High Stability, High Activity Pt/ITO Oxygen Reduction Electrocatalysts Ying Liu and William E. Mustain* Department of Chemical, Materials and Biomolecular Engineering, University of Connecticut, 191 Auditorium Drive, Storrs, Connecticut 06269, United States S1
Synthesis of Sn-doped indum oxide (ITO) nanoparticles A white slurry of both metal precursors, indium acetylacetonate (In(acac) 3 ) (99.99+%, Sigmaaldrich) and tin bis(acetylacetonate) dichloride (Sn(acac) 2 Cl 2 ) (98%, Sigma-aldrich) were prepared with various nominal concentrations of Sn and In by mixing in at various ratios (Table S1) in 4 ml oleylamine (98+%, Sigma-aldrich). The slurries with were heated at 250 C for 3h in air to produce a dark yellow suspension. The ITO nanoparticles were precipitated by adding ethanol (40 ml) to the suspension and the supernatant was removed by centrifugation at 2500 rpm for 6 min. Washing with ethanol was repeated 3 times and produced a white powder of ITO nanoparticles. The ITO was heated at 80 C under vacuum for 24h and then heated at 500 C for 3h in the furnace in the air. Table S1. Preparation of ITO nanoparticles with various Sn content Nominal Sn content in ITO nanoparticles (atom %) In(acac) 3 (mmol) Sn(acac) 2 Cl 2 (mmol) 1 0.721 0.00728 5 0.692 0.0364 10 0.655 0.0728 Deposition of Pt on ITO Pt/ITO electrocatalysts were prepared by depositing Pt nanoparticles on ITO via galvanic displacement of a Cu layer by Pt in a three-electrode electrochemical cell in N 2 atmosphere to avoid the oxidation of Cu atoms in contact with O 2. First, the working electrode was covered with a Cu layer by cycling the electrode potential three times between 0.0 and 0.3 V in aqueous 0.05 M H 2 SO 4 /0.05 M CuSO 4 at a sweep rate of 50 mv/s. Then, the electrode was rinsed with copious amounts of 18 MΩ Millipore water to remove Cu 2+ ions from the electrode. Next, the electrode was placed into a 0.001 M K 2 PtCl 4 (Acros Organics), 0.05 M H 2 SO 4 aqueous solution for 3 min. To quantify the Pt loading on ITO, the difference in area under the cathodic (reduction) and anodic (oxidation) scans were considered. Since the total charge that is passed during an electrochemical S2
experiment is simply the time integral, we can determine the number of coulombs passed in an electrochemical experiment, Q, by: Q = Tf Ti i(t)dt However, the data is collected by scanning the voltage at a certain scanrate, ν, which is related to time by: dv = νdt Thus: Q = 1 ν Vf Vi i(t)dv This allows researchers to independently calculate the charge passed during the anodic and cathodic scan, Q a and Q c, repectively. The difference between these two values, Q c -Q a, is equivalent to the total amount of Cu that was deposited on the electrode surface per cycle. The total number of moles of Cu that were deposited, N Cu, can be found by: N Cu = Q c Q nf a Where n is the electron equivalence and F is Faraday s constant. Since the Pt displacement mechanism is equimolar with respect to Cu and Pt, Pt 2+ + Cu Pt + Cu 2+, N Pt =N Cu and: N Pt = scan3 scan1 Q c Q nf a In this study, we performed several trials to deposit Pt on the ITO surface and found that the average number of coulombs of Pt deposited was 2.04 *10-3 C, which was 1.06 * 10-8 mol or 2.06 µg. The average mass of ITO on the surface was 7.34 µg. This estimates that the Pt loading on the ITO in this work was 22%, which nominally agreed with the EDX results. S3
Preparation of Pt/C electrocatalyst 120 mg of NaOH was added to 25 ml of ethylene glycol (EG) under vigorous stirring at 100 C for 1h until dissolved completely. Then, 70 mg of carbon black (Vulcan XC-72R) was added to solution, and the platinum precursor, hexachloroplatinic acid, H 2 Cl 6 Pt.6H 2 O, Sigma-Aldrich), was added to the mixture to produce 20 wt% of Pt/C. The resulting suspension was stirred for 1h at room temperature followed by heating under reflux at 160 C for 3h. The solution was allowed to cool down to room temperature and then the ph was adjusted to 1.5 using 0.5 M H 2 SO 4. After 48 h, Pt/C was filtered and washed with acetone and deionized water. Transmission electron microscopy characterization Samples for high resolution transmission electron microscopy (HETEM) characterization were prepared by ultrasonically dispersing ITO and Pt/ITO particles in ethanol and drying them on holey carbon/copper grids. These samples were observed under a JEOL 2010 FasTEM with 200 kv thermionic electron source. XRD and XPS analysis The bulk composition of the electrocatalyst supports was confirmed by XRD with a Bruker D8 Advance diffractometer system. XPS analysis was done using a PHI multiprobe system with twin anode XPS using unmonochromatized Al K% radiation (1486.6 ev) operated at 250 W and 15 kv. The pressure in the analysis chamber was always ~10 8 Torr or less. The full survey was taken at 100 ev pass energy with a scan rate of 1 ev/s and the high resolution scans were conducted at 20 ev pass energy with a scan rate of 0.1 ev/s. The spectra were calibrated with respect to graphitic C 1s electron bond at 284.6 ev. The backgrounds were determined using the Shirley-type background correction and the curves were fitted with Gaussian/Lorentzian product functions. S4
Electrochemical analysis Electrochemical experiments were conducted in a custom-built three-electrode electrochemical cell with a Luggin capillary (Adams & Chittenden Scientific Glass). All studies were carried out in 0.1 M HClO 4 at room temperature using an Autolab PGSTAT302N potentiostat, with a platinum flag as the counter electrode and reversible hydrogen electrode (RHE) as the reference electrode. Results are reported in this paper with reference to RHE. Pt/C was ultrasonicated for 0.5 h with DI water to make dispersions. 20 µl of the suspension was applied to an inverted 5 mm glassy carbon electrode (GCE) that was polished to a 0.05 µm mirror finish (alumina, Buehler) before each experiment. The water was evaporated from the suspension by rotating the electrode at 300 rpm in air. The dry samples were fixed onto the GCE by dropping 20 µl of diluted DE-520 Nafion solution (1/100, 5 wt %, DuPont). To evaluate the oxygen reduction reaction (ORR), the electrolyte was bubbled with O 2 for 1 h and at least 5 cyclic voltammograms (CVs) were run at 50 mv/s to condition the electrode before the data were recorded. For the rotating disk electrode (RDE) measurements, the working electrode was scanned cathodically at a rate of 5 mv s -1 with varying rotating speed from 400 rpm to 2500 rpm. Koutecky- Levich plots (J -1 vs. ω -1 ) in Figure 3(b) were analyzed at various electrode potentials. The slope of the best linear fit line was used to calculate the number of electrons transferred (n) on the basis of the Koutecky- Levich equation: 1 1 1 1 1 / B 0.62 nf / / Where J is the measured current density, J k and J D are the kinetic and diffusion-limiting current densities, respectively, ω is the angular velocity, n is the electron equivalence, F is Faraday s constant, C o is the bulk concentration of O 2, and υ is the kinematic viscosity of the electrolyte. S5
To obtain the Tafel plot in Figure 3(e), the experimental data was mass transport corrected. To accomplish this, the kinetic current was calculated at all points using the observed current density and diffusion limited current density by the following equation: The ORR activity was taken as the mass transport corrected current at 0.9 V vs. NHE for all catalysts. Each experiment was done at least five times to ensure reproducibility and the average value is reported along with the standard deviation. A characteristic linear sweep voltammogram for the ORR on Pt/ITO was shown in Figure 3c. The Pt/ITO catalyst exhibited a specific activity of 0.750 ±.04 ma/cm 2 at 0.9 V, less than a 5% deviation, which was 3 times greater than Pt/C (0.235 0.01 ± ma/cm 2 ). Normalizing to the loading amount of Pt metal, the mass activity of Pt/ITO catalyst was found to be 621 ± 31 ma/mg Pt, which was 4 times greater than Pt/C (156 ± 9 ma/mg Pt ). Supplementary figures Figure S1. XPS spectra of Sn 3d 5/2 in tin-doped indium oxide nanparticles with various Sn concentrations. S6
Pt clusters were deposited on the ITO nanoparticles through galvanic displacement by Pt of the Cu layer (Figure S2(d)). TEM images of the product Pt/ITO reveal that a number of well-dispersed Pt nanoparticles were deposited on the ITO surface (Figure S2 (a)), which was also confirmed by XPS (Figure S3). These ITO particles have the average size of 17.8 nm, while the Pt nanoparticles have the average size of 2.4 nm, as shown in Figure S2 (c). S7
Figure S2. (a) TEM images of Pt nanoparticle supported on ITO prepared by heating to 500 C, with 5 at % Sn; (b) ITO particle size distribution, obtained from measurements of 200 particles; (b) Pt particle size distribution, obtained from measurements of 120 particles. (d) CV curves for Cu deposition on ITO recorded at room temperature in an N 2 -purged 0.05 M H 2 SO 4 /0.05 M CuSO 4 solution. Sweep rate, 50 mv/s. Figure S3. XPS spectras of ITO (5% Sn) and fresh-prepraed Pt supported ITO (5% Sn). By changing the synthesis temperature, ITO nanoparticles with the same Sn content (5%) showed significantly different morphology. It is well known that higher temperatures can accelerate the growth of the small nanoparticles with high surface energy, resulting in larger nanoparticles or agglomerates than those treated under the lower temperature with the same time period. The size of the ITO nanoparticles heated at 600 C were around 30-40 nm. And those heated at 700 C showed the rectangle structure with the size of 25-35 nm (Figure S4), which suggests that instead of the random aggregation of ITO nuclei S8
under the lower temperature (e.g. 600 C), the rectangle tin-doped indium oxide nanoparticles resulted from the direct growth of certain specific crystal phases continuously. Among the ITO particles with various Sn content prepared under different temperatures, the 5% Sn-doped In 2 O 3 prepared at 500 C with smaller size and higher surface area was selected as Pt support to avoid the surface SnO/SnO 2 oxidation/reduction which may impact the long-term stability of the catalysts. Figure S4. TEM images of tin-doped indium oxide nanoparticles: (a) Heated at 600 C, 5 at%; (b) Heated at 700 C, 5 at%; (c), (d): HRTEM images of (a), (b), respectively. Figure S5 shows cyclic voltammograms (CVs) for Pt/ITO and Pt/C catalysts (20% wt. Pt on Vulcan XC-72) recorded at room temperature in N 2 -purged 0.1M HClO 4 solution at a sweep rate of 50 mv/s. All potentials are discussed relative to the reversible hydrogen electrode (RHE). S9
Figure S5. CV curves for Pt/ITO and Pt/C electrocatalysts recorded at room temperature in an N 2 -purged 0.1 M HClO 4 solution. Sweep rate, 50 mv/s. The stability of the Pt/C catalysts was determined by CV between 0.0 and 1.4 V. The stability test was conducted by applying potential sweeps at the rate of 10 mv/s to a thin-film rotating disk electrode in O 2 -saturated 0.1 M HClO 4 solution at room temperature. After 1000 cycles, changes in the Pt ESCA and electrocatalytic activity of the ORR were determined. The stability of the Pt/ITO catalyst was far superior to Pt/C. The catalytic activity of Pt/ITO, measured by the linear potential sweeps, showed only a 4 mv degradation in half-wave potential over the cycling period (shown in Figure 3c of the primary manuscript). In contrast, the shift in the half-wave potential for Pt/C was 20 mv, which is shown in Figure S6a. Integrating the charge between 0 and 0.35 V associated with H adsorption for Pt/ITO shows almost no change, indicating no recordable loss of Pt ECSA (Figure 3d in the primary manuscript). However, for Pt/C, only ~65% of the original ECSA remained after potential cycling (Figure S6b). S10
Figure S6. (a) Polarization curves for the O 2 reduction reaction on Pt/C catalysts on a rotating disk electrode, before and after 1,000 cycles. Sweep rate, 10 mv/s; rotating rate, 1600 rpm. (b) Voltammetry curves for Pt/C before and after 1,000 cycles. Sweep rate, 10 mv/s. Table S2. Surface In/Sn concentrations of Pt/ITO fresh-prepared and after 1000 cycles Pt/ITO In 2 O 3 (%) SnO 2 (%) SnO (%) Fresh-prepared 100 91 9 After 1000 cycles 100 28 72 S11