Unraveling the Role of the Stoichiometry of Atomic Layer Deposited Nickel Cobalt Oxides on the Oxygen Evolution Reaction

Abstract

Nickel cobalt oxides (NCOs) are promising, non-precious oxygen evolution reaction (OER) electrocatalysts. However, the stoichiometry-dependent electrochemical behavior makes it crucial to understand the structure-OER relationship. In this work, NCO thin film model systems are prepared using atomic layer deposition. In-depth film characterization shows the phase transition from Ni-rich rock-salt films to Co-rich spinel films. Electrochemical analysis in 1 m KOH reveals a synergistic effect between Co and Ni with optimal performance for the 30 at.% Co film after 500 CV cycles. Electrochemical activation correlates with film composition, specifically increasing activation is observed for more Ni-rich films as its bulk transitions to the active (oxy)hydroxide phase. In parallel to this transition, the electrochemical surface area (ECSA) increases up to a factor 8. Using an original approach, the changes in ECSA are decoupled from intrinsic OER activity, leading to the conclusion that 70 at.% Co spinel phase NCO films are intrinsically the most active. The studies point to a chemical composition dependent OER mechanism: Co-rich spinel films show instantly high activities, while the more sustainable Ni-rich rock-salt films require extended activation to increase the ECSA and OER performance. The results highlight the added value of working with model systems to disclose structure-performance mechanisms.

Keywords: atomic layer deposition; cobalt nickel oxides; electrochemical activation; oxygen evolution reaction; thin film characterization.

Figure 1

X‐ray spectroscopy measurements at the a) Ni3p and Co3p, b) O1s, c) Ni2p and d) Co2p spectra of the deposited films. The legend of (a) is valid for all figures.

Figure 2

a) X‐ray diffraction pattern of NCO films deposited on FTO glass. The vertical lines refer to the ICSD reference measurements of NiO (grey, dotted), Co₃O₄ (brown, solid) and SnO₂ (orange, dashed) (ICSD cards 9866, 36 256 and 9163, respectively). b) Shift of the angle at 44°. The legend of (a) is also valid for (b).

Figure 3

Redox features and onset potential for a) NiO investigated at 50 mV s⁻¹; b) rock‐salt films, activated for 500 cycles at 20 mV s⁻¹, with different cobalt relative content. The redox feature of 30 at.% Co film is highlighted using an offset. c) The variation in integrated charge, obtained from the 50 mV s⁻¹ CV curves after 100 and 500 CV cycles. d) Onset potential of activated (500 cycles) spinel films as a function of Co at.% at 20 mV s⁻¹.

Figure 4

a) The evolution of the overpotential as a function of number of CV cycles for different Co at.% films. b) The change in overpotential, defined as the difference between the overpotential in cycle 1 with respect to cycle 500 such that a large positive value describes an extended activation, as a function of the Co at.%. Note that the overpotential is not defined for less than 100 CV cycles activated NiO, since the current density of 10 mA cm⁻² _geo was not reached below 1.8 V vs RHE.

Figure 5

a) XPS of the O1s spectra of the films after 500 CV cycle activation, normalized to the maximum intensity of each trace. b) Ratio of hydroxide to (oxide + hydroxide) contributions derived from (a).

Figure 6

a) The change in ECSA, defined as the adsorbate capacitance of the activated film after 500 CV cycles (C_activated) divided by the adsorbate capacitance of the pristine film (C_pristine), as a function of the change in overpotential Δη = η_activated−η_pristine. b) The corrected overpotential at a current density of 5 mA cm⁻² adapted to the change in ECSA as a function of the Co at.%.

Scheme 1

Schematic illustration of the ALD process employed for the NCO films. The precursor dosing is followed by an Ar purge/pump of 5s/1s, whilst the plasma dosing is followed by an O₂ purge/pump of 1s/3s.