Water electrolysis on La1−xSrxCoO3−δ perovskite electrocatalysts

Perovskite oxides are attractive candidates as catalysts for the electrolysis of water in alkaline energy storage and conversion systems. However, the rational design of active catalysts has been hampered by the lack of understanding of the mechanism of water electrolysis on perovskite surfaces. Key parameters that have been overlooked include the role of oxygen vacancies, B–O bond covalency, and redox activity of lattice oxygen species. Here we present a series of cobaltite perovskites where the covalency of the Co–O bond and the concentration of oxygen vacancies are controlled through Sr2+ substitution into La1−xSrxCoO3−δ. We attempt to rationalize the high activities of La1−xSrxCoO3−δ through the electronic structure and participation of lattice oxygen in the mechanism of water electrolysis as revealed through ab initio modelling. Using this approach, we report a material, SrCoO2.7, with a high, room temperature-specific activity and mass activity towards alkaline water electrolysis.

is based on a bounded 3-dimensional diffusion model, with a diagram of the model shown in the inset of (b) where the intersection of i vs. t -1/2 at i=0 corresponds to λ=a/√Dt, where λ is a dimensionless shape factor, a is the radius of the particle, D is the diffusion rate, and t -1/2 . 1-3 Given a, and choosing an appropriate shape factor, λ, the diffusion rate can thus be easily measured. For a full derivation of the model, see Supplementary  In this case, λ was chosen as 2, which is representative of a rounded paralelipid, halfway between the values for a sphere (λ=1.77) and a cube (λ=2. 26). The value of a=150 nm was used for all materials based on our previous work, and the relation of surface area to particle size for a sphere: SA=6/2aρ, where ρ is the density of LSCO (i.e. ρ=7.235 g cm -3 for LaCoO3). 4 Further, it has been shown that particle size distribution does not affect the measured diffusion rate Computed Pourbaix diagrams of Fe, Co and Ni in aqueous environment. 5 The reaction formation energy, ΔG, is equal to ΔE+ΔZPE-TΔS-URHE, where ΔE is the formation enthalpy; ΔZPE is the difference of zero point energies; ΔS is the formation entropy, and URHE is the reversible hydrogen electrode potential.
Details can be found in Supplementary Reference 6. 6 Supplementary Table 1: Lattice parameters of the LCO, LSCO and SCO samples, weight fraction of Co3O4 and reliability factors after Rietveld refinement from PXRD data. For the samples with x = 0 -0.4 the R-3c model was applied instead of the orbital-ordered monoclinic I2/a structure because detection of this weak distortion is beyond the resolution of our PXRD experiment. [16][17][18] In spite of the tetragonal superstructure detected by SAED in the x = 0.8 sample, the refinement has been performed with the cubic Pm-3m model because neither reflection splitting nor superstructure reflections are detected in the PXRD pattern. Scherrer analysis was performed on all samples on the peak at 2θ ≈ 47°, using the following formula: = , where τ is the mean crystallite domain size (nm), K is a dimensionless shape factor (0.9 for spherical particles), λ is the X-ray wavelength (0.15418 nm), β is the line broadening at FWHM (radians), and θ is the Bragg angle (°). (1) (7) 15.519(2) not present n/a, 0.111 15.7 * -refined as cubic Pm-3m, tetragonal ap x ap x 2ap detected from SAED. ** -unit cell parameters refined using Le Bail procedure (see Fig. S1).

Supplementary
ǂ -mean crystallite domain size, derived from Scherrer Analysis of the peak at 2θ ≈ 47° Supplementary Table 2: Brunauer-Emmett-Teller (BET) N2 adsorption surface areas for oxides in this study. The surface areas were determined using a 7-point fit of the linear region of the adsorption isotherm.
x in La1-xSrxCoO3-δ BET Surface Area (m 2 g -1 ) 14.5 Supplementary Table 3: Computed reaction enthalpy changes of both LOM and AEM on SrCoO3 for the construction Figure 8d. ΔEi is labeled in the same order as shown in Figure 8d. As can be seen in Supplementary Figure 8, the role of carbon has a negligible effect on the activity of IrO2 both in geometric current density and in specific current density, yielding activities that corroborate the activities found in 3-electrode thin film RDE testing described in the literature. However, as has been noted in a number of papers, the low conductivities of perovskites require the use of carbon additives to help wire the catalysts to the electrode surface and to provide good electrical conductivity to the oxides. We have optimized our electrode preparation to maximize the gravimetric current density by using a composite of 30 wt% perovskite supported on a number of carbons and a total mass loading of 51 ugtotal/cm 2 geom. We have also tested our catalysts using the protocol described in Supplementary Reference 20 (shown in the plots as "85wt%", and have found that the high electrode mass loadings (~300 ugtotal/cm 2 geom) and relatively low amount of carbon (15 wt%) detract from the material performance and therefore do not describe the "true" activities of the catalysts due to inadequate wiring of the catalysts and poor utilization of the catalysts from increased mass transfer constraints in the thicker films. 20 In addition, we use a technique of spin coating our catalysts inks to provide uniform electrode films based on a study of electrode drying techniques. 21 It should be noted that in   Figure 9). 13,23,24,[26][27][28][29][30][31][32][33] The high activities of Ba0.5Sr0.5Co0.8Fe0.2O3-δ (δ = 0.4) as well as other layered double perovskite cobaltites can also be rationalized in terms of this vacancy mediated mechanism due to their high oxygen vacancy concentrations and high electronegativity of the active Co site. 8,9 Further, the vacancy mediated mechanism is not exclusive to perovskites nor alkaline electrolytes. Through oxygen isotope exchange measurements, lattice oxygen has been demonstrated to be involved during the OER on the precious metal-oxides, IrO2 and RuO2, as well as layered-oxides, such as LiNiO2, and spinel oxides, like Co3O4. [34][35][36][37][38][39] Therefore the covalency parameter and the vacancy content should both be taken into account in the rational design of oxygen evolution catalysts. It also consistent with previous models of OER activity scaling with d-electron filling and bulk oxide formation energies, as the covalency of the M-O bond is expected to increase with the electronegativity of the transition metal. 23,30,40,41 Supplementary Methods:

General:
All chemicals were used as received.

Surface Atomic Characterization:
The SCO sample was handled in an Ar-filled glove box. TEM specimen was prepared in the glove Rotation rates of ω = 400, 800, 1200, 1600, and 2000 rpm were used.

Oxygen Diffusion Rate Measurements:
Glassy carbon rotating disk electrodes with a thin layer of perovskite/VC were initially cycled in Ar saturated 1M KOH at 20 mV s -1 for 2 cycles. The E1/2 of oxygen intercalation/de-intercalation was determined from the halfway potential between the anodic and cathodic peaks. For materials with low oxygen vacancy concentration (i.e. x ≤ 0.4) the E1/2 was determined as the open-circuit potential. After being cycled, the oxygen ion diffusion rate was measured chronoamperometrically, using the same electrode, by applying a potential 50 mV more anodic of the E1/2. The electrodes were spun at 1600 rpm to get rid of electrolyte based mass transfer effects, and the current was measured as a function of time for 4 hrs. The current was plotted versus t -1/2 and the linear section of the curve was fit to find the intercept with the t -1/2 axis.

Measurement of RHE Potential:
Initially   Figure 7, Main Text), further increasing the relative stability of I1 to I0 and thus do not qualitatively change our results.
We consider [001] surfaces, as these are observed experimentally. Along the (001) Figure 8a). We note that LSCO is observed to undergo phase changes with increased oxygen deficiency. Thus, the rhombohedral LaCoO3 and the SrCoO2.75 are also investigated (Supplementary Figure 8b-d). We find that neither the phase nor the bulk oxygen stoichiometry changes the relative stability of I1 to I0 for the whole range of LSCO compositions.
To elucidate the lattice-oxygen-participated (LOM) OER mechanism on LSCO, we begin by comparing the free energy (stability) of each adsorbate-evolution-mechanism (AEM) intermediate