Ultrafine TaO_x/CB Oxygen Reduction Electrocatalyst Operating in Both Acidic and Alkaline Media

Abstract

The high activity of non-platinum electrocatalysts for oxygen reduction reaction (ORR) in alkaline media is necessary for applications in energy conversion devices such as fuel cells and metal-air batteries. Herein, we present the electrocatalytic activity of TaO_x/carbon black (CB) nanoparticles for the ORR in an alkaline atmosphere as well as in an acidic electrolyte. Ultrafine TaO_x nanoparticles 1–2 nm in size and uniformly dispersed on CB supports were prepared by potentiostatic electrodeposition in a nonaqueous electrolyte and subsequent annealing treatment in an H₂ flow. The TaO_x/CB nanoparticles largely catalyzed the ORR with an onset potential of 1.03 V_RHE in an O₂-saturated 0.1 M KOH solution comparable to that of a commercial Pt/CB catalyst. ORR activity was also observed in 0.1 M H₂SO₄ solution. According to the rotating ring disk electrode measurement results, the oxide nanoparticles partly produced H₂O₂ during the ORR in 0.1 M KOH, and the ORR process was dominated by both the two- and four-electron reductions of oxygen in a diffusion-limited potential region. The Tafel slope of −120 mV dec⁻¹ in low and high current densities revealed the surface stability of the oxide nanoparticles during the ORR. Therefore, these results demonstrated that the TaO_x/CB nanoparticles were electroactive for the ORR in both acidic and alkaline electrolytes.

1. Introduction

Oxygen reduction reaction (ORR) is a critical step in renewable energy conversion systems such as fuel cells and metal-air batteries because the reaction is driven by sluggish and complex four-electron transfer over electrocatalyst surfaces determining the reaction kinetics and activity [1,2]. A Pt catalyst supported on carbon black (Pt/CB) is very active in ORR in acidic and oxidizing environments for typical polymer electrolyte fuel cells (PEFCs) [2,3,4,5]. However, because of the low catalytic durability and high cost of Pt/CB catalysts, many attempts have been made to suggest nonprecious alternatives such as Fe- and co-based catalysts coordinated with nitrogen and N-doped metal-free carbon catalysts applicable to PEFCs [6,7,8,9,10,11]. The non-platinum catalysts showing high ORR activity, comparable to that of Pt/CB catalysts, were limited to long-term operations in acidic conditions.

Group IV or V compounds based on Ta, Nb, and Zr, e.g., nitrides, carbides, and oxides are chemically stable in acidic conditions, so they have been investigated as candidate ORR electrocatalysts [12,13,14,15,16]. The transition metals are relatively cheaper than a commercial Pt catalyst [17]. Despite their chemical durability and low cost, their low electroconductivity led to relatively poor catalytic activity in ORR. The doping of N or C elements in the compounds or heat treatments in various gas environments enhanced electrocatalysis [12,13,14,15]. The oxide nanoparticles based on the group IV-V metals readily agglomerate during high-temperature calcination. It is difficult to decrease the size of the oxide particles by conventional preparations such as impregnation. The electrodeposition method is very effective in controlling the size of Pt or metal particles [18,19,20]. In fact, the nanotechnology of tantalum oxide particles (TaO_x) via potentiostatic electrodeposition in nonaqueous electrolytes significantly promoted an ORR, resulting in an onset potential of 0.95 V_RHE in a 0.1 M H₂SO₄ electrolyte [21,22,23,24]. The ORR activity over TaO_x nanoparticles supported on CB (TaO_x/CB) was also retained to approximately 94% of the initial current density during an accelerated long-term operation test for 10,000 cycles, demonstrating its high catalytic stability in an acidic electrolyte. The current density for an ORR was strongly dependent upon the size of the TaO_x nanoparticles, and oxide particles of 1~2 nm in size drove the ORR process via a four-electron transfer pathway, comparable to that of the commercial Pt/CB catalyst [23]. These results indicated that TaO_x/CB nanoparticles were a promising electrocatalyst for an ORR in acidic media.

In addition to the acidic atmosphere, the ORR process over nonprecious electrocatalysts besides commercial Pt/CB has been reported in alkaline atmospheres for applications to anion-exchange PEFCs and metal-air batteries [1,25]. However, the catalytic activity of group IV or V oxide nanoparticles has rarely been studied in an alkaline solution. In this study, we first present ORR activity over ultrafine TaO_x particles highly dispersed on CB (designated TaO_x/CB thereafter) in a 0.1 M KOH electrolyte. The TaO_x/CB particles were successfully synthesized by potentiostatic electrodeposition in a nonaqueous TaCl₅-based electrolyte, followed by annealing treatment in an H₂ flow. The ORR activity was compared to that in an acidic electrolyte. In addition, a kinetic study of the oxide nanoparticles was performed using a rotating (ring) disk electrode (R(R)DE).

2. Results and Discussion

Ultrafine TaO_x nanoparticles were deposited on CB supports by potentiostatic electrodeposition in a nonaqueous Ta-based solution, followed by annealing treatment in an H₂ flow. The synthesis route of the TaO_x/CB catalyst was consistent with that in a previous report, although the annealing temperature and the H₂ gas flow rate, in consideration of the high purity (4N) of TaCl₅ chemical, were modified in this study [23]. Figure 1a exhibits a TEM image of the prepared Ta-based catalyst. Black fine particles, an element heavier than carbon, were observed at the surfaces of the CB particles, approximately 30–40 nm in size. For validation, the STEM image of the catalyst was also measured, as shown in Figure 1b. Figure 1c displays the size distribution of the fine white particles, indicating that the particles 1–2 nm in size (average 1.4 nm) were uniformly loaded on ball-like CB particles, consistent with the TEM result. The size of the Ta-based particles was almost similar to that in a previous report [23]. However, the loading amount of the Ta species in the prepared catalyst determined by ICP-OES analysis was increased to 8.6 wt%, compared to 6.4 wt% in the former work. In this study, the higher purity of the Ta precursor might have increased the electrodeposited amount. Although the crystal structure of the annealed Ta-based powder was investigated by X-ray diffraction, there were no distinguishable peaks of crystallized Ta-based oxides. Lattice fringes of the Ta-based nanoparticles were also not observed in the STEM images, demonstrating that the electrodeposited nanoparticles were amorphous. This result was inconsistent with the previous result, and the difference could be that the high purity of the Ta precursor caused the loading of more concentrated Ta inside a nanoparticle during electrodeposition [18,22]. Figure 1d shows an elemental mapping analysis of the Ta-based nanoparticles displayed in Figure 1b. The Ta, O, and C elements were homogeneously distributed on the catalyst surface, proving that the electrodeposited catalyst was composed of Ta, O, and C species. Therefore, these results revealed that ultrafine, amorphous Ta-based oxide particles 1–2 nm in size were synthesized on the CB supports by the electrodeposition and subsequent heat treatment.

The chemical states of the Ta, O, and C species on the prepared catalyst were estimated by XPS analysis, as shown in Figure 2. The narrow-scan Ta 4f XPS spectrum was shifted from the binding energy in fully oxidized Ta₂O₅ to that of a lower valence state. For best fitting, the weak, broad peak of the spectrum was de-convoluted by Ta⁴⁺ and Ta⁵⁺, indicating that the electrodeposited Ta species were composed of two different valence oxides [22,26,27]. In the O 1s XPS spectrum, the broad peak was also deconvoluted with an oxygen anion in the nanoparticles and an OH group in the adsorbed H₂O on the catalyst surface. From the results, the chemical identity of the electrodeposited nanoparticles was determined to be TaO_x on CB supports, designated as TaO_x/CB thereafter.

A previous study demonstrated that the TaO_x/CB catalyst was active in an ORR in 0.1 M H₂SO₄ acidic solution, in which its activity was comparable to that of a commercial Pt/CB catalyst [23]. In this study, the ORR activity over the prepared TaO_x/CB nanoparticles was investigated in alkaline media. Figure 3 presents the CVs for the TaO_x/CB catalyst in Ar- and O₂-saturated 0.1 M KOH aqueous solutions at a scan rate of 100 mV s⁻¹. The RDE electrode was not rotated in these measurements. In the O₂-purged environment, the cathodic current was significantly increased from a potential of approximately 0.9 V_RHE compared to that in the Ar-saturated condition. The difference in the cathodic current indicated the participation of gaseous O₂ as a reactant in the reduction reaction, demonstrating that the TaO_x/CB catalyst drove the ORR in 0.1 M KOH alkaline solution. Therefore, the TaO_x/CB nanoparticles prepared by electrodeposition and subsequent annealing were electroactive for an ORR in alkaline media as well as in acidic conditions.

Figure 4a shows the LSVs data for the TaO_x/CB nanoparticles in 0.1 M KOH solution at 5 mV s⁻¹ measured using an RDE electrode at 1600 rpm. The current density for the actual ORR (i_orr) was estimated by the correction of the current density measured in Ar-saturated solution. For comparison, the ORR activity over 10 and 20 wt% commercial Pt/CB catalysts was also evaluated in the same manner. The TaO_x/CB catalyst showed significantly high ORR activity with an onset potential of 1.03 V_RHE (for −2 μA cm⁻²). The activity was comparable to that of commercial Pt/CB catalysts even though the loading amount of Ta in the oxide catalyst was lower. Moreover, the ORR activity over the TaO_x/CB catalyst in 0.1 M KOH was compared to that in 0.1 M H₂SO₄ solution, as shown in Figure 4b. The ORR activity of bare CB particles was also measured in both pH environments. The current density of the ORR in the acidic electrolyte began to generate from a potential of 0.94 V_RHE (for −2 μA cm⁻²), close to that reported in previous work. The ORR activity of the TaO_x/CB nanoparticles in 0.1 M KOH was higher than that in 0.1 M H₂SO₄, judged by the onset potential and current density at all the potentials [23]. Hence, the ORR activity over the CB supports was very distinct between the different pH conditions. In 0.1 M H₂SO₄, the CB particles were nearly inactive at the potentials measured, while they catalyzed the ORR in 0.1 M KOH. It has known that CB drives the ORR in alkaline media by a two-electron transfer pathway from O₂ to OOH⁻ [9,28,29]. Similarly, the CB in this study also produced some current density for the ORR, although the activity was much lower than that of the TaO_x/CB nanoparticles. The results showed that the TaO_x nanoparticles played the exclusive role of an electrocatalyst in the ORR in 0.1 M H₂SO₄, whereas both the oxide nanoparticles and CB supports were employed as active sites under the 0.1 M KOH electrolyte condition. Nevertheless, a current density at high potentials of 0.7–1.0 V_RHE was generated by the TaO_x nanoparticles, definitely indicating that the oxide was a major contributor to ORR activity in the alkaline solution.

The Tafel slopes of the TaO_x/CB nanoparticles and the 20 wt% Pt/CB catalyst were estimated at low and high current densities (at −0.1 and −1 mA cm⁻²), as shown in Figure 4c. The slopes of the Pt/CB catalyst varied from −60 to −100 mV dec⁻¹, consistent with previous studies [5,30,31]. It was known that a change in the slope from −2.303RT/F (−60 mV dec⁻¹) to approximately −2 × 2.303RT/F (−120 mV dec⁻¹), when the transfer coefficient α is assumed to be 0.5, is attributed to the change in the adsorption of ORR intermediates, i.e., from Temkin to Langmuir isotherms [5,23,31]. The slope of the CB was constant at −60 mV dec⁻¹. Tafel slopes lower than −120 mV dec⁻¹ for the ORR were typically obtained on oxide-free catalysts such as Pt/CB due to the oxide coverage by OH⁻ at the catalyst surfaces [5,14,15,16,23,30,31]. In contrast, the slopes for the TaO_x/CB nanoparticles of approximately −120 mV dec⁻¹ were almost unchanged at most current densities. The Tafel slope of −120 mV dec⁻¹ indicated that the surface states of the oxide nanoparticles were not varied, and the one-electron transfer reaction was a rate-determining step during the ORR in 0.1 M KOH solution. A similar Tafel slope was also observed in the ORR over the TaO_x/CB nanoparticles in 0.1 M H₂SO₄ solution. Therefore, these findings revealed that the electrochemical reaction sites at the surfaces of TaO_x/CB were considerably stable during the ORR in a kinetically limited region at high potentials, regardless of the pH condition. Figure 4d shows a chronoamperometry curve of the TaO_x/CB nanoparticles applied at 0.7 V_RHE in O₂-saturated 0.1 M KOH aqueous solution for 5 h. The ORR current was gradually increased and then maintained at −1.2 mA cm⁻² in 5 h reaction. This result indicates that the oxide nanoparticles were durable during the long-period ORR in 0.1 M KOH. The ORR activity of TaO_x/CB was also stable in acidic solution [23]. Therefore, the result demonstrated that TaO_x/CB catalyst showed stable ORR activity in both alkaline and acidic media.

RRDE measurements were carried out to identify the ORR pathway over the TaO_x/CB nanoparticles in alkaline media. Figure 5 shows the LSVs of the oxide nanoparticles in O₂-saturated 0.1 M KOH solution, which was compared to that in 0.1 M H₂SO₄ solution. A potential of 1.2 V_RHE was applied at the Pt ring electrode to oxidize the H₂O₂ product, swept from the glassy carbon disk electrode during the ORR, to O₂ again. In the disk current, the limiting current by the diffusion of O₂ reactant at the potentials, prior to approximately 0.3 V_RHE beginning to generate hydrogen underpotential deposition (UPD), was observed in 0.1 M KOH, whereas it was not saturated in 0.1 M H₂SO₄. From the limiting current, the half-wave potential, E_1/2, of the TaO_x/CB catalyst in the alkaline media was estimated to be 0.66 V_RHE. The CB supports were also active at low potentials during the ORR in 0.1 M KOH, as shown in Figure 4. The current density at low potentials might be increased by the additional activity of CB besides that of the TaO_x nanoparticles as the major active catalyst. In the ring current, the anodic current was obtained during the ORR in 0.1 M KOH, indicating that H₂O₂ was partly produced via the two-electron transfer pathway. This significantly contrasts with the result that the ring current was nearly negligible in the 0.1 M H₂SO₄. The H₂O₂ production was estimated by the equation.

$$\% {\mathrm{H}}_2{\mathrm{O}}_2=\frac{2i_{ring}/N}{-i_{disk}+i_{ring}/N}\times 100$$

(1)

where i_disk is disk current, i_ring is ring current, and N is the current collection efficiency of the RRDE. N was estimated to be 0.420 by the reduction in K₃[Fe(CN)]₆. At a potential of 0.4 V_RHE, the H₂O₂ yield resulting from the ORR over the TaO_x/CB nanoparticles in 0.1 M KOH was estimated to be 37.9%, much greater than 0.2% in 0.1 M H₂SO₄. A considerable proportion of two-electron transfer reactions occurred in the alkaline atmosphere. The bare CB particles were found to produce approximately 70% H₂O₂ at all the potentials where they were active during the ORR in 0.1 M NaOH solution [15]. Interestingly, in this study, the ring current for the oxidation of H₂O₂ was increased from a potential of 0.73 V_RHE, quite closer to the onset potential of bare CB particles for the ORR in 0.1 M KOH. Therefore, these results indicated that the production of H₂O₂ was mainly by the ORR over bare CB surfaces.

In addition, hydrodynamic voltammetry was carried out using RDE to establish the ORR kinetics of the TaO_x/CB catalyst in alkaline media. Figure 6a displays the LSVs of the nanoparticles in O₂-purged 0.1 M KOH at different rotation speeds of 100, 400, 900, 1600, and 2500 rpm. The ORR cathodic current was saturated to a constant value at low potentials before the hydrogen UPD region and gradually enhanced by increasing the rotation rate. This indicated that the ORR current at the low potentials over the catalyst was dominated by the diffusion-limiting current. Figure 6b illustrates the Koutecky–Levich plots, i⁻¹ versus ω^−1/2, for the TaO_x/CB nanoparticles, determined by the Koutecky–Levich equation.

$$\frac{1}{i}=\frac{1}{i_{\mathrm{k}}}+\frac{1}{i_{\mathrm{Lev}}}$$

(2)

where i is the measured disk current, i_k is the kinetic current, and i_Lev is the Levich current. The diffusion-limiting current. i_Lev can be expressed by the Levich equation:

$$i_{\mathrm{Lev}}=0.62nFA{D}_{\mathrm{O}2}^{2/3}{\nu }^{-1/6}{C}_{\mathrm{O}2}{\omega }^{1/2}$$

(3)

where n is the number of electrons transferred for the ORR, F is the Faraday constant, A is the electrode area, D is the diffusivity coefficient of O₂ in 0.1 M KOH (1.9 × 10⁻⁵ cm² s⁻¹ at 298 K) [32,33], ν is the kinematic viscosity of the solution (1.0 × 10⁻² cm² s⁻¹) [32,33], C_O2 is the concentration of dissolved O₂ in the electrolyte (1.2 × 10⁻³ mol L⁻¹ at 298 K) [32,33], and ω is the rotational angular speed of the electrode (rad s⁻¹). The Koutecky–Levich plots exhibited a linear relation between i⁻¹ versus ω^−1/2 in the potential range of 0.2–0.6 V_RHE. The slopes of the straight lines were compared to two theoretical lines indicating the two- and four-electron transfer pathways. In the potential range, the line slopes were a mixed n-electron transfer between two- and four-electron transfer lines, revealing that the ORR mechanism on the TaO_x/CB nanoparticles in the diffusion-limited region was governed by both two and four-electron reductions of oxygen [34]. Kinetic parameters of the TaO_x/CB nanoparticles obtained from the Koutecky–Levich plots were summarized in

Table 1. Kinetic parameters estimated from the Koutecky–Levich plots of the ultrafine TaO_x/CB nanoparticles shown in Figure 6.

Potential/VRHE	ik/mA cm−2	Number of Electrons Transferred
0.6	7.08	3.57
0.4	31.84	3.91
0.2	28.24	3.46

. These results were consistent with the RRDE results showing that H₂O₂ was partly generated during the ORR. The ORR kinetics over the TaO_x nanoparticles in acidic media was nearly all via the four-electron transfer pathway [23]. In this study, no ring current was observed at kinetically limited potentials over 0.75 V_RHE, verifying the four-electron reduction in oxygen over only the TaO_x nanoparticles. Thus, it is likely that the TaO_x nanoparticles and the CB support in the TaO_x/CB catalyst were the main contributors to the four- and two-electron reductions of oxygen in alkaline solution, respectively. Therefore, these electrochemical results clearly demonstrated that the TaO_x/CB nanoparticles were active in ORR in both acidic and alkaline electrolytes.

3. Experimental

3.1. Preparation of TaO_x/CB Electrocatalyst

TaO_x nanoparticles were loaded on the CB surfaces by potentiostatic electrodeposition in a nonaqueous electrolyte. A total of 10 mg of CB powder (Vulcan-XC72, Cabot Corporation, Cilegon, Indonesia) was positioned between two carbon papers (AvCarb MGL190, AvCarb Material Solutions, Lowell, MA, USA). This sandwich-type CB assembly as the working electrode was inserted into a homemade electrodeposition cell as in a previous study [23]. A typical three-electrode system was used for the electrodeposition using a potentiostat (CHI700E, CH Instruments, Inc., Austin, TX, USA). A carbon rod and Ag/Ag⁺ electrode were mounted as the counter and reference electrodes, respectively. The nonaqueous Ta-based plating solution consisted of 34 mM TaCl₅ (99.99%, Sigma-Aldrich, St. Louis, MO, USA), 10 mM NH₄Cl (99%, Kanto Chemical, Tokyo, Japan) as the supporting electrolyte, and anhydrous ethanol. To deposit Ta species on the CB, a constant potential at −0.5 V_Ag/Ag+ was applied to the working electrode for 10 s in a stirred Ar-saturated TaCl₅-NH₄Cl ethanol solution at 298 K. The electrodeposited Ta-based CB powder was washed by ethanol to remove the remaining Ta precursor and dried naturally. Then, the resulting TaO_x/CB powder was annealed in a pure H₂ (5N grade) flow of 50 mL min⁻¹ at 1073 K for 2 h at a ramp rate of 10 K min⁻¹.

3.2. Electrochemical Measurements

Catalyst ink using the TaO_x/CB powder was prepared for the ORR measurements. A total of 2 mg of the TaO_x/CB powder and 12 µL of a Nafion solution (5 wt% in water/aliphatic alcohols, Sigma-Aldrich, St. Louis, MO, USA) were dispersed in 300 µL of isopropyl alcohol (99.9%, Samchun Chemical, Seoul, Korea). The mixture was mounted in an ultrasonic bath and then on a magnetic stirrer to enhance the dispersion of the powder, and this process was repeated three times. Subsequently, 2.5 µL of the catalyst ink was cast on a glassy carbon RDE (4 mm in diameter) and dried at 298 K. The loading amount of the catalyst was approximately 128 µg cm⁻². For comparison, commercial 10 and 20 wt% Pt/CB (Premetek Co., Pt on the Vulcan XC-72, Cherry Hill, NJ, USA) catalyst ink was also dropped on the RDE in the same procedure.

Electrochemical measurements were carried out using the RDE at 1600 rpm, the Hg/HgO reference, and carbon rod counter electrodes. Cyclic and linear sweep voltammograms (CVs and LSVs) were obtained over a potential range of 1.2 and 0.1 V_RHE at a scan rate of 5 mV s⁻¹ in Ar- and O₂-purged 0.1 M H₂SO₄ and 0.1 M KOH aqueous electrolytes. The difference in current density between the O₂- and Ar-saturated conditions was considered the real ORR current, the i_ORR. The Hg/HgO reference electrode was calibrated to a reversible hydrogen electrode (RHE). For RRDE measurements, the glassy carbon disk electrode (4.0 mm in diameter) was scanned at a sweep rate of 5 mV s⁻¹, and a constant potential at 1.2 V_RHE was applied at the Pt ring electrode (5.0 mm in inner diameter; 7.0 mm in outer diameter).

3.3. Surface and Physical Characterizations

The surface morphology of the TaO_x/CB nanoparticles was characterized using a field emission transmission electron microscope (FE-TEM; JEM-2200FS, JEOL, Tokyo, Japan) equipped with energy-dispersive X-ray spectroscopy (EDS) capability. The amount of Ta loading in TaO_x/CB was estimated by inductively coupled plasma optical emission spectroscopy (ICP-OES; PerkinElmer, Avio 500, Waltham, MA, USA). The chemical states of the Ta and O species on the TaO_x/CB surface were determined by high-performance X-ray photoelectron spectroscopy (HP-XPS; K-Alpha+, Thermo Scientific, Waltham, MA, USA), using a monochromatic X-ray source producing Al Kα emission with a current of 6 mA, and an acceleration voltage of 12 kV.

4. Conclusions

In this study, we demonstrated the catalytic activity of TaO_x/CB nanoparticles in the ORR in both acidic and alkaline media. Highly dispersed, ultrafine TaO_x nanoparticles 1–2 nm in size on CB supports were synthesized by potentiostatic electrodeposition in a nonaqueous electrolyte and subsequent annealing treatment in an H₂ flow. The TaO_x/CB nanoparticles largely catalyzed the ORR in O₂-saturated 0.1 M KOH solution, comparable to that over a commercial Pt/CB catalyst. The ORR activity was higher than that in the 0.1 M H₂SO₄ solution because the CB supports were also active in the reaction via the two-electron transfer pathway. The participation of CB in addition to TaO_x nanoparticles in the ORR caused an approximately 38% H₂O₂ production and the mixed n-electron transfer reduction in oxygen in a diffusion-limited potential region. Nevertheless, the high activity of the TaO_x nanoparticles with an onset potential of 1.03 V_RHE indicated that the oxide nanoparticles were a major catalyst in the ORR in the alkaline solution. The Tafel slope of −120 mV dec⁻¹ in low and high current densities also revealed the surface stability of the oxide nanoparticles during the ORR. Therefore, the study findings suggest that the TaO_x/CB nanoparticles are promising nonprecious oxygen catalyst candidates for applications to energy conversion systems operable at any pH conditions such as fuel cells and metal-air batteries.