NL2014577B1

NL2014577B1 - Electrocatalysts for Efficient Water Electrolysis

Info

Publication number: NL2014577B1
Application number: NL2014577A
Authority: NL
Inventors: Theodorus Maria Koper Marcus; Johannes Raaijman Stefan; Alfonso Diaz Morales Oscar; Fu Wentian
Original assignee: Univ Leiden
Priority date: 2015-04-02
Filing date: 2015-04-02
Publication date: 2017-01-11
Also published as: NL2014577A; WO2016156599A1

Abstract

The present invention relates to an electrocatalyst suitable for oxygen evolution in electrolytic processes, such as the electrolysis of water, comprising a perovskite-like double perovskite compound having the following general formula: An+1 (BB')n/2 03n+1, wherein B represents Iridium, Ruthenium, Osmium, Rhenium, Rhodium, and/or combinations thereof, and to methods of its preparation and use.

Description

Electrocatalysts for Efficient Water Electrolysis

The present invention relates to novel electrocatalysts for the electrolysis of water, their use, and to electrolytic cells and systems comprising the novel electrocatalyst materials.

Background of the Invention

Modern society and its energy demands are expanding at an unprecedented rate, whilst fossil fuel reserves dwindle and the carbon dioxide mediated greenhouse effect flourishes. These events are foundation to the recent heightened interest in renewable energy sources. Promising green energy candidates such as solar, wind, and hydro energy all have their energy generation decoupled from demand and as such require a way for storing excess energy.

Currently available means are not adequate for large scale long term energy storage due to a variety of reasons amongst which efficiency and cost are of most significance. One promising way of storing energy with high power density is the electrochemical formation of combustible or convertible fuels such as hydrogen (H2) or hydrocarbons (CxHy).

However, the efficiency of many sustainable energy conversion and storage solutions, for instance metal-air batteries, fuel cells, or electrolyzers is inherently linked to, and limited by, the slow kinetics of the oxygen evolution reaction (further referred to herein as OER). Due to this, a high overpotential (η) is typically required to reach acceptable reaction rates, resulting in lower efficiency for these storage or energy conversion solutions. Water oxidation is the efficiency-limiting redox reaction of water electrolysis, and corresponds to the conversion according to the following formula : 2H20 —► 4-H+ + 4-e' + 02. Water oxidation may provide a particularly good access to protons (H+) and electrons (e') required for the storage or energy conversion solutions, as water is in abundance.

Generally, under conditions typical for water oxidation, i.e. at a potential of above 1.23 V vs. RHE, the majority of metallic surfaces acting as electrocatalysts will be covered by a metal oxide layer.

Presently used catalysts for water oxidation typically comprise precious metal oxides, such as Ir02 or Ru02. These catalysts, while typically offering an acceptable overpotential, are based on very rare resources and therefore quite expensive. Hence, the majority of research in recent years has focused on metal oxides of more abundant and cheaper 3rd row transition metals instead of the highly active precious metals, such as iridium and ruthenium.

Due to the inherent chemical instability of these metal ions, alkaline media are strongly favoured, as so far only under such conditions, suitable chemical stability and prolonged catalyst lifetime could be ensured. In acidic media, however, the 3rd row transition metal catalysts are highly unstable, and hence not useful as replacement for Ru02 and Ir02.

An electrocatalyst system has now surprisingly been found that has a strongly enhanced activity, is stable and in an acidic environment, and when compared to the Ru02 and Ir02 electrocatalysts described above, has a strongly reduced precious metal content.

Summary of the Invention

Accordingly, in a first aspect, the present invention pertains to an electrocatalyst suitable for oxygen evolution in electrolytic processes, comprising a perovskite-like compound having the following general formula: An+i (BB’)n/2 03n+i, wherein B represents Iridium, Ruthenium, Osmium, Rhenium, Rhodium, and/or combinations thereof.

In a second aspect, the present invention also relates to the preparation of electrocatalysts according to the present invention. Accordingly, the present invention relates to a method for the production of an electrocatalyst material, the method comprising: (a) providing a mixture of suitable precursor metal salts in a particle size and form suitable for the preparation; (b) subjecting the mixture to a temperature in the range of from 850°C to 950°C for a suitable period, preferably at least 2 hours, and (c) sintering the resultant powder at a temperature in the range of from 1200°C to 1300°C, preferably 1225°C to 1275 °C, for a suitable period, preferably at least 48 hours, optionally accompanied by frequent intermittent regrinding, and (d) furnace-cooling the obtained samples to room temperature. In a third aspect, the present invention also relates to systems comprising the electrocatalysts according to the present invention.

In an alternative method, the preparation of the electrocatalysts comprises (i) the preparation of a first solution containing the precursor of the materials , and (ii) depositing the material precursor from the solution onto a metal substrate by an induced solubility change, and (iii) subjecting the thus formed deposited substrate to a thermal treatment at a temperature in the range of from 300 °C to 1300 °C, wherein the material exhibiting at least in part a double perovskite structure is formed.

In a fourth aspect, the present invention also relates to the use of the electrocatalysts according to the present invention in electrolysis reactions, preferably for the oxidation of water, or as catalytic electrocatalyst material in solid/air cells or batteries.

The present electrocatalyst materials preferably comprises a perovskite-like double perovskite having the stoichiometric formula An+i (BB’)n/2 03n+i.

Without wishing to be bound to any particular theory, the three-dimensional arrangement of corner-share octahedra resulted in both in unprecedented catalytic activity and stability.

The double perovskite structure is closely related to the perovskite structure. The generic formula for a perovskite is AB03, where A is a large cation and B a smaller cation. Due to the difference in size, the coordination number for the A-cation is different than for the B-cation. In the case of a double perovskite, there are two similar cations on the B-site in a 1:1 ratio, resulting in ΑΒο.5Β’0.5θ3, or the generic form: A2BB’06.

Many quaternary oxides with the general formula A2BB’06 belong to the ordered double perovskites. Depending on the tolerance factor, t = (rA + r0)/2(r(B,B’) + r0), where γ(β,β) is the averaged ionic radii of B and B’-cations, the structure of double perovskites is either cubic (t«1), with the lattice parameter double of that of the primitive cubic perovskite and the space group Fm'3m, or distorted (t>1 or t < 1) with lower symmetry. This type of ordering of B and B’ is also referred to as the rock-salt sublattice. Due to the difference in ionic radii between B and B’, the structure may preferably attain the order as depicted in Figure 1. The double perovskite according to this invention preferably comprise alkaline earth metal ions to occupy the A-site, while transition metals tend to occupy the B-site. However, this is by no means absolute, with many different combinations of A and B cations known.

Preferably, A represents = (AA’)/2, wherein A and A’ each independently may represents a metal ion selected from alkali or earth alkali metals. Preferred earth alkali metal ions comprise Ba, Sr, and Ca, with a preference for Ba or Sr, most preferred being Sr due to the good results obtained therewith.

In the formula, B represents Iridium, Ruthenium, Osmium, Rhenium,

Rhodium, and/or combinations thereof. Preferably, B represents a metal ion selected from Iridium, Ruthenium, or combinations thereof, the most preferred being Iridium.

Preferably, B’ represents a metal ion selected selected from one or more rare earth metals such as lanthanides or actinides. Preferred are lanthanides such as La, Ce, Pr, Nd, Tb, Y, or Lu. More preferred are Cerium, Neodymium , Prasodinium, Lanthanum, and/or Lutetium. Most preferred, these ions are preferably at the 3+ or 4+ oxidation stage.

Preferably, each B’ atom is arranged in an octrahedral oxygen coordination.

In this way, the electrocatalyst comprises at least in part a layered perovskite crystalline structure, or a double perovskite crystalline structure comprising at least corner-shared octahedral oxygen coordination. It was found that layered structures, but preferably fully crystalline compounds with double perovskite structure resulted in both in unprecedented catalytic activity and stability.

Preferably, the electrocatalyst according to the invention is selected from the group comprising Ba2Celr06, Ba2Ndlr06, Ba2Tblr06 and Sr2Ylr06, Ba2Ylr06; Sr2lr04, or a combination thereof, due to the high activity in acidic media, and the low amount of iridium present in the material. Particularly preferred double perovskites include Ba2Prlr06 (containing lr4+); Ba2Ndlr06 (containing lr5+); Ba2Ylr06 (containing a nonlanthanide at the B-site) and Sr2Ylr06 (containing strontium at the A-site).

Advantageously, the electrocatalyst further comprises a metal substrate, and an external catalytic layer comprising the electrocatalyst material. Any suitable metal substrate may be employed. Typically, Ti or Pb substrates are employed due to their inherent stability and easy availability. The electrocatalyst layer can advantageously be applied as set out herein below.

Short Description of the Figures

In Figure 1 the schematic representation of a generic A2BB’06 double perovskite is depicted.

Figure 2 discloses the Ohmic drop corrected specific activity (j) of commercial Ir02 and Ir02 NPs (Nanoparticles) as a function of potential in 0.1 M HCI04 saturated with air. Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements. Catalyst loading was equal to 14.3 pg0Xide / cm2 disk, whilst for Ir02 NPs a mass loading could not be determined.

Figure 3 illustrates the Ohmic drop corrected specific activity 0 of commercial Ι1Ό2, lr02NPs, and Ba2l_nlr06(l_n = La, Ce, Pr, Nd, Tb) as a function of potential in 0.1 M HCIO4 saturated with air. Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements. Catalyst loading for plotted compounds was equal to 14.3 pg0Xide / cm2 disk.

Figure 4 discloses the Ohmic drop corrected mass activity (i) of commercial Ir02, Ir02 NPs (from Lee et al.p,]), and Ba2LnlrOe (Ln = La, Ce, Pr, Nd, Tb) as a function of potential in 0.1 M HCIO4 saturated with air. Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements. Catalyst loading for plotted compounds was equal to 14.3 pg0Xide / cm2 disk-

Figure 5 discloses the Ohmic drop corrected specific activity (j) of commercial I1O2, lr02NPs, Ba2PrlrOe, Ba2YlrOeand S^YIrOeas a function of potential in 0.1 M HCIO4 saturated with air, to allow for comparison between compounds of the invention with known compounds of differing Iridium levels. Data points are acquired via controlled-potential chrono-amperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements. Catalyst loading for plotted compounds was equal to 14.3 pg0Xide / cm2 diSk.

Figure 6 discloses the Ohmic drop corrected specific activity 0 of Ba2PrlrOeas a function of potential in 0.1 M NaOH (pH 13), 0.001 M HCI04(pH 3), 0.01 M HCI04(pH 2) and 0.1 M HCI04(pH 1). The total concentration of perchlorate (CIO4-) in each sample is held constant at 0.2 M. Data points are acquired via controlled potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements, except in the case of pH 13, where only 1 sample was measured. Catalyst loading for plotted compounds was equal to 14.3 pgoxide / cm2 diSk.

Figure 7 discloses the Observed and calculated profiles of SrTi03:lr(25%) in the space group Pm3m . Tick marks (|) indicate allowed reflections. Insert shows a close-up of the high 2Θ part. At the bottom, a difference curve (/obs- /caic) is shown.

Figure 8 discloses the Ohmic drop corrected specific activity (j) of commercial I1O2, lr02NPs and SrTiOs: lr(25%) as a function of potential in 0.1 M HCIO4 saturated with air. Data points are acquired via cyclic voltammetry scanning at 10 mV/s, whilst rotating at 1500 rpm. Insert shows a close-up of the low specific current region. Catalyst loading for plotted compounds was equal to 71.5pg0xide/ cm2 disk. The oxidation peaks in the 1st scans at 1.5 V, correspond to the oxidation of the Au substrate.

Figure 9 discloses the observed and calculated profiles of Ba2Pr(Nbo.5lro.5)C>6for the space groups Fm3m and 12 lm. Tick marks (|for Ba2Pr(ln-xNbx)Osand |for Ba2Pr(Nbi-ylry)C>6) indicate allowed reflections. Insert shows a close-up of the high 2Θ part. At the bottom, a difference curve (/obs- /cai) is shown.

Figure 10 discloses the observed and calculated profiles of Ba2Pr(Tao.5lro.5)C>6in the space group 12 lm. Tick marks (|) indicate allowed reflections. Insert shows a close-up of the high 2Θ part. At the

Figure 11 discloses the Ohmic drop corrected specific activity (j) of commercial I1O2, lr02NPs and Ba2Pr(Lno.5lro.5)06(Ln = Ta , Nb) as a function of potential in 0.1 M HCIO4 saturated with air. Data points are acquired via cyclic voltammetry scanning at 10 mV/s, whilst rotating at 1500 rpm. Catalyst loading for plotted compounds was equal to 71.5 pg0Xide / cm2 disk. The oxidation peaks in 1st scans at 1.5 V, correspond to the oxidation of the Au substrate.

Figure 12 discloses the Ohmic drop corrected specific activity (j) of Ba2PrRuOe and Ba2PrlrOeas a function of potential in 0.1 M HCIO4 saturated with air. Data points are acquired via cyclic voltammetry scanning at 10 mV/s, whilst rotating at 1500 rpm.

Catalyst loading for Ba2PrRuOewas to 14.3 pgoxide/ cm2disk, whilst the loading for Ba2PrlrOewas 15.1 pgoWcrm disk.

Figure 13 discloses the Ohmic drop corrected specific activity (j) of Ba2PrlrO6as a function of potential in 0.1 Μ MOH (M = Li, Na, K, Rb, Cs). Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements, except in the case of NaOH, where they represent two independent measurements. Insert shows a Tafel plot close up of onset region. Catalyst loading for plotted compounds was equal to 71.5 pgoxide/ cm2 disk.

Figure 14 discloses the Ohmic drop corrected specific activity 0 of commercial Ι1Ό2,11O2NPS and Ba2PrlrO6as a function of potential in 0.1 M HCIO4, together with Ba2Prlr06in 0.1 M RbOH. Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements. Catalyst loading for plotted compounds was equal to 14.3 pgoxide/ cm2 disk, except in the case of Ba2PrlrO6 where it was 71.5 pgoxide / cm2 disk.

Figure 15 discloses the Ohmic drop corrected specific activity 0 of commercial I1O2, lr02NPs, Ba2PrlrO6, Sr2lr04and Pr3lr07as a function of potential in 0.1 M HCIO4. Data points are acquired via controlled-potential chronoamperometric steps, whilst rotating at 1500 rpm. Error bars represent the standard deviations of at least three independent measurements, except in the case of Sr2lr04, where they represent two independent measurements. Catalyst loading for plotted compounds was equal to 14.3 pgoxide / cm2 disk. The results indicate that at least a two-dimensional perovskite layer is suitably present.

Detailed Description of the Invention

The present invention relates to the use of a perovskite-like or double perovskite material comprising iridium alongside with a number of other metals.

Figure 1 shows a schematic representation of the crystal structure of a generic A2BB’06 double perovskite, as visualized by the VESTA software, showing the B06 (dark shaded) and BO6 (light shaded) octahedra, oxygen anions (black), and A cations (dark grey). In this structure, iridium ions are typically surrounded by six oxygen ions in an octahedral fashion. Each of these iridium ions is surrounded by six B-site ions, also in an octahedral fashion, as illustrated in Figure 1. The structure may advantageously accommodate a wide range of ions in different valence states with little lattice deformation, allowing to leverage this versatility to fine tune parameters such as catalytic performance and catalyst electrochemical stability.

Applicants found that such iridium containing double perovskites according to the invention exhibit superior electrochemical activity, at an iridium content typically in the range of from 25 to 35 wt%, which is considerably lower than the 85 wt% present in Ir02.

The double perovskite structure may advantageously be determined by any suitable means, including but not limited to X-ray crystallography on powder diffraction patterns,a s se out herein below.

Electrocatalysts comprising the double perovskites according to the invention were found to exhibit a considerably higher intrinsic catalytic activity towards water oxidation in acidic media than the best catalysts reported for this reaction (Ir02). Simultaneously, the iridium weight content in these materials is typically a factor 3 lower than in the respective oxide while the catalytic activity of the compound is at least 10 times higher than Ir02, as illustrated by Figure 3.

In the structure set out above, A preferably represents an alkali or earth alkali metal, preferably Ba or Sr.

In the structure, B preferably represents a rare earth metal, preferably from the Lanthanides, such as La, Ce, Pr, Nd, Tb, and/or Y, or mixtures thereof.

Preferred materials include Ba2Celr06, Ba2NdlrC>6, Ba2Ïblr06and S^YIrOe, Ba2YlrOe; Sr2lrC>4, Pr3lrO, SrTio.75lro.25O3, Ba2PrTao.5lro.5O6and Ba2PrNbo.5lro.5O6.

Applicants found that the materials are readily available, and can be prepared in various suitable ways, including solid state chemistry at high temperatures, but also by (co)-precipitation and calcination steps.

One method involves the following steps: (a) providing a mixture of suitable precursor metal salts in a particle size and form suitable for the preparation; (b) subjecting the mixture to a temperature in the range of from 850°C to 950°C for a suitable period, and (c) sintering the resultant powder at a temperature in the range of from 1200°C to 1300°C, preferably 1225°C to 1275 °C, fora suitable period of time, optionally accompanied by frequent intermittent regrinding, and (d) furnace-cooling the obtained samples to room temperature. In a third aspect, the present invention also relates to systems comprising the electrocatalysts according to the present invention.

Step (a) may be performed by any suitable manner. Step (b) is performed by heating the mixture to the temperature for a suitable period of time, preferably for at least 2 hours. Step (c) then involves heating the thus obtained first heated mixture to sintering conditions. It may involve regrinding, while the heating is performed for a suitable time period. It was found that 48 hours generally resulted in good sintered materials. In step (d), the sintered materials are allowed to cool down slowly to room temperature, thereby avoiding any thermal shocks.

This process involves the co-precipitation of materials, but may effectively directly result in the desired coating, upon calcinations in step (iii).

The following, non-limiting examples illustrate an embodiment of the invention.

Experimental part Synthesis and characterization

Samples of Ba2Mlr06 (M = La, Ce, Pr, Nd, Tb, Y), Ba2PrRu06 and Sr2Ylr06 were synthesized using a standard solid-state reaction as described in Fu et al., Fu, W.T. and D.J.W. Ijdo, Re-examination of the structure of Ba2Mlr06 (M = La, Y): space group revised. Journal of Alloys and Compounds, 2005. 394(1-2): p. L5-L8; Fu, W.T. and D.J.W. Ijdo, On the space group of the double perovskite Ba2Prlr06. Journal of Solid State Chemistry, 2005. 178(4): p. 1312-1316; and Fu, W.T. and D.J.W. Ijdo, On the symmetry and structure of the double perovskites Ba(2)LnRuO(6) (Ln = La, Prand Nd). Solid State Communications, 2005. 136(8): p. 456-461.

Samples of Ba2Prlr06 were prepared from BaC03, Pr6On and Ir metal in platinum crucibles using the standard solid-state reaction. The well-ground mixtures were first reacted at 900 °C for a few hours. The resultant powders were then sintered at 1250 °C for several days, accompanied by frequent intermittent regrinding, and finally the samples were furnace-cooled to room temperature. All synthesis were carried out in air.

Ba2CelrC>6, Ba2NdlrOe, Ba2TblrOeand S^YIrOe prepared in a fashion similar to Ba2YlrOe, using the processes as disclosed by Fu et al., and described above. lrC>2 nanoparticles (NPs) were synthesized as described by Zhao et al., Zhao, Y.X., et al., A High Yield Synthesis of Ligand-Free Iridium Oxide Nanoparticles with High Electrocatalytic Activity. Journal of Physical Chemistry Letters, 2011.2(5): p.402-406. Ir02was used as commercially available from Aldrich, in 99.99% t.m.b., trace metals basis purity.

Samples of Sr2lrÜ4, Pr3lr07, SrTio.75lro.25O3, Ba2PrTao.5lro.5O6and Ba2PrNbo.5lro.5O6 were prepared from Sr(N03)2 (Acros organics, 99%), iridium metal (Johnson, Matthey & co.), PreOn (Elcomat-Lmf, 99.9%), T1O2 (Riedel-de Haën AG, 99.8%), Ba02(Cerac, 99%), Ta2Os(Fluka) and Nb2Os (Aldrich, 99.9% t.m.b) in alumina crucibles using the standard solid-state reaction.

The well ground mixtures were first reacted at 900°C for a few hours. Samples of Sr2lrC>4and P^lrOwere then sintered at 1200°C for 8 hours. Samples of SrTio.75lro.25O3 were sintered for 18 hours at 1100 °C; followed by 36 hours at 1250 °C.

Samples of Ba2PrTao.5lro.5O6 and Ba2PrNbo.5lro.5O6 were sintered for 48 hours at 1250 °C; followed by 12 hours at 1350 °C; followed by 12 hours at 1450 °C.

All samples were ground intermittently, and furnace-cooled to room temperature. All reactions were carried out in air. X-ray powder diffraction patterns were collected on a Philips X’Pert diffractometer, equipped with the X’Celerator, using Cu-Κα radiation in steps of 0.020° (2Θ) with 10 s counting time in the range 10° < 20 <100°. Calculations were performed by the Rietveld method, using the software Rietica. The profiles were fitted using a Pseudo-Voigt function, while the Chebyshev polynominal function with 12 parameters was used to fit the background.

Reagents: All electrolyte solutions were prepared by dissolving the appropriate amount of chemicals in 18.2 M_/cm water from Millipore Milli-Q. LiOH· H2O (Aldrich, 99.95% t.m.b.), NaOH (Sigma-Aldrich, 99.99% t.m.b.), KOH (Sigma-Aldrich, 99.99% t.m.b.), RbOH solution (Aldrich, 99.9% t.m.b.), CsOH· H2O (Aldrich, 99.95% t.m.b.), HCIO4 solution (Merck, EMSURE), HNO3 solution (Sigma-Aldrich, ACS reagent grade), NaCI04· H2O (Fluka, 99% for HPLC), Nafion solution (Aldrich) and argon (purity grade 6.0) were used as received.

Na-exchanged Nafion. The immobilizing agent (Nafion) was ion-exchanged with sodium hydroxide to avoid possible corrosion of the oxide powders when mixed together. Nafion was mixed with a 0.05 M NaOH solution in a 1:1 volumetric ratio to yield near pH neutral ion-exchanged Nafion.

Electrocatalyst preparation. Catalyst inks were prepared by sonication (Bransonic 2510, Branson) of appropriate amounts of oxide catalyst, ion-exchanged Nafion and ethanol (Sigma-Aldrich, 99.8%) to yield inks with final concentrations of 1 mgoxide/1 mhnk or 5 mgoxide /1 mhnk and 0.7 mgNaüon/1 mUnk. Next, 2.5 pl_ ink was drop-cast onto an Au electrode (0.17 cm2 geometrical area) surrounded by a Pt ring, which was polished to a mirror finish with 0.04 pm alumina slurry (Kemet) on a polishing cloth (Buehler), and sonicated for 5 minutes prior to drop-casting. After drop-casting, the ethanol was evaporated under vacuum. The final catalyst loading was 14.3 pgoxide/crrPdisk ΟΓ 71.5 pgoxide/cm2 disk with 2.2 pgNafion /cm2 disk.

Loading concentrations are specified for each graph shown. Prior to making catalyst inks, oxide powders were dry ground using a pestle and mortar to increase homogeneity of the final inks. lrC>2 NPs could not be drop-cast due to resulting agglomeration issues when dried. Therefore, they were electrodeposited on the electrode directly from solution. The particles were electrodeposited via dipping the electrode in a mildly acidic solution of NPs (2 mM precursor), and applying a constant potential of 1.3 V vs. Ag/AgCI for 120 seconds.

Surface area determination: The electrochemically active surface was estimated from double-layer capacitance using cyclic voltammetry (CV). A 100 mV window where capacitive charging is the only source of current is verified, as the so called double-layer region. Then the CVs in this window, 0.8 - 0.9 V, at different scan rates, namely v = 0.025; 0.050; 0.075; 0.1; 0.15; 0.2; 0.3; 0.4 and 0.5 V/s. From the middle of the potential window one then extract the charging current (ic) for each scan rate v.

These parameters are related to the double-layer capacitance (tC) via equation 1:

A plot of charging current versus scan rate should yield a straight line whose slope is equal to the double-layer capacitance. The double-layer capacitance is related to the electrochemically active surface area (As) via the specific capacitance (C*) as exemplified in equation 2:

where C* = 60 pF/crmis used in this work, this refers to a theoretical value calculated for oxides in general, unrelated to crystal structure.

Different methods for determining surface area such as Brunauer-Emmett-Teller (BET), Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) do exist, to name but a few important ones. However, the equipment necessary for these methods is not available to the authors, nor would they yield values closer to the ‘real’ active surface area, since they rely on particle size and related surface area prior to incorporation in thin-films. As such, the actual electrochemically active surface will be different as compared to surface areas determined via these methods.

Electrochemistry: Glassware was cleaned via boiling in a 3:1 mixture of sulfuric acid and nitric acid, and stored in permanganate solution between experiments. Prior to each experiment glassware was cleaned by boiling in Milli-Q water. Electrolyte solutions were prepared by dissolving the appropriate amount of chemicals in 200 ml. Milli-Q water. Where stated, the electrolyte was purged of oxygen by bubbling argon for 20 minutes, or saturated with air by bubbling with air for 20 minutes. Air was first bubbled through a 6 Μ KOH solution to remove impurities. All electrochemical measurements were carried out in a two compartment electrochemical cell with the reference electrode separated by a Luggin capillary. Measurements were performed using a rotator (Electrocraft Motomatic Motor Generator) to which the working electrode, loaded with thin-film catalyst, was attached. A gold wire was used as counter electrode. OER activities were obtained whilst rotating at 1500 rpm. A reversible hydrogen electrode (RHE) was employed as reference electrode, and potentials were controlled using a potentiostat/galvanostat (Autolab PGSTAT12). Electrocatalytic behaviour at room temperature was explored via cyclic voltammetry as well as via a series of controlled-potential chronoamperometric steps. In the latter, potential was held constant for 30 seconds and the steady-state measured current is plotted versus the applied potential.

Electrolyte resistance. The ohmic resistance of an electrochemical cell filled with electrolyte is given in equation 3:

In this equation, κ is the specific conductance of the electrolyte, while I* A'1 can be considered a pseudoconstant of the set-up, dependant on the geometry, size, and spacing of the electrodes. To determine the constant I· A'1 for the set-up used in this work, the resistance of the cell (filled with 0.1 M HCIO4) was extracted from impedance spectra obtained by Electrochemical Impedance Spectroscopy (EIS) in the frequency range 50 kHz - 0.5 Hz at 0.8, 0.9 and 1.0 V vs. RHE. The specific conductance of the electrolyte was then measured using a MeterLab. Table 1 shows the obtained resistances for electrolyte solutions.

CDM230 conductivity meter: The constant I· A'1 was then determined to be 1.1 cm'1 via equation 3.

All resistances used in this work were obtained by measuring κ and converting this to the resistance. The obtained values for electrolyte resistances are listed in Table 1 and these are the values used when correcting for ohmic drop, where stated.

Detection of oxygen formation: A Rotating Ring Disk Electrode (RRDE) set-up was used for determining the formation of oxygen. Briefly, cyclic voltammetry was performed on the Au disk loaded with thin-film catalyst, whilst rotating at 1500 rpm. The formed oxygen, if any, was then partially reduced back to water at the Pt ring surrounding the disk, held at 0.45V vs. RHE (02 + 4· H++ 4· e"· H20). Therefore, a decrease in ring current corresponds to the formation of oxygen at the disk.

Faradaic efficiency: The efficiency of Ba2PrlrOe was measured using the same setup as described in the previous section; in a 0.1 M HCI04 solution purged with Ar. Ring background currents were observed to be below 0.5 μΑ. To determine efficiency, one needs to know the collection factor, which is defined as stated in equation 4 (for a catalyst with 100% efficiency). For this equation to hold true, water is assumed to be oxidized to oxygen on the disk, and oxygen is assumed to be reduced to water on the Pt ring:

Linear sweep voltammetry of Ι1Ό2 NPs electodeposited on the disk in the region 1.525 < E < 1.75 V vs. RHE, while keeping the ring constant at 0.4 V vs. RHE, was used to determine the collection factor. Ir02was used because its efficiency is known to be (very near) 100%. The calculated value for N was 0.20 ± 0.01. This value is in close agreement with the collection factor as calibrated with the [Fe(CN)e]3 / [Fe(CN)e]4 couple (N = 0.23). The slight difference between calculated collection factors may be attributed to the non-ideal outward flow of O2.

The faradaic efficiency of Ba2PrlrOefor OER was then determined by dropcasting 14.3 pgoxide/ cm2 disk and doing linear sweep voltammetry in the potential window 1.25 < E < 1.75 V vs. RHE, while again keeping the ring at 0.4 V vs. RHE.

The efficiency was then calculated via equation 5:

The calculated specific area at the used loading concentration is very similar to the electrode geometrical area. Due to the fact that the calculated specific area of the oxide and the geometrical area of the electrode are comparable, the faradaic efficiency was calculated for the potential window: 1.525 < E < 1.575 V vs. RHE, where the double perovskites generally show a current density of 1 mA/crri2 oxide (see also Figure 3).

In this potential window, the faradaic efficiency of Ba2PrlrOefor OER was determined to be 82.5 ± 13 %. The large error in this value is mainly due to the fact that oxygen bubbles tend to form very easily on the surface of these particular double perovskites, much more so than in the case of e.g. electrodeposited lr02NPs. These bubbles trap part of the formed oxygen, resulting in bubble growth and lowering of the effective collection factor. As they reach a certain size, they will dissociate from the surface, but will not be detected on the ring as only dissolved oxygen is detected. This formation and subsequent release of oxygen bubbles yields a noisy ring response. However, due to the fact that the used potential window is already very close to the onset potential (about 1.45 V vs. RHE), using a lower potential window would give rise to other difficulties, e.g. ring background current interference.

To illustrate relative activity of the compounds measured in this work, all data was compared to the catalytic activity of both commercial Ir02 and freshly prepared Ir02 nanoparticles (NPs), measured under the same conditions (Figure 2).

Activity of commercially sourced iridium oxide was found to differ from that of a nanosized variant, despite both being normalized by surface area.

The nanoparticles presented in this work were prepared as exemplified by Zhao et al., Zhao, Y.X., et al., A High Yield Synthesis of Ligand-Free Iridium Oxide Nanoparticles with High Electrocatalytic Activity. Journal of Physical Chemistry Letters, 2011.2(5): p. 402-406.

However, they exhibit catalytic activity comparable to the nanoparticles of Lee et al., Synthesis and Activities of Rutile Ir02 and Ru02 Nanoparticles for Oxygen Evolution in Acid and Alkaline Solutions, Journal of Physical Chemistry Letters, 2012. 3(3): p. 399-404, even though those were synthesized via a different route. Activity of both variants of Ir02 is added as a reference.

All double perovskites accoding to the invention, as shown in Figure 3, exhibit a higher intrinsic catalytic activity for water oxidation than commercial Ir02 and Ir02 NPs.

Neither the cation present at the B-site, nor the oxidation state of iridium seems to significantly affect catalytic activity (i.e. their error bars overlap). Particularly high catalytic activity was found for a preferred double perovskite containing neodymium on the B-site, which exhibits higher catalytic activity than all other double perovskites.

However, the catalytic activity for all plotted compounds is at least one order of magnitude higher than that of commercial and nanosized Ir02. As such, the cation residing on the A- and/or B-site seems not to be the crucial component for the improved catalytic performance with respect to (nanosized) Ir02. The change in Tafel slope observed at high overpotentials appears to be caused by the formation and growth of bubbles on the surface, which adhere tightly to the surface, even at rotating speeds of 1500 rpm.

In Figure 4, the same compounds as shown in Figure 3 are depicted, only now normalized per gram of iridium.

For the internally prepared lrO2NPs as comparative examples, mass activity could not be calculated due to the way they were prepared and deposited. However, since they compare well to the nanoparticles reported by Lee et al., see also Figure 3, their data is added instead. Even though the double perovskites contain less iridium, the difference in activity is a lot less pronounced in this plot, with some double perovskites even exhibiting comparable activity (e.g. Ba2CelrOe). This is primarily due to the fact that nanoparticles exhibit an increased surface to weight ratio as compared to microsized particles, resulting in more active sites per unit mass and thus higher mass activity.

Effect of the A-site and non-lanthanides on the B-site: The effect of cations other than lanthanides on the B-site as well as the effect of the cation on the A-site was investigated via comparing S^YIrOeand Ba2YlrOe. The measured catalytic activity of these two double perovskites is plotted in Figure 5. Ba2PrlrC>6is also plotted, together with, as comparative examples, commercial and nanosized lrC>2 as a reference.

Both compounds show catalytic behaviour comparable to that of Ba2l_nIrOe-type double perovskites, despite not containing a lanthanide on the B-site (Ba2YlrOe) and/or barium on the A-site (S^YIrOe). These results indicate that the cation residing on the B-site does not necessarily have to be a lanthanide, and that A-site cation substitution may significantly affect catalytic activity, with S^YIrOe showing higher intrinsic catalytic activity than Ba2YlrC>6. However, as with the B-site cation, the cation present on the A-site appears not to be responsible for the large shift in activity as compared to Ir02. Interestingly, though S^YIrOe shows higher catalytic activity than Ba2YlrC>6, the latter proves to be more stable, see also Table 2. Electrochemical stability of double perovskites: Short term electrochemical stability for the compounds discussed so far was tested via cyclic voltammetry. Briefly, catalytic activity was measured via cyclic voltammetry in the region 1.25 - 1.8 V vs. RHE at a scan rate of 10 mV/s, before and after 500 cycles in the same potential window at 0.5 V/s.

Thereafter, the relative decrease in catalytic activity (Δ) was calculated via equation 6:

(6) wherein

is the specific current at the highest ohmic drop corrected potential in the 1st scan, and wherein

is the specific current at the same potential after 500 cycles.

In Table 2 this relative decrease in activity is given for commercial and nanosized I1O2, together with a selection of double perovskites: Ba2PrlrOe (containing IΓ4+);

Ba2NdlrOe (containing Irs+); Ba2YlrOe (containing a nonlanthanide at the B-site) and S^YIrOe (containing strontium at the A-site). 'TaMe 2.. Relative decreche in activity for W}2 and a selection of double perovskites.

It must be noted that this is quite a crude method for determining stability, with a number of factors that may not be primarily related to stability also affecting the outcome, e.g. possible film tearing, catalyst dissociation, and/or surface roughening/smoothing.

Yet, as conditions for each oxide were comparable, it is considered as a sutibale indication of relative stability and how stability differs between samples.

As can be seen in Table 2, relative stability is actually highly variable between different double perovskites, with observed values ranging between 17% and 80%. The values obtained for Ι1Ό2 however do not differ too much between commercial and nanosized Ir02at 74% vs. 81%.

To test electrochemical stability, a current of 0.25 A was passed through a cylindrical pellet of Ba2Prlr06with radius 6.5 mm and height 3 mm for 2 hours. This experiment was conducted in a two electrode set-up, where the potential was not controlled nor monitored, with the purpose of seeing which ions are most prone to dissolution. During the course of the experiment a purplish film was deposited at the cathode, which was separated from the electrolyte via centrifugation and dissolved in concentrated nitric acid, after which electrolyte and dissolved film were analyzed for dissolved metal species via Inductively Coupled Plasma (ICP).

The obtained results are summarized in Table 3:

Table 3. Double peroiraMte cation, dissolution during water oxidation.

As can be seen in Table 3, total barium (Asite cation) and praseodymium (B-site cation) concentrations exceed iridium (B’-site cation) concentrations by three orders of magnitude. This indicates that valuable iridium is lost only in small amounts during water oxidation, although activity does decrease with time (see table 2). The ratio of dissolved barium vs. praseodymium is 2.35:1, which roughly resembles their ratio in the double perovskite structure (2:1). Therefore, preferably, to increase catalyst stability, as the most readily solvable ions can actually be replaced by other metal ions.

Effect of pH on catalytic activity. Catalytic activity was observed to be vastly different in alkaline media as compared to acidic media; therefore the effect of pH on catalytic activity was investigated, while keeping the ionic strength constant. A pH series consisting of pH 1,2,3 and 13 was prepared, where the concentration of perchlorate (CIO4 -) was kept constant at 0.2 M via the addition of appropriate amounts of NaCICk The catalytic activity of Ba2PrlrOe under these conditions was measured, and the measured activities are plotted in Figure 6.

As can be seen in Figure 6, activity seems to increase with decreasing pH. However, further on in this work it is shown that cationic species affect catalytic activity in alkaline media. If such an effect is also present in acidic media, this increase in activity with decreasing values for pH could possibly be explained by the fact that Na+ concentrations decrease with decreasing pH, as fewer NaCI04 is needed to yield the same ionic strength. Values for pH lower than 1 were not measured, due to the fact that the solubility of perchlorate salts then comes into play.

Effect of B-/B’-site ordering on catalytic activity: Most double perovskites contain B- and B’- octahedra which are arranged in an orderly fashion. However, if the B- and B’-site cation are similar enough, they can form a randomly distributed perovskite, wherein no such ordering is present. The effect of the ordering of the B-/B’-site cations (rocksalt type & random) was investigated by synthesizing and investigating an iridium containing perovskite with randomly distributed B-/B’-octahedra. For this, part of the titanium cations in the cubic perovskite SrTiOswere replaced by iridium to yield SrTio75lro.2s03 (hereafter denoted SrTi03:lr(25%)).

Rietveld refinements using the X-ray powder diffraction data of SrTiC>3:lr(25%) carried out for Pm3m yielded satisfactory results for the incorporation of iridium on the titanium sites. On convergence, the agreement factors read RwP= 5.86%, RP= 4.20% and X2= 1.80, with a refined cell parameter of a = 3.91413(9) A. The observed and refined profiles are plotted in Figure 7.

The refined lattice parameter is slightly larger than the known lattice parameter for SrTi03 without iridium (a = 3.845 A). This difference can be attributed to the fact that the ionic radius of iridium is larger than that of titanium, and thus incorporation of iridium in the lattice increases the cell constant.

It was also found that iridium is not easily incorporated in a randomly distributed pero vs kite-structure, with many attempts at forming a disordered perovskite (e.g. BaPbi-xlrx03, BaZn-xlrxC>3, SrPbi-xlrxC>3, and BaTh-xIrxCb) being unsuccessful. Increasing the iridium content beyond 25% in SrTh-xIrxCbwas also unsuccessful, with the formation of a contaminant Sr2lrC>4-phase observed for higher iridium levels. As such, SrTiC>3:lr(25%) is the only randomly distributed iridium containing perovskite reported in this work. The measured catalytic activity of SrTiC>3:lr(25%) is plotted in Figure 8. Due to the evidently lower activity than double perovskites, only cyclic voltammetry was utilized to explore the catalytic behaviour. CVs of both commercial and nanosized lrC>2are added as references.

As can be seen in Figure 8, SrTiC>3:lr(25%) shows activity comparable to that of commercial lrC>2, but inferior to that of lrC>2 NPs. Although the number of iridium ions per mole is half of that in the double perovskites previously plotted, this appears not sufficient enough a change to be responsible this steep decrease in observed catalytic activity.

Effect of iridium site doping on catalytic activity: In an attempt to further decrease the iridium content in double perovskites, B’-site doping was attempted. Substitution of half of the iridium was attempted with both niobium and tantalum.

Rietveld refinement was carried out for these compounds in order to determine the degree of mixing that had occurred. For Ba2Pr(Nbo.5lro.5)C>6, all peaks in the XRD pattern were observed to have a shoulder, indicative of an additional phase. Therefore, the Rietveld refinement was carried out for a mixture of two phases:

Phase 1: Ba2Pr(lri-xNbx)C>6(spacegroup Fm3m)

Phase 2: Ba2Pr(Nbi-ylry)06(spacegroup 121m)

Ba2PrlrOeis known to form the first phase, while Ba2PrNbOeis known to form the second phase. This refinement was found to satisfactory describe the experimentally obtained XRD pattern on convergence. In the end, the agreement factors read: RwP = 5.73%, RP= 4.29% and χ2= 0.95. In Figure 9 the observed and calculated profiles of Ba2Pr(Nbo.5lro.5)C>6are shown.

For Phase 1, a value of 0.5 was found for x, whilst for Phase 2, y was determined to be 0.35. This indicates that a phase transition occurs from the lower symmetry / 2 /m phase to the higher symmetry Fm3m phase with increasing iridium content. Even though doping with niobium did not result in a single phase product, the two main phases present both contain a mixture of B’-site cations (i.e. neither phase contains solely iridium or niobium on the B’-site).

For Ba2Pr(Tao.5lro.5)Oea single phase was found to be sufficient to describe the obtained XRD pattern. Due to the observed peak splitting, the lower symmetry / 2 Im phase was used for the refinement, yielding satisfactory results. On convergence, the agreement factors read: RwP = 7.03%, RP= 4.96% and χ2= 1.52.

In Figure 10 the observed and calculated profiles of Ba2Pr(Tao.5lro.5)06are shown. Comparing Ba2Pr(Nbo 5lro5)Oewith Ba2Pr(Tao.5lro.5)06, it looks that tantalum mixes more readily than niobium, with tantalum forming a single homogeneous mixed phase while niobium forms a mixture of phases, each with slightly different ratios of Nb and Ir. The catalytic activity of both these compounds was explored via cyclic voltammetry, and the obtained CVs are plotted in Figure 11, with commercial and nanosized Ι1Ό2 added as references.

Both Ta and Nb doped double perovskites show a maximum at 1.675 V (see Figure 11, denoted *), with activity decreasing again at higher potentials. This peak is related to oxygen production, but a maximum is only seen during the first scan. After the first cycle, catalytic activity is observed to decreases drastically. This behaviour suggests that preferably, iridium is not partially substituted with other elements, as this may negatively affects the stability of double perovskites.

Similar to SrTiC>3:lr(25%), catalytic activity is comparable with commercial Ir02, but inferior with respect to nanosized I1O2. Again, the relative decrease in iridium content is not large enough to be responsible for the large observed decrease in catalytic activity. Even though B-/B’-site ordering is present in these compounds, a disorder of a different nature is introduced with the partial substitution of iridium. Namely that iridium and the substitute cation are now randomly distributed on the B’-sites. Possibly, this statistical distribution of cations at the B’-sites is cause for the observed decrease in activity.

Effect of iridium replacement on catalytic activity: In order to determine whether iridium is the active site, a double perovskite where iridium was replaced with ruthenium (Ba2PrRuOe) was also tested for its catalytic properties and compared to the iridium containing variant (Ba2Prlr06). Even though RuC^is slightly more active than I1O2 in acidic media, the ruthenium containing double perovskite shows considerately lower catalytic activity for OER than the iridium containing double perovskite, as can be seen in Figure 12. The fact that surface normalized catalytic activity is so different makes it clear that iridium must be the active site.

Double perovskites in alkaline media: During the course of this work, it was observed that catalytic activity of Ba2PrlrO6was different when measured in either 0.1 M NaOH or 0.1 Μ KOH. Therefore a series of 0.1 Μ MOH solutions with different complementary cationic species (M = Li, Na, K, Rb, Cs) was prepared, and catalytic activity of Ba2PrlrC>6was determined in each of these media.

This resulted in the data shown in Figure 13. In Figure 14, the specific activity of Ba2PrlrOein 0.1 M RbOH is compared to the specific activity of Ba2PrlrC>6in 0.1 M HCIO4, with commercial and nanosized lrÜ2 added as references. Catalytic activity of Ba2PrlrOe in 0.1 M RbOH falls in between the activity of commercial and nanosized Ir02. However the onset of oxygen evolution is delayed as compared to acidic media. The delay in onset, together with a higher Tafel slope, results in a somewhat reduced catalytic performance. It must be noted that iridium oxide is known to exhibit slightly higher catalytic performance in acidic media as compared to alkaline media, so this comparison is slightly biased.

However, catalytic performance of double perovskites in even the most active alkaline medium is lower as compared to the catalytic activity of the same compound in acidic media. This behaviour makes acidic media more interesting than alkaline media, when one is interested in high OER activity.

Further comparative examples: Related compounds in acidic media: Two iridium containing metal oxides with different structures were synthesized and investigated for their catalytic properties. Sr2lr04was selected because of the resemblance of the crystal structure to that of double perovskites. This compound belongs to the An-iA’2Bn03n+i (Ruddlesden-Popper) family, consisting of 2D alternating perovskite and rock-salt planes.

Meanwhile, Pr3lr07 adopts a Fluorite-like structure, which bears little, if any, resemblance to the double perovskite structure. This structure contains octahedrally surrounded iridium ions, which are comer-linked to only two other octahedra each, resulting in one-dimensional chains. This as opposed to Sr2lr04, which contains two-dimensional layers of comer-linked iridium octahedra, and Α2ΒΒΌ6 double perovskites, which consist of a three dimensional space filled with comer-linked octahedra. Together, these compounds were utilized to probe the effect of the crystal structure on the catalytic activity of iridium containing double perovskites. The catalytic activity of both compounds is plotted in Figure 15, and compared to the catalytic activity of Ba2Prlr06, commercial Ireland IrChNPs. Interestingly, Sr2lr04shows catalytic activity for OER comparable to that of double perovskites, suggesting that a two dimensional perovskite layer is sufficient for the increased performance displayed by double perovskites. And while Pr3lr07 exhibits activity which is comparable to the activity of nanosized Ι1Ό2, it is considerably less active than double perovskites. However, ultimately both compounds are very unstable under electrochemical conditions in acid, which results in a steep drop in catalytic activity within seconds.

Summarising, applicants have shown convincingly the good electrochemical properties of ΑςΒΙγΟθ double perovskites

In the subject electrocatalysts, the transition metal, Iridium, Ruthenium,

Osmium, Rhenium, Rhodium, and/or combinations thereof, preferably Iridium or ruthenium, was found to be the active site for water oxidation with catalytic activity dependant on the pH of the electrolyte (lower pH higher activity), and the complementary cationic species (in alkaline media). B-site cation substitution showed only a minor effect on catalytic activity, and is therefore considered as not responsible for the remarkable jump in activity as compared to I1O2.

The oxidation state of the metal, e.g. iridium, which is linked to the cation residing on the B-site, does not exhibit superior activity in either the 4+ or 5+ state. Changing the cation residing on the A-site was shown to allow to fine tune the catalytic activity.

However, as was the case for the cation residing on the B-site, the cation present at the A-site is not responsible for the jump in activity as compared to Ι1Ό2. Albeit not responsible for the remarkable shift in catalytic activity, the cations present at the A- and B-site were found to have a prominent effect on catalyst stability and relative catalytic activity.

Claims

An electrocatalyst suitable for the production of oxygen in electrolytic processes, a perovskite-like double perovskite compound comprising the following general formula: An + 1 (BB ') n / 203rl + i, wherein B stands for iridium, ruthenium, osmium, rhenium, rhodium, and / or combinations of the foregoing.

The electrocatalyst according to claim 1, wherein each B 'atom is arranged in an octahedral oxygen coordination.

The electrocatalyst according to claim 1 or claim 2, wherein the compound is at least partially in possession of a layered crystalline perovskite structure or of a double crystalline perovskite structure.

The electrocatalyst according to any of claims 1 to 3, wherein B is selected from the rare earths, such as the lanthanides or actinides, or mixtures thereof.

The electrocatalyst according to claim 4, wherein B is selected from La, Ce, Pr, Nd, Tb, and / or Y.

The electrocatalyst according to claim 4, wherein B is preferably present in the 3+ or 4+ step.

The electrocatalyst according to any of claims 4 to 6, wherein B is selected from the group consisting of Ce, Nd, Pr, La, Lu, or combinations thereof.

An electrocatalyst according to any one of the preceding claims, wherein A is (AA ") / 2.

The electrocatalyst according to any one of the preceding claims, wherein A represents alkali or alkaline earth metals, preferably Ba, Ca and / or Sr.

An electrocatalyst according to any one of the preceding claims, further comprising a metal substrate.

The electrocatalyst of claim 10, wherein a layer of a perovskite-like double perovskite electrocatalytic material is provided on the outside of the metal substrate.

The electrocatalyst of claim 10 or claim 11, wherein the metal substrate comprises titanium or lead.

The electrocatalyst according to any one of claims 1 to 11, wherein the perovskite-like double perovskite is selected from the group consisting of Ba2 CellR06, Ba2 Ndlr06, Ba2 Tblr06 and Sr2 Ylr06, Ba2 Ylr06, Sr2lr04, or any combination thereof.

A method for the production of an electrocatalyst material according to any of claims 1 to 12, the method comprising: a) providing a mixture of suitable precursor metal salts with a particle size and a shape suitable for preparation ; b) exposing the mixture to a temperature in the range of 850 ° C to 950 ° C for a suitable period, and c) sintering the resulting powder at a temperature in the range from 1200 ° C to 1300 ° C, and which is better located in the range of 1225 ° C to 1275 ° C, and this for a suitable period, optionally accompanied by the frequent intermediate regrinding, and d) cooling the obtained samples to room temperature.

A method for the production of an electrocatalyst material according to any one of claims 1 to 12, wherein the method comprises: i. preparing a first solution comprising the precursor of the materials, and ii. depositing the material precursor on a metal substrate from the solution by means of an induced change in solubility, preferably by means of a cathodic electro-deposit, and iii. subjecting the thus formed molded substrate to a heat treatment at a temperature in the range of 300 ° C to 1300 ° C, thereby forming the material that is at least partially in possession of a double perovskite structure.

Use of a perovskite-like double perovskite AA'BB'06 according to any of claims 1 to 13, or the composition obtainable by the method according to any of claims 14 or 15, as a catalyst.

A method for splitting water into oxygen and protons by electrolysis, wherein a composition according to any of claims 1 to 13 is used in an electrode.

The method of claim 17, wherein the pH of the process medium is in the range of 0 to 5.5.

A method according to claim 17 or claim 18, wherein the required electricity for the electrolytic process is supplied by using a renewable source of electrical energy selected from wind energy, solar energy, wave energy or tidal energy.