AN OXYGEN EVOLUTION REACTION ELECTRODE CATALYST ASSEMBLY, ITS USE AND A METHOD TO PRODUCE SAID ASSEMBLY

Field of the disclosure

The present disclosure relates to an electrode catalyst assembly suitable for catalyzing the oxygen evolution reaction, a method for producing such electrode catalyst assembly and use of such an electrode catalyst assembly in the oxygen evolution reaction.

Background of the disclosure

The oxygen evolution reaction (OER) is one of the most relevant anodic reactions within electrochemical cells, where it is coupled to the hydrogen evolution reaction (HER) or the CO ₂ reduction reaction (CO ₂ RR) to energy-dense carbon compounds at the cathode. It is thus of high relevance for electrochemical energy conversion and storage technologies. Oxygen generation from water oxidation at the anode is typically carried out in acidic or alkaline conditions. Operation in alkaline conditions allows the use of cheap, efficient and stable non-precious-metal catalysts, in contrast to acidic conditions, in which only expensive and scarce noble metal-based catalysts such as lrO ₂ and RuO ₂ exhibit significant stability. In alkaline conditions, the best performance and highest stabilities were observed for Ni-based multimetallic catalysts, which led to their widespread use as OER catalysts, as notably indicated in the study of Schalenbach M. et al., entitled “A perspective on low-temperature water electrolysis - challenges in alkaline and acidic technology” (J nt. J. Electrochem. Sci., 2018, 13, 1173-1226).

However, sluggish kinetics of the four-electron oxygen evolution reaction requires a significant anodic overpotential to achieve relevant geometric current densities, reducing the efficiency of the conversion of electrical to chemical energy.

Hence identifying efficient, cheap and stable OER catalysts comprising earth-abundant elements is of fundamental importance and has been a prominent field of research during the last 20 years. Among non-noble multi-metallic metal-based, OER catalysts reported so far, mixed nickel/iron/cobalt oxides, in particular, have shown stable low overpotentials at relevant geometric current densities.

In the study of Trotochaud L. et al., entitled “Solution-cast metal oxide thin film electrocatalyst for oxygen evolution" (J. Am. Chem. Soc., 2012, 134, 17253-17261), it is reported that OER electrocatalysts deposited on Au/Ti perform water oxidation at a current density of 10 mA cm ^-2 with observed overpotentials r| higher than 300 mV. Similar overpotentials were reported in the studies of McCrory C. C. L. et al., entitled “Benchmarking heterogeneous electrocatalyst for the oxygen evolution reaction” (J. Am. Chem. Soc., 2013, 135, 16977-16987) and “Benchmarking hydrogen evolving reaction and oxygen evolving reaction electrocatalysts for solar water splitting devices” (J. Am. Chem. Soc., 2015, 137, 4347-4357) in which OER electrocatalysts are deposited on glassy carbon electrodes.

The study of Cai C., et al., entitled “Engineering ordered dendrite-like nickel selenide as electrocatalyst’ (Electrochim. Acta, 2019, 295, 92-98) discloses a structure in NiSe ₂ forming dendrites on a glassy carbon electrode. This catalyst, when used in an oxygen evolution reaction, achieves an overpotential of about 299 mV at a current density of 10 mA cm ^-2 .

The study of Xu X., et al., entitled “A nickel iron diselenide-derived efficient oxygen-evolution catalyst’ (Nature Comm., 2016, 7, 12324) has shown that nickel iron diselenide is entirely converted into nickel iron oxide catalyst under oxygen-evolution conditions. In this study, this selenide-derived oxide catalyst drives oxygen evolution with an overpotential of 195 mV at a current density of 10 mA cm ^-2 .

On the other hand, nickel foam (NF) support is an efficient current collector and good support for active material deposition, due to its conductivity, mechanical strengths, relative inertness at alkaline pH and low cost. Nickel foam shows extended geometric surface areas and fine three-dimensional structures which make it attractive as a support for heterogeneous catalyst (see the study of Chaudhari N. K., et al., entitled “Nanostructured material on 3D nickel foam as electrocatalysts for water splitting”, Nanoscale, 2017, 9, 12231-12247).

Thus, the study of Zhan J-Y., et al., entitled “Rational design of cobalt-iron selenides for highly efficient electrochemical oxidation" (ACS Appl. Mater. Interfaces, 2017, 9 (39), 33833-33840) shows that Co04Fe06Se2 nanosheets on nickel foam exhibit an overpotential of 217 mV at a current density of 10 mA cm ^-2 .

The study of Wu M.-S. et al., entitled “Hollow mesoporous nickel dendrites grown on porous nickel foam for electrochemical oxidation of urea” (Electrochim. Acta, 2019, 304, 131-137) describes a microporous nickel foam with attached hollow mesoporous nickel dendrites prepared through electrochemical deposition of copper-nickel dendrites on the nickel foam followed by selective removal of the copper cores. The nickel dendrites with hollow porous structure offer transport networks in the three-dimensional conductive skeleton for fast movement of electrolyte species and electrons. In this study, the electrocatalytic oxidation of urea occurred at the catalyst-electrolyte interface. The porous nature of the nickel foam with nickel dendrites offers a large number of interconnected pores which permit the penetration and the percolation of electrolyte species into the inner region of hollow dendrites, facilitating the redox reaction of urea. Moreover, more active sites are offered for urea adsorption.

The study of Lofti N. et al., entitled “Surface modification of Ni foam by the dendrite Ni-Cu electrode for hydrogen evolution reaction in an alkaline solution” (J. Electroanal. Chem., 2019, 848, 113350) describes the nickel foam modified with nickel dendrites can be used as a catalyst in a hydrogen evolution reaction. Such electrode provides an overpotential of 202 mV at a current density of 10 mA cm ^-2 and a Tafel slope of 82 mV decade ^-1 .

The study of Wei W. et al., entitled “A bio-inspired 3D quasi-fractal nanostructure for an improved oxygen evolution reaction” (Chem. Comm., 2019, 55, 357-360) describes a catalyst having an active phase comprising Fe and Ni which is deposited on Fe-doped nanoarrays of dendritic nickel trees. The catalyst has been used in oxygen evolution reaction and presents an overpotential of 250 mv at a current density of 50 mA cm ^-2 or an overpotential of 263 mV at a current density of 100 mA cm ^-2 .

The need for OER electrocatalysts exhibiting low overpotential has not yet been fulfilled, especially under conditions close to industrial application, namely at elevated current density. There is still a search to enhance the efficiency of conversion of the electrical energy into chemical energy.

Summary of the disclosure

One or more of the above needs can be fulfilled by the oxygen evolution reaction electrode catalyst assembly according to the present disclosure wherein a catalyst comprising one or more iron oxides is deposited on a support comprising nickel foam that shows a high electrochemically surface area (ECSA); for example, on a support comprising nickel foam that shows a dendrite morphology.

According to a first aspect, the present disclosure provides an oxygen evolution reaction electrode catalyst assembly (i.e. OER electrode catalyst assembly) comprising a multi-metallic catalyst and a support, with the catalyst being deposited on the support, the catalyst comprising one or more iron oxides, the oxygen evolution reaction electrode catalyst assembly is remarkable in that it has a double layer capacitance of at least 6.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE (reversible hydrogen electrode), and in that the support comprises dendritic nickel foam comprising nickel dendrites. It has been found that the combination of a support being a nickel foam that shows a high ECSA, (i.e.. , a support being or comprising a dendritic nickel foam) with one or more iron oxides shows an improved double-layer capacitance by comparison to the regular supports. It has also been found that such improved double-layer capacitance allows a reduction of the overpotential values during an oxygen evolution reaction. It is preferred that the nickel foam that shows a high ECSA is or comprises dendritic nickel foam with nickel dendrites that are devoid of dopants. The dendritic nickel foam is a nickel foam that shows a dendrite morphology evidenced by scanning electron microscopy with dendrites.

The presence of the dendritic morphology on the nickel foam is reflected by an improved doublelayer capacitance of at least 2.0 mF of the support; preferably at least 4.0 mF. A synergy is obtained between the support and the catalyst that leads to a substantial improvement in the performance of these OER electrode catalyst assembly.

Indeed, the OER electrode catalyst assembly can show overpotential values below 260 mV when the oxygen evolution reaction is carried out at a current density of 100 mA cm ^-2 , at pH 14 and in 1 .0 M electrolyte solution; for example, below 255 mV; for example, at most 250 mV; for example, below 250 mV. For example, the OER electrode catalyst assembly can show overpotential values of at least 200 mV when the oxygen evolution reaction is carried out at a current density of 100 mA cm ^-2 , at pH 14 and in 1 .0 M electrolyte solution; for example, at least 210 mV or at least 220 mV.

Also, the OER electrode catalyst assembly can show overpotential values below 210 mV when the oxygen evolution reaction is carried out at a current density of 10 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution; for example, below 205 mV; for example, at most 200 mV; for example, below 200 mV. For example, the OER electrode catalyst assembly can show overpotential values of at least 180 mV when the oxygen evolution reaction is carried out at a current density of 10 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution; for example, at least 190 mV or at least 195 mV.

For example, the oxygen evolution reaction electrode catalyst assembly has a double layer capacitance of at least 6.0 mF or at least 7.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1 .05 V versus RHE; for example, of at least 7.5 mF, or at least 8.0 mF, or at least 8.5 mF, or at least 9.0 mF or at least 9.3 mF, or at least 9.5 mF, or at least 9.8 mF, or at least 10.0 mF. For example, the oxygen evolution reaction electrode catalyst assembly has a double layer capacitance of at most 15.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE; for example, at most 14.0 mF; or at most 13.0 mF; or at most 12.5 mF; or at most 12.0 mF.

For example, the one or more iron oxides are or comprise MeFeOx with Me being one or more transition metals. With preference, the one or more iron oxides are or comprise MeFeOx MeFeOx wherein Me is a transition metal selected from Co, Ni, V, Mo, and any mixture thereof; more preferably, the catalyst is MeFeOx MeFeOx wherein Me is a transition metal selected from Co, Ni, V, Mo, and any mixture thereof. For example, the atomic content of Me is equal to or greater than the atomic content of Fe; for example the Me: Fe ratio is at least 3.1 ; for example, the Me: Fe ratio is at least 4:1 .

For example, the catalyst further comprises at least one transition metal selected from Co, Ni, Cu, V, Mo, and any mixture thereof. Preferably, the catalyst further comprises at least one transition metal selected from Co, Ni, V, Mo, and any mixture thereof ; for example, selected from Ni and/or Co. Thus, the catalyst may be or comprise MeFeOx wherein Me is Ni and/or Co. In a preferred embodiment, the catalyst is devoid of Cu.

With preference, the catalyst further comprises at least one transition metal selected from Ni and/or Co with the atomic content of Ni and/or Co being equal to or greater than the atomic content of Fe.

For example, the catalyst or the one or more iron oxides are or comprise NiFe-OOH.

For example, the catalyst or the one or more iron oxides are or comprise one or more iron selenidederived oxides; with preference, the one or more iron selenide-derived oxides are selected from NiFeSe-dO and/or CoFeSe-dO. For example, the one or more iron selenide-derived oxides are selected from NiFeSe-dO with a Ni: Fe ratio being about 4:1. For example, the one or more iron selenide-derived oxides are selected from CoFeSe-dO with a Co: Fe ratio being about 4:1.

In an embodiment, the catalyst or the one or more iron oxides comprise one or more nickel-iron selenide-derived oxides and/or cobalt iron selenide-derived oxides wherein the atomic content of iron is lower than the atomic content of nickel and/or of cobalt respectively.

In an embodiment, the catalyst is or comprises Ni _x Fei.xSe ₂ -dO with x ranging between 0.1 and 1 , preferably x is ranging from 0.4 to 1 ; or from 0.6 to 0.9; or from 0.7 to 0.9; more preferably x = 0.8. Thus, for example, said catalyst is or comprises Nio8Feo ₂ Se ₂ -dO. Additionally, or alternatively, the catalyst is or comprises Co _y Fei.ySe ₂ -dO with y ranging between 0.1 and 1 , preferably y is ranging from 0.4 to 1 ; or from 0.6 to 0.9; or from 0.7 to 0.9; more preferably y = 0.8. Thus, for example, said catalyst is or comprise Coo 8Feo ₂ Se ₂ -dO.

For example, said oxygen evolution reaction electrode catalyst assembly has a Tafel slope of at most 70 mV decade ^-1 as determined by chronopotentiometry measurements conducted in an aqueous 1 M solution of KOH, more preferably of at most 65 mV decade ^-1 , even more preferably of at most 64 mV decade ^-1 .

For example, the metal content of the catalyst is at most 160 pmol as measured by inductively coupled plasma - mass spectroscopy (ICP-MS) on the catalyst with a 1 cm ² geometric area; for example, at most 150 pmol; for example, at most 145 pmol; for example, at most 140 pmol. For example, the metal content of the catalyst is at least 20 pmol as measured by inductively coupled plasma - mass spectroscopy (ICP-MS) on the catalyst with a 1 cm ² geometric area; for example, at least 30 pmol; for example, at least 40 pmol; for example, at least 50 pmol.

For example, said catalyst has a metal molar activity of at least 60 mA cm ^-2 mmol ^-1 ; preferably at least 80 mA cm ^-2 mmol ^-1 , more preferably at least 100 mA cm ^-2 mmol ^-1 , even more preferably at least 500 mA cm ^-2 mmol ^-1 .

For example, the mass loading of the catalyst over the support is ranging between 0.5 mg cm ^-2 and 40 mg cm ^-2 . The mass loading is determined by weighing the material before and after catalyst deposition.

Advantageously, the nickel foam is porous with a pore size diameter ranging from 100 pm to 1000 pm as determined by scanning electron microscopy, for example, ranging from 200 to 900 pm; for example, ranging from 300 to 800 pm; for example, ranging from 350 to 700 pm; for example, ranging 400 pm to 600 pm.

The dendrite morphology can be evidenced by scanning electron microscopy or reflected by an improved electrochemically active surface area of the nickel foam, or reflected by an improved double-layer capacitance.

In an embodiment, the support has a double-layer capacitance of at least 2.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1 .05 V versus RHE; for example, at least 2.5 mF; for example, at least 3.0 mF; for example, at least 3.5 mF; for example, at least 4.0 mF; for example, at least 4.5 mF. For example, the support has a double-layer capacitance of at most 8.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE; for example, at most 7.0 mF; or at most 6.0 mF; or at most 5.5 mF; or at most 5.0 mF.

Thus, in an embodiment, said dendritic nickel foam has an electrochemically active surface area of at least 70 cm ² cm _geo ' ² , preferably at least 75 cm ² cm _geo ' ² , more preferably at least 80 cm ² cm _geo ' ² . The electrochemically active surface area (ECSA) was determined from the double-layer capacitance using C _s = 60 [ F cm~ ² for the specific capacitance of a nickel-based material and the using following relationship ECSA = C _D JC _S .

In an embodiment, the nickel foam shows a dendrite morphology evidenced by scanning electron microscopy with nickel dendrites; wherein, the one or more of the following is true: the one or more dendrites are porous with a pore size diameter ranging between 0.5 pm and 40 pm as determined by scanning electron microscopy; for example, ranging from 1 to 30 pm; for example, ranging from 2 to 20 pm; for example, ranging from 3 to 10 pm; for example, ranging 4 pm to 8 pm; and/or the one or more dendrites are free of dopant; and/or the one or more dendrites are free of Fe.

According to a second aspect, the present disclosure provides an oxygen evolution reaction electrode catalyst assembly obtained by depositing an oxygen evolution reaction electrode catalyst comprising one or more iron oxides on a support being dendritic nickel foam that shows a dendrite morphology evidenced by scanning electron microscopy with nickel dendrites wherein the dendritic nickel foam is selected to have a double-layer capacitance of at least 4.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1 .05 V versus RHE.

With preference, said step of depositing comprises hydrothermal formation of layered double hydroxides of said one or more iron oxides, followed by selenisation and subsequent oxidation.

With preference, the dendritic nickel foam is obtained by electrodeposition of nickel on said nickel foam.

According to a third aspect, the present disclosure provides a method for producing the oxygen evolution reaction electrode catalyst assembly according to the first aspect, remarkable in that said method comprises a step (a) of providing a support being dendritic nickel foam that shows a dendrite morphology evidenced by scanning electron microscopy with nickel dendrites wherein the dendritic nickel foam is selected to have a double-layer capacitance of at least 4.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE followed by a step (b) of depositing a catalyst comprising one or more iron oxides on said support.

With preference, step (b) comprises the hydrothermal formation of layered double hydroxides, followed by selenisation and subsequent oxidation.

For example, step (a) comprises providing nickel foam followed by a step of electrodeposition of nickel on said nickel foam to obtain a dendritic nickel foam.

For example, the support being dendritic nickel foam provided in step (a) is selected to have an electrochemically active surface area of at least 70 cm ² cm _ge o' ² as determined using the following relationship ECSA = CDL/C _S ; wherein CDL is the double-layer capacitance in mF as determined by cyclic voltammetry in the range between +0.95 V and +1 .05 V versus RHE and C _s = 60 [ F cm~ ² .

For example, the method further comprises the step (c) of recovering the oxygen evolution reaction electrode catalyst assembly as defined in the first aspect.

According to a fourth aspect, the present disclosure provides the use of the oxygen evolution reaction electrode catalyst assembly as defined in the first or in the second aspect or as produced by the method according to the third aspect in an oxygen evolution reaction carried out under alkaline conditions.

According to a fifth aspect, the disclosure provides for a process for generating molecular oxygen by an oxygen evolution reaction, the process comprising a step of providing water and a step of water oxidation in presence of an oxygen evolution reaction electrode catalyst assembly; the process is remarkable in that the oxygen evolution reaction electrode catalyst assembly is according to the first or the second aspect or is an oxygen evolution reaction electrode catalyst assembly produced according to the third aspect.

With preference, said water oxidation is carried out at a current density ranging between 10 mA cnv ² and 100 mA cm ^-2 , or between 50 mA cm ^-2 and 100 mA cm ^-2 .

With preference, said water oxidation is carried out at a pH ranging from 11.5 to 15; for example, from 12 to 14.

With preference, said water oxidation is carried out in presence of an electrolyte solution comprising one or more basic electrolytes. With preference, the one or more basic electrolytes are or comprise at least one selected from KOH and NaOH. With preference, the one or more basic electrolytes are or comprise KOH.

With preference, said water oxidation is carried out in presence of an electrolyte solution of a concentration ranging from 0.1 to 5.0 M; for example, from 0.8 to 4.5 M. The water oxidation can be carried out under static condition or flux conditions. For example, the oxygen evolution reaction electrode catalyst assembly is selected to show overpotential values below 210 mV at a current density of 10 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution; and/or overpotential values below 260 mV at a current density of 100 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution. Description of the figures

- Figure 1 : A CDL measurement was performed on a 1 cm ² Ni plate, to estimate the specific capacitance of a nickel-based material. Cyclic voltammetry scans were performed at different scan rates from 0.95 to 1.05 V vs RHE, which is a non-Faradaic region for nickel. The objective was to evaluate the active surface area of a Ni foam with a 1 cm ² geometry. It was measured using:

- Figure 2: A Double-layer capacitance of the commercial Ni foam, NiNF, and of NiFeSe deposited on both supports. B Double-layer capacitances of the commercial Ni foam, NiNF, and of CoFeSe deposited on both supports. Measured in 1 M KOH, without stirring at room temperature

- Figure 3: Scanning Electron Microscopy images of OER catalysts A the bare nickel foam, B Cu-O, C NiMoFe-O, D CoPi, E FeCoW, F CoV-OOH, G NiFe-OOH, H NiFeSe, I CoFeSe, J CoV-O. Scale bar: 20 pm.

- Figure 4: Current density versus overpotential (J-r) profiles obtained from CP steps analysis. Data collected in 1 M KOH, under stirring, at room temperature, using an 85% IR-correction.

Figure 5: The stability of the potential under galvanostatic conditions was tested by running electrolysis at j = 50 mA cm-2 for 30 minutes. The potential response of all the samples was very stable in 1 M KOH. The potential stability % was calculated by measuring the potential increase between t = 0 min and t = 30 min, according to this equation. All the samples show stabilities > 97 %. The slight increase in the potential is likely to be due to the accumulation of bubbles in the sample, which is responsible for an increase in the resistance. 100

- Figure 6: SEM image of the dendritic Ni foam (NiNF).

- Figure 7: SEM image of NiFeSe-dO catalyst deposited on dendritic Ni foam (NiNF).

- Figure 8: SEM image of CoFeSe-dO catalyst deposited on dendritic Ni foam (NiNF).

- Figure 9: A Tafel plots calculated from CP steps data recorded in 1 M KOH. Each r/ value was plotted against log) (squares), the Tafel slope b was given by the linear fit r/ = a + b log ) (lines). B Tafel slopes b calculated from data in Figure A.

- Figure 10: Improvement in the double layer capacitance (CDL) of different catalysts reached via a simple modification of the Ni foam support: a dendritic Ni foam support was used (NiNF), leading to an increase in the CDL value for all the catalysts studied.

- Figure 11 : Impact of the NiNF support on the catalytic activity of the NiFeSe-dO catalyst.

- Figure 12: Impact of the NiNF support on the catalytic activity of the CoFeSe-dO catalyst.

- Figure 13: Flow electrochemical setup for the evaluation of catalysts’ stability. The catalyst was loaded in a two-compartment cell, separated by a Nation® membrane. A platinum mesh stacked with a platinum foil was used as the cathode. 100 mL of 1 M KOH solution was recirculated from an electrolyte container in each compartment at a 9 mL min ^-1 flow rate. A constant current density of 100 mA cm ^-2 was applied for 10 hours. The gas produced in the anodic compartment was analysed by gas chromatography. An aliquot of anolyte was collected every hour. The aliquots collected were analysed in Inductively coupled plasma mass spectroscopy (ICP-MS).

- Figure 14: Stability of Ni _x Fei. _x Se2/NiNF in 1 M KOH at 100 mA cm ^-2 . Potential measured during the test in flow conditions at 100 mA cm ^-2 using Ni _x Fei. _x Se2 (x=0.8)/NiNF as the anode. The potential is compared to the potential obtained with bare NF as the anode. A 1 M KOH solution was recirculated from a container in each compartment. The faradaic efficiency measured overtime for O2 production during the test in flow conditions using NF or Ni _x Fei. _x Se2/NiNF is displayed on the right axis. The bottom plot shows Ni, Co and Fe concentrations (in pg L ^-1 ) in the aqueous solution of 1 M KOH electrolyte flowing in the anode compartment during the stability test with NF (black) or Ni _x Fei. _x Se2 (x=0.8)/NiNF (grey) used as the anode. Such concentrations were determined by ICP-MS.

- Figure 15: Stability of Co _y Fei. _y Se2 (y=0.4)/NiNF in 1 M KOH at 100 mA cm- ² . Potential measured during the test in flow conditions at 100 mA cm ^-2 using Co _y Fei. _y Se2/NiNF as the anode. The potential is compared to the potential obtained with bare NF as the anode. A 1 M KOH solution was recirculated from a container in each compartment. The faradaic efficiency measured overtime for O2 production during the test in flow conditions using NF or Co _y Fei. _y Se2 (y=0.4)/NiNF is displayed on the right axis. The bottom plot shows Ni, Co and Fe concentrations (in pg L ^-1 ) in the aqueous solution of 1 M KOH electrolyte flowing in the anode compartment during the stability test with NF (black) or Co _y Fei. _y Se2/NiNF (grey) used as the anode. Such concentrations were determined by ICP-MS.

- Figure 16: The metal content (HM, in pmol) measured in catalysts with a 1 cm ² geometric area. The surface of the catalysts was removed from the NF support, dissolved in nitric acid and analysed in inductively coupled plasma - mass spectroscopy (ICP-MS). The surface layer could not be removed in the case of NiFe-OOH and CoV-O. nw is the total number of metal atoms, regardless of their nature.

- Figure 17: Empty bars represent the current densities measured at a fixed overpotential of 250 mV (7,7=250 mv) in a 1 M KOH electrolyte, at room temperature, with an 85% iR-correction. Dashed bars are the metal molar activities of the catalysts, which correspond to j^so m values normalized by nw.

Detailed description For the disclosure, the following definitions are given:

The terms "comprising", "comprises" and "comprised of" as used herein are synonymous with "including", "includes" or "containing", "contains", and are inclusive or open-ended and do not exclude additional, non-recited members, elements or method steps. The terms "comprising", "comprises" and "comprised of" also include the term “consisting of”.

The recitation of numerical ranges by endpoints includes all integer numbers and, where appropriate, fractions subsumed within that range (e.g. 1 to 5 can include 1 , 2, 3, 4, 5 when referring to, for example, a number of elements, and can also include 1.5, 2, 2.75 and 3.80, when referring to, for example, measurements). The recitation of endpoints also includes the recited endpoint values themselves (e.g. from 1.0 to 5.0 includes both 1.0 and 5.0). Any numerical range recited herein is intended to include all sub-ranges subsumed therein.

The particular features, structures, characteristics or embodiments may be combined in any suitable manner, as would be apparent to a person skilled in the art from this disclosure, in one or more embodiments.

In the context of the disclosure, and for the convenience of the readers, NiFeSe and NiFeSe- dO are used as a synonymous respectively for Ni _x Fei. _x Se2 and Ni _x Fei. _x Se2-dO, with x ranging between 0.1 and 1. Similarly, CoFeSe and CoFeSe-dO are used as a synonymous respectively for Co _y Fei. _y Se2 and Co _y Fei. _y Se2-dO, with y ranging between 0.1 and 1. Also, Pj stands for inorganic phosphate.

The present disclosure relates to an oxygen evolution reaction electrode catalyst assembly (i.e. OER electrode catalyst assembly), said OER electrode catalyst assembly comprising a multi-metallic catalyst and a support with the catalyst being deposited on the support, the catalyst comprising one or more iron oxides, the OER electrode catalyst assembly is remarkable in that the support comprises dendritic nickel foam. According to a definition, oxygen evolution reaction electrode catalyst assembly of the present disclosure is comprising a multi-metallic catalyst and a support, with the catalyst being deposited on the support, the catalyst comprising one or more iron oxides, the electrode catalyst assembly is remarkable in that it has a double layer capacitance of at least 6.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE and in that the support comprises dendritic nickel foam comprising nickel dendrites.

For example, the one or more iron oxides are or comprise MeFeOx with Me being a transition metal. For example, the catalyst further comprises at least one transition metal selected from Co, Ni, Cu, V, Mo and any mixture thereof; preferably, selected from Co, Ni, Cu, V, Mo and any mixture thereof; more preferably selected from Ni and/or Co. Thus, the catalyst may be or comprise MeFeOx wherein Me is a transition metal selected from Co, Ni, V, Mo, and any mixture thereof; preferably is Ni and/or Co. For example, the MeFeOx wherein Me is Ni and/or Co and further wherein the Me: Fe ratio is at least 3:1 ; preferably at least 4:1. In a prefered embodiment, the catalyst is devoid of Cu.

For example, the catalyst or the one or more iron oxides are or comprise NiFe-OOH.

As shown in the examples, when NiFe-OOH is used as catalyst and is deposited on the support being nickel foam with dendritic morphology, the OER electrode catalyst assembly shows an overpotential of 250 mV when the oxygen evolution reaction is carried out at a current density of 100 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution.

For example, the catalyst or the one or more iron oxides are or comprise one or more iron selenide-derived oxides. With preference, the one or more iron selenide-derived oxides are selected from NiFeSe-dO and/or CoFeSe-dO. For example, the one or more iron selenidederived oxides are selected from NiFeSe-dO with a Ni: Fe ratio being at least 4:1. For example, the one or more iron selenide-derived oxides are selected from CoFeSe-dO with a Co: Fe ratio being at least 4:1.

In an embodiment, the catalyst or the one or more iron oxides comprise one or more nickel iron selenide-derived oxides and/or cobalt iron selenide-derived oxides wherein the atomic content of iron is lower than the atomic content of nickel and/or of cobalt respectively.

In an embodiment, the catalyst is or comprises Ni _x Fei. _x Se2-dO with x ranging between 0.1 and 1 , preferably x is ranging from 0.4 to 1 ; or from 0.6 to 0.9; or from 0.7 to 0.9; more preferably x = 0.8. Thus, for example, said catalyst is or comprise Nio.8Feo.2Se2--dO.

As shown by the examples when Nio.8Feo.2Se2-dO is used as catalyst and is deposited on the support being nickel foam with dendritic morphology, the OER electrode catalyst assembly shows an overpotential of 198 mV when the oxygen evolution reaction is carried out at a current density of 10 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution; and an overpotential of 247 mV when the oxygen evolution reaction is carried out at a current density of 100 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution.

Additionally, or alternatively, the catalyst is or comprises Co _y Fei- _y Se2--dO with y ranging between 0.1 and 1 , preferably y is ranging from 0.4 to 1 ; or from 0.6 to 0.9; or from 0.7 to 0.9; more preferably y = 0.8. Thus, for example, said catalyst is or comprises Coo.8Feo.2Se2-dO. As shown by the examples when Coo.8Feo.2Se2--dO is used as catalyst and is deposited on the support being nickel foam with dendritic morphology, the OER electrode catalyst assembly shows an overpotential of 195 mV when the oxygen evolution reaction is carried out at a current density of 10 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution; and an overpotential of 247 mV when the oxygen evolution reaction is carried out at a current density of 100 mA cm ^-2 at pH 14 and in 1.0 M electrolyte solution.

In iron selenide-derived oxides, the substitution of selenium by oxygen allows more active sites to be exposed, in line with significantly higher densities of available active sites for the OER electrode catalyst assembly comprising selenium in comparison with OER electrode catalyst assembly devoid of selenium. The presence of selenium atoms can be used as a piece of evidence related to the method of preparation of the OER electrode catalyst assembly wherein a selenisation step is performed. For example, the selenium is present in the catalyst at an atomic content of at most 4.0 % as determined by EDX (excitation voltage 15 kV); for example, at most 3.0 %; for example, at most 2.5 %; for example, at most 2.0 %. For example, the selenium is present in the catalyst at an atomic content ranging from 0.5 to 2.5% % as determined by EDX (excitation voltage 15 kV).

The OER electrode catalyst assembly of the present disclosure shows an increased structuration that leads to significant improvement of the tested catalysts. The dendritic morphology of the nickel foam support allows having a greater electrochemically active surface area. For example, said support has a surface area of at least 70 cm ² cm _geo ' ² , preferably at least 75 cm ² cm _geo ' ² , more preferably at least 80 cm ² cm _geo ' ² . The electrochemically active surface area (ECSA) was determined from the double-layer capacitance using C _s = 60 .F cm~ ² for the specific capacitance of a nickel-based material and the using following relationship ECSA = CDL/CS.

Advantageously, the dendritic morphology of the nickel foam is or comprises nickel dendrites and/or the dendritic morphology of the nickel foam is free of dopant. For example, the nickel dendrites are free of dopant, preferably, the nickel dendrites are free of Fe.

Advantageously, the nickel foam of said support is porous with a pore size ranging from 100 pm to 1000 pm as determined by scanning electron microscopy, for example, ranging from 200 to 900 pm; for example, ranging from 300 to 800 pm; for example, ranging from 350 to 700 pm; for example, ranging 400 pm to 600 pm.

Advantageously, the one or more dendrites of the nickel foam are also porous with a pore size ranging from 0.5 pm to 40 pm as determined by scanning electron microscopy, for example, ranging from 1 to 30 pm; for example, ranging from 2 to 20 pm; for example, ranging from 3 to 10 pm; for example, ranging 4 pm to 8 pm.

The porosity of the OER electrode catalyst assembly of the present disclosure advantageously further shows the presence of smaller pores resulting from the layered structure of the OER electrode catalyst assembly. Such smaller pores have a size ranging between 30 nm and 100 nm as determined by scanning electron microscopy, preferably ranging between 35 nm and 95 nm, more preferably ranging between 40 nm and 90 nm.

For example, said OER electrode catalyst assembly has a Tafel slope of at most 70 mV decade ^-1 as determined by chronopotentiometry measurements conducted in an aqueous 1 M solution of KOH, more preferably of at most 65 mV decade ^-1 , even more preferably of at most 64 mV decade ^-1 . The lower the Tafel slope of an OER electrode catalyst assembly, the slower the increase in the overpotential with an increasing current density. Also, a small value for the Tafel slope is expected when one deals with a highly active electrocatalyst.

For example, said OER electrode catalyst assembly has a double layer capacitance of at least 6.0 mF or at least 7.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1 .05 V versus RHE, more preferably of at least 7.5 mF, even more preferably of at least 8.0 mF, most preferably of at least 8.5 mF, even most preferably of at least 9.0 mF or at least 9.3 mF, or at least 9.5 mF, or at least 9.8 mF, or at least 10.0 mF. The higher is the doublelayer capacitance of a catalyst, the higher is the density of the accessible active sites. For example, the OER electrode catalyst assembly has a double layer capacitance of at most 15.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE; for example, at most 14.0 mF; or at most 13.0 mF; or at most 12.5 mF; or at most 12.0 mF.

For example, said OER electrode catalyst assembly has a metal molar activity of at least 60 mA cm ^-2 mmol ^-1 , or at least 70 mA cm ^-2 mmol ^-1 , or at least 80 mA cm ^-2 mmol ^-1 , or at least 90 mA cm ^-2 mmol ^-1 , or at least 100 mA cm ^-2 mmol ^-1 or at least 250 mA cm ^-2 mmol ^-1 , or at least 500 mA cm ^-2 mmol ^-1 , or at least 510 mA cm ^-2 mmol ^-1 , or at least 515 mA cm ^-2 mmol ^-1 .

For example, the metal content of the catalyst is at most 160 pmol as measured by inductively coupled plasma - mass spectroscopy (ICP-MS) on the catalyst with a 1 cm ² geometric area; for example, at most 150 pmol; for example, at most 145 pmol; for example, at most 140 pmol. For example, the metal content of the catalyst is at least 20 pmol as measured by inductively coupled plasma - mass spectroscopy (ICP-MS) on the catalyst with a 1 cm ² geometric area; for example, at least 30 pmol; for example, at least 40 pmol; for example, at least 50 pmol. The method to produce the OER electrode catalyst assembly according to the disclosure comprises providing a support comprising dendritic nickel foam followed by the deposition of a catalyst comprising one or more iron oxides on said support.

For example, providing a support comprising dendritic nickel foam comprises providing nickel foam followed by a step of electrodeposition of nickel on a nickel foam to form a support comprising dendritic nickel foam; with preference, the nickel foam is pretreated before performing the electrodeposition step and/or the electrodeposition step is performed using an aqueous solution of NiCh.

Without being bound by a theory, performing a pretreatment may help the formation of Ni seeds at the surface of NF, which then improves the electrodeposition step. The pretreatment can comprise soaking the nickel foam in a solution of Nickel (II) chloride at a concentration ranging from 0.5 to 5.0 M.

For example, the deposition of a catalyst on said support comprises the hydrothermal formation of layered double hydroxides; with preference, followed by selenisation and subsequent oxidation.

This allows to provide an oxygen evolution reaction electrode catalyst assembly obtained by depositing an oxygen evolution reaction electrode catalyst comprising one or more iron oxides on a support being nickel foam selected to have a double-layer capacitance of at least 4.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE. More particularly, this allows for obtaining NiFeSe and CoFeSe -derived oxides. In the said method, the deposition of the catalyst on the support is performed during the hydrothermal step. The support is introduced in the autoclave, and the catalyst (i.e. the active phase) grows directly on its surface. It is understood that the catalyst is an oxide, with a low content of selenium. The catalyst is called “Se-derived oxides” as the selenide precursor has an important impact on the activity of the catalysts.

For example, said hydrothermal formation is carried out at a temperature ranging between 100°C and 150°C, preferably between 110°C and 140°C.

For example, said selenisation comprises providing selenium powder and at least one reducing agent, such as hydrazine. With preference, said selenisation is carried out at a temperature of at least 150°C, preferably of at least 160°C, more preferably of at least For example, said subsequent oxidation is carried out electrochemically, advantageously at a current density of at least 3 mA cm ^-2 , preferably at least 4 mA cm ^-2 , more preferably at least 5 mA cm ^-2 . With preference, said subsequent oxidation is carried out under alkaline conditions, for example at a pH ranging from 11.5 to 15, or from 12 to 14. Also, said subsequent oxidation is advantageously carried out in presence of an electrolyte solution comprising one or more basic electrolytes. With preference, the one or more basic electrolytes are or comprise at least one selected from KOH and NaOH. More preferably, the one or more basic electrolytes are or comprise KOH. For example, said subsequent oxidation is carried out in presence of an electrolyte solution of a concentration ranging from 0.1 to 5.0 M; for example, from 0.8 to 4.5 M.

For example, before the step of oxidation, the selenium is present in the catalyst at an atomic content of at least 40 % as determined by EDX (excitation voltage 15 kV); for example, at least 45 %; for example, at least 50%; for example, at least 55%.

For example, after the step of oxidation, the selenium is present in the catalyst at an atomic content of at most 4.0 % as determined by EDX (excitation voltage 15 kV); for example, at most 3.0 %; for example, at most 2.5 %; for example, at most 2.0 %.

For example, the dendritic nickel foam is selected to have a double-layer capacitance of at least 4.0 mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE and/or an electrochemically active surface area of at least 70 cm ² cm _ge o' ² as determined using the following relationship ECSA = CDL/CS; wherein CDL is the double-layer capacitance in mF as determined by cyclic voltammetry in the range between +0.95 V and +1.05 V versus RHE and C _s = 60 ^iF cm~ ² .

The OER electrode catalyst assembly was then used in an oxygen evolution reaction preferably carried out under alkaline conditions.

Thus, the disclosure provides for a process for generating molecular oxygen by an oxygen evolution reaction, the process comprising a step of providing water and a step of water oxidation in presence of an OER electrode catalyst assembly; the process is remarkable in that the OER electrode catalyst assembly is comprising a multi-metallic catalyst and a support with the catalyst being deposited on the support, the catalyst comprising one or more iron oxides, and the support comprises dendritic nickel foam.

With preference, the alkaline conditions of the oxygen evolution reaction comprise the use of a basic electrolyte, such as a solution of KOH. The concentration of KOH is ranging between 0.8 M and 1.2 M. ICP-MS studies of the electrolyte reveal that the catalyst, Ni _x Fei. _x Se2 with x ranging between 0.1 and 1 , preferably x = 0.8; and/or Co _y Fei. _y Se2 with y ranging between 0.1 and 1 , preferably y = 0.8, are stable since at a current density of 100 mA cm ^-2 , extremely small amounts of transition metals were leached from the catalyst’s surfaces, resulting in Ni, Co and Fe concentrations comprised between 100 ppb and 700 ppb in the electrolyte solution.

With preference, said oxygen evolution reaction is carried out at a current density ranging between 10 mA cm ^-2 and 100 mA cm ^-2 , or between 50 mA cm ^-2 and 100 mA cm ^-2 .

Advantageously, the current density of the said oxygen evolution reaction is at least 50 mA cm ^-2 .

With preference, said water oxidation is carried out at a pH ranging from 11.5 to 15; for example, from 12 to 14.

With preference, said water oxidation is carried out in presence of an electrolyte solution comprising one or more basic electrolytes. With preference, the one or more basic electrolytes are or comprise at least one selected from KOH and NaOH. With preference, the one or more basic electrolytes are or comprise KOH.

With preference, said water oxidation is carried out in presence of an electrolyte solution of a concentration ranging from 0.1 to 5.0 M; for example, from 0.8 to 4.5 M.

The water oxidation can be carried out under static condition or flux conditions. For example, the flux conditions comprise an electrolyte flow ranging between 5 ml min ^-1 and 15 ml min ^-1 , more preferably ranging between 7 ml min ^-1 and 13 ml min ^-1 .

Test and determination methods

Determination of surface area

Cyclic voltammetry scans with different scan rates were conducted between +0.95 and +1 .05 V vs. RHE and the difference between the forward and reverse scan current plotted against the scan rate to obtain double-layer capacitance, which was then related to the specific capacitance of the metal (Ni) using the following relationship: ECSA = CDL/CS.

Scanning Electron Microscopy (SEM)

SEM images were collected on a Sll-70 Hitachi FEGSEM. SEM measurements have been used for examining the catalyst surface morphologies. The powder sample is first dissolved in an aqueous nitric acid solution before the measurements.

Chronopotentiometry measurements Chronopotentiometry measurements are used to determine the OER activity and Tafel slope of the catalyst. Chrono-potentiometric steps (CP steps) were performed at different fixed current densities (j = 0, 5, 10, 25, 50 and 100 mA cm ^-2 ) for 5 minutes each with stirring. An 85% iR-correction was applied to avoid the contribution from the resistance of the testing system.

Cyclic voltammetry

Cyclic voltammetry measurements are used to determine the double-layer capacitance of the catalyst.

Determination of the mass loading of the catalyst over the support

Weighing before and after catalyst deposition.

Inductively coupled plasma - mass spectroscopy (ICP-MS)

ICP-MS measurements were performed on an ICP-QMS 7900 Agilent apparatus. ICP-MS measurements allow evaluating the metal content. They were used to determine the metal content of the catalyst or the metal content in the electrolyte solution.

Scanning Electron Microscopy (SEM) combined with Energy Dispersive X-Ray (EDX) Spectroscopy

SEM images were collected on a Sll-70 Hitachi FEGSEM eguipped with an X-Max 50 mm ² Oxford spectrometer for EDX measurements. The combination of SEM-EDX allows for the determination of the elemental composition of the catalysts. The powder sample is first dissolved in an agueous nitric acid solution before the measurements.

Inductively coupled plasma - optical emission spectrometry (ICP-OES)

ICP-OES measurements were performed on a Thermo Scientific iCAP 6300 duo device. ICP- OES measurements allow determining the elemental composition of the catalysts. The powder sample is first dissolved in an agueous nitric acid solution before the measurements.

Electrochemical characterization A two-compartment cell separated by a glass frit was used for electrochemical measurements. The electrolyte was an aqueous solution of 1 M KOH. A three-electrode arrangement using a platinum mesh counter electrode (Goodfellow, 2.25 cm ² ) and an Ag/AgCI/KCI _sa t reference electrode (BioLogic), which was very regularly calibrated against potassium ferrocyanide to ensure the absence of any shift in its potential. The potentials were reported vs RHE according to the following equation: ERHE = EAg/Agci + 0.197 + 0.059pH. The working electrode was positioned in the cell to minimize the distance to the reference electrode (» 1 mm), thus avoiding a large contribution from the cell in the ohmic drop (the resistance was always between 0.1 and 0.25 Q). Before each set of experiments, O2 flowed through the working electrode’s compartment for 20-30 minutes. This is an important step as it prevents any contribution from the O2 partial pressure P02 in the thermodynamic potential calculation: ENemst = EH ₂ O/O2 - 0.059 log pH + 0.015 log po ₂ - In an 02-saturated solution po ₂ =1 , and the thermodynamic potential is given by the following equation: E _N emst = EH ₂ O/O ₂ - 0.059log pH.

Each material was characterized in 10 ml_ of 1 M KOH aqueous solution following a precise protocol divided into 3 steps. Step 1 : Consecutive linear sweep voltammetry (LSV) scans were performed at a scan rate of 10 mV s ^-1 until the response was stable. Step 2: To study the oxygen evolution reaction kinetics, it is important to avoid any transient oxidation process such as the oxidation of Ni(OH)2 to NiOOH. For this purpose, chrono- potentiometric steps (CP steps) were performed at different fixed current densities (j = 0, 5, 10, 25, 50 and 100 mA cm ^-2 ) for 5 minutes each with stirring. In some cases, a stable potential was not obtained after 5 minutes, so the CP steps were extended by 5 additional minutes. The (/, E _; ) data points were collected. The overpotential at a given current density j (qj) was calculated according to the following equation r/j = E _; -1.23 with E, the potential measured at the current density j, in V vs RHE. The (j, j) points were plotted in a j-q graph. Tafel slopes were obtained by plotting q against log j. The linear fit of these plots: q = a + b log j gives the Tafel slope b. Step 3: The short-term stability of the different samples in each electrolyte was tested by running electrolysis at a fixed current density j of 50 mA cm ^-2 for 30 minutes under stirring. The pH of the electrolyte in the anode compartment did not change during electrolysis.

Electrochemically active surface area evaluation: The double-layer capacitance CDL values were determined electrochemically in an aqueous solution of 1 M KOH. All measurements were conducted in the voltage range +0.95 - +1.05 V vs RHE as it is a non-Faradaic region for most of the studied samples as well as for the NF support. An exception was made for Cu-O, which shows a Faradaic process in this region and therefore the double layer capacitance was measured in the range +0.66 - +0.76 V vs RHE for this sample. The difference between the anodic and cathodic charging currents A was obtained from CV scans at different scan rates (from 20 to 600 mV s“ ¹ ). The double-layer capacitance is given by Aj/2 = V- CDL where v is the scan rate. Electrochemically active surface areas (ECSAs) could theoretically be obtained using the relation ECSA = CDL/CS where Cs is the specific capacitance of the sample, which corresponds to the capacitance of an atomically smooth planar surface of the same material per unit area under identical electrolyte conditions. However, it is impossible to determine the reliable values of Cs for each sample. In the studies of McCrory C. C. L. et al. (cf. supra), for determination of ECSAs of various OER catalysts, the same value of Cs = 0.040 mF cm ^-2 in 1 M NaOH, based on typical reported values for metallic surfaces, was applied. However, one should be aware that this only gives an estimation of the ECSAs since, as in in the studies of McCrory C. C. L. et al., entitled “Benchmarking heterogeneous electrocatalyst for the oxygen evolution reaction" (J. Am. Chem. Soc., 2013, 135, 16977-16987), the Cs value varies significantly from one material to another.

Metal content evaluation Each catalyst deposited on NF with a 1 cm ² geometric area was carefully removed from the support. Note that this operation was not possible in the case of NiFe-OOH and CoV-O. The collected powders were dissolved in 65% HNO3 at room temperature for 10 days in Teflon tubes. The solutions were diluted with 2% HNO3 and analysed using inductively coupled plasma - mass spectroscopy (ICP-MS). The quantity of each metal in each catalyst was calculated. The total number of moles of metals DM is the sum of the number of moles of each metal contained in the catalyst. For instance, in the case of NiMoFe-O, HM = r?Ni + HMO + npe. This assumes that all the metals are potentially OER active, which is only a rough estimation as the exact nature/com position of the active sites is unknown.

Stability test in flow conditions Stability measurements was performed in a two-electrode electrochemical flow cell FLC-Standard purchased from Sphere Energy (Figure 13). The potential was measured using a leak-free Ag/AgCI/KCl3.4M micro reference (Innovative Instruments Ltd.). The gas produced in the anodic compartment was analysed by gas chromatography every 30 minutes using an SRI 8610C gas chromatograph equipped with a packed Molecular Sieve 5 A column for permanent separation. Argon (Linde 5.0) was used as the carrier gas; the flow rate was regulated using a mass flow controller (Bronkhorst). A thermal conductivity detector (TCD) was used to quantify O2. The Faradaic efficiency (FE02) was calculated by dividing the measured amount of oxygen by the theoretical amount of O2 expected. FE02 ⁼ no2, measured/nc>2, expected ⁼ nc>2, measured '4F7Q Where 0o2, measured and 0o2, expected are the measured and expected amounts of O2, Q the charge passed and Fthe Faraday constant. An aliquot of anolyte was collected every hour. The aliquots collected were analysed via inductively coupled plasma - mass spectrometry (ICP-MS).

Examples

The embodiments of the present disclosure will be better understood by looking at the examples below.

Materials Chemical reagents were purchased in reagent grade from Alfa Aesar and Merck. Nickel foam (1.6 mm in thickness, purity 99.5%%, density 0.45 g cm ^-3 , 95% porosity, 20 pores cm ^-2 ) was purchased from Goodfellow. Oxygen 5.0 was purchased from Linde. All electrochemical experiments were performed with a VSP300 BioLogic Potentiostat and the Biologic EC-Lab software was used for data analysis. Hydrothermal syntheses were performed in a Carbolite Gero CWF1213 furnace.

In particular, H2SO4, dimethylformamide, CUSO4.5H2O, tetrabutylammonium tetrafluoroborate TBABF ₄ , NH ₄ F, NaOH, KOH, (NH ₄ ) ₂ CO3, N ₂ H ₂ H ₂ O, Ni(NO ₃ ) ₂ -6H ₂ O, FeSO ₄ '7H ₂ O, Na ₂ MoO4'2H2O, Co(NO ₃ ) ₂ 6H2O, CoCI ₂ 6H ₂ O, CoCI ₂ , VCI3, NH ₄ VO ₃ WCI ₆ , Nation® (5 wt.% aqueous solution) were purchased from Sigma Aldrich. NH4CI, NiSO ₄ 6H ₂ O, NiCI ₂ 6H ₂ O, KsCeHsOr HzO, CH ₄ N ₂ O, FeCI ₃ , CH ₃ P(OH) ₂ , C ₃ H ₆ O, NasCeHsO? were purchased from Alfa Aesar.

Example 1- Analysis of the support for the catalytic assemblies

Comparative support: Nickel foam (NF)

NF was used as support for the comparative catalyst assemblies. 1 cm ² square foams were cut, and an additional section was left for electrical contact. This area was partially covered with epoxy glue to delimit the 1 cm ² area as precisely as possible. These foams were pretreated by soaking in a 3 M HCI solution for 10 minutes, to remove the nickel oxide layer formed at the surface when in contact with air. Then the substrates were sonicated for 5 minutes in ethanol, 5 minutes in water and dried with compressed air before use.

The NF support benefits from a relatively high ECSA of approximately 15 cm ² cm _ge o ^-2 (per geometric square centimetre) estimated using CDL measurement for NF and Cs measurement using a Ni plate electrode (Figure 1).

New support: Dendritic Nickel foam (NiNF)

The dendritic Ni foam support was synthesized as follows. First, a pre-treated Ni foam was soaked in a 0.1 M NiCI ₂ 6H2O aqueous solution for4 hours. An aqueous solution of NiCI ₂ 6H2O (0.1 M) and NH4CI (2 M) was prepared and used for electrodeposition. The electrodeposition was performed in a three-necked round-bottomed flask (50 mL solution) using a three- electrode arrangement with a Pt mesh counter electrode and an Ag/AgCI/KCI _sa t reference electrode. A constant current of -2.0 A cm ^-2 was applied for 100 s at the working electrode. The gas produced during deposition was delivered into a gas trap. After deposition, the sample was rinsed thoroughly with water and dried in air.

The deposition of the metallic branched structures in the presence of protons at very high current density generates H2 bubbles at the surface of the electrode, creating a porous dendritic morphology, as shown on the SEM image on figure 6. As a result of this increased structuration, as it can be seen on figure 2, the double-layer capacitance of the support was greatly increased from 0.9 mF (for NF) to 4.9 mF (for NiNF) leading to an estimated ECSA of approximately 82 cm ² cm _ge o ^-2

Example 2- Synthesis of PER electrode catalyst assemblies comprising iron-oxide catalyst deposited of a support comprising nickel foam or dendritic nickel foam

The synthesis procedures of the different PER electrode catalyst assemblies. A mass loading between 0.5 and 40 mg cm ^-2 was obtained.

Nine catalysts were chosen on the basis that they display low /710 overpotentials (below 300 mV). However, the Co-based catalyst CoPi (catalyst #2) was also chosen, despite it exhibits a /]io value greater than 300 mV in 1 M KPH, since it is widely used in literature.

Catalyst #1 - Cu-O: Cu dendrites were deposited from a 0.1 M CUSP4 5H2P and 1.5 M H2SP4 agueous solution (50 mL), according to a procedure of Angew. Chem. Int. Ed., 2017, 56, 4792- 4796. Electrodeposition was performed in a three-necked round-bottomed flask in a three- electrode arrangement using a Pt mesh counter electrode and an Ag/AgCI/KCI _sa t reference electrode (BioLogic). A constant current of -3.0 A was applied to the working electrode for 80 seconds under stirring (» 200 rpm). A dark red layer formed during deposition. The sample was rinsed with water and ethanol and dried in the air before annealing in the air (300 °C, 30 minutes, 10 °C min ^-1 ramp rate). Copper oxide nanoparticles were electrodeposited on the surface from a 0.2 mM Cu(imidazole)2Cl2 solution prepared by dissolving the complex in MeCN: H2O with a 1 :0.03 volume ratio, containing a 0.103 M TBABF4 supporting electrolyte. Electrodeposition was performed in a three-necked round-bottomed flask with 40 mL solution in a three-electrode arrangement using a Ag/AgCI/KCI _sa t reference electrode and a platinum wire as the counter electrode. Four consecutive cyclic voltammetry scans were performed in the range -0.5 V to 1.0 V vs Ag/AgCI with a scan rate of 50 mVs ^-1 . The sample was rinsed with water and dried in air. The mass loading was 24 mg cm ^-2 .

Catalyst #2 - CoPi: The procedure was adapted from the Energy Environ. Sci., 2011 , 4, 499- 504. A 0.1 M agueous methylphosphonic acid (MePi) solution was prepared and adjusted to pH 8.5 with KOH. This solution was used due to the higher solubility of the cobalt salt in MePi. CO(NO3)2'6H2O (10 mM total concentration) was added to form the deposition solution. Electrochemical deposition was performed from 100 mL of this solution in a three-necked round-bottomed flask using a three-electrode arrangement with a Pt mesh counter electrode and an Ag/AgCI/KCI _sa t reference electrode (BioLogic) under rapid stirring (» 300 rpm). A potential of 1.1 V vs Ag/AgCI was applied to the working electrode for 6 hours. A thick black deposit was obtained. We observed that the film cracked and became very fragile and brittle upon drying. Conseguently, all the electrochemical characterizations were performed directly after synthesis. The sample was rinsed with distilled water, but never dried. The mass loading measured after characterization and subsequent air drying was 35 mg cm ^-2 .

Catalyst #3 - NiMoFe-O: Synthesis was adapted from Int. J. Electrochem. Sci., 2008, 3, 908-917 and J. Am. Chem. Soc., 2015, 137, 4347-4357. Ni foams were pre-treated following the method described above. A mixture of NiSC>4-6H2O (1.157 mmol, 0.304 g), NiCl2'6H2O (1.350 mmol, 0.321 g), Na ₂ MoO4-2H ₂ O (0.207 mmol, 0.050 g), FeSO ₄ '7H ₂ O (0.180 mmol, 0.05 g), K3C6H5O7 H2O (0.925 mmol, 0.300 g) and (N^ COs (4.163 mmol, 0.400 g) was prepared in 50 mL H2O, pH 7. Electrochemical deposition was performed from 50 mL of this solution in a three-necked round-bottomed flask using a three-electrode arrangement with a Pt mesh counter electrode and an Ag/AgCI/KCI _sa t reference electrode (BioLogic) under rapid stirring (» 200 rpm). A constant current of -861 mA was maintained for 10 minutes to form a dark grey deposit. The sample was rinsed with water and dried in air. The mass loading was 18 mg cm ^-2 .

Catalyst #4 - NiFe-OOH: The procedure was reproduced from Energy Environ. Sci., 2018, 77, 2858-2864. FeCl3'6H2O (0.814 mmol) was dissolved in 60 mL of absolute ethanol and sonicated for 5 minutes to form a yellow solution. A second solution was prepared by adding FeCls.6H2O (0.814 mmol) and NH4HCO3 (4.46 mmol) in 60 mL of absolute ethanol. The pretreated NF substrate was immersed in the FeCh-EtOH solution overnight. After this step, the substratewas immersed inthe second solutionfor6 hours understirring. The film was rinsed with H2O and dried in air. The etched mass was -14 mg cm ^-2 . This negative loading is due to a galvanic replacement reaction occurring between Ni° at the surface of the substrate and Fe ³⁺ ions in solution.

Table 1 : Atomic composition of NiFeOOH obtained using EDX (excitation voltage 15 kV) with comparison to the reported values on Ni foam of the reference work.

The values on dendritic nickel foam are expected to be the same since EDX provides an analysis of the catalyst (since it is deposited on the surface of the support) not of the support. Catalyst #5 - NiFeSe-dO: The procedure was adapted from Nat. Commun., 2016, 4, 499- 504.

Step 1 : Synthesis of NiFe-LDH An aqueous solution containing Ni(NOs)2-6H2O (0.91 mmol), FeSC>4-7H2O (0.23 mmol), NH4F (4.55 mmol) and CH4N2O (11.36 mmol) was prepared. 18.2 mL of this solution was added to a 23 mL Teflon-lined autoclave reactor and a piece of pretreated Ni foam was added. The autoclave was heated to 120°C for 16 hours to form a light green deposit on the nickel foam. The sample was rinsed with EtOH and H2O and dried in air.

Step 2: Synthesis of NiFeSe: Se powder (1.70 mmol), NaOH (3.4 mmol) and N2H4 (0.7 mmol, 36.6 L) were dissolved in DMF (11.3 mL). The previously obtained NiFe-LDH sample was added to a 23 mL Teflon-lined autoclave and the solution was added. The autoclave was heated to 180°C for 1 hour to obtain a black deposit of NiFeSe, which was rinsed with water and dried in air.

Step 3: Synthesis of NiFeSe-dO: NiFeSe was activated electrochemically in 1 M KOH, by applying a constant current of 5 mA cm ^-2 for 2 hours, until a stable potential was observed. This led to the formation of the selenide-derived oxide NiFeSe-dO. The mass loading was 6 mg cm ^-2 .

Table 2: Atomic composition of NiFeSe-dO obtained using EDX (excitation voltage 15 kV), before and after activation with comparison to the reported values on Ni foam of the reference work

From the results, it can be seen that the NiFeSe-dO is Nio.8Feo.2Se2-dO

Also, NiFeSe-dO has a metal content of 133 pmol as measured by inductively coupled plasma - mass spectroscopy (ICP-MS) on the OER electrode catalyst assembly with a 1 cm ² geometric area

Catalyst #6 - CoFeSe-dO: The procedure was adapted from ACS Appl. Mater. Interfaces, 2017, 9, 33833-33840. Step 1 : Synthesis of CoFe-LDH An aqueous solution containing Co(NC>3)2-6H ₂ O (0.900 mmol), FeSO ₄ -7H ₂ O (0.227 mmol), NH4F (4.45 mmol) and CH ₄ N ₂ O (11.16 mmol) was prepared. 18.2 mL of this solution was added to a 23 mL Teflon-lined autoclave reactor and a piece of pre-treated Ni foam was added. The autoclave was closed and heated to 120°C for 16 hours to form a red deposit on the nickel foam. The sample was rinsed with EtOH and H ₂ O and dried in air.

Step 2: Synthesis of CoFeSe

Se powder (1.70 mmol), NaOH (3.4 mmol) and N ₂ H ₄ (0.7 mmol, 36.6 L) were dissolved in DMF (11.3 mL). The previously obtained CoFe-LDH sample was added to a 23 mL Teflon- lined autoclave and the solution was added. The autoclave was heated to 200°C for 6 hours to obtain a black deposit of CoFeSe, which was rinsed with water and dried in air.

Step 3: Synthesis of CoFeSe-dO: CoFeSe was activated electrochemically in 1 M KOH, by applying a constant current of 5 mA cm ^-2 for 2 hours, until a stable potential was observed. This led to the formation of the selenide-derived oxide CoFeSe-dO. The mass loading was 7 mg cm ^-2 .

Table 3: Atomic composition of CoFeSe-dO obtained using EDX (excitation voltage 15 kV), before and after activation with comparison to the reported values on Ni foam of the reference work.

Catalyst #7 - FeCoW: This material was synthesized according to Science, 2016, 352, 333- 337. For solution A, FeCh (146.0 mg, 0.9 mmol), CoCI ₂ (116.9 mg, 0.9 mmol) and WCh (356.9 mg, 0.9 mmol) were dissolved in 2 mL ethanol. The as-prepared solution was cooled in an ice bath for 2 hours. Solution B consisted of 2 mL of ethanol and 0.18 mL of water, which was cooled in an ice bath for 2 hours. Solution B was added to solution A, and 1 mL of propylene oxide was slowly added to this mixture. After a few minutes, a dark green gel was formed. This gel was aged for 3 days in ambient conditions and became dark red/brown. The gel was immersed in ethanol and agitated to form a colloidal suspension. This suspension was washed three times with ethanol. 5 mL of the suspension in ethanol were kept and mixed with 250 L of 5 wt.% aqueous Nation® solution. The solution was sonicated for at least 30 minutes and a homogeneous dispersion was obtained. The ink was drop-cast onto the pre-treated nickel foam (40 mg cm ^-2 ) and dried in air. Note that we also performed CO2 supercritical drying on another batch but the activity of the catalyst was lower than with the undried colloidal material.

Catalyst #8 - CoV-OOH: The synthesis was adapted from Energy Environ. Sc/., 2018, 11, 1736-1741. COCI2.6H2O (0.5 mmol) was added to 50 mL H2O under stirring. The flask was heated to 30°C in a water bath. A solution of NaOH (1 mmol) and NH4VO3 (0.17 mmol) in 20 mL of deionized water was prepared and added dropwise to the C0CI2 solution. The solution was maintained at 30°C under stirring for a further 15 min to obtain a green-brown precipitate. The precipitate was washed three times with water and two times with ethanol with centrifugation and disposal of the supernatant between washing steps. The sample was dispersed in 5 mL of ethanol and an ink was prepared by mixing this nanoparticles suspension with 250 L of 5 wt.% aqueous Nation® solution. The solution was sonicated for at least 30 minutes to obtain a homogeneous dispersion. The ink was drop-cast onto the pre-treated nickel foam (25 mg cm ^-2 ) and dried in air.

Catalyst #9 - CoV-O: The synthesis was adapted from a reported procedure described in ACS Catal., 2018, 8, 644-650. C0CI2 (70.11 mg, 0.54 mmol), VCI ₃ (28.31 mg, 0.18 mmol), urea (68.11 mg, 1.13 mmol) and trisodium citrate TSC (37,2 mg, 0.14 mmol) were added to 18 mL H2O. As the precursors are air and light-sensitive, the solution was kept in the dark and degassed with Ar for 30 minutes. The solution (18 mL) was transferred into a Teflon-lined autoclave (23 mL capacity) and a piece of the pre-treated Ni foam was added. The autoclave was heated to 150 °C for 14 hours to obtain a very thin deposit. The mass loading was 0.5 mg cm ^-2 .

Deposition on the supports

For all NF-catalyst assemblies, the same Ni foam was exclusively used as the conductive support. The syntheses followed the reported procedures as strictly as possible with slight adaptations to enable deposition on a 1 cm ² nickel foam (NF) support. Figure 3A shows an SEM image of the bare NF support.

CuO (catalyst #1) and NiMoFe-0 (catalyst #3) were cathodically electrodeposited on Ni foam from aqueous solutions of the metallic precursors. This simple method resulted in the formation of dendritic structures as shown by SEM analysis (see figures 3B and 3C respectively). NiMoFe-0 shows a particularly strong adhesion to the substrate because the presence of nickel in the as-deposited catalyst ensures a continuous interface with NF and thus a small lattice mismatch. EDX was used to confirm the elemental composition. With regards to NiMoFe-O, the use of NF as the support led to small modifications in the chemical composition of the film.

CoPi (catalyst #2) was anodically electrodeposited, which resulted in the slow formation of large dendrites with poor adhesion to the NF support (see figure 3D).

FeCoW (catalyst #7) (see figure 3E) and CoV-OOH (catalyst #8) (see figure 3F) were synthesized from nanomaterial dispersions mixed with a Nation ink and drop-cast and dried onto the surface of the support to form thick layers, which remained well attached to the support despite large cracks. The major limitation of this method is the hydrophobicity of dry Nation which promotes the formation of an air film trapped at the catalyst/electrolyte interface and limits their contact area. To reach a stable OER activity, these catalysts must be kept under oxidative conditions (10 min at 5 mA cm ^-2 ) in an aqueous electrolyte for the Nation’s hydrophilic domains to swell and become predominant in the bulk, as described in the study of Zhao Q., et al., entitled “Diffusion and interfacial transport of water in Nation" (J. Phys. Chem. B, 2011 , 115, 2717-2727). A modification in the nanostructure of FeCoW was observed after this process, revealing the formation of a thin layer structure. Therefore, once the Nation network is fully hydrated, it can be concluded that the metallic sites are exposed to the electrolyte. This allows dissolution/precipitation equilibria at the interface, resulting in the formation of highly nanostructured surfaces. The tungsten content in FeCoW is lower than expected.

NiFe-OOH (catalyst #4) (see figure 3G) was made through the galvanic exchange of the Ni-based substrate with Fe ³⁺ precursor, followed by the deposition of a bimetallic oxyhydroxide. As the support is the only source of nickel, it is etched during the reaction. This ensures a very high adhesion of the catalyst on the support, but also slightly modifies the material by etching its Ni backbone. Nevertheless, this method is particularly interesting as it is very simple.

NiFeSe and CoFeSe -derived oxides (catalysts #5 and #6, respectively) were synthesized by a more complex three-step procedure involving hydrothermal formation of layered double hydroxides (LDHs), followed by selenisation and subsequent reoxidation. The SEM images and EDX elemental analysis of the resulting Nio.8Feo.2Se2-dO (see figure 3H) and Coo.8Feo.2Se2-dO (see figure 3I) catalysts revealed the formation of very dense, thick and mechanically stable deposits at the surface of NF with fine nanostructures in the range of 30-100 nm. However, the formation of a selenide involves hazardous synthesis steps and its reoxidation leads to toxic selenite waste products. CoV-0 (catalyst #9) was synthesized by a simple one-step hydrothermal procedure involving the coprecipitation of Co and V in a mixed-phase composed of a fine LDH nanostructure (see figure 3J). This method is simple but results in a very low loading on nickel foam.

For NiNF-catalyst assemblies, the above procedure was repeated but the support was changed to be NiNF for all catalyst assemblies.

Example 3- Catalytic performance of PER catalyst

The catalytic activity of each material was measured under alkaline conditions, in an aqueous 1 M KOH electrolyte solution. Chrono-potentiometric steps (CP steps) were performed at different fixed current densities (j = 0, 5, 10, 25, 50 and 100 mA cm ^-2 ) for 5 minutes each under stirring. This method is better than linear sweep voltammetry as it ensures that exclusively the oxygen evolution reaction response is measured and other contributions to the current are eliminated, such as the oxidation of Ni(OH)2 to NiOOH in Ni-containing materials. Furthermore, these CP steps give some information regarding the stability of the potential measured at different current densities on a short time scale. A possible drawback resides in some additional ohmic drop associated with the accumulation of oxygen bubbles at the surface of the electrode, as described in the study of Angulo A., et al., entitled “Influence of bubbles on the energy conversion efficiency of electrochemical reactors" (Joule, 2020, 4, 555-579).

The j-r] profiles and overpotentials at j = 10 and 100 mA cm ^-2 of the tested catalysts are displayed in figure 4. Table 4 report the overpotential /710 and r]ioo obtained from CP steps. Data collected in 1 M KOH, under stirring, at room temperature, using an 85% IR-correction. The overpotential /710 value is considered as a figure of merit for OER catalysts. The pH was 14.0 (+/- 0.2).

Oxygen evolution must be carried out at higher current densities to meet the requirements for the electrochemical conversion and storage of renewable energy, such as solar-driven water splitting and CO2 reduction technologies. The focus was therefore directed on the catalytic activities of the nine catalysts at 100 mA cm ^-2 on each support (see figure 4 and table 4). While NF requires an overpotential as high as 563 mV, the nine catalysts enable an important drop in /j oas compared to their support. The results are provided in Table 4.

Additionally, the electrodes were held at a fixed current density (j = 50 mA cm ^-2 ) for 30 minutes in static conditions to verify that measured potentials were stable over a longer reaction time. In all cases, this confirmed no clear degradative reactions (see figure 5 for stability data). Table 4: Overpotentials obtained in 1 M KOH for the catalysts deposited on NF or NiNF. /710 and /]ioo (in mV) are the overpotentials at 10 and 100 mA cm ^-2 respectively. AE represents the difference between existing between the composition with support NF and NiNF. The pH was 14.0 (+/- 0.2). All the catalyst assemblies with NF support were found to have /710 values between 211 and 347 mV, a significant 136 mV range. The lowest overpotentials were obtained for NiFeSe- dO/NF (/]io = 211 mV) and CoFeSe-dO/NF (/710 = 212 mV). The highest overpotentials were obtained for CoV-O/NF (/710 = 331 mV) and Cu-O/NF (7710 = 347 mV).

Despite differences in the slopes, the trend in OER activity is the same at 10 and 100 mA cm ^-2 for assemblies using NF support. NiFeSe-dO/NF remains at the head of the group with a low overpotential of /7100 = 264 mV, followed by CoFeSe-dO/NF and NiFe-OOH/NF with the same overpotential value of 289 mV, and by FeCoW/NF with r^oo = 293 mV. CoV- O/NF (/7IOO =397 mV) and Cu-O/NF (<7100 =432 mV) show the lowest performances at all current densities.

Significantly higher activity was observed for catalyst assemblies having dendritic nickel foam (NiNF) compared to the ones with NF support. For the support alone, 80 mV decrease of its /710 value (/7io|NiNF = 331 mV), and a 129 mV decrease of its r|ioo value (/7ioo|NiNF = 434 mV) were measured. This marked increase in activity makes NiNF a highly interesting anodic support material for electrolytic cells.

The results obtained with the most active catalysts, namely NiFeSe-dO (catalyst #5) and CoFeSe-dO (catalyst #6), deposited on NiNF are here detailed. From the SEM images displayed respectively on figures 7 and 8, it appears that the nanostructure of NiFeSe-dO and CoFeSe-dO was maintained on NiNF support, resulting in a hierarchical porous structure composed of three levels of porosities: the large pores of the nickel foam (» 500 pm), the pores formed by the nickel dendrites (» 1-10 pm) and the meso-/macropores resulting from the layered structure of the catalysts (» 30-100 nm)

As shown by the results on table 4, the use of NiNF as a support obtains overpotentials in 1 M KOH at 100 mA cm ^-2 that are below 260 mV for 3 catalyst assemblies, namely CoFeSe-dO/NiNF, NiFeSe-dO/NiNF and NiFe-OOH/NiNF, while this was not achieved with the comparative support NF. The 3 catalyst assemblies have an overpotential in 1 M KOH at 100 mA cm ^-2 of at most 250 mV, and for CoFeSe-dO/NiNF and NiFeSe- dO/NiNF is below 250 mV.

Also, the use of NiNF as support allows obtaining overpotentials in 1 M KOH at 10 mA cm ^-2 that is at most 230 mV for 4 catalyst assemblies, namely CoFeSe-dO/NiNF, NiFeSe- dO/NiNF, FeCoW/NiNF and NiFe-OOH/NiNF while this was achieved only for the selenide-derived oxides containing catalyst assemblies with the comparative support NF. The use of NiNF as support in the selenide-derived oxides catalyst assemblies allows achieving an overpotential in 1 M KOH at 10 mA cm ^-2 of below 200 mV. The /710 obtained for CoFeSe-dO and NiFeSe-dO are among the lowest value reported in the literature so far.

Example 4 - Tafel analysis

The dependence of the OER kinetics on the applied potential for the catalysts is well illustrated through Tafel analysis (Figure 9). The fastest increase in current density upon potential increase occurs at the surface of NiFe-OOH, with a slope as low as 36 mV decade ^-1 . NiFeSe-dO and FeCoW also showed low Tafel slope values of 55 and 56 mV decade ^-1 . The obtained values for the other catalysts ranged from 63 to 83 mV decade ^-1 . Larger Tafel slopes were systematically observed for reported ones, except for NiFe-OOH. This is likely because CP steps were used to measure Tafel slopes in place of LSV scans, which is the general methodology employed in the recent literature. Also, as a consequence of the CP steps, accumulation of O2 bubbles at the surface of the electrodes contributes some extra resistance, especially at higher current densities, which might be a source of increased Tafel slope values.

The Tafel slope decreased slopes by 1 to 30 mV decade ^-1 was observed except for Cu-0 (table 5). For example, the decrease was from 55 to 54 mV decade ^-1 in the case of NiFeSe-dO, and from 72 to 63 mV decade ^-1 in the case of CoFeSe-dO. The gain of 20 to 30 mV decade ^-1 in most cases is significant as it induces a large decrease in the overpotential at high current densities. In particular, NiFe-OOH displays a remarkably low Tafel slope of 20 mV decade ^-1 when deposited on NiNF.

Table 5: Tafel slopes b obtained for our catalysts deposited either on NF or NiNF Tafel slopes b (in mV decade ^-1 ) were measured in 1 M KOH. The corresponding change in b (Ab) arises from the microstructured substrate.

Example 5 - Double-layer capacitance determination

Comparing catalysts with different morphologies in terms of their intrinsic activities requires determination of the density of electrochemically active sites, a very important yet challenging analysis. Depending on parameters such as their nanostructure, porosity and lattice structure, catalysts can show very different interactions with the surrounding electrolyte. The density of electrochemically active and accessible sites can vary a lot from one catalyst to another. A range of techniques can be used to relate the total OER activity of a catalyst to the intrinsic activity of each active site. The density of accessible active sites can be obtained through the determination of the electrochemically active surface area (ECSA) by means of the double-layer capacitance determination.

OER catalysts behave as capacitors: upon application of a potential, a charge build-up is observed at the catalyst-electrolyte interface. The capacitance of a catalyst in the absence of any Faradaic process is the double layer capacitance CDL. ECSAS can theoretically be calculated from CDL values, however, they are difficult to obtain accurately. Indeed, ECSA = CDL /CS where Cs is the specific capacitance of the material, which corresponds to the capacitance of an atomically smooth planar surface of the same material per unit area under identical electrolyte conditions. While CDL can be experimentally determined by measuring the non-faradaic capacitive current associated with double-layer charging from the scan-rate dependence of the cyclic voltammograms (CVs),

The increase in conductivity upon application of anodic potentials is expected to impact the measured CDL: this value must be measured in a conductive range to be representative of the capacitance of the active material. In the present case, CDL was measured by cyclic voltammetry in the range from +0.95 to +1.05 V vs RHE, which is a non-faradaic and conductive region for all the catalysts (except Cu-0 which was characterized between +0.66 and +0.76 V vs RHE).

Here the CDL values for the nine OER catalysts and the NF support are reported, measured in 1 M KOH using electrodes with 1 cm ² geometric areas (see figure 10). Large differences were indeed observed: some of the studied catalysts have CDL values in the range of 1 mF, slightly larger than that of the Ni foam (0.9 mF), while CoFeSe-dO, NiMoFe- O and NiFeSe-dO have much larger CDL values of 3.65, 2.40 and 2.35 mF, respectively. These results are in line with the observation of extremely fine nanostructures for these three catalysts (see respectively figure 31, 3H and 3C).

The CDL value of NiFeSe-dO on NiNF was a factor of four greater than on the NF, reaching a high CDL value of 9.6 mF (Figure 2A). The CDL value for CoFeSe-dO was roughly three times higher, giving an extremely large CDL value of 10.3 mF (Figure 2B). Therefore, the use of NiNF support enables a significant increase in the density of accessible active sites.

To determine how the increase in CDL impacted the catalytic activity, the NiNF-deposited catalysts were evaluated using our standard electrochemical characterisation procedure detailed above. The impact of the NiNF support on the catalytic activity of NiFeSe-dO and CoFeSe-dO is shown in figure 11 and 12 respectively. The j-q profiles of figures 11 and 12 are obtained by recording chrono-potentiometric steps in 1 M KOH, under stirring at room temperature, using an 85% iR-correction. For both catalysts, the /710 and /7100 values decreased upon substitution of NF with NiNF. For NiFeSe-dO, replacing NF with NiNF decreased /] from 211 mV to 198 mV and /] ofrom 264 mV to 247 mV. With CoFeSe-dO, the improvement was even more significant with /710 decreasing from 212 mV to 195 mV and /7100 decreasing from 289 mV to 247 mV. This represents a substantial improvement in the performance of these OER catalysts.

For the other catalysts, a large increase in the CDL was also observed by shifting from NF to NiNF as the support (Figure 10). In all cases, apart from FeCoW, the CDL values increased by a factor between 2.4 and 5, reflecting the improvement in the specific surface area provided by NiNF. FeCoW shows a unique 9.7-fold increase in its CDL, larger than that of the support itself. This is likely due to altered morphology of this catalyst when moving from NF to NiNF, leading to an increased density of accessible active sites.

Example 6 - Water splitting flow experiment

Stable operation under continuous flow conditions at high current density is an important property of OER systems. A water-splitting experiment under flow conditions (Figure 13) was designed to test our best catalysts (NiFeSe-dO/NiNF and CoFeSe-dO/NiNF) under conditions closer to industrial applications. The catalyst was loaded in a two-compartment cell separated by a Nation® membrane with a platinum mesh-based cathode. The anolyte and catholyte were 1 M KOH aqueous solutions. These solutions were recirculated in each compartment from electrolyte containers. A current density of 100 mA cm ^-2 was applied for 8 hours and the potential response as well as FEo ₂ were measured over time. Online monitoring of the elements presents in solution during electrolysis also allowed quantitative measurement of metal leaching from the catalysts. The flow experiment described above was performed using NF, NiFeSe-dO/NiNF (figure 14) and CoFeSe-dO/NiNF (figure 15) as the anodes. It can be seen that these two catalysts show extremely stable potentials over 8 hours of electrolysis at a high current density of 100 mA cm ^-2 . NF performs oxygen evolution at 1.96 ± 0.06 V vs RHE. The initial increase in potential is attributed to the oxidation of nickel, which occurs at the surface of the foam but also reaches its subsurface during the first hour of electrolysis under such a high current density. Both NiFeSe-dO/NiNF and CoFeSe-dO/NiNF perform oxygen evolution at potentials as low as 1.58 ± 0.02 V vs RHE, with outstanding stability, which represents a major improvement as compared to NF. The faradaic efficiency for O2 production (FEo ₂ ) was evaluated by measuring the amount of oxygen produced at the anode. During the first hour of the experiment, the headspace of the anolyte container was saturated in gas. After this equilibration period, FEo ₂ was very stable, with mean values of 98.0 ± 1.5 %, 97.8 ± 3.0 % and 98.5 ± 1.9 % for NF, NiFeSe-dO/NiNF and CoFeSe-dO/NiNF respectively. This confirms that oxygen evolution was the only process occurring at the surface of these catalysts.

The concentrations of Ni, Co and Fe in the anolyte were measured every hour by ICP-MS (figures 14 and 15, bottom plot). In the case of NF, the Ni concentration is in the range 40-120 ppb. This concentration does not increase over time, which proves the very high stability of this support in 1 M KOH under a high current density. The concentration of Fe is comprised between 250 and 450 ppb and is stable over time. This Fe content in a KOH electrolyte is common (see for example the study of Spanos I. et al., entitled “Facile protocol for alkaline electrolyte purification and its influence on a Ni-Co oxide catalyst for the oxygen evolution reaction”, ACS Catal., 2019, 9, 8165-8170). In the case of NiFeSe-dO/NiNF, a small increase in the Ni concentration was observed at the beginning of the experiment, as it reaches 300 ppb. After this small increase, the concentration slowly stabilises at around 100 ppb, which corresponds to the background concentration measured in the case of NF. No additional Ni was dissolved over the course of the reaction. As a result, NiFeSe-dO/NiNF is an extremely stable catalyst in these conditions. As expected, no Co was detected for NF or NiFeSe-dO /NiNF. In the case of CoFeSe-dO/NiNF, Ni and Co concentrations in the range 200-700 ppb were measured. This means that some Co and some Ni are dissolved from the surface of the catalyst at the beginning of the experiment, but do not accumulate in the solution, thus revealing the absence of continuous dissolution over the course of the electrolysis. The Fe concentrations remained constant for both catalysts and correspond to the background concentration measured with bare NF. In conclusion, these two catalysts have shown a high chemical and mechanical stability over 8 hours under continuous flow conditions, at a high current density of 100 mA cm ^-2 and an electrolyte flow of 9 mL min ^-1 .

Finally, the evaluation of the total number of moles of metal atoms (DM) in a catalyst on a 1 cm ² geometric area electrode can provide access to another useful information, namely the mass activity or molar activity. A high molar activity effectively translates into a lower cost, as a lower number of metal atoms are required to perform OER catalysis at a given overpotential.

An value can be obtained through the dissolution of the catalyst layer and analysis using ICP-MS. The metal content of the nine catalysts is displayed in figure 16. Ni _x Fei- _x Se2 and Co _y Fei. _y Se2 have the lowest metal content whereas FeCoW has the highest.

The metals molar activity was calculated by dividing the current density at r) = 250 mV by the metal content of each catalyst (see figure 17).

The data clearly show that not only Ni _x Fei. _x Se2 and Co _y Fei. _y Se2 display the largest current densities but they do it with the lowest amount of metals, therefore they show extremely high metal molar activities (520 and 634 mA cm ^-2 mmol ^-1 , respectively). In contrast, all other catalysts have much lower molar activities, in the 20-50 mA cm ^-2 mmol ^-1 range. In the case of FeCoW, while the number of metals is high, only a small fraction is involved in the OER. Consequently, an improved exposition of the metal sites to the electrolyte might be key in the enhancement of OER activity of this catalyst.