Effect of activation range on CE dissolution
In most recently reported studies related to HER, surface activation or cleaning is the first step performed in electrochemical measurements. It is widely accepted that as the current increases, the potential on the counter electrode increases4. This encouraged us to study the effect of different potential windows during the activation step and its correlation with the dissolution rate of the counter electrode. Thus, we explored two ranges for CV cycles; narrow turn-on range, NRA, (0.06 V to -0.2 vs RHE) and wide turn-on range, WRA, (0.06 V to -0.34 V vs RHE). When expanding the potential window towards more negative potentials, the current begins to increase as the number of cycles increases, Fig. 1a, b. This should increase the rate of Pt dissolution. Dissolved Pt would redeposit on the surface of the working electrode, resulting in a misleading enhancement of the recorded catalytic activity of the working electrode. Figure 1c shows the shift in line-scan voltammograms (LSV) recorded before and after electrochemical activation at -10 mA/cm2. When a narrow CV range sweep is applied in the non-Faradic region, an offset of 170 mV was recorded after 300 CV cycles, while an offset of 400 mV was observed after 300 CV cycles in the wide range of activation. EDX analysis of the two samples after electrochemical measurements (inset in Fig. 1c) showed a much higher intensity of the Pt peak of the sample tested in the wide activation range than that tested in the narrow range. Additionally, SEM images of the sample before electrochemical activation (Fig. 1d) and after application of narrow (Fig. 1e) and wide (Fig. 1f) activation were investigated. Note the presence of a small number of small Pt nanoparticles when applying a narrow activation range. However, upon activation over a wide range, the surface was heavily covered with Pt nanoparticles. The corresponding EDX maps of the surface of the working electrode after the application of NRA and WRA are shown in Fig. 1g, h, respectively. This signifies the importance of choosing the activation range which does not involve any increase in current to avoid CE dissolution. Note that electrode polarization usually takes place when high current is reached or if the electron transport rate is greater than the reaction rate4.
Effect of working electrode area on CE dissolution rate
Counter electrodes in electrochemical cells are used to maintain the rate of the half reaction occurring on their surfaces faster than the other half reaction on the WE. Therefore, if the CE half-reaction is slower than the complementary on the WE, the recorded current would be misleading as the CE dictates the current response. To overcome this problem, CE must possess high electrical conductivity and larger surface area (typically ten times) than WE to ensure fast reaction kinetics on CE.9. To elucidate such an impact, the area of the Pt CE was kept constant while varying the area of the WE (0.25 cm2 and 1cm2). As the area of the WE increases, the overvoltage is significantly reduced (Fig. 2a), which can be attributed to the higher rate of dissolution and redeposition of Pt on the working electrode. This was confirmed by EDX analysis, inset in Figure 2a, where the intensity of the Pt peak increases significantly as the area increases. Moreover, SEM images further confirm the correlation between working electrode area and CE dissolution rate, Fig. 2b, c. When the WE area was 0.25 cm2, small Pt nanoparticles begin to appear on the surface of the WE. On the other hand, when the WE area was 1 cm2, very dense Pt particles were observed on the WE surface, starting to form Pt clusters. The corresponding EDX mapping of the working electrode surface after the electrochemical measurements is shown in Fig. 2d, e, respectively.
is there an exit?
The above findings showed the dissolution of Pt when used as CE for HER in acidic media. Therefore, more materials were tested to elucidate their stability in hopes of identifying a stable material to use as CE. In particular, a Pt sheet, a gold coil, a glassy carbon rod and a titanium mesh roll were tested as counter electrodes in the HER 3-electrode system. LSV experiments were performed by setting the area of the working electrode to 0.25 cm2. The overvoltages were measured for the different counter electrodes before and after the application of 700 cycles of activation in the narrow activation range (0.06 V to −0.2 V vs RHE) at a slew rate of 5 mVs−1 in 0.5MH2SO4. Figure 3a shows the LSV scans of the four counter electrodes before activation, revealing almost the same overvoltage for all four counter electrodes. Figure 3b shows the LSV scans of the four counter electrodes after activation. Note that there is a substantial lag in the recorded surge. The surges recorded are in ascending order as gold coil, Pt foil, titanium mesh and glassy carbon rod, with the values of -0.35 V, -0.5, -0.69 and – 0.84 V, respectively. While the lowest overvoltage was recorded when using a gold coil as CE, using a glassy carbon electrode resulted in the highest overvoltage. Tafel slopes were calculated to study the kinetics of HERten, as shown in Figure 3c. When the gold coil was used as the CE, the lowest Tafel slope was obtained (39.8), while the Tafel slope of the Pt, Ti mesh and the CE glassy carbon rod was 65.9, 121.1 and 258.4, respectively. The improvement in Tafel overvoltage and slope values when gold coil and Pt foil were used as EC indicates that this is likely due to dissolution and redeposition of gold and platinum on the working electrode. We expect that the improvement in reaction kinetics is due to the deposition of gold and Pt species on the surface of the working electrode, resulting in a reduced overvoltage, consistent with the LSV measurements previously discussed. Moreover, the shift observed after activation suggests that electrochemical dissolution occurs mainly during activation cycles. EDX analysis (Fig. 3d) revealed the presence of gold and platinum on the working electrode, while no Ti was detected.
Moreover, the XPS analysis confirmed the deposition of Pt and Au nanoparticles on the surface of the counter electrodes, as shown in Fig. 4, when using Pt foil and Au coil as counter electrodes. Spin-orbit coupling of Au 4f7/2 and Au 4f5/2 is 84 eV and 87.5 eV, indicating the presence of elemental Au0 on the surface of the working electrode11.12. Also, Pt 4f7/2 and Pt 4f5/2 peaks were observed at 71.6 eV and 74.9 eV respectively4,13,14. On the other hand, when using the Ti mesh as CE, no Ti-related signal was detected. This indicates that the gold coil and the platinum foil are not stable. On the other hand, although the Ti mesh and the glassy carbon rod are stable, a large overvoltage lag is evident, which is probably due to the small surface area of the glassy carbon rod compared to the large surface area of the Ti mesh roll. Yi et al. demonstrated that glassy carbon in acidic media undergoes degradation as the acid catalyzes the formation of surface oxides followed by ring opening in the graphitic structure, and finally bulk oxidation. Thus, the glassy carbon rod does not appear to be effective in elucidating the catalytic activity of the working electrode and this could result in a misleading overvoltage.15.