Alkaline Electrolyte and Fe Impurity Effects on the Performance and Active-Phase Structure of NiOOH Thin Films for OER Catalysis Applications
John D. Michael, Ethan L. Demeter, Steven M. Illes, Qingqi Fan, Jacob R. Boes and John R. Kitchin, “Alkaline Electrolyte and Fe Impurity Effects on the Performance and Active-Phase Structure of NiOOH Thin Films for OER Catalysis Applications”, The Journal of Physical Chemistry C 119(21), 11475-11481 (2015). https://doi.org/10.1021/acs.jpcc.5b02458
This summary was written by a language model from the paper and has not been checked line by line. Rely on the published paper for details.
Question
Nickel oxyhydroxide (NiOOH), especially when it contains Fe, is an inexpensive oxygen evolution reaction (OER) catalyst in alkaline electrolyte. Prior work had shown that even ppb-level Fe impurities in commercial hydroxide electrolytes strongly promote NiOOH, and that alkali cations intercalate into Ni(OH)2/NiOOH films during redox cycling. The authors asked whether the choice of alkali hydroxide (LiOH, NaOH, KOH or CsOH) changes NiOOH OER activity, how that interacts with deliberately controlled Fe levels, and whether any activity differences can be traced to changes in the active-phase structure (Michael et al. 2015).
Approach
Ni(OH)2 thin films (about 30 nm, roughly 1.55 × 10-9 mol Ni from a redox charge near 150 μC) were electrodeposited on roughened Au electrodes. Electrolytes were 0.1 M LiOH, NaOH, KOH or CsOH, either purified of Fe by soaking with freshly precipitated Ni(OH)2 as an Fe adsorbent, or “Fe-saturated” by adding Fe(NO3)3 to about 1 ppm until Fe hydroxide precipitated. Films were cycled to a stable state and then measured by linear sweep voltammetry (LSV) at 1 mV/s against Hg/HgO while the electrolyte was swapped back and forth on the same electrode, which tested whether the cation effect is reversible. iR correction was judged unnecessary at these low currents. For LiOH and CsOH, in situ Raman spectra (532 nm excitation) were recorded during LSV at overpotentials of 240, 340 and 440 mV, fit to sums of Gaussians, and compared by mean positions of the O-Ni-O bending band near 480 cm-1 and stretching band near 560 cm-1. The Python fitting code was provided in the Supporting Information.
Key findings
In purified electrolyte, CsOH gave roughly double the OER current density of LiOH wherever OER was appreciable (for example about 1 versus 0.5 mA/cm2 at 0.5 V overpotential). Above about 1 mA/cm2 the ordering was Cs+ > K+ ≈ Na+ ≈ Li+, though overlapping error bars left Cs+ > Li+ as the only unambiguous result.
Adding Fe dominated everything else: onset overpotential dropped by at least 100 mV and current density rose by nearly an order of magnitude in all four electrolytes. With Fe present the ordering became K+ ≈ Na+ > Cs+ > Li+ at all current densities; at 0.3 V overpotential KOH and NaOH gave about 7 mA/cm2, CsOH about 4.5 and LiOH about 2.5 mA/cm2. CsOH still beat LiOH by about 50%. Based on the position of the Ni redox peak, the Fe-saturated films were estimated to contain 5 to 25% Fe, a range over which prior work found nearly identical activity, so the authors argue that differences in Fe uptake cannot explain why KOH and NaOH became best. They describe the combined result as a “double promotion”: activity can be tuned by both the cation and the Fe level, with the cation effect much smaller than the Fe effect.
Raman bands in CsOH sat at lower wavenumbers than in LiOH: by about 1.5 and 5.0 cm-1 (bending, stretching) in purified electrolyte and about 1.5 and 4.5 cm-1 in Fe-saturated electrolyte, with fitted uncertainties of a few tenths to about 1 cm-1. Using the Hardcastle-Wachs relation between Raman shift and metal-oxygen bond length, this indicates slightly longer Ni-O bonds in CsOH, which correlates with its higher activity. Fe lowered only the bending band, by about 1 cm-1, so Fe had a smaller Raman-visible structural effect than Cs+; the authors attribute the Fe promotion instead to Fe acting as active sites in the lattice, consistent with a cited X-ray absorption study. Peak positions did not shift significantly with overpotential. The authors state that Raman alone cannot assign a definite γ- or β-NiOOH phase, and they suggest (without testing) that longer Ni-O bonds might open the lattice and ease transport to active sites.
Significance
The paper shows that electrolyte identity and trace Fe are both experimental variables, not background conditions, when benchmarking Ni-based OER catalysts, and that their effects interact (Fe changed which cation was best). It also offers a simple, scriptable Gaussian-fitting procedure for quantifying small Raman shifts in NiOOH under reaction conditions (Michael et al. 2015).