetic survey studies of applied catalysts, as for instance performed with the thin-film rotating disk electrode technique (TF-RDE) [2, 14–16]. Such electrochemical model studies enable the determination of true kinetic current densities even of porous materials without complex interference with mass-transport or catalyst utilization effects, which are additional important factors for the final performance in electrochemical reactors [17]. In order to perform these ORR half-cell measurements accurately, certain practical guidelines have been shown to be essential. These include especially the cleanliness of the experimental setup [18], catalyst film preparation on the electrode [14], appropriate potentiostat sampling mode, evaluation of kinetic data [15], IR compensation [19], correction for true reversible hydrogen electrode (RHE) potential, and background currents, later especially at high scan rates [15]. Interestingly, although already reported in literature for polycrystalline and high surface area Pt catalysts [2, 16, 20], the effect of the voltage scan rate on the activity determination is often underestimated or not clearly separated from the effect of impurities. In the present work, the impact of the voltage scan rate on the activity determination of a high surface area Pt catalyst is systematically studied. We take special care in order to work extra clean and avoid the effect of impurities. We confirm that the relatively slow platinum surface oxidation process significantly influences the kinetic evaluation [20, 21], which is one reason behind varying ORR-specific activities in literature. Moreover, we describe the scan rate-dependent influence of chloride ions, as well as discuss the relationship to surface coverage and the implications of the results for practical applications.
Electrocatalysis
Materials and Methods A three-compartment electrochemical Teflon cell with a graphite rod and an Ag/AgCl, 3 M KCl (Metrohm) housed in a Tschurl modification [18] were used as counter and reference electrode, respectively. The 0.1 M HClO4 electrolyte solution was prepared using ultrapure water (18 MΩ, Millipore®) and analytical grade chemicals (Suprapur®, Merck). A Gamry Reference 600 potentiostat was used for biasing. Electrochemical experiments were run utilizing a home-written lab-view program, which enables fully automated and thus highly reproducible measurements [22]. A platinum high surface area catalyst with a metal loading of 50.8 % on carbon and a mean particle size of 4.8 nm (denoted as BPt 5 nm^) supplied by Tanaka Kikinzoku Kogyo (Japan) was utilized as a reference catalyst. The catalyst was suspended in water in various concentrations, and aliquots of 20 μL were drop casted on glassy carbon discs of 5-mm diameter (area 0.196 cm2) to achieve the desired loadings. The real Pt surface area was measured via a CO stripping protocol at 50 mV s−1. The activity for the ORR, the correction for the background from double-layer capacity and adsorption processes, and the normalization to the active surface area was performed as reported several times before [7].
Results and Discussion Standard activity measurements for the ORR on the Pt 5-nm catalyst are performed in oxygen-purged 0.1 M HClO4. The background-corrected cyclic voltammograms show the expected behavior for Pt, with a diffusion-limited current below 0.7 VRHE depending on the electrode rotation rate, a kinetic region in between 0.7 and 1.0 VRHE, and a hysteresis between positive and negative going scan direction in the kinetic region due to irreversible surface oxidation (Fig. 1a). The LevichKoutecky plot in the insert of Fig. 1b confirms the number of transferred electrons in the ORR as 4.0. The specific activity (SA) in the kinetic region obtained after surface normalization and mass-transport correction is independent from the rotation rates (Fig. 1b), which is a first indication that minority of species possibly present in the electrolyte do not affect the measurements. Moreover, as summarized in Fig. 2c, the SA is also independent from the catalyst loading within roughly 10 %, which would not be the case if spectator species (impurities) have an impact on the activity determination [23, 24]. Finally, a comparison to SA values from literature confirms that the effect of impurities is negligible in these experiments [7, 8], which sets the basis for the further study. Figure 2a shows a positive potential shift of ORR polarization curves with the increase of scan rate during both, the positive and the negative scan. The SA for the ORR derived from Tafel plots of the positive sweep demonstrates an activity
Fig. 1 a Background-corrected ORR polarization curves of Pt/C 5 nm with a loading of 20 μgPt cm−2 measured with 10 mV s−1 in 0.1 M HClO4 at different rotation rates. b Tafel plot of curves in a, with a LevichKoutecky plot at 0.4 VRHE in the inset
gain from 0.28 to 0.61 mA cm−2Pt when increasing from 5 to 200 mV s−1 at 0.9 VRHE. The scan rates affect not only the ORR but also the cyclic voltammograms in Ar-purged electrolyte, as seen in the normalized currents in Fig. 2b. While the Hupd region remains constant as expected for fast adsorption/ desorption processes, the surface oxidation and reduction charge significantly decrease with an increase in the scan rate. This suggests that the electrosorption of O/OH on Pt (note that the exact potential dependent nature of this process is not yet resolved) contains a rather slow step and the surface condition does not reach equilibrium during a poteniodynamic study. This was already noticed in the study of polycrystalline Pt [20, 25] and in real fuel cell measurement [21]. The same is true for the reduction of the surface, which is confirmed by the increase in extent of the positive potential shift of the electrochemically irreversible peak in the negative scan with decreasing scan rate. The reason why we observe lower ORR activity for slower scan rates in the negative scan is however the increased amount of oxide that grows at longer exposures to oxidative potentials (Fig. 2b). The specific ORR activity at
Electrocatalysis
from the scan rate, implying an inverse, direct relationship between SA and oxide coverage. This is well in line with the common perception of the ORR limitations, as a higher coverage with blocking oxygenated species (ΘOH), which is directly correlated to QOH, leads to a decrease in activity [4–6, 9–13]. Note that the decrease in Pt oxidation rate with time [26–28] leads to a gradual decay, but not a complete termination, of the observed trend in increased QOH at slower scan rates. In a further experiment, the electrolyte was deliberately spiked with 10−4 M NaCl, in order to investigate the effect of chloride additives (here referred as impurities) which besides activity also influence Pt catalyst stability [29]. As seen in Fig. 3, this leads to major deviations compared to experiments free of impurities. In contrast to the logarithmic dependence in Fig. 2c, the SA for the ORR scales linearly with the scan rate. This suggests that the surface coverage of Cl− is directly proportional to the scan rate. The longer the accumulation time, the more impurities can reach and specifically adsorb to the surface [30, 31] and thus block the ORR [18,
Fig. 2 a Polarization curves for Pt/C 5 nm with a loading of 20 μgPt cm−2 measured at different scan rates from 5 to 500 mV s−1 in O2-saturated 0.1 M HClO 4 , with Tafel plots in the inset. b Series of cyclic voltammograms of Pt/C 5 nm with loading of 60 μgPt cm−2 in Arsaturated 0.1 M HClO4, normalized to the scan rates. c Summary of the SA for the ORR at 0.9 VRHE obtained at different catalyst loadings (10– 30 μgPt cm−2) and the charge for platinum oxidation obtained from the integration of the positive scan cycles between 0.7 and 0.9 VRHE in Fig. 2b as a function of scan rate
0.9 VRHE and the surface oxidation data extracted from positive scan, arbitrarily denominated by the oxidation charge between 0.7 and 0.9 VRHE that has been corrected for double-layer charging, are summarized in Fig. 2c. It ranges from 40 μC cm−2 for 5 mV s−1 to 68 μC cm−2 for 200 mV s−1 which is in agreement with the literature [8]. Interestingly, both the SA for the ORR and the QOH show logarithmic-type dependence
Fig. 3 a SA for the ORR at 0.9 VRHE for Pt/C 5 nm with loading of 10 μgPt cm2 measured at different scan rates in O2-saturated 0.1 M HClO4 spiked with 10−4 M NaCl. The inset shows the complete Tafel plots of the kinetic region. b Comparison of the ORR SA with and without the addition of 10−4 M NaCl to the electrolyte as a function of scan rate (same loading of 10 μgPt cm2), with the ratio between the respective values indicated by the axis on the right side (red)
Electrocatalysis
23, 24]. This is also supported by the Tafel slope of ca. 120 mV dec−1 typical for a Langmuir isotherm in the presence of Cl−, which renders the heat of adsorption of oxygenated species coverage independent. This is in contrast to the Temkin isotherm conditions with a slope close to 60 mV dec−1 at higher OH coverages in the absence of Cl− impurities [32–34]. As a consequence of this different time dependency, the SA determined at slow scan rates is much more affected by the presence of minor impurities in the electrolyte. Nevertheless, although the ratio between the SA in chloride-free and chloride containing electrolyte significantly drops from 10 to 2 when going from slow to fast scan rates, the effect can never be completely avoided (Fig. 3b). The presented study is not only a further fragment for improving the understanding of the ORR and its limitations but also poses important consequences for the quantitative evaluation of catalysts. The results in Fig. 2 show that the equilibrium surface state has not been reached even at the lowest scan rate of 5 mV s−1; actually, the SA is expected to further decrease logarithmically with timescale. This means that no transient measurement can capture the SA as accurately as a steady-state experiment, and thus, half-cell studies are typically overestimating catalyst activity compared to MEA or fuel cell stack data that is typically obtained over extended times [1, 2, 21]. This however does not necessarily imply that fundamental studies should be performed at steady-state, since then (i) the experiments would become too time-consuming and (ii) the unavoidable influence of minor spectator species/ impurities would become dominant (see Fig. 3). It is sufficient to be aware of this effect and quantitatively compare catalysts only under similar conditions. Note that, although not shown here, the effect is more general and the above also applies to measurements in different electrolytes of various pH as well as to Pt-alloy catalysts, as would be expected from the ORR behavior.
differences in acquired data of more than a factor of 2, which has to be considered when comparing different literature data also on other catalysts in both acidic and alkaline electrolyte. Acknowledgments Nejc Hodnik would like to acknowledge the FP7PEOPLE-2013-IEF Marie Curie Intra-European Fellowship (project 625462 - ElWBinsTEM). Conflict of Interest The authors declare that they have no conflict of interest.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
Conclusions
17.
We report that the SA for the ORR of a reference Pt high surface area electrocatalyst indeed scales with the voltage scan rate in potentiodynamic studies. The reason for this effect is however not the presence of impurities but the time dependence of the establishment of equilibrium conditions at the surface, which is dominated by the relatively slow process of Pt surface oxidation as was already reported by Pasti et. al. for bulk platinum [20]. Lower scan rates allow more time for oxidation, leading to enhanced blockage of active sites for ORR and consequently lower apparent kinetic rates. While it is in principal desirable to study activity as close to equilibrium as possible, a compromise has to be found regarding experiment length and the associated impact of minor impurities in the electrolyte. The extent of the scan rate results in
18. 19. 20. 21. 22. 23. 24. 25. 26.
M.K. Debe, Nature 486, 43 (2012) H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B Environ. 56, 9 (2005) I. Katsounaros, S. Cherevko, A.R. Zeradjanin, K.J.J. Mayrhofer, Angew. Chem. Int. Ed. 53, 102 (2014) N. Markovic, H. Gasteiger, P.N. Ross, J. Electrochem. Soc. 144, 1591 (1997) N.M. Marković, P.N. Ross Jr., Surf. Sci. Rep. 45, 117 (2002) N. Hodnik, M. Bele, S. Hočevar, Electrochem. Commun. 23, 125 (2012) M. Nesselberger, S. Ashton, J.C. Meier, I. Katsounaros, K.J.J. Mayrhofer, M. Arenz, J. Am. Chem. Soc. 133, 17428 (2011) W. Sheng, S. Chen, E. Vescovo, Y. Shao-Horn, J. Electrochem. Soc. 159, B96 (2011) N. Hodnik. Chem. Eng. 41 (2014) V.R. Stamenkovic, B. Fowler, B.S. Mun, G. Wang, P.N. Ross, C.A. Lucas, N.M. Marković, Science 315, 493 (2007) M. Wakisaka, H. Suzuki, S. Mitsui, H. Uchida, M. Watanabe, J. Phys. Chem. C 112, 2750 (2008) J. Zhang, F.H.B. Lima, M.H. Shao, K. Sasaki, J.X. Wang, J. Hanson, R.R. Adzic, J. Phys. Chem. B 109, 22701 (2005) V.S. Murthi, R.C. Urian, S. Mukerjee, J. Phys. Chem. B 108, 11011 (2004) T.J. Schmidt, H.A. Gasteiger, G.D. Stäb, P.M. Urban, D.M. Kolb, R.J. Behm, J. Electrochem. Soc. 145, 2354 (1998) K.J.J. Mayrhofer, D. Strmcnik, B.B. Blizanac, V. Stamenkovic, M. Arenz, N.M. Markovic, Electrochim. Acta 53, 3181 (2008) Y. Garsany, O.A. Baturina, K.E. Swider-Lyons, S.S. Kocha, Anal. Chem. 82, 6321 (2010) A.J. Bard, L.R. Faulkner, Electrochemical methods: fundamentals and applications (Wiley, New York, 2000), pp. 83–136 K.J.J. Mayrhofer, S.J. Ashton, J. Kreuzer, M. Arenz, Int. J. Electrochem. Sci. 4, 1 (2009) D. van der Vliet, D.S. Strmcnik, C. Wang, V.R. Stamenkovic, N.M. Markovic, M.T.M. Koper, J. Electroanal. Chem. 647, 29 (2010) I.A. Pasti, N.M. Gavrilov, S.V. Mentus, Int. J. Electrochem. Sci. 7, 11076 (2012) F.A. Uribe, T.A. Zawodzinski Jr., Electrochim. Acta 47, 3799 (2002) A. A. Topalov, I. Katsounaros, J. C. Meier, S. O. Klemm, K. J. J. Mayrhofer. Rev. Sci. Instrum. 82 (2011) I. Katsounaros, J.C. Meier, K.J.J. Mayrhofer, Electrochim. Acta 110, 790 (2013) T.J. Schmidt, U.A. Paulus, H.A. Gasteiger, R.J. Behm, J. Electroanal. Chem. 508, 41 (2001) G.K.H. Wiberg, M. Arenz, J. Power Sources 217, 262 (2012) B.E. Conway, B. Barnett, H. Angerstein-Kozlowska, B.V. Tilak, J Chem Phys 93, 8361 (1990)
Electrocatalysis 27.
M. Alsabet, M. Grden, G. Jerkiewicz, J. Electroanal. Chem. 589, 120 (2006) 28. B.E. Conway, Prog. Surf. Sci. 49, 331 (1995) 29. A. Pavlisic, P. Jovanovic, V.S. Selih, M. Sala, N. Hodnik, S. Hocevar, M. Gaberscek, Chem. Commun. 50, 3732 (2014) 30. D.M. Novak, B.E. Conway, J. Chem. Soc. Faraday Trans. 77, 2341 (1981)
31.
A. Zolfaghari, B.E. Conway, G. Jerkiewicz, Electrochim. Acta 47, 1173 (2002) 32. D.B. Sepa, M.V. Vojnovic, A. Damjanovic, Electrochim. Acta 26, 781 (1981) 33. N.M. Marković, H.A. Gasteiger, B.N. Grgur, P.N. Ross, J. Electroanal. Chem. 467, 157 (1999) 34. A. Holewinski, S. Linic, J. Electrochem. Soc. 159, H864 (2012)
