La-based Transition Metal Oxide Perovskites as Electrocatalysts for Electrochemical Carbon Dioxide Reduction

We report here the feasibility of using LaTMO3-based perovskites (TM = Co, Cr, Fe, Mn, Ni, i.e., non-Cu 3d transition metals) as electrocatalysts for electrochemical CO2 reduction reaction (eCO2RR). Phase pure LaTMO3 perovskites, having TM-ions in multiple oxidation states for all and O-defects for LaFeO3 and LaNiO3, have been synthesized and tested as electrocatalysts for eCO2RR in flow cell type set-up. The above characteristics of the La-TM-oxides have been found to influence the current densities during eCO2RR at the various applied potentials, with favorable effects of the presence of O-defects (as for LaFeO3 and LaNiO3). Upon eCO2RR, both C1 and C2 liquid products have been obtained, including ethanol, with a partial current density of −2.66 mA cm−2 at −1.2 V vs RHE (for LaFeO3). The types of products and the faradic efficiencies have been found to depend on the TM-ion present (in the LaTMO3); in particular, the oxidation state(s), associated O-defect(s) and electronic conductivity. Furthermore, the electrocatalysts have been found to be stable during eCO2RR. Overall, the present work highlights the potential of La-TM-oxide perovskites for usage as stable electrocatalysts for eCO2RR, and also provides insights into the proper selection of “TM” and reaction conditions for obtaining the desired product(s).

Mitigation of unwanted carbon dioxide to reduce global warming and balancing the atmospheric gas concentrations in the environment is a challenge.Among the various routes for reducing carbon dioxide concentration, electrochemical route has significance due to the possibility of formation of valuable chemicals and fuels as products of the reduction process upon the usage of proper electro-catalysts and suitable tuning of the electrochemical conditions. 1 Usage of metallic electro-catalysts, such as Au, Ag, Cu, Sn, Zn etc, [2][3][4] has been more common; but restricts the range of products that can be formed and also suffers from long term stability. 57][8][9] In this regard, transition metal (T M ) oxides may be considered as promising candidates, especially due to the possibility of the T M changing oxidation states and/or creating oxygen defects [10][11][12] which may further aid the electrochemical CO 2 reduction process; thus, enhancing the current density, faradic efficiency associated with the electrochemical CO 2 reduction reactions (eCO 2 RR), as well as the formation of a range of C 2 , C 3 and higher organic molecules as products.However, it is usually Cucontaining oxides that have been associated with the formation of such higher order products (i.e., C 2 , C 3 products) upon eCO 2 RR, 1,5,6,[13][14][15][16][17] often with a focus on the effects of pre-reduction of the oxide(s) to metallic Cu (for example, CuO-derived Cu), as also revealed in our previously published work. 18In the context of non-Cu T M -based electrocatalysts, in a couple of earlier works, Ni-Zn and Ni-Ga alloys have been reported to yield C 2 products. 19,20n the above contexts, in 1993 Schwartz et al. 13 first reported that perovskite structured Cu-containing La 1.8 Sr 0.2 CuO 4 can reduce CO 2 to methanol, ethanol and n-propanol, with total faradic efficiency of 40%, using 0.5 M KOH electrolyte; with La 1.8 Sr 0.2 CuO 4 being found to be structurally stable during eCO 2 RR.More recently, Mignard et al. 14 observed the formation of ethylene, methane and CO upon the usage of La 1.8 Sr 0.2 CuO 4 as electrocatalyst for eCO 2 RR in 0.5 M KOH electrolyte.Whittingham et al. 15 reported that undoped La 2 CuO 4 can produce formate and acetate only using 1.0 M KOH electrolyte, whereas Singh et al. 16 reported Faradic efficiencies of 40.3% and 4.1% towards the formation of ethylene and methane, respectively, using 0.5 M NaHCO 3 electrolyte at a low overpotential of −0.4 V vs RHE, with the presence of Cu in its reduced state (i.e., Cu + , as opposed to the starting Cu 2+ ) after eCO 2 RR.Chen et al. 21reported that La 2 CuO 4 can reduce CO 2 to methane with faradic efficiency of 56.3% at a potential of −1.4 V vs RHE, with the Cu/La 2 CuO 4 interface being deemed responsible for the methanation.Furthermore, Wang et al. 17 reported that the Cucontaining perovskite structured La 2 CuO 4 , with nano-bamboo type morphology, with accrued grain boundary region, can produce ethylene with faradic efficiency of 60% with no change in oxidation state of Cu during eCO 2 RR.
However, the present literature base contains only a very few reports concerning the usage of T M -oxides, in particular, non-Cu containing La-based T M -oxides, sans doping/modification, and those having perovskite structure (LaT M O 3 ), as electrocatalysts for CO 2 reduction under ambient conditions. 22,23For example, Hwang et al. 22 reported that LaCoO 3 can produce methane at low overpotentials of −0.6 to −0.8 V vs RHE, while maintaining structural stability during eCO 2 RR.In a different work, Cardona et al. 23 reported that La 0.5 Ba 0.5 CoO 3 can produce CO (preferentially on the (011) surface plane) with faradic efficiency of 85% at −1.2 V vs SHE and formate (preferentially on the (011) and (110) surface planes) with faradic efficiency of 30% at −0.9 V vs SHE.
From a generic perspective, perovskites are usually stable in terms of the structure, [24][25][26][27][28] with the possibility of tuning the A (here, La) and B (here, T M ) sites with suitable elements, [13][14][15][16][17][21][22][23] leading to a variation of the features/properties; which can be interesting to explore in the context of eCO 2 RR. Furthemore, the effects of usage of different set-ups and electrolytes for conducting electrochemical CO 2 reduction with such electrocatalysts also need to be looked into in the practical context.
Against this backdrop, we have developed La-based non-Cucontaining T M -oxides (T M = Co, Cr, Fe, Mn, Ni), possessing perovskite structure, and investigated them as potential electrocatalysts for eCO 2 RR using custom-made flow cell type set-up.0][31][32][33][34] The focus of this work has been primarily z E-mail: a.sarkar@iitb.ac.inECS Advances, 2024 3 020502 on the influence of the T M -ions, the mixed valence state of the same and the presence of oxygen defects (such as O-vacancies and/or "oxidized oxygen" species) 32,33,35,36 on the current density and type of products that form upon electrochemical CO 2 reduction.The results and inferences, as obtained in this work, not only establish the potential of La-T M -oxide perovskites as compositionally, structurally and electrochemically stable electrocatalysts for electrochemical CO 2 reduction, but also provide useful information regarding the selection of La-T M -oxide and conditions for obtaining the desired product type(s).as a fuel in 1:1 molar ratio.The resultant solution was placed on a hot plate, with continuous stirring @ 250 rpm, at 120 °C till the formation of gel.Then the temperature was increased to 300 °C and the gel was expanded to form "ash." This synthesis process is known as "citrate solution combustion."The as-synthesized "ash" was ground in an agate mortar and then taken in a quartz boat for calcination in air atmosphere at 850 °C for 6 h, with the heating rate being 10 °C min −1 .The calcined powders were again ground using agate mortar.

Experimental
Materials characterizations.-Thephase/structure of the Labased T M -oxides (T M = Co, Cr, Fe, Mn, Ni) before and after electrochemical CO 2 reduction were checked with X-ray diffraction (XRD; EMPYREAN PANalytical diffractometer, operating at 30 mA and 40 kV) using Cu Kα radiation and scan rate of 1°/min.Inorganic Crystal Structure Database (ICSD) was used for the phase analysis.Fullprof suite software 37 was used for Rietveld refinement and Vesta software 38 for generating a view of the crystal structure.The elemental compositions of the catalyst powders were quantitatively analyzed using energy dispersive spectroscopy (EDS) in QUANTA 200 scanning electron microscopy (SEM).Morphologies of the powder particles were observed using FEG-SEM (JEOL JSM-7600F) and the specific surface areas measured by BET (Brunauer-Emmett-Teller) using Micromeritics 3 FLEX system.X-ray photoelectron spectroscopy (XPS) was conducted with Al Kα monochromatic radiations in Kratos AXIS-supra analytical system.Surface charge correction of the binding energies of the elements was done with adventitious carbon C 1 s (C-C) at the binding energy of 284.8 eV.The binding energies of the survey and high-resolution scans were deconvoluted and determined based on XPSPEAK41 software and the NIST XPS database.
The electronic conductivities (σ e ) of the La-based T M -oxides (T M = Co, Cr, Fe, Mn, Ni) were estimated via DC polarization.The electronic conductivities were measured by applying a constant voltage (∼0.5 V) using Keithley system source meter (Model 2635B) and estimated as per σ e = tI/SΔE, where, t is the thickness of the LaT M O 3 pellet (0.1 cm), I is the steady state or residual current, S is the surface area in contact with the Ag electrodes (0.785 cm 2 ) and ΔE (0.5 V) is the applied voltage.
Electrochemical set-up and electrochemical CO 2 reduction.-Theset-up, i.e., electrochemical cell, as used here for performing the electrochemical carbon dioxide reduction reactions (eCO 2 RR) with the La-T M -oxide electrocatalysts is a custom-made flow cell type electrolytic cell set-up (as presented in Fig. 1).The "ink" for the working electrode was prepared using 800 μl of ultra-pure water, 200 μl of acetone, 14 mg of La-T M -oxide catalyst and 60 μl of basic nafion (binder; from Merck); which were taken in a centrifuge tube and sonicated for 30 min.Cyclic voltamograms (CVs) were performed at a scan rate of 50 mV s −1 in both N 2 and CO 2 saturated 0.5 M NaHCO 3 solution.For CVs, 6 μl of the same "ink" was dropcasted onto the polished glassy carbon rod electrode (3 mm diameter) and dried under IR lamp.The pHs of the 0.5 M NaHCO 3 solution after N 2 and CO 2 purging for 30 min were 8.5 and 7.2, respectively.The CVs have been reported in terms of current density (μA/cm 2 ) vs potential (V) (without iR correction).
In the flow cell type set-up (as in Fig. S1a in SI) three electrodes, viz., cathode, anode and reference (see Fig. S1b in SI) were used during the experiment.Gas diffusion electrode (see Fig. S1c in SI), 2.5 cm × 2.5 cm in size, was used as the cathode, Pt-mesh of 2.5 cm × 2.5 cm size was used as the anode and Hg/HgO was used as the reference electrode.The difference between Hg/HgO and RHE was found to be −0.90V. Hence, the potential applied using Hg/HgO reference electrode was converted to RHE scale using the relationship; E RHE = E Hg/HgO + 0.9 V v/s RHE.500 μl of IPA (isopropyl alcohol), 50 mg of electrocatalyst powder and 50 μl of 5 wt% basic nafion (binder, from Sigma-Aldrich) were taken in a glassware and sonicated for at least 15 min.The as-prepared solution, i.e., the "ink," was brush painted on a 2.5 cm × 2.5 cm PTFE (polytetrafluoroethylene) coated carbon paper (from Sainergy Fuel Cell India Pvt.Ltd) in orthogonal directions, alternatively.The electrode/GDE (see Fig. S1c in SI) was then kept in a petri dish for 1 day for drying under ambient condition.
A 0.50 M potassium hydroxide (KOH, 84%; from EMPLURA, Merck) solution was prepared by taking required amounts of millipore water (18 MΩ cm) and KOH pellets in a 500 ml volumetric flask.The pH of the as-prepared electrolyte was 13.69.High purity carbon dioxide was purged into the electrolyte for at least 30 min before performing the electrochemical carbon dioxide reduction reaction, with the pH of the electrolyte being 13.56.The cathode and anode sides were supported by 5 × 5 cm sized square stainless-steel plates, with flow channels designed onto them (see Fig. 1a).Carbon dioxide was passed through the flow channel, which diffused to the back side of the gas diffusion electrode (GDE) to reach the catalyst surface.The carbon dioxide flow rate was maintained at 35 ml min −1 .A membrane (from Celgard, North Carolina, USA) was placed on the catalyst surface for separating the gas diffusion layer from the gasket (5 mm thickness).The celgard separator was used to avoid electrolyte flooding of the GDE. 39The electrolyte was passed onto the cathode side through the fiber filled capillary.The other side of the capillary was connected to the reference electrode.The electrolyte flow from a storage tank to the cathode side was maintained by a peristaltic pump at 60 ml h −1 .The PTFE plate was physically separated from the anode side by a 5 × 5 cm and 130 μm thick PEEK-reinforced anion exchange membrane (FUMATECH BWT GmbH).The electrolyte from the anode side was re-circulated using the peristaltic pump at 80 ml min −1 .The PTFE plate was filled with 2.5 × 2.5 cm sized PPS fibers (polyphenylene sulphide; from TORCON TM by TORAY).The product got soaked onto the fibers during the reduction reaction and came out from the flow cell type electrolytic cell set-up (see Figs. S1a, S1b in SI) as liquid product.The mechanism of the flow cell type electrolytic cell operation during eCO 2 RR (as depicted in Fig. 1b and the following reactions) involves reduction of CO 2 to products, formation of H 2 and OH -ions at the cathode side.The OH - ions traverse to the anode side through the anion exchange membrane and produce O 2 and H 2 O.
A few electrochemical CO 2 reduction reactions (eCO 2 RR) at the cathode side (i.e., at the working electrode) are shown below; Oxygen evolution reaction (OER) at the anode side (or counter electrode): Using this set-up, the eCO 2 RR was performed via chronoamperometry using Biologic potentiostat (Model VSP 300), with applied potentials being −0.2, −0.4,−0.6, −0.8, −1.0 and −1.2 V vs RHE ("iR effect" not corrected) for 3 h each.The current (mA) vs time (min) was extracted, plotted and analyzed using EC-Lab 10.40 software.The partial current density (mA/cm 2 ) was estimated by multiplying the current density and product faradic efficiency at each of the applied potentials.
Product analysis.-Theliquid products contained in the electrolyte, were analyzed using high performance liquid chromatography (HPLC; Agilent 1260 infinity series model) with Aminex HPX-87H column (supplied by BIO-RAD, USA).The mobile phase of the HPLC was 13 mM H 2 SO 4 and the pH was kept at 2. Before injecting the product into the HPLC, a required amount of 1 M H 2 SO 4 was added to it in order to maintain the pH of 2. All the samples were measured three times to obtain the mean and standard deviation.The hydrogen was detected by gas chromatography-thermal conductivity detector (GC-TCD) using Nucon 5700 gas chromatography at a temperature of 60 °C and current of 100 mA, with the carrier gas being Ar.

Results and Discussion
Phase, morphology and surface area of the electrocatalysts.-TheXRD scans recorded with the calcined La-T M -oxides (T M = Co, Cr, Fe, Mn, Ni) indicate the development of phase pure materials, possessing the desired perovskite structures, for LaCoO 3 , LaCrO 3 , LaFeO 3 , LaMnO 3 and LaNiO 3 ; see Fig. 2. The space groups of LaCoO 3 (ICSD 17668), LaCrO 3 (ICSD 91271), LaMnO 3 (ICSD 96038), LaNiO 3 (ICSD 84933) were found to be R-3cH and that for LaFeO 3 was Pnma (ICSD 164083).Rietveld refinement of the XRD data (as presented in Fig. S3 of SI) was performed for obtaining the crystallographic parameters, viz., lattice parameters (a, b, c) and unit cell volume (V), which have been presented in Table I, along with the associated R-factors (R p , R wp , χ 2 ).The crystallographic parameters of all the five LaT M O 3 s (where, T M = Co, Cr, Fe, Mn and Ni) agree fairly well with those reported in the literature. 40,41The crystal systems have also been presented in Fig. S4 of SI.Compositional analysis using EDS (see Table II), as well as XPS survey scans between 0 to 1200 eV (see Table S1 of SI), confirm the attainment of the desired stoichiometry.
Oxidation states of the transition metal ions and oxygen defects.-XPS,which yields the chemical state of ions, being a surface sensitive technique is particularly relevant to electrocatalysis since catalysis also involves primarily surface reactions.XPS scans obtained with the as-synthesized La-T M -oxide (perovskite) based electrocatalysts revealed that the binding energy of La 3d 5/2 in the case of all the five types of perovskite based electrocatalyst powders particles is ∼836.12eV (see Fig. S5 of SI).The adventitious carbon peak position in C 1 s is at ∼284.8 eV, as per NIST standard database, which has also been used here as a reference for charge correction.
Deconvolution of the high-resolution XPS spectra for the T M -ions indicate the presence of mixed valence (oxidation) states, i.e., (a) 53% Co 3+ (at 779.69 eV; Co 2p 3/2 ) and 47% of Co 4+ (at 780.87 eV; Co 2p 3/2 ) in the case of LaCoO 3 , (b) 24% Cr 3+ (at 575.91 eV; Cr 2p 3/2 ), 59% Cr 4+ (576.97 eV; Cr 2p 3/2 ) and 17% Cr 6+ (at 579.66 eV; Cr 2p 3/2 ) in the case of LaCrO 3 , (c) 76% Fe 2+ (at 709.56 eV; Fe 2p 3/2 ) and 24% Fe 3+ (at 711.57[41][42][43][44] While the T M -ions in four of the La-T M -oxide based perovskites are present in two oxidation states (viz., as for LaCoO 3 , LaFeO 3 , LaMnO 3 and LaNiO 3 ), with LaFeO 3 and LaNiO 3 having the respective T M -ions partly in the lower oxidation state of +2, 32,33 Cr in LaCrO 3 exhibits three oxidation states (i.e., +3, +4 and +6). 44t may be mentioned here that, as per the stoichiometry of the perovskites, the T M -ions, which form the B-site of the ABO 3 -type perovskite composition, are expected to possess an average oxidation state of +3.In this regard, along with the mixed valence states of T M -ions, which is known to be good towards electronic conductivity of T M -oxides (see Table S1 of SI), the presence of T M -ions in oxidation states lower than +3 (i.e., oxidation state of +2 for LaFeO 3 and LaNiO 3 ) is also likely to induce (or be the result of) oxygen defects (such as, O-vacancies and/or "oxidized oxygen" species), which are also reported to be beneficial towards electrocatalytic activities. 32,33Deconvolution of the high-resolution O 1 s XPS spectra indicates the presence of lattice oxygen (viz., as oxygenmetal bond or O 2− ) and absorbed O-containing species (viz., hydroxyl group and adsorbed water) for all the five perovskites and, in the cases of LaFeO
The Electrocatalyst Target atomic % As-obtained atomic % (as per EDS) Particle sizes with standard deviations (nm) BET surface area (m 2 /g) indicates that the contents of O − /O 2 − /O vacancies are ∼21.4% and ∼6.8% for LaFeO 3 and LaNiO 3 , respectively, which tend to agree with the relatively greater content of the corresponding T M -ion in the +2 oxidation state for the former (see Fig. S6 in SI).Overall, the mixed valence state of T M -ions and the presence of oxygen defects (only for LaFeO 3 and LaNiO 3 ) 32,33 are deemed to be helpful towards their application as electrocatalysts.
Here, it may be noted that the XPS results and the associated discussion solely indicates the conditions of the electrocatalysts in the as-synthesized state (i.e., the starting conditions of the catalysts for eCO 2 RR).In general, the perovskite structure and oxidation states of the T M -ions in the structure are believed to be fairly stable under most electrochemical conditions (including for eCO 2 RR).Nevertheless, it is not possible to rule out that right at the negative potentials used for eCO 2 RR there may be some change in the oxidation state of the transition metal ions; but that is expected to be over and above the oxidation states and presence/absence of Ovacancies/defects in the pristine or starting conditions.Furthermore, the eCO 2 RR conditions used in this work are exactly the same for all the catalysts and, hence, the changes in oxidation state(s), if any, during eCO 2 RR are not expected to cause too much deviation in the context of the relative ranking across the catalysts as compared to in the pristine condition.Hence, it is strongly believed here that the starting/pristine condition of the electrocatalysts is expected to have significant impact(s) on the electrochemistry associated with the eCO 2 RR process and, concomitantly, on the performance(s) of the electrocatalysts (as presented in Electrochemical behavior and CO 2 reduction section below).The current densities obtained in the CO 2 saturated 0.5 M NaHCO 3 solutions are higher than the current densities obtained in the N 2 saturated 0.5 M NaHCO 3 solutions for all the La-T M -oxide electrocatalysts.Furthermore, as an example based on chronoamperometry experiments in the flow cell type set-up, it can be seen from Fig. S9 in SI that the current densities are significantly higher at all the applied potentials of −0.2, −0.4,−0.6, −0.8, −1.0, −1.2 V (vs RHE) when CO 2 purged 0.5 M KOH solution is used, as compared to when Ar purged 0.5 M KOH solution is used.Also, at all these potentials, only H 2 has been observed as the product when Ar purged 0.5 M KOH solution, while other eCO 2 RR products (along with H 2 ) have been obtained when using CO 2 purged 0.5 M KOH solution.Hence, in a way, all these indicate that the perovskite structured La-T M -oxides under consideration here are efficient as electrocatalysts for electrochemical CO 2 reduction.
Current densities obtained during eCO 2 RR.-Electrochemical CO 2 reduction reactions (eCO 2 RR) were performed and associated current densities measured at potentials of −0.2, −0.4,−0.6, −0.8, −1.0, −1.2 V vs RHE using the flow cell type set-up.As expected, in general, the current densities increased upon moving towards higher potentials.Overall, higher current densities were obtained with the LaFeO 3 , LaNiO 3 electrocatalysts, as compared to the other LaT M O 3 s used here under the same conditions (as shown in Fig. 5).The maximum current density was found to be ∼38.8mA cm −2 , as for LaFeO 3 at the potential of −1.2 V vs RHE.
Tafel(-type) plots based on the current densities (j) (as log j) obtained at the applied potentials (V) of −0.2, −0.4,−0.6, −0.8, −1.0, −1.2 V vs RHE during eCO 2 RR with the five La-T M -oxides (T M = Co, Cr, Fe, Mn, Ni) have been presented in Fig. 6a.An immediate observation is that log j is maximum at all the six applied potentials in the case of LaFeO 3 , followed by those for LaNiO 3 .Furthermore, the slope of the fit to the Log j vs V data points is the least in the case of LaFeO 3 (viz., ∼525 mV/dec), followed by that for LaNiO 3 (viz., ∼659 mV/dec), which implies that the current densities increase to a greater extent for a given change in applied potential (i.e., overpotential) in the cases of these two La-T M -oxide electrocatalysts and, thus, point towards favorable eCO 2 RR kinetics bestowed by LaFeO 3 and LaNiO 3 .In the above context, it is also encouraging to see that among the five LaT M O 3 s under consideration here, LaNiO 3 and LaFeO 3 possess considerably higher electronic conductivities than the rest (as presented in Fig. 6b).
Interestingly, the current densities obtained with LaMnO 3 were found to be on the lower side (as compared to the other electrocatalysts), despite the electrocatalyst possessing the highest specific surface area among all the counterparts.This indicates that it is the inherent characteristics of the as-developed electrocatalyst materials, in terms of the oxidation state(s) of the T M -ions, the stability of the same, O-defect(s), concomitant electronic conductivity and/or adsorption properties of the gaseous reactants/products, and not the specific surface area per se, that dominate the performance as electrocatalysts for CO 2 reduction.For example, the instability of Mn-ions in its +3 oxidation state is well known 5 and the electronic conductivity of LaMnO 3 has been found to be the second lowest among the five LaT M O 3 perovskites (see Fig. 6b).Similarly, the presence of Cr 6+ , as found here, is known to negatively influence the electronic conductivity 45 and, not surprisingly, our measurements also reveal that LaCrO 3 indeed possesses the lowest electronic conductivity among the LaT M O 3 s under consideration here (see Fig. 6b).Hence, in addition to LaMnO 3 , the current densities with LaCrO 3 have also been found to be on the lower side.On the other hand, the current density is consistently high for LaFeO 3 , with it being the highest at all the potentials between −0.4 and −1.2 V.The presence of O-defects (viz., O − /O 2 − /O vacancies ), that too in considerable amounts (as mentioned in Oxidation states of the transition metal ions and oxygen defects section), as well as enhanced electronic conductivity (as shown in Fig. 6b) in such a mixed valence state, is believed to be an important factor that contributes to the high current densities obtained with LaFeO 3 .In the context of Ovacancy/defect, it may be recalled here that, while LaNiO 3 and LaFeO 3 are the two catalysts which have been found to possess oxygen vacancies/defects, the O-vacancy/defect content is greater in the case of LaFeO 3 (viz., ∼21.4%) than for LaNiO 3 (viz., ∼6.8%).On a different note, the relative variations of the current densities between the La-T M -oxide based electrocatalysts used here also tend to qualitatively agree with those reported in possibly the only other prior work with similar perovskite structured La-T M -oxide electrocatalysts for eCO 2 RR 22 ; which also indicates that it is the materials, or in more specific terms, the T M -ions that primarily influence the electrocatalytic activity.
Products obtained during eCO 2 RR.-The products obtained upon eCO 2 RR were found to be strongly dependent on the electrocatalyst, i.e., the transition metal ion present, the associated oxidation states and the presence/absence of O-defects/vacancies (at/close to the surface), which have been evaluated using XPS, which is primarily a surface characterization technique.Of course, the possible additional influence(s) of the specific surface area and the electronic conductivity, which have also been found to be different for the different electrocatalysts under consideration here, are not ruled out.In this flow cell type set-up (having 0.5 M KOH as electrolyte; pH of 13.6), the five electrocatalysts could not reduce CO 2 to any useful product at −0.2 V vs RHE since the current densities were too low; but with HCOO − , CH 3 COO − , C 2 H 5 OH and H 2 obtained as the products at the higher potentials depending on the electrocatalysts but rigorous quantification was performed only for liquid samples.Only LaMnO 3 led to the formation of all the three types of products.LaCoO 3 and LaNiO 3 produced formate (HCOO − ) and acetate (CH 3 COO − ), while LaCrO 3 produced only formate and LaFeO 3 produced only ethanol (see Fig. 7).A maximum faradic efficiency of ∼6.8% was obtained for the formation of ethanol using the O − /O 2 − /O vacancy -containing LaFeO 3 at a potential of −1.2 V vs RHE (see Fig. 7c), while the next best faradic efficiency was obtained for the formation of HCOO − ( i.e., ∼2%) with LaCrO 3 as the electrocatalyst at the potential of −0.6 V vs RHE (see Fig. 7a).
Figure 7d presents the partial current densities (viz., product of current density and faradic efficiency) corresponding to the formation of ethanol upon usage of LaFeO 3 as the electrocatalyst.The partial current densities can be observed to be −0.64 and −2.66 mA cm −2 at the applied potentials of −1.0 and −1.2 V vs RHE respectively.One may also observe that, LaCrO 3 favors the formation of HCOO − , whereas LaMnO 3 favors CH 3 COO − .LaMnO 3 is also effective in terms of formation of multi-products at relatively lower potentials, since it led to the formation of acetate and ethanol at −0.4 V vs RHE (see Figs. 7b and 7c).We may recall here that the BET surface area is the highest for LaMnO 3 and relatively lower for LaFeO 3 .Despite that, LaFeO 3 has been found to produce C 2 products with higher faradic efficiencies.Here, again, the presence of Fe ion in its +2 oxidation state, the associated O-defects (viz., O − /O 2 − /O vacancies ) (at least at the surface; as revealed by XPS) which possibly aids in favorable surface interaction of the catalyst with CO 2 and related intermediates, as well as the higher electronic conductivity, are believed to be the important contributing factors.
All the liquid products obtained upon eCO 2 RR using the Labased transition metal oxide electrocatalysts (viz., LaCoO 3 , LaCrO 3 , LaFeO 3 , LaMnO 3 , LaNiO 3 ), under consideration here, have been  S2 in SI.Table S3 in SI further presents a comparison of the eCO 2 RR products obtained with the present LaT M O 3 electrocatalysts, with the perovskite-type electrocatalysts reported in the literature (both Cu-and non-Cu containing).The capability of the LaT M O 3 s towards the formation of various useful chemicals, including C 2 products, under certain electrochemical conditions (viz., electrolyte-type, set-up etc), despite being undoped and not containing Cu, can be immediately noted.Interestingly, the only two products (other than H 2 ) reported by Hwang et al. 22 upon using similar La-T M -oxide perovskites for eCO 2 RR happen to be C 1 products, viz., CO and CH 4 .Furthermore, the faradic efficiencies corresponding to all the C 1 and C 2 products that were detected in the present work are considerably greater than those reported in Ref. 22  which, to the best of the author's knowledge, happens to be the only other work that reports the usage of LaT M O 3 perovskites as electrocatalysts for eCO 2 RR.Further studies on such perovskitebased electrocatalysts, with varied electrochemical conditions for eCO 2 RR and materials characteristics will aid obtaining more  ECS Advances, 2024 3 020502 insights into the effects of T M -type, oxidation state(s), oxygen defects, electronic conductivities and T M -O covalency on the electrocatalytic behaviour/performance.Electrochemical, structural, compositional stability of the electrocatalysts (post-eCO 2 RR).-In order to obtain insights into the electrochemical stability during eCO 2 RR upon usage of the LaCoO 3 , LaCrO 3, LaFeO 3, LaMnO 3 , LaNiO 3 electrocatalysts at the chronoamperometry potentials of −0.2, −0.4,−0.6, −0.8, −1.0 and −1.2 V vs RHE, the current densities (as averages of data obtained for 30 min) have been plotted as a function of time for 3 h and presented in Fig. 8.The plots reveal good stability of the current densities during eCO 2 RR with all the five electrocatalysts.
In order to check for the structural stability of the electrocatalysts upon being used for eCO 2 RR, XRD was performed with the five electrocatalysts after eCO 2 RR at each of the applied potentials, as shown in Fig. 9.Other than minor peaks from the carbon paper (for all) and deposited KOH (visible for LaMnO 3 , LaNiO 3 ), only the peaks corresponding to those of the electrocatalysts could be detected in the XRD patterns obtained post eCO 2 RR for LaCoO 3 , LaCrO 3, LaFeO 3, LaMnO 3 , LaNiO 3 .Hence, all the La-T M -oxides retain the perovskite structure upon being used for eCO 2 RR via chronoamperometry at the various potentials.From a practical perspective this is an important result since it indicates the stability of the electrocatalysts, in terms of retaining the overall structure, during the electrochemical CO 2 reduction process.Only one additional strong peak for fluorine (i.e., F 1 S) was found at a binding energy of ∼690 eV post electrolysis, which arises from the presence of some nafion binder which was used for the preparation of the gas diffusion electrode.The post eCO 2 RR XPS data also confirms that the oxidation states of all the elements in the LaCoO 3 , LaCrO 3, LaFeO 3, LaMnO 3 , LaNiO 3 electrocatalysts get maintained at the same levels as in the pre-eCO 2 RR conditions.Overall, the above post-eCO 2 RR characterizations confirm the stability of the perovskite structured La-T M -oxides under consideration here.

Figure 1 .
Figure 1.(a) Schematic representation of the flow cell type electrolytic cell, as used for electrochemical CO 2 reduction reaction via chronoamperometry; (b) Mechanism of eCO 2 RR using the flow cell type electrolytic cell set up.
3 and LaNiO 3 , oxidized oxygen species (O − /O 2 − )/oxygen vacancies (O vacancies ) as oxygen defects (Odefect).In more specific terms, in the deconvoluted O 1 s spectra, the presence of O-La can be noted at binding energies of ∼528.54--529.23eV, O-T M at the binding energies of ∼529.33-529.64 eV, OH − at the binding energies of ∼ 531.05-531.77eV, adsorbed H 2 O at the binding energies of ∼532.01-533.18eV and oxygen defects at the binding energies of ∼530.05-530.08 eV (only for LaFeO 3 and LaNiO 3 ) (see Fig. S6 in SI).The presence of oxygen defects (Odefects) in the forms of oxidized oxygen species (O − /O 2 − ) and/or oxygen vacancies (O vacancies ) only for LaFeO 3 and LaNiO 3 agrees with the presence of Fe-and Ni-ions in their +2 oxidation state in the respective perovskites.Furthermore, qualitative analysis

Figure 7 .
Figure 7. Faradaic efficiencies corresponding to the formation of (a) HCOO − , (b) CH 3 COO − and (c) C 2 H 5 OH as the products of electrochemical CO 2 reduction in the flow cell type set-up, upon using LaCoO 3 , LaCrO 3 , LaFeO 3 , LaMnO 3 and LaNiO 3 as electrocatalysts at the applied potentials of −0.4,−0.6, −0.8, −1.0 and −1.2 V vs RHE.(d) Partial current density for the formation of ethanol at the applied potential of −1.0 and −1.2 V vs RHE upon usage of LaFeO 3 .